Version leap to 1.0.2

This shall avoid the risk of false positives as long as the test is new.

COPYRIGHT

@@ -1,12 +1,12 @@
 Vreixo Formoso <metalpain2002@yahoo.es>,
 Mario Danic <mario.danic@gmail.com>, 
 Thomas Schmitt <scdbackup@gmx.net>
-Copyright (C) 2007-2008 Vreixo Formoso, Mario Danic, Thomas Schmitt
+Copyright (C) 2007-2010 Vreixo Formoso, Mario Danic, Thomas Schmitt
 
 
     This program is free software; you can redistribute it and/or modify
-    it under the terms of the GNU General Public License version 2 as
+    it under the terms of the GNU General Public License version 2 or later
-    published by the Free Software Foundation.
+    as published by the Free Software Foundation.
 
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of

ChangeLog

@@ -1 +1,235 @@
+bzr branch lp:libisofs/for-libisoburn (to become libisofs-1.0.2.tar.gz)
+===============================================================================
+- no novelties yet
+
+libisofs-1.0.2.tar.gz Tue Wed 23 2011
+===============================================================================
+* Bug fix: iso_write_opts_set_aaip(opts, 1) could cause fatal miscalculation
+           of the root directory size. This eventually truncated directory
+           tree and spoiled all data file content.
+* Bug fix: Volume Descriptor Set Terminator contained non-zero bytes in
+           the reserved field (ECMA-119 8.3.4). The bytes stem from the
+           previously written Volume Descriptor.
+* New API calls iso_tree_clone(), iso_stream_clone.
+* New IsoFileSourceIface version 2 with method clone_src().
+* New IsoStreamIface version 4 with method clone_stream().
+* New public function prototype iso_node_xinfo_cloner.
+* New API calls iso_node_xinfo_make_clonable(), iso_node_xinfo_get_cloner().
+* New public iso_node_xinfo_cloner instance aaip_xinfo_cloner().
+* New API calls iso_node_get_next_xinfo(), iso_node_remove_all_xinfo().
+* New API call iso_node_remove_tree().
+* New API call iso_write_opts_set_old_empty().
+
+libisofs-1.0.0.tar.gz Mon Jan 17 2011
+===============================================================================
+* Bug fix: ECMA-119 directory names were truncated to 8 characters if
+           lowercase characters or full ASCII are allowed.
+* New API call iso_write_opts_set_untranslated_name_len()
+* New API call iso_write_opts_set_allow_dir_id_ext()
+* New API call iso_memory_stream_new(). (Was formely a private call.)
+
+libisofs-0.6.40.tar.gz Fri Dec 10 2010
+===============================================================================
+* New API call iso_write_opts_set_disc_label(), new system area type
+  3 = SUN Disk Label for booting SUN SPARC systems.
+* New API call iso_write_opts_set_will_cancel() avoids start of write thread
+  and is to be used to inquire the future image size.
+* New error reply code ISO_DISPLACE_ROLLOVER for external data sources with
+  address displacement.
+
+libisofs-0.6.38.tar.gz Sat Oct 23 2010
+===============================================================================
+* New API calls iso_write_opts_attach_jte() and iso_write_opts_detach_jte()
+  allow to use libjte for jigdo production.
+* New API call iso_write_opts_set_tail_blocks() for tail padding inside
+  ISO image.
+* New API call iso_image_generator_is_running() to learn when the write thread
+  is done.
+* New API calls iso_image_add_mips_boot_file(),
+  iso_image_get_mips_boot_files(), iso_image_give_up_mips_boot().
+* New API call iso_write_opts_set_partition_img() for appending e.g. a small
+  empty FAT12 filesystem which may be used on USB stick.
+
+libisofs-0.6.36.tar.gz Wed Sep 15 2010
+===============================================================================
+* New API function iso_write_opts_set_part_offset() controls creation of
+  an MBR with a first partiton table entry that bears non-zero start address.
+  A second set of volume descriptors and directory tree+tables gets created
+  which can be used to mount the image at the partition start.
+* Hiding all non-API symbols from the linker by use of --version-script
+* Automatic C++ detection in libisofs.h by using macro __cplusplus
+* Corrected several memory leaks and potential NULL pointer evaluations
+  in case of memory shortage.
+* Now with history of release notes in ./ChangeLog file.
+
+libisofs-0.6.34.tar.gz Tue Jun 29 2010
+===============================================================================
+* New API call iso_image_set_boot_catalog_hidden()
+* New API call iso_node_get_hidden()
+* New IsoHideNodeFlag bit LIBISO_HIDE_BUT_WRITE
+* New error code ISO_BOOT_NO_CATALOG
+* Opportunity to reduce compile line length by including "config.h"
+
+libisofs-0.6.32.tar.gz Mon May 03 2010
+===============================================================================
+* New API call iso_image_set_boot_catalog_weight()
+* New API call iso_image_add_boot_image()
+* New API calls el_torito_set_boot_platform_id(), el_torito_set_id_string(),
+  el_torito_set_selection_crit()
+* New API calls iso_image_get_all_boot_imgs(), el_torito_get_boot_platform_id(), 
+  el_torito_get_load_seg(), el_torito_get_load_size(), el_torito_get_bootable(), 
+  el_torito_get_id_string(), el_torito_get_selection_crit(),
+  el_torito_get_isolinux_options(), el_torito_get_boot_media_type()
+* New API call el_torito_seems_boot_info_table()
+
+libisofs-0.6.30.tar.gz Sat Apr 17 2010
+===============================================================================
+* New API call iso_write_opts_set_system_area() acts like mkisofs option -G.
+* New API call iso_write_opts_set_pvd_times().
+* Now able to produce a bootable System Area from an ISOLINUX mbr/isohdp
+  [fp]x*.bin file and an ISOLINUX El Torito bootable image (isolinux.bin).
+* Now able to produce the same Joliet names as mkisofs.
+* New API calls iso_read_opts_load_system_area() and 
+  iso_image_get_system_area() for multi-session handling of MBRs.
+
+libisofs-0.6.28.tar.gz Wed Feb 10 2010
+===============================================================================
+* Bug fix: Random checksum index could sneak in via boot catalog node 
+  and cause a SIGSEGV.
+* Improved compilability out of the box on FreeBSD.
+
+libisofs-0.6.26.tar.gz Wed Jan 20 2010
+===============================================================================
+* Bug fix: Invalid old checksum tags were preserved with
+  iso_write_opts_set_overwrite_buf(), if the new session produced no checksums.
+* The checksum buffer for the emerging image gets now marked as invalid if
+  image generation is canceled.
+* More graceful reaction on filesystems where ACL are not enabled but
+  nevertheless requested by the application.
+* Adaptions to problems reported by Debian buildd.
+
+libisofs-0.6.24.tar.gz Thu Oct 08 2009
+===============================================================================
+* Bug fix: Short Rock Ridge names got stripped of trailing blanks when loaded
+  and written again to a follow-up session. Long names could lose inner blanks.
+* Bug fix: Avoided to return NULL or single blanks as content of id strings by
+  API calls iso_image_get_volset_id() ... iso_image_get_biblio_file_id().
+* New API call iso_write_opts_set_scdbackup_tag().
+
+libisofs-0.6.22.tar.gz Tue Aug 25 2009
+===============================================================================
+* New API call iso_write_opts_set_record_md5() for writing MD5 sums.
+* New API call iso_read_opts_set_no_md5() for importing MD5 sums.
+* New API calls iso_image_get_session_md5() and iso_file_get_md5().
+* New API calls iso_md5_start(), iso_md5_compute(), iso_md5_clone(),
+  iso_md5_end(), iso_md5_match() for own MD5 computations.
+* New API call iso_util_decode_md5_tag() to recognize and parse checksum tags.
+* New API call iso_file_make_md5() to equip old file nodes with MD5.
+* Improvements with ./configure and its help text.
+
+libisofs-0.6.20.tar.gz Sun May 30 2009
+===============================================================================
+* Optional automatic detection and recording of hard link 
+  relations between files.
+* Support for restoring hard link relations by the app.
+
+libisofs-0.6.18.tar.gz Fri Apr 17 2009
+===============================================================================
+* Opportunity to set the input charset automatically from an eventual xattr
+  "isofs.cs" of the image root node.
+* New general filter API to inquire and remove filters.
+* Specialized APIs for installing filters which are based on external processes
+  or based on zlib.
+* New API call to inquire the original source path of a data file in an
+  emerging image.
+
+libisofs-0.6.16.tar.gz Wed Mar 11
+===============================================================================
+* Bug fix: The ".." directory record pointed to the same data block as the "."
+  entry.
+* Bug fix: The use of iso_write_opts_set_rrip_version_1_10() caused a wrong
+  size announcement in the CE entry which points to the ER signature 
+  of the image root.
+* New API call iso_write_opts_get_data_start() inquires the start address of
+  the data section of an emerging ISO image.
+* ISO image generation does not absolutely depend on the availability of
+  character set "WCHAR_T" with iconv_open(3) any more.
+
+libisofs-0.6.14.tar.gz Sat Feb 28 2009
+===============================================================================
+* New API calls iso_image_set_ignore_aclea(), iso_read_opts_set_no_aaip()
+  control import of ACL and xattr.
+* New API calls iso_write_opts_set_aaip(), iso_write_opts_set_aaip_susp_1_10()
+  control output of ACL and xattr into generated ISO image.
+* New API call iso_file_source_get_aa_string(), new function member
+  get_aa_string() in IsoFileSource_Iface allow to access opaquely encoded ACL 
+  and xattr. New function handle aaip_xinfo_func attaches aa_strings to 
+  IsoNode objects.
+* New API calls iso_node_get_acl_text(), iso_node_set_acl_text(),
+  iso_node_get_perms_wo_acl() allow inquiry and manipulation of ACLs in 
+  IsoNode objects.
+* New API calls iso_node_get_attrs(), iso_node_set_attrs() allow inquiry and
+  manipulation of xattr in IsoNode objects.
+
+libisofs-0.6.12.tar.gz Wed Nov 26 2008
+===============================================================================
+* New API calls iso_set_local_charset() and iso_get_local_charset()
+* New API calls iso_write_opts_set_rrip_version_1_10() and
+  iso_write_opts_set_dir_rec_mtime()
+* New API call el_torito_set_isolinux_options() allows to patch ISOLINUX boot
+  images and to generate a isohybrid MBR on the fly. Such an MBR makes the ISO
+  image bootable from disk-like hardware, e.g. from USB stick. The ISOLINUX 
+  boot image has to be of syslinux 3.72 or later to allow MBR generation.
+* Old API call el_torito_patch_isolinux_image() is deprecated now.
+
+libisofs-0.6.10.pl01.tar.gz Wed Nov 19 2008
+===============================================================================
+* Bug fix: If images generated by mkisofs were loaded then files of size 0
+  could share their size information with files that contain data. Ticket #144.
+* Bug fix: ISOLINUX boot images were patched suitable for El Torito but not for
+  an eventual MBR added by SYSLINUX script isohybrid.
+
+libisofs 0.6.10  Mon Oct 6 2008:
+===============================================================================
+* Bug fix: Patching of existing ISOLINUX boot images led to a SIGSEGV.
+* Bug fix: Adding a new ISOLINUX boot image or patching of an existing one
+  caused a read operation although writing had already begun.
+
+libisofs-0.6.8.tar.gz Thu Sep 18 2008
+===============================================================================
+* Support for very large data files in the ISO 9660 image 
+  (Level 3, multi-extent)
+* Bug fix: it was assumed that isolinux images were always a multiple of 4
+  bytes
+* New API call iso_image_update_sizes() to refresh recorded file sizes
+  immediately before image generation begins
+
+libisofs-0.6.6.tar.gz Sun Jun 1 2008
+===============================================================================
+* Bug fix: major,minor numbers of device files were not read properly from
+  existing images
+* Bug fix: iso_tree_path_to_node() returned 1 if a directory path component was
+  a non-directory file
+* New API call iso_special_get_dev() retrieves major, minor numbers of device
+  files
+
+libisofs-0.6.4.tar.gz Sun Apr 27 2008
+===============================================================================
+* Extended information: iso_node_add_xinfo()
+* New node iteration: iso_dir_find_children()
+* Custom image file content via iso_tree_add_new_file()
+* Missing feature added to map a disk file to an arbitrary image file path via
+  iso_tree_add_new_node()
+* Obtain image path of a node object via iso_tree_get_node_path()
+* Various bugfixes
+
+libisofs-0.6.2.1.tar.gz Thu Feb 14 2008
+===============================================================================
+* FIX: missing buffer.h preventing build from succeeding
+
+Libisofs 0.6.2
+===============================================================================
+* Initial release of new generation libisofs
+* Completely new API
+* Long term commitment to ABI libisofs.so.6
 

Makefile.am

@@ -16,7 +16,9 @@ libisofs_libisofs_la_LDFLAGS = \
 
 # Eventually enabling system adapters for ACL and EA.
 # ts A90409: Eventually enabling use of zlib.
-libisofs_libisofs_la_CFLAGS = $(LIBACL_DEF) $(XATTR_DEF) $(ZLIB_DEF)
+# ts B00927: Eventually enabling use of libjte (Jigdo Template Extraction)
+libisofs_libisofs_la_CFLAGS = $(LIBACL_DEF) $(XATTR_DEF) $(ZLIB_DEF) \
+                              $(LIBJTE_DEF)
 
 # ts A90114 : added aaip_0_2.*
 
@@ -42,7 +44,6 @@ libisofs_libisofs_la_SOURCES = \
 	libisofs/stream.c \
 	libisofs/filter.h \
 	libisofs/filter.c \
-	libisofs/filters/xor_encrypt.c \
 	libisofs/filters/external.c \
 	libisofs/filters/zisofs.c \
 	libisofs/filters/gzip.c \
@@ -73,81 +74,117 @@ libisofs_libisofs_la_SOURCES = \
 	libisofs/iso1999.c \
 	libisofs/data_source.c \
 	libisofs/aaip_0_2.h \
-	libisofs/aaip_0_2.c
+	libisofs/aaip_0_2.c \
+	libisofs/md5.h \
+	libisofs/md5.c
 libisofs_libisofs_la_LIBADD= \
 	$(THREAD_LIBS)
 libinclude_HEADERS = \
 	libisofs/libisofs.h
 
+install-exec-hook:
+	$(LIBBURNIA_LDCONFIG_CMD) "$(DESTDIR)$(libdir)" || echo 'NOTE: Explicite dynamic library configuration failed. If needed, configure manually for:' "$(DESTDIR)$(libdir)"
+
 ## ========================================================================= ##
 
 ## Build demo applications
 noinst_PROGRAMS = \
-	demo/lsl \
+	demo/demo 
-	demo/cat \
-	demo/catbuffer \
-	demo/tree \
-	demo/find \
-	demo/ecma119tree \
-	demo/iso \
-	demo/isoread \
-	demo/isocat \
-	demo/isomodify \
-	demo/isoms
 
+#	demo/tree \
+#	demo/find \
+#	demo/iso \
+#	demo/isoread \
+#	demo/isocat \
+#	demo/isomodify \
+#	demo/isoms
+
+#	demo/ecma119tree \
+#	demo/lsl \
+#	demo/cat \
+#	demo/catbuffer \
 #	demo/isogrow
 
-demo_lsl_CPPFLAGS = -Ilibisofs
-demo_lsl_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
-demo_lsl_SOURCES = demo/lsl.c
 
-demo_cat_CPPFLAGS = -Ilibisofs
+# ts A90807
-demo_cat_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
+# Consolidated demo code for having less linker mesages with a make run.
-demo_cat_SOURCES = demo/cat.c
+demo_demo_CPPFLAGS = -Ilibisofs
+demo_demo_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
+demo_demo_SOURCES = demo/demo.c
 
-demo_catbuffer_CPPFLAGS = -Ilibisofs
+# ts A90806
-demo_catbuffer_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# This includes fsource.h and thus is no API demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_lsl_CPPFLAGS = -Ilibisofs
-demo_catbuffer_SOURCES = demo/cat_buffer.c
+# demo_lsl_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
+# demo_lsl_SOURCES = demo/lsl.c
 
-demo_tree_CPPFLAGS = -Ilibisofs
+# ts A90806
-demo_tree_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# This includes fsource.h and thus is no API demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_cat_CPPFLAGS = -Ilibisofs
-demo_tree_SOURCES = demo/tree.c
+# demo_cat_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
+# demo_cat_SOURCES = demo/cat.c
 
-demo_find_CPPFLAGS = -Ilibisofs
+# ts A90806
-demo_find_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# This inlcudes buffer.h and thus is no API demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_catbuffer_CPPFLAGS = -Ilibisofs
-demo_find_SOURCES = demo/find.c
+# demo_catbuffer_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_catbuffer_SOURCES = demo/cat_buffer.c
 
-demo_ecma119tree_CPPFLAGS = -Ilibisofs
+# ts A90807
-demo_ecma119tree_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# Consolidated in demo/demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_tree_CPPFLAGS = -Ilibisofs
-demo_ecma119tree_SOURCES = demo/ecma119_tree.c
+# demo_tree_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_tree_SOURCES = demo/tree.c
 
-demo_iso_CPPFLAGS = -Ilibisofs
+# ts A90807
-demo_iso_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
+# Consolidated in demo/demo
-demo_iso_SOURCES = demo/iso.c
+# demo_find_CPPFLAGS = -Ilibisofs
+# demo_find_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_find_SOURCES = demo/find.c
 
-demo_isoread_CPPFLAGS = -Ilibisofs
+# ts A90806
-demo_isoread_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# This inlcudes lots of internal .h files and thus is no API demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_ecma119tree_CPPFLAGS = -Ilibisofs
-demo_isoread_SOURCES = demo/iso_read.c
+# demo_ecma119tree_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_ecma119tree_SOURCES = demo/ecma119_tree.c
 
-demo_isocat_CPPFLAGS = -Ilibisofs
+# ts A90807
-demo_isocat_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# Consolidated in demo/demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_iso_CPPFLAGS = -Ilibisofs
-demo_isocat_SOURCES = demo/iso_cat.c
+# demo_iso_LDADD = $(libisofs_libisofs_la_OBJECTS) $(libisofs_libisofs_la_LIBADD)
+# demo_iso_SOURCES = demo/iso.c
 
-demo_isomodify_CPPFLAGS = -Ilibisofs
+# ts A90807
-demo_isomodify_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# Consolidated in demo/demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_isoread_CPPFLAGS = -Ilibisofs
-demo_isomodify_SOURCES = demo/iso_modify.c
+# demo_isoread_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_isoread_SOURCES = demo/iso_read.c
 
-demo_isoms_CPPFLAGS = -Ilibisofs
+# ts A90807
-demo_isoms_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# Consolidated in demo/demo
-	$(libisofs_libisofs_la_LIBADD)
+# demo_isocat_CPPFLAGS = -Ilibisofs
-demo_isoms_SOURCES = demo/iso_ms.c
+# demo_isocat_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_isocat_SOURCES = demo/iso_cat.c
+
+# ts A90807
+# Consolidated in demo/demo
+# demo_isomodify_CPPFLAGS = -Ilibisofs
+# demo_isomodify_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_isomodify_SOURCES = demo/iso_modify.c
+
+# ts A90807
+# Consolidated in demo/demo
+# demo_isoms_CPPFLAGS = -Ilibisofs
+# demo_isoms_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# 	$(libisofs_libisofs_la_LIBADD)
+# demo_isoms_SOURCES = demo/iso_ms.c
 
 # demo_isogrow_CPPFLAGS = -Ilibisofs -Ilibburn
 # demo_isogrow_LDADD = $(libisofs_libisofs_la_OBJECTS) \
@@ -156,27 +193,31 @@ demo_isoms_SOURCES = demo/iso_ms.c
 # demo_isogrow_SOURCES = demo/iso_grow.c
 
 
-## Build unit test
+## ts A90428 , ticket 147, The test code does not use the API and is totally
+##                         outdated in its creation of mocked objects.
+##                         A volunteer is needed to rewrite it using the API. 
+
+# ## Build unit test
  
-check_PROGRAMS = \
+# check_PROGRAMS = \
-	test/test
+# 	test/test
-	
+# 	
-test_test_CPPFLAGS = -Ilibisofs
+# test_test_CPPFLAGS = -Ilibisofs
-test_test_LDADD = $(libisofs_libisofs_la_OBJECTS) \
+# test_test_LDADD = $(libisofs_libisofs_la_OBJECTS) \
-	$(libisofs_libisofs_la_LIBADD) -lcunit
+# 	$(libisofs_libisofs_la_LIBADD) -lcunit
-test_test_LDFLAGS = -L.. -lm
+# test_test_LDFLAGS = -L.. -lm
-	
+# 	
-test_test_SOURCES = \
+# test_test_SOURCES = \
-	test/test.h \
+# 	test/test.h \
-	test/test.c \
+# 	test/test.c \
-	test/test_node.c \
+# 	test/test_node.c \
-	test/test_image.c \
+# 	test/test_image.c \
-	test/test_tree.c \
+# 	test/test_tree.c \
-	test/test_util.c \
+# 	test/test_util.c \
-	test/test_rockridge.c \
+# 	test/test_rockridge.c \
-	test/test_stream.c \
+# 	test/test_stream.c \
-	test/mocked_fsrc.h \
+# 	test/mocked_fsrc.h \
-	test/mocked_fsrc.c
+# 	test/mocked_fsrc.c
 
 ## ========================================================================= ##
 
@@ -224,6 +265,8 @@ EXTRA_DIST = \
 	doc/susp_aaip_2_0.txt \
 	doc/susp_aaip_isofs_names.txt \
 	doc/zisofs_format.txt \
+	doc/checksums.txt \
+	libisofs/libisofs.ver \
 	libisofs/aaip-os-dummy.c \
 	libisofs/aaip-os-linux.c \
 	libisofs/aaip-os-freebsd.c

README

@@ -4,15 +4,68 @@
 
 Released under GPL (see COPYING file for details).
 
-Copyright (C) 2008 Vreixo Formoso, Mario Danic, Thomas Schmitt
+Copyright (C) 2008 - 2010 Vreixo Formoso, Mario Danic, Thomas Schmitt
 
 libisofs is part of the libburnia project (libburnia-project.org)
 ------------------------------------------------------------------------------
 
-libisofs is a library to create an ISO-9660 filesystem, and supports extensions
+                        Download, Build and Installation
-like RockRidge or Joliet. It is also a full featured ISO-9660 editor, allowing 
+
-you to modify an ISO image or multisession disc, including file addition and 
+libisofs code is mantained in a Bazaar repository at Launchpad 
-removal, change of file names and attributes, etc
+(https://launchpad.net/libisofs/). You can download it with:
+
+$ bzr branch lp:libisofs/for-libisoburn
+
+Our build system is based on autotools. For preparing the build you will need 
+autotools of at least version 1.7. If you have download the code from the
+repository, first of all you need to execute
+
+   ./autogen.sh
+   
+on toplevel dir to execute autotools.
+
+Alternatively you may unpack a release tarball for which you do not need
+autotools installed. For the most recent release of libisofs see:
+   http://libburnia-project.org/wiki/Releases
+
+To build libisofs it should be sufficient to go into its toplevel directory 
+and execute
+
+   ./configure --prefix=/usr
+   make
+
+To make the libraries accessible for running resp. developing applications 
+   make install
+
+On GNU/Linux it will try to run program ldconfig with the library installation
+directory as only argument. Failure to do so will not abort installation.
+One may disable ldconfig by ./configure option --disable-ldconfig-at-install .
+
+By use of a version script, the libisofs.so library exposes no other function
+names but those of the API definitions in <libisofs/libisofs.h>.
+If -Wl,--version-script=... makes problems with the local compiler, then
+disable this encapsulation feature by
+    ./configure --disable-versioned-libs
+    make clean ; make
+
+The ./configure script checks for the availability of supporting libraries.
+If found, they will become mandatory for the emerging libisofs.so and all
+applications which use it. This dependency can be avoided by configure options
+    --disable-libacl   avoid use of ACL functions like acl_to_text()
+    --disable-xattr    avoid use of xattr functions like listxattr()
+    --disable-zlib     avoid use of zlib functions like compress2()
+
+See INSTALL file for general options of ./configure.
+
+
+------------------------------------------------------------------------------
+
+libisofs is a library to create an ISO-9660 filesystem, supports extensions
+like RockRidge or Joliet, and introduces an own extension AAIP.
+It is a full featured ISO-9660 editor which composes and changes the directory
+tree of an ISO image. This tree and its newly imported data file contents get
+then written as independent single-session image or as add-on session for the
+image from where the tree was originally loaded.
 
 Features:
 ---------
@@ -20,32 +73,54 @@ Features:
 - Image creation
 	- Creates ISO-9660 images from local files.
 	- Support for RockRidge and Joliet extensions.
-	- Support for ISO-9660:1999 (version 2)
+	- Support for ISO-9660:1999 (version 2).
-	- Support for El-Torito bootable images.
+	- Support for El-Torito bootable images. Tested are: PC-BIOS and EFI.
-	- Full featured edition of file names and attributes on the image.
+	- Support for multi-extent data files up to 400 GB (level 3).
-	- Several options to relax ISO-9660 constraints.
-	- Special options for images intended for distribution (suitable default
-	  modes for files, hiding of real timestamps...)
-- Multisession
-	- Support for growing an existing image
 	- Full-featured edition of the image files, including: addition of new
 	  files, removing of existent files, moving files, renaming files,
 	  change file attributes (permissions, timestamps...)
-	- Support for "emulated multisession" or image growing, suitable for non
+	- Optional recording per file of non-ISO 9660 features:
-	  multisession media such as DVD+RW
+	  ACL, xattr, content MD5, hard link relations.
+	  They do not hamper image readability by operating systems but can be
+	  retrieved only via libisofs.
+	- Optional zisofs compression, gzip compression, external filter
+	  processes.
+	- Several options to relax ISO-9660 constraints.
+	- Special options for images intended for distribution (suitable
+	  default modes for files, hiding of real timestamps...).
+- Image reading
+	- Image tree and data heap can be verified by stream reading and
+	  eventually recorded MD5 tags.
+	- Directory tree and file attributes of ISO 9660 session get loaded
+	  into memory for editing or for extraction into local filesystem.
+	- File content can be read by applications.
+	- Automatic zisofs decompression.
+	- Optional application of gzip decompression or external filter
+	  processes.
+	- Eventually recorded MD5 of data file can be obtained, MD5 of data
+	  stream can be computed and compared.
+	- Helper functions for restoring ACL and/or xattr to the local
+	  filesystem.
+- Multisession
+	- Support for growing an existing image on multi-session media.
+	- Support for "emulated multisession" on overwriteable media such as
+	  DVD+RW, USB sticks, regular files.
+	- Support for blindly prepared add-on sessions (mkisofs style -M -C)
+	  suitable for pipes which lead to an external burn program.
 - Image modification
-	- It can create a completely new image from files on another image.
+	- Creates a completely new image from files out of another image and
-	- Full-featured edition of image contents
+	  eventual editing operations. Suitable for any target medium.
 - Others
-	- Handling of different input and output charset
+	- Handling of different input and output charset.
 	- Good integration with libburn for image burning.
 	- Reliable, good handling of different kind of errors.
 
 Requirements:
 -------------
 
-- libburn 0.4.2 headers must be installed at compile time. It is not required
+- iconv() functions for character set conversion must be available.
-  at runtime.
+  Either implicitely as in Linux or by a separate library like libiconv
+  on FreeBSD.
   
 Know bugs:
 ----------
@@ -55,32 +130,20 @@ Multisession and image growing can lead to undesired results in several cases:
 a) Images with unsupported features, such as:
 	- UDF.
 	- HSF/HFS+ or other Mac extensions.
-	- El-Torito with multiple entries.
+	- ECMA-119 Extended attributes.
-	- ECMA-119 with extended attributes, multiple extends per file.
-	- Non El-Torito boot info.
-	- zisofs compressed images.
 	- ...
    In all these cases, the resulting new image (or new session) could lack some
-   features of the original image.
+   features of the original image. Nevertheless, the ECMA-119 System Area with
+   an eventual Master Boot Record gets preserved by default.
    In some cases libisofs will issue warning messages, or even refuse to grow
    or modify the image. Others remain undetected. Images created with libisofs
    do not have this problems.
 
-b) Bootable El-Torito images may have several problems, that result in a new
+b) Bootable El-Torito images may have problems, that result in a new image that
-   image that is not bootable, or that boots from an outdated session. In many
+   is not bootable, or that boots from an outdated session. In some cases it
-   cases it is recommended to add boot info again in the new session.
+   might be necessary to add boot info again in a new first session.
-   
+    - There is no safe way to modify hidden boot images, as the size of the 
-    - isolinux images won't be bootable after a modify. This is because 
+      boot image can't be figured out.
-      isolinux images need to have hardcoded the root dir lba. libisofs cannot 
-      know whether an image is an isolinux image or not, so the user is 
-      responsible to tell libisofs that it must patch the image, with the
-      el_torito_patch_isolinux_image() function. This problem could also exists
-      on other boot images.
-	- Most boot images are highly dependent of the image contents, so if the 
-	  user moves or removes some files on image it is possible they won't boot 
-	  anymore.
-	- There is no safer way to modify hidden boot images, as the size of the 
-	  boot image can't be figured out.
 
 c) Generated images could have different ECMA-119 low level names, due to 
    different way to mangle names, to new files added that force old files to 
@@ -91,210 +154,11 @@ c) Generated images could have different ECMA-119 low level names, due to
    relaxed contraints), otherwise libisofs might arbitrarily change the names.
 
 
-------------------------------------------------------------------------------
-
-                        Download, Build and Installation
-
-libisofs code is mantained in a Bazaar repository at Launchpad 
-(https://launchpad.net/libisofs/). You can download it with:
-
-$ bzr branch lp:libisofs
-
-Our build system is based on autotools. For preparing the build you will need 
-autotools of at least version 1.7. If you have download the code from the
-repository, first of all you need to execute
-
-   ./autogen.sh
-   
-on toplevel dir to execute autotools.
-
-Alternatively you may unpack a release tarball for which you do not need
-autotools installed.
-
-To build libisofs it should be sufficient to go into its toplevel directory 
-and execute
-
-   ./configure --prefix=/usr
-   make
-
-To make the libraries accessible for running resp. developing applications 
-   make install
-
-See INSTALL file for further details.
-
-
-------------------------------------------------------------------------------
-
-                        Overview of libburnia-project.org
-
-libburnia-project.org is an open-source software project for reading, mastering
-and writing optical discs.
-For now this means only CD media and all single layer DVD media except DVD+R.
-
-The project comprises of several more or less interdependent parts which
-together strive to be a usable foundation for application development.
-These are libraries, language bindings, and middleware binaries which emulate
-classical (and valuable) Linux tools.
-
-Our scope is currently Linux 2.4 and 2.6 only. For ports to other systems
-we would need : login on a development machine resp. a live OS on CD or DVD,
-advise from a system person about the equivalent of Linux sg or FreeBSD CAM,
-volunteers for testing of realistic use cases.
-
-We have a workable code base for burning CD and most single layer DVD.
-The burn API is quite comprehensively documented and can be used to build a
-presentable application.
-We have a functional binary which emulates parts of cdrecord in order to
-prove that usability, and in order to allow you to explore libburnia's scope
-by help of existing cdrecord frontends.
-
-The project components (list subject to growth, hopefully):
-
-- libburn  is the library by which preformatted data get onto optical media.
-           It uses either /dev/sgN (e.g. on kernel 2.4 with ide-scsi) or
-           /dev/hdX (e.g. on kernel 2.6).
-           libburn is the foundation of our cdrecord emulation. Its code is
-           independent of cdrecord. Its DVD capabilities are learned from
-           studying the code of dvd+rw-tools and MMC-5 specs. No code but only
-           the pure SCSI knowledge has been taken from dvd+rw-tools, though.
-
-- libisofs is the library to pack up hard disk files and directories into a
-           ISO 9660 disk image. This may then be brought to media via libburn.
-           libisofs is to be the foundation of our upcoming mkisofs emulation.
-
-- cdrskin  is a limited cdrecord compatibility wrapper for libburn.
-           Cdrecord is a powerful GPL'ed burn program included in Joerg
-           Schilling's cdrtools. cdrskin strives to be a second source for
-           the services traditionally provided by cdrecord. Additionally it
-           provides libburn's DVD capabilities, where only -sao is compatible
-           with cdrecord.
-           cdrskin does not contain any bytes copied from cdrecord's sources.
-           Many bytes have been copied from the message output of cdrecord
-           runs, though.
-           See cdrskin/README and man cdrskin/cdrskin.1 for more.
-
-- test     is a collection of application gestures and examples given by the
-           authors of the library features. The main API example for libburn
-           is test/libburner.c .
-           Explore these examples if you look for inspiration.
-
-We plan to be a responsive upstream. Bear with us. We are still practicing.
-
-
-------------------------------------------------------------------------------
-Project history as far as known to me:
-
-- Founded in 2002 as it seems. See mailing list archives
-     http://lists.freedesktop.org/archives/libburn/
-  The site of this founder team is reachable and offers download of a
-  (somewhat outdated) tarball and from CVS :
-     http://icculus.org/burn/
-  Copyright holders and most probably founders:
-  Derek Foreman and Ben Jansens.
-
-- I came to using libburn in 2005. Founded the cdrskin project and submitted
-  necessary patches which were accepted or implemented better. Except one
-  remaining patch which prevented cdrskin from using vanilla libburn from CVS.
-  The cdrskin project site is reachable and offers download of the heavily
-  patched (elsewise outdated) tarball under the name  cdrskin-0.1.2  :
-     http://scdbackup.sourceforge.net/cdrskin_eng.html
-  It has meanwhile moved to use vanilla libburn.pykix.org , though.
-  Version 0.1.4 constitutes the first release of this kind.
-
-- In July 2006 our team mate Mario Danic announced a revival of libburn
-  which by about nearly everybody else was perceived as unfriendly fork.
-  Derek Foreman four days later posted a message which expressed his
-  discontent.
-  The situation first caused me to publically regret it and then - after i
-  got the opportunity to move in with cdrskin - gave me true reason to
-  personally apologize to Derek Foreman, Ben Jansens and the contibutors at 
-  icculus.org/burn. Posted to both projects:
-    http://lists.freedesktop.org/archives/libburn/2006-August/000446.html
-    http://mailman-mail1.webfaction.com/pipermail/libburn-hackers/2006-August/000024.html
-
-- Mid August 2006 project cdrskin established a branch office in 
-  libburn.pykix.org so that all maintainers of our tools have one single place
-  to get the current (at least slightely) usable coordinated versions of
-  everything.
-  Project cdrskin will live forth independendly for a while but it is committed
-  to stay in sync with libburn.pykix.org (or some successor, if ever).
-  cdrskin is also committed to support  icculus.org/burn  if the pending fork
-  is made reality by content changes in that project. It will cease to maintain
-  a patched version of  icculus.org/burn  though. Precondition for a new
-  release of cdrskin on base of  icculus.org/burn  would be the pending
-  "whitelist patch" therefore.
-  I would rather prefer if both projects find consense and merge, or at least
-  cooperate. I have not given up hope totally, yet.
-  I, personally, will honor any approach.
-
-- 2nd September 2006 the decision is made to strive for a consolidation of
-  copyright and a commitment to GPL in a reasonable and open minded way.
-  This is to avoid long term problems with code of unknown origin and
-  with finding consense among the not so clearly defined group of copyright
-  claimers and -holders.
-  libisofs is already claimed sole copyright Mario Danic.
-  cdrskin and libburner are already claimed sole copyright Thomas Schmitt.
-  Rewrites of other components will follow and concluded by claiming full
-  copyright within the group of libburn.pykix.org-copyright holders.
-
-- 16th September 2006 feature freeze for release of libburn-0.2.2 .
-
-- 20th September 2006 release of libburn-0.2.2 .
-
-- 26th October 2006 feature freeze for cdrskin-0.2.4 based on libburn-0.2.3 .
-  This version of cdrskin is much more cdrecord compatible in repect
-  to drive addressing and audio features.
-
-- 30th October 2006 release of cdrskin-0.2.4 .
-
-- 13th November 2006 splitting releases of libburn+cdrskin from libisofs.
-
-- 24th November 2006 release of libburn-0.2.6 and cdrskin-0.2.6 . cdrskin has
-  become suitable for unaware frontends as long as they perform only the core 
-  of cdrecord use cases (including open-ended input streams, audio, and
-  multi-session).
-
-- 28th November 2006 the umbrella project which encloses both, libisofs and
-  libburn, is now called libburnia. For the origin of this name, see 
-  http://en.wikipedia.org/wiki/Liburnians .
-
-- 16th January 2007 release of libburn-0.3.0 and cdrskin-0.3.0 . Now the scope
-  is widened to a first class of DVD media: overwriteable single layer types
-  DVD-RAM, DVD+RW, DVD-RW. This is not a cdrecord emulation but rather inspired
-  by dvd+rw-tools' "poor man" writing facility for this class of media.
-  Taking a bow towards Andy Polyakov.
-
-- 11th February 2007 version 0.3.2 covers sequential DVD-RW and DVD-R with
-  multi-session and with DAO.
-
-- 12th March 2007 version 0.3.4 supports DVD+R and thus covers all single layer
-  DVD media. Code for double layer DVD+/-R is implemented but awaits a tester
-  yet.
-
-- 23th April 2007 version 0.3.6 follows the unanimous opinion of Linux kernel
-  people that one should not use /dev/sg on kernel 2.6.
-
-- 31st July 2007 version 0.3.8 marks the first anniversary of libburn revival.
-  We look back on improved stability, a substantially extended list of media
-  and write modes, and better protection against typical user mishaps.
-
-- 24th October 2007 version 0.4.0 is the foundation of new library libisoburn
-  and an upcomming integrated application for manipulating and writing
-  ISO 9660 + Rock Ridge images. cdrskin-0.4.0 got capabilities like growisofs
-  by these enhancements: growing of overwriteable media and disk files.
-  Taking again a bow towards Andy Polyakov.
-
-- 26th Januar 2008 version 0.4.2 rectifies the version numbering so that we
-  reliably release libburn.so.4 as should have been done since libburn-0.3.2.
-  cdrskin now is by default linked dynamically and does a runtime check
-  to ensure not to be started with a libburn which is older than itself.
-
-
 ------------------------------------------------------------------------------
 
     This program is free software; you can redistribute it and/or modify
-    it under the terms of the GNU General Public License as published by
+    it under the terms of the GNU General Public License version 2 or later
-    the Free Software Foundation. To be exact: version 2 of that License.
+    as published by the Free Software Foundation.
 
     This program is distributed in the hope that it will be useful,
     but WITHOUT ANY WARRANTY; without even the implied warranty of
@@ -309,19 +173,19 @@ Project history as far as known to me:
 Clarification in my name and in the name of Mario Danic, upcoming copyright
 holders on toplevel of libburnia. To be fully in effect after the remaining
 other copyrighted code has been replaced by ours and by copyright-free
-contributions of our friends:
+contributions of our friends.
+
+Note:
+In the particular case of libisofs there is no foreign copyright involved.
+As of 2010 foreign copyright is only in component libburn.
 ------------------------------------------------------------------------------
 
-We, the copyright holders, agree on the interpretation that
+We will not raise any legal protest to dynamic linking of our libraries
-dynamical linking of our libraries constitutes "use of" and
+with applications that are not under GPL, as long as they fulfill
-not "derivation from" our work in the sense of GPL, provided
+the condition of offering the library source code used, whether
-those libraries are compiled from our unaltered code.
+altered or unaltered, under the GPLv2+, along with the application.
-
+Nevertheless, the safest legal position is not to link libburn with
-Thus you may link our libraries dynamically with applications
+non-GPL compatible programs.
-which are not under GPL. You may distribute our libraries and
-application tools in binary form, if you fulfill the usual
-condition of GPL to offer a copy of the source code -altered
-or unaltered- under GPL.
 
 We ask you politely to use our work in open source spirit
 and with the due reference to the entire open source community.
@@ -335,7 +199,10 @@ It is the open source idea of responsible freedom which will be
 decisive and you will have to prove that you exhausted all own
 means to qualify for GPL.
 
-For now we are firmly committed to maintain one single license: GPL.
+We are firmly committed to allow GPLv2+ now and with future releases.
 
-signed: Mario Danic, Thomas Schmitt
+Signed: Mario Danic, Thomas Schmitt
+Agreement joined later by: Vreixo Formoso
+
+Public contact: <libburn-hackers@pykix.org>
 

acinclude.m4

@@ -1,3 +1,14 @@
+AC_DEFUN([LIBBURNIA_SET_FLAGS],
+[
+case $target_os in
+freebsd*)
+        LDFLAGS="$LDFLAGS -L/usr/local/lib"
+        CPPFLAGS="$CPPFLAGS -I/usr/local/include"
+        ;;
+esac
+])
+
+
 AC_DEFUN([TARGET_SHIZZLE],
 [
   ARCH=""
@@ -5,14 +16,18 @@ AC_DEFUN([TARGET_SHIZZLE],
 
   AC_MSG_CHECKING([target operating system])
 
+  LIBBURNIA_LDCONFIG_CMD="echo 'No ldconfig run performed. If needed, configure manually for:'"
   case $target in
     *-*-linux*)
       ARCH=linux
       LIBBURN_ARCH_LIBS=
+      LIBBURNIA_LDCONFIG_CMD=ldconfig
       ;;
     *-*-freebsd*)
       ARCH=freebsd
       LIBBURN_ARCH_LIBS=-lcam
+
+      # This may later be overridden by configure --enable-libdir-pkgconfig
       LIBBURNIA_PKGCONFDIR=$(echo "$libdir" | sed 's/\/lib$/\/libdata/')/pkgconfig
       ;;
     *)
@@ -24,3 +39,162 @@ AC_DEFUN([TARGET_SHIZZLE],
 
   AC_MSG_RESULT([$ARCH])
 ])
+
+
+dnl LIBBURNIA_CHECK_ICONV is by Thomas Schmitt, libburnia project
+dnl It is based on gestures from:
+dnl iconv.m4 serial AM7 (gettext-0.18)
+dnl Copyright (C) 2000-2002, 2007-2009 Free Software Foundation, Inc.
+dnl This file is free software; the Free Software Foundation
+dnl gives unlimited permission to copy and/or distribute it,
+dnl with or without modifications, as long as this notice is preserved.
+dnl From Bruno Haible.
+dnl
+AC_DEFUN([LIBBURNIA_CHECK_ICONV],
+[
+
+  dnl Check whether it is allowed to link with -liconv
+  AC_MSG_CHECKING([for iconv() in separate -liconv ])
+  libburnia_liconv="no"
+  libburnia_save_LIBS="$LIBS"
+  LIBS="$LIBS -liconv"
+  AC_TRY_LINK([#include <stdlib.h>
+#include <iconv.h>],
+    [iconv_t cd = iconv_open("","");
+     iconv(cd,NULL,NULL,NULL,NULL);
+     iconv_close(cd);],
+     [libburnia_liconv="yes"],
+     [LIBS="$libburnia_save_LIBS"]
+  )
+  AC_MSG_RESULT([$libburnia_liconv])
+
+  if test x"$libburnia_save_LIBS" = x"$LIBS"
+  then
+    dnl GNU iconv has no function iconv() but libiconv() and a macro iconv()
+    dnl It is not tested whether this is detected by above macro.
+    AC_CHECK_LIB(iconv, libiconv, , )
+  fi
+
+  dnl Check for iconv(..., const char **inbuf, ...)
+  AC_MSG_CHECKING([for const qualifier with iconv() ])
+  AC_TRY_COMPILE([
+#include <stdlib.h>
+#include <iconv.h>
+size_t iconv (iconv_t cd, char * *inbuf, size_t *inbytesleft, char * *outbuf, size_t *outbytesleft);
+], [], [libburnia_iconv_const=""], [libburnia_iconv_const="const"]
+  )
+  AC_DEFINE_UNQUOTED([ICONV_CONST], [$libburnia_iconv_const])
+  test -z "$libburnia_iconv_const" && libburnia_iconv_const="no"
+  AC_MSG_RESULT([$libburnia_iconv_const])
+])
+
+
+dnl LIBBURNIA_ASSERT_ICONV is by Thomas Schmitt, libburnia project
+dnl 
+AC_DEFUN([LIBBURNIA_ASSERT_ICONV],
+[
+  if test x$LIBISOFS_ASSUME_ICONV = x
+  then
+    dnl Check for the essential gestures of libisofs/util.c
+    AC_MSG_CHECKING([for iconv() to be accessible now ])
+    AC_TRY_LINK([
+#include <stdlib.h>
+#include <wchar.h>
+#include <string.h>
+#include <errno.h>
+#include <ctype.h>
+#include <stdio.h>
+#include <limits.h>
+#include <iconv.h>
+#include <locale.h>
+#include <langinfo.h>
+#include <unistd.h>],
+[iconv_t cd = iconv_open("","");
+iconv(cd,NULL,NULL,NULL,NULL);
+iconv_close(cd);
+], [iconv_test="yes"], [iconv_test="no"]
+    )
+    AC_MSG_RESULT([$iconv_test])
+    if test x$iconv_test = xno
+    then
+      echo >&2
+      echo "Cannot get function iconv() to work. Configuration aborted." >&2
+      echo "Check whether your system needs a separate libiconv installed." >&2
+      echo "If it is installed but not found, try something like" >&2
+      echo '  export LDFLAGS="$LDFLAGS -L/usr/local/lib"' >&2 
+      echo '  export CPPFLAGS="$CPPFLAGS -I/usr/local/include"' >&2
+      echo '  export LIBS="$LIBS -liconv"' >&2
+      echo "You may override this test by exporting variable" >&2
+      echo "  LIBISOFS_ASSUME_ICONV=yes" >&2
+      echo >&2
+      (exit 1); exit 1;
+    fi
+  fi
+])
+
+
+dnl LIBISOFS_ASSERT_VERS_LIBS is by Thomas Schmitt, libburnia project
+dnl It tests whether -Wl,--version-script=... works with the compiler
+AC_DEFUN([LIBISOFS_ASSERT_VERS_LIBS],
+[
+    libburnia_save_LDFLAGS="$LDFLAGS"
+    LDFLAGS="$LDFLAGS -Wl,--version-script=libisofs/libisofs.ver"
+    AC_TRY_LINK([#include <stdio.h>], [printf("Hello\n");],
+                [vers_libs_test="yes"], [vers_libs_test="no"])
+    if test x$vers_libs_test = xno
+    then
+        LDFLAGS="$libburnia_save_LDFLAGS"
+    fi
+])
+
+
+dnl LIBBURNIA_SET_PKGCONFIG determines the install directory for the *.pc file.
+dnl Important: Must be performed _after_ TARGET_SHIZZLE
+dnl
+AC_DEFUN([LIBBURNIA_SET_PKGCONFIG],
+[
+### for testing --enable-libdir-pkgconfig on Linux
+### LIBBURNIA_PKGCONFDIR="$libdir"data/pkgconfig
+
+if test "x$LIBBURNIA_PKGCONFDIR" = "x$libdir"/pkgconfig
+then
+  dummy=dummy
+else
+  AC_ARG_ENABLE(libdir-pkgconfig,
+  [  --enable-libdir-pkgconfig  Install to $libdir/pkgconfig on any OS, default=no],
+   , enable_libdir_pkgconfig="no")
+  AC_MSG_CHECKING([for --enable-libdir-pkgconfig])
+  if test "x$enable_libdir_pkgconfig" = xyes
+  then
+    LIBBURNIA_PKGCONFDIR="$libdir"/pkgconfig
+  fi
+  AC_MSG_RESULT([$enable_libdir_pkgconfig])
+fi
+
+libburnia_pkgconfig_override="no"
+AC_ARG_ENABLE(pkgconfig-path,
+[  --enable-pkgconfig-path=DIR  Absolute path of directory for libisofs-*.pc],
+libburnia_pkgconfig_override="yes" , enable_pkgconfig_path="none")
+AC_MSG_CHECKING([for overridden pkgconfig directory path])
+if test "x$enable_pkgconfig_path" = xno
+then
+  libburnia_pkgconfig_override="no"
+fi
+if test "x$enable_pkgconfig_path" = x -o "x$enable_pkgconfig_path" = xyes
+then
+  libburnia_pkgconfig_override="invalid argument"
+fi
+if test "x$libburnia_pkgconfig_override" = xyes
+then
+  LIBBURNIA_PKGCONFDIR="$enable_pkgconfig_path"
+  AC_MSG_RESULT([$LIBBURNIA_PKGCONFDIR])
+else
+  AC_MSG_RESULT([$libburnia_pkgconfig_override])
+fi
+AC_SUBST(LIBBURNIA_PKGCONFDIR)
+
+dnl For debugging only
+### AC_MSG_RESULT([LIBBURNIA_PKGCONFDIR = $LIBBURNIA_PKGCONFDIR])
+ 
+])
+

configure.ac

@@ -1,17 +1,13 @@
-AC_INIT([libisofs], [0.6.18], [http://libburnia-project.org])
+AC_INIT([libisofs], [1.0.2], [http://libburnia-project.org])
 AC_PREREQ([2.50])
 dnl AC_CONFIG_HEADER([config.h])	
 
 AC_CANONICAL_HOST
 AC_CANONICAL_TARGET
 
-AM_INIT_AUTOMAKE([subdir-objects])
+LIBBURNIA_SET_FLAGS
 
-dnl A61101 This breaks Linux build (makes 32 bit off_t)
+AM_INIT_AUTOMAKE([subdir-objects])
-dnl        http://sourceware.org/autobook/autobook/autobook_96.html says
-dnl        one must include some config.h and this was a pitfall.
-dnl        So why dig the pit at all ?
-dnl AM_CONFIG_HEADER(config.h)
 
 dnl
 dnl if MAJOR or MINOR version changes, be sure to change AC_INIT above to match
@@ -37,14 +33,14 @@ dnl iso_lib_version(). It returns the major, minor and micro revision of the
 dnl library. This means LIBISOFS_*_VERSION kept its second job which does not  
 dnl comply to the usual ways of configure.ac . I.e. now *officially* this is 
 dnl the source code release version as announced to the public. It has no 
-dnl conection to SONAME or libtool version numbering.
+dnl connection to SONAME or libtool version numbering.
 dnl It rather feeds the API function iso_lib_version().
 dnl
 dnl If LIBISOFS_*_VERSION changes, be sure to change AC_INIT above to match.
 dnl
-LIBISOFS_MAJOR_VERSION=0
+LIBISOFS_MAJOR_VERSION=1
-LIBISOFS_MINOR_VERSION=6
+LIBISOFS_MINOR_VERSION=0
-LIBISOFS_MICRO_VERSION=18
+LIBISOFS_MICRO_VERSION=2
 LIBISOFS_VERSION=$LIBISOFS_MAJOR_VERSION.$LIBISOFS_MINOR_VERSION.$LIBISOFS_MICRO_VERSION
 
 AC_SUBST(LIBISOFS_MAJOR_VERSION)
@@ -54,11 +50,11 @@ AC_SUBST(LIBISOFS_VERSION)
 
 dnl Libtool versioning
 LT_RELEASE=$LIBISOFS_MAJOR_VERSION.$LIBISOFS_MINOR_VERSION
-# 2009.04.15: development jump has not yet happened
+# 2011.02.23  development jump has not yet happened
-# SONAME = 20 - 14 = 6 . Library name = libisofs.6.14.0
+# SONAME = 46 - 40 = 6 . Library name = libisofs.6.40.0
-LT_CURRENT=20
+LT_CURRENT=46
+LT_AGE=40
 LT_REVISION=0
-LT_AGE=14
 LT_CURRENT_MINUS_AGE=`expr $LT_CURRENT - $LT_AGE`
 
 AC_SUBST(LT_RELEASE)
@@ -87,7 +83,11 @@ fi
 
 dnl If iconv(3) is in an extra lib, then it gets added to variable LIBS.
 dnl If not, then no -liconv will be added.
-AC_CHECK_LIB(iconv, iconv, , )
+LIBBURNIA_CHECK_ICONV
+
+dnl To abort configuration if iconv() still cannot be compiled
+LIBBURNIA_ASSERT_ICONV
+
 
 AC_PROG_LIBTOOL
 AC_SUBST(LIBTOOL_DEPS)
@@ -124,39 +124,44 @@ AC_SUBST(THREAD_LIBS)
 
 TARGET_SHIZZLE
 AC_SUBST(ARCH)
-AC_SUBST(LIBBURNIA_PKGCONFDIR)
 AC_SUBST(LIBBURN_ARCH_LIBS)
 
-dnl Add compiler-specific flags
 
 dnl See if the user wants aggressive optimizations of the code
 AC_ARG_ENABLE(debug,
-[  --enable-debug          Disable aggressive optimizations [default=yes]],
+[  --enable-debug          Disable aggressive optimizations, default=yes],
               , enable_debug=yes)
 if test x$enable_debug != xyes; then
     if test x$GCC = xyes; then
-        CFLAGS="$CFLAGS -O3"
+        CFLAGS="-O3 $CFLAGS"
-        CFLAGS="$CFLAGS -fexpensive-optimizations"
+        CFLAGS="-fexpensive-optimizations $CFLAGS"
     fi
-    CFLAGS="$CFLAGS -DNDEBUG"
+    CFLAGS="-DNDEBUG $CFLAGS"
 else
     if test x$GCC = xyes; then
-        CFLAGS="$CFLAGS -g -pedantic -Wall"
+        CFLAGS="-g -pedantic -Wall $CFLAGS"
     fi
-    CFLAGS="$CFLAGS -DDEBUG"
+    CFLAGS="-DDEBUG $CFLAGS"
 fi
 
 dnl Verbose debug to make libisofs issue more debug messages
 AC_ARG_ENABLE(verbose-debug,
-[  --enable-verbose-debug          Enable verbose debug messages [default=no]],
+[  --enable-verbose-debug  Enable verbose debug messages, default=no],
 	AC_DEFINE(LIBISOFS_VERBOSE_DEBUG, 1))
 
 
-dnl ts A90123
+dnl Determine target directory for libisofs-*.pc
+dnl Important: Must be performed _after_ TARGET_SHIZZLE
+dnl
+LIBBURNIA_SET_PKGCONFIG
+
+
+dnl Add compiler-specific flags
+
 AC_ARG_ENABLE(libacl,
-[  --enable-libacl    Enable use of libacl by libisofs, default=yes],
+[  --enable-libacl         Enable use of libacl by libisofs, default=yes],
                , enable_libacl=yes)
-if test x$enable_libacl = xyes; then
+if test "x$enable_libacl" = xyes; then
 dnl Check whether there is libacl-devel and libacl-runtime.
 dnl If not, erase this macro which would enable use of acl_to_text and others
     LIBACL_DEF="-DLibisofs_with_aaip_acL"
@@ -170,9 +175,9 @@ AC_SUBST(LIBACL_DEF)
 
 dnl ts A90123
 AC_ARG_ENABLE(xattr,
-[  --enable-xattr     Enable use of xattr by libisofs, default=yes],
+[  --enable-xattr          Enable use of xattr by libisofs, default=yes],
                , enable_xattr=yes)
-if test x$enable_xattr = xyes; then
+if test "x$enable_xattr" = xyes; then
 dnl Check whether there is the header for Linux xattr. 
 dnl If not, erase this macro which would enable use of listxattr and others
     XATTR_DEF="-DLibisofs_with_aaip_xattR"
@@ -185,19 +190,63 @@ AC_SUBST(XATTR_DEF)
 
 dnl ts A90409
 AC_ARG_ENABLE(zlib,
-[  --enable-zlib     Enable use of zlib by libisofs, default=yes],
+[  --enable-zlib           Enable use of zlib by libisofs, default=yes],
                , enable_zlib=yes)
-if test x$enable_zlib = xyes; then
+if test "x$enable_zlib" = xyes; then
 dnl Check whether there is the header for zlib. 
 dnl If not, erase this macro which would enable use of compress2() and others.
-dnl The empty parameter after "compress2" causes -lz.
+dnl Linking fails on SuSE 9.0 because zlib has compress2() but lacks  
+dnl compressBound(). So compressBound is the more modern thing to test.
+dnl The empty parameter after "compressBound" causes -lz.
     ZLIB_DEF="-DLibisofs_with_zliB"
-    AC_CHECK_HEADER(zlib.h, AC_CHECK_LIB(z, compress2, , ZLIB_DEF= ), ZLIB_DEF= )
+    AC_CHECK_HEADER(zlib.h, AC_CHECK_LIB(z, compressBound, , ZLIB_DEF= ), ZLIB_DEF= )
 else
     ZLIB_DEF=
 fi
 AC_SUBST(ZLIB_DEF)
 
+dnl ts B00927
+AC_ARG_ENABLE(libjte,
+[  --enable-libjte           Enable use of libjte by libisofs, default=yes],
+               , enable_libjte=yes)
+if test "x$enable_libjte" = xyes; then
+    LIBJTE_DEF="-DLibisofs_with_libjtE"
+    AC_CHECK_HEADER(libjte/libjte.h, AC_CHECK_LIB(jte, libjte_new, , LIBJTE_DEF= ), LIBJTE_DEF= )
+else
+    LIBJTE_DEF=
+fi
+AC_SUBST(LIBJTE_DEF)
+
+# Library versioning normally serves a complex purpose.
+# Since libisofs obeys strict ABI backward compatibility, it needs only the
+# simple feature to declare function names "global:" or "local:". Only the
+# global ones are visible to applications at library load time.
+AC_ARG_ENABLE(versioned-libs,
+[  --enable-versioned-libs Enable strict symbol encapsulation , default=yes],
+               , enable_versioned_libs=yes)
+if test x$enable_versioned_libs = xyes; then
+    vers_libs_test=no
+    LIBISOFS_ASSERT_VERS_LIBS
+    if test x$vers_libs_test = xno
+    then 
+        echo "disabled strict symbol encapsulation (test failed)"
+    else
+        echo "enabled  strict symbol encapsulation"
+    fi
+else
+    echo "disabled strict symbol encapsulation"
+fi
+
+AC_ARG_ENABLE(ldconfig-at-install,
+[  --enable-ldconfig-at-install    On GNU/Linux run ldconfig, default=yes],
+               , ldconfig_at_install=yes)
+if test x$ldconfig_at_install = xyes; then 
+    dummy=dummy
+else
+    LIBBURNIA_LDCONFIG_CMD="echo 'NOTE: ldconfig is disabled. If needed, configure manually for:'"
+    echo "disabled run of ldconfig during installation on GNU/Linux"
+fi  
+AC_SUBST(LIBBURNIA_LDCONFIG_CMD)
 
 AC_CONFIG_FILES([
 	Makefile

demo/cat.c

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #include "libisofs.h"

demo/cat_buffer.c

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #include "libisofs.h"

demo/demo.c

@@ -0,0 +1,867 @@
+
+/*
+ * Copyright (c) 2007 - 2009 Vreixo Formoso, Thomas Schmitt
+ * 
+ * This file is part of the libisofs project; you can redistribute it and/or 
+ * modify it under the terms of the GNU General Public License version 2 
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
+ */
+
+static char helptext[][80] = {
+"",
+"This is a collection of libisofs gestures which formerly were distinct",
+"programs. The first argument chooses the gesture:",
+"  -tree  absolute_directory_path",
+"               Import a directory and print the resulting iso tree.",
+"  -find  absolute_directory_path",      
+"               Import a directory, find matching nodes and print the",
+"               resulting iso tree.",
+"  -iso  [options] directory output_file",
+"               Create an iso image from a local directory. For options see",
+"               output of -iso -h",
+"  -iso_read  image_file",
+"               Output the contents of an iso image.",
+"  -iso_cat  image_file path_in_image",
+"               Extract a file from a given ISO image and put out its content",
+"               to stdout. The file is addressed by path_in_image.",
+"  -iso_modify  image_file absolute_directory_path output_file",
+"               Load an iso image, add a directory, and write complete image.",
+"  -iso_ms  image_lba nwa image_file directory_path output_file",
+"               Load an iso image, add a directory, and write as add-on",
+"               session which shall be appended to the old image.",
+"               image_lba gives the block address of the start of the most",
+"               recent session in the image_file. nwa gives the block address",
+"               where the add-on session will be appended to the image.",
+"@"
+};
+
+
+#define LIBISOFS_WITHOUT_LIBBURN yes
+#include "libisofs.h"
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
+#include <getopt.h>
+#include <fcntl.h>
+#include <err.h>
+#include <limits.h>
+
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
+/* ------------------------- from demo/tree.c ----------------------- */
+
+static void
+print_permissions(mode_t mode)
+{
+    char perm[10];
+    
+    /* TODO suid, sticky... */
+    
+    perm[9] = '\0';
+    perm[8] = mode & S_IXOTH ? 'x' : '-';
+    perm[7] = mode & S_IWOTH ? 'w' : '-';
+    perm[6] = mode & S_IROTH ? 'r' : '-';
+    perm[5] = mode & S_IXGRP ? 'x' : '-';
+    perm[4] = mode & S_IWGRP ? 'w' : '-';
+    perm[3] = mode & S_IRGRP ? 'r' : '-';
+    perm[2] = mode & S_IXUSR ? 'x' : '-';
+    perm[1] = mode & S_IWUSR ? 'w' : '-';
+    perm[0] = mode & S_IRUSR ? 'r' : '-';
+    printf("[%s]",perm);
+}
+
+static void 
+tree_print_dir(IsoDir *dir, int level) 
+{
+    int i;
+    IsoDirIter *iter;
+    IsoNode *node;
+    char *sp;
+
+    sp = calloc(1, level * 2 + 1);
+    
+    for (i = 0; i < level * 2; i += 2) {
+        sp[i] = '|';
+        sp[i+1] = ' ';
+    }
+    
+    sp[level * 2-1] = '-';
+    sp[level * 2] = '\0';
+    
+    iso_dir_get_children(dir, &iter);
+    while (iso_dir_iter_next(iter, &node) == 1) {
+        
+        if (ISO_NODE_IS_DIR(node)) {
+            printf("%s+[D] ", sp);
+            print_permissions(iso_node_get_permissions(node));
+            printf(" %s\n", iso_node_get_name(node));
+            tree_print_dir(ISO_DIR(node), level+1);
+        } else if (ISO_NODE_IS_FILE(node)) {
+            printf("%s-[F] ", sp);
+            print_permissions(iso_node_get_permissions(node));
+            printf(" %s\n", iso_node_get_name(node) );
+        } else if (ISO_NODE_IS_SYMLINK(node)) {
+            printf("%s-[L] ", sp);
+            print_permissions(iso_node_get_permissions(node));
+            printf(" %s -> %s \n", iso_node_get_name(node),
+                   iso_symlink_get_dest(ISO_SYMLINK(node)) );
+        } else {
+            printf("%s-[C] ", sp);
+            print_permissions(iso_node_get_permissions(node));
+            printf(" %s\n", iso_node_get_name(node) );
+        } 
+    }
+    iso_dir_iter_free(iter);
+    free(sp);
+}
+
+int gesture_tree(int argc, char **argv)
+{
+    int result;
+    IsoImage *image;
+    
+    if (argc != 2) {
+need_abs_path:;
+        fprintf (stderr, "You need to specify a valid absolute path\n");
+        return 1;
+    }
+    if (argv[1][0] != '/')
+        goto need_abs_path;
+
+    iso_init();
+    iso_set_msgs_severities("NEVER", "ALL", "");
+    
+    result = iso_image_new("volume_id", &image);
+    if (result < 0) {
+        printf ("Error creating image\n");
+        return 1;
+    }
+    
+    result = iso_tree_add_dir_rec(image, iso_image_get_root(image), argv[1]);
+    if (result < 0) {
+        printf ("Error adding directory %d\n", result);
+        return 1;
+    }
+    
+    printf("================= IMAGE =================\n");
+    tree_print_dir(iso_image_get_root(image), 0);
+    printf("\n\n");
+    
+    iso_image_unref(image);
+    iso_finish();
+    return 0;
+}
+
+
+/* ------------------------- from demo/find.c ----------------------- */
+
+static void 
+find_print_dir(IsoDir *dir) 
+{
+    IsoDirIter *iter;
+    IsoNode *node;
+    IsoFindCondition *cond, *c1, *c2;
+    
+    c1 = iso_new_find_conditions_name("*a*");
+    c2 = iso_new_find_conditions_mode(S_IFREG);
+    cond = iso_new_find_conditions_and(c1, c2);
+    iso_dir_find_children(dir, cond, &iter);
+    while (iso_dir_iter_next(iter, &node) == 1) {
+        char *path = iso_tree_get_node_path(node);
+        printf(" %s\n", path);
+        free(path);
+    }
+    iso_dir_iter_free(iter);
+}
+
+int gesture_find(int argc, char **argv)
+{
+    int result;
+    IsoImage *image;
+    
+    if (argc != 2) {
+need_abs_path:;
+        fprintf (stderr, "You need to specify a valid absolute path\n");
+        return 1;
+    }
+    if (argv[1][0] != '/')
+        goto need_abs_path;
+
+    iso_init();
+    iso_set_msgs_severities("NEVER", "ALL", "");
+    
+    result = iso_image_new("volume_id", &image);
+    if (result < 0) {
+        printf ("Error creating image\n");
+        return 1;
+    }
+    
+    result = iso_tree_add_dir_rec(image, iso_image_get_root(image), argv[1]);
+    if (result < 0) {
+        printf ("Error adding directory %d\n", result);
+        return 1;
+    }
+    
+    find_print_dir(iso_image_get_root(image));
+    
+    iso_image_unref(image);
+    iso_finish();
+    return 0;
+}
+
+
+/* ------------------------- from demo/iso.c ----------------------- */
+
+
+static const char * const optstring = "JRIL:b:hV:";
+extern char *optarg;
+extern int optind;
+
+void iso_usage(char **argv)
+{
+    printf("%s [OPTIONS] DIRECTORY OUTPUT\n", argv[0]);
+}
+
+void iso_help()
+{
+    printf(
+        "Options:\n"
+        "  -J        Add Joliet support\n"
+        "  -R        Add Rock Ridge support\n"
+        "  -I        Add ISO 9660:1999 support\n"
+        "  -V label  Volume Label\n"
+        "  -L <num>  Set the ISO level (1 or 2)\n"
+        "  -b file   Specifies a boot image to add to image\n"
+        "  -h        Print this message\n"
+    );
+}
+
+int iso_callback(IsoFileSource *src)
+{
+    char *path = iso_file_source_get_path(src);
+    printf("CALLBACK: %s\n", path);
+    free(path);
+    return 1;
+}
+
+int gesture_iso(int argc, char **argv)
+{
+    int result;
+    int c;
+    IsoImage *image;
+    struct burn_source *burn_src;
+    unsigned char buf[2048];
+    FILE *fp = NULL;
+    IsoWriteOpts *opts;
+    char *volid = "VOLID";
+    char *boot_img = NULL;
+    int rr = 0, j = 0, iso1999 = 0, level = 1;
+
+    while ((c = getopt(argc, argv, optstring)) != -1) {
+        switch(c) {
+        case 'h':
+            iso_usage(argv);
+            iso_help();
+            goto ex;
+            break;
+        case 'J':
+            j = 1;
+            break;
+        case 'R':
+            rr = 1;
+            break;
+        case 'I':
+            iso1999 = 1;
+            break;
+        case 'L':
+            level = atoi(optarg);
+            break;
+        case 'b':
+            boot_img = optarg;
+            break;
+        case 'V':
+            volid = optarg;
+            break;
+        case '?':
+            iso_usage(argv);
+            goto ex;
+            break;
+        }
+    }
+
+    if (argc < 2) {
+        printf ("Please pass directory from which to build ISO\n");
+        iso_usage(argv);
+        goto ex;
+    }
+    if (argc < 3) {
+        printf ("Please supply output file\n");
+        iso_usage(argv);
+        goto ex;
+    }
+
+    fp = fopen(argv[optind+1], "w");
+    if (fp == NULL) {
+        err(1, "error opening output file");
+        goto ex;
+    }
+
+    result = iso_init();
+    if (result < 0) {
+        printf ("Can't initialize libisofs\n");
+        goto ex;
+    }
+    iso_set_msgs_severities("NEVER", "ALL", "");
+
+    result = iso_image_new(volid, &image);
+    if (result < 0) {
+        printf ("Error creating image\n");
+        goto ex;
+    }
+    iso_tree_set_follow_symlinks(image, 0);
+    iso_tree_set_ignore_hidden(image, 0);
+    iso_tree_set_ignore_special(image, 0);
+    iso_set_abort_severity("SORRY");
+    /*iso_tree_set_report_callback(image, callback);*/
+
+    result = iso_tree_add_dir_rec(image, iso_image_get_root(image),
+                                  argv[optind]);
+    if (result < 0) {
+        printf ("Error adding directory %d\n", result);
+        goto ex;
+    }
+
+    if (boot_img) {
+        /* adds El-Torito boot info. Tunned for isolinux */
+        ElToritoBootImage *bootimg;
+        result = iso_image_set_boot_image(image, boot_img, ELTORITO_NO_EMUL,
+                                     "/isolinux/boot.cat", &bootimg);
+        if (result < 0) {
+            printf ("Error adding boot image %d\n", result);
+            goto ex;
+        }
+        el_torito_set_load_size(bootimg, 4);
+        el_torito_patch_isolinux_image(bootimg);
+    }
+
+    result = iso_write_opts_new(&opts, 0);
+    if (result < 0) {
+        printf ("Cant create write opts, error %d\n", result);
+        goto ex;
+    }
+    iso_write_opts_set_iso_level(opts, level);
+    iso_write_opts_set_rockridge(opts, rr);
+    iso_write_opts_set_joliet(opts, j);
+    iso_write_opts_set_iso1999(opts, iso1999);
+
+    result = iso_image_create_burn_source(image, opts, &burn_src);
+    if (result < 0) {
+        printf ("Cant create image, error %d\n", result);
+        goto ex;
+    }
+
+    iso_write_opts_free(opts);
+
+    while (burn_src->read_xt(burn_src, buf, 2048) == 2048) {
+        fwrite(buf, 1, 2048, fp);
+    }
+    fclose(fp);
+    burn_src->free_data(burn_src);
+    free(burn_src);
+
+    iso_image_unref(image);
+    iso_finish();
+    return 0;
+ex:;
+    if (fp != NULL)
+      fclose(fp);
+    return 1;
+}
+
+
+/* ------------------------- from demo/iso_read.c ----------------------- */
+
+
+static void
+iso_read_print_type(mode_t mode)
+{
+    switch(mode & S_IFMT) {
+    case S_IFSOCK: printf("[S] "); break;
+    case S_IFLNK: printf("[L] "); break;
+    case S_IFREG: printf("[R] "); break;
+    case S_IFBLK: printf("[B] "); break;
+    case S_IFDIR: printf("[D] "); break;
+    case S_IFIFO: printf("[F] "); break;
+    }
+}
+
+static void
+iso_read_print_file_src(IsoFileSource *file)
+{
+    struct stat info;
+    char *name;
+    iso_file_source_lstat(file, &info);
+    iso_read_print_type(info.st_mode);
+    print_permissions(info.st_mode);
+    printf(" %10.f ", (double) info.st_size);
+    /* printf(" {%ld,%ld} ", (long)info.st_dev, (long)info.st_ino); */
+    name = iso_file_source_get_name(file);
+    printf(" %s", name);
+    free(name);
+    if (S_ISLNK(info.st_mode)) {
+        char buf[PATH_MAX];
+        iso_file_source_readlink(file, buf, PATH_MAX);
+        printf(" -> %s\n", buf);
+    }
+    printf("\n");
+}
+
+static void
+iso_read_print_dir(IsoFileSource *dir, int level)
+{
+    int ret, i;
+    IsoFileSource *file;
+    struct stat info;
+    char *sp;
+
+    sp = calloc(1, level * 2 + 1);
+
+    for (i = 0; i < level * 2; i += 2) {
+        sp[i] = '|';
+        sp[i+1] = ' ';
+    }
+
+    sp[level * 2-1] = '-';
+    sp[level * 2] = '\0';
+
+    ret = iso_file_source_open(dir);
+    if (ret < 0) {
+        printf ("Can't open dir %d\n", ret);
+    }
+    while ((ret = iso_file_source_readdir(dir, &file)) == 1) {
+        printf("%s", sp);
+        iso_read_print_file_src(file);
+        ret = iso_file_source_lstat(file, &info);
+        if (ret < 0) {
+            break;
+        }
+        if (S_ISDIR(info.st_mode)) {
+            iso_read_print_dir(file, level + 1);
+        }
+        iso_file_source_unref(file);
+    }
+    iso_file_source_close(dir);
+    if (ret < 0) {
+        printf ("Can't print dir\n");
+    }
+    free(sp);
+}
+
+int gesture_iso_read(int argc, char **argv)
+{
+    int result;
+    IsoImageFilesystem *fs;
+    IsoDataSource *src;
+    IsoFileSource *root;
+    IsoReadOpts *ropts;
+
+    if (argc != 2) {
+        printf ("You need to specify a valid path\n");
+        return 1;
+    }
+
+    iso_init();
+    iso_set_msgs_severities("NEVER", "ALL", "");
+
+    result = iso_data_source_new_from_file(argv[1], &src);
+    if (result < 0) {
+        printf ("Error creating data source\n");
+        return 1;
+    }
+
+    result = iso_read_opts_new(&ropts, 0);
+    if (result < 0) {
+        fprintf(stderr, "Error creating read options\n");
+        return 1;
+    }
+    result = iso_image_filesystem_new(src, ropts, 1, &fs);
+    iso_read_opts_free(ropts);
+    if (result < 0) {
+        printf ("Error creating filesystem\n");
+        return 1;
+    }
+
+    printf("\nVOLUME INFORMATION\n");
+    printf("==================\n\n");
+
+    printf("Vol. id: %s\n", iso_image_fs_get_volume_id(fs));
+    printf("Publisher: %s\n", iso_image_fs_get_publisher_id(fs));
+    printf("Data preparer: %s\n", iso_image_fs_get_data_preparer_id(fs));
+    printf("System: %s\n", iso_image_fs_get_system_id(fs));
+    printf("Application: %s\n", iso_image_fs_get_application_id(fs));
+    printf("Copyright: %s\n", iso_image_fs_get_copyright_file_id(fs));
+    printf("Abstract: %s\n", iso_image_fs_get_abstract_file_id(fs));
+    printf("Biblio: %s\n", iso_image_fs_get_biblio_file_id(fs));
+
+    printf("\nDIRECTORY TREE\n");
+    printf("==============\n");
+
+    result = fs->get_root(fs, &root);
+    if (result < 0) {
+        printf ("Can't get root %d\n", result);
+        return 1;
+    }
+    /* iso_read_print_file_src(root); */
+    iso_read_print_dir(root, 0);
+    iso_file_source_unref(root);
+
+    fs->close(fs);
+    iso_filesystem_unref((IsoFilesystem*)fs);
+    iso_data_source_unref(src);
+    iso_finish();
+    return 0;
+}
+
+
+/* ------------------------- from demo/iso_cat.c ----------------------- */
+
+
+int gesture_iso_cat(int argc, char **argv)
+{
+    int res;
+    IsoFilesystem *fs;
+    IsoFileSource *file;
+    struct stat info;
+    IsoDataSource *src;
+    IsoReadOpts *opts;
+
+    if (argc != 3) {
+        fprintf(stderr, "Usage: isocat /path/to/image /path/to/file\n");
+        return 1;
+    }
+
+    res = iso_init();
+    if (res < 0) {
+        fprintf(stderr, "Can't init libisofs\n");
+        return 1;
+    }
+    
+    res = iso_data_source_new_from_file(argv[1], &src);
+    if (res < 0) {
+        fprintf(stderr, "Error creating data source\n");
+        return 1;
+    }
+
+    res = iso_read_opts_new(&opts, 0);
+    if (res < 0) {
+        fprintf(stderr, "Error creating read options\n");
+        return 1;
+    }
+    res = iso_image_filesystem_new(src, opts, 1, &fs);
+    if (res < 0) {
+        fprintf(stderr, "Error creating filesystem\n");
+        return 1;
+    }
+    iso_read_opts_free(opts);
+
+    res = fs->get_by_path(fs, argv[2], &file);
+    if (res < 0) {
+        fprintf(stderr, "Can't get file, err = %d\n", res);
+        return 1;
+    }
+
+    res = iso_file_source_lstat(file, &info);
+    if (res < 0) {
+        fprintf(stderr, "Can't stat file, err = %d\n", res);
+        return 1;
+    }
+
+    if (S_ISDIR(info.st_mode)) {
+        fprintf(stderr, "Path refers to a directory!!\n");
+        return 1;
+    } else {
+        char buf[1024];
+        res = iso_file_source_open(file);
+        if (res < 0) {
+            fprintf(stderr, "Can't open file, err = %d\n", res);
+            return 1;
+        }
+        while ((res = iso_file_source_read(file, buf, 1024)) > 0) {
+            fwrite(buf, 1, res, stdout);
+        }
+        if (res < 0) {
+            fprintf(stderr, "Error reading, err = %d\n", res);
+            return 1;
+        }
+        iso_file_source_close(file);
+    }
+    
+    iso_file_source_unref(file);
+    iso_filesystem_unref(fs);
+    iso_data_source_unref(src);
+    iso_finish();
+    return 0;
+}
+
+
+/* ------------------------- from demo/iso_modify.c ----------------------- */
+
+
+void iso_modify_usage(char **argv)
+{
+    printf("%s IMAGE DIRECTORY OUTPUT\n", argv[0]);
+}
+
+int gesture_iso_modify(int argc, char **argv)
+{
+    int result;
+    IsoImage *image;
+    IsoDataSource *src;
+    struct burn_source *burn_src;
+    unsigned char buf[2048];
+    FILE *fp = NULL;
+    IsoWriteOpts *opts;
+    IsoReadOpts *ropts;
+	
+    if (argc < 4) {
+        iso_modify_usage(argv);
+        goto ex;
+    }
+    
+    fp = fopen(argv[3], "w");
+    if (fp == NULL) {
+        err(1, "error opening output file");
+        goto ex;
+    }
+
+    iso_init();
+    iso_set_msgs_severities("NEVER", "ALL", "");
+    
+    /* create the data source to accesss previous image */
+    result = iso_data_source_new_from_file(argv[1], &src);
+    if (result < 0) {
+        printf ("Error creating data source\n");
+        goto ex;
+    }
+    
+    /* create the image context */
+    result = iso_image_new("volume_id", &image);
+    if (result < 0) {
+        printf ("Error creating image\n");
+        goto ex;
+    }
+    iso_tree_set_follow_symlinks(image, 0);
+    iso_tree_set_ignore_hidden(image, 0);
+    
+    /* import previous image */
+    result = iso_read_opts_new(&ropts, 0);
+    if (result < 0) {
+        fprintf(stderr, "Error creating read options\n");
+        goto ex;
+    }
+    result = iso_image_import(image, src, ropts, NULL);
+    iso_read_opts_free(ropts);
+    iso_data_source_unref(src);
+    if (result < 0) {
+        printf ("Error importing previous session %d\n", result);
+        goto ex;
+    }
+    
+    /* add new dir */
+    result = iso_tree_add_dir_rec(image, iso_image_get_root(image), argv[2]);
+    if (result < 0) {
+        printf ("Error adding directory %d\n", result);
+        goto ex;
+    }
+    
+    /* generate a new image with both previous and added contents */
+    result = iso_write_opts_new(&opts, 1);
+    if (result < 0) {
+        printf("Cant create write opts, error %d\n", result);
+        goto ex;
+    }
+    /* for isolinux: iso_write_opts_set_allow_full_ascii(opts, 1); */
+    
+    result = iso_image_create_burn_source(image, opts, &burn_src);
+    if (result < 0) {
+        printf ("Cant create image, error %d\n", result);
+        goto ex;
+    }
+    
+    iso_write_opts_free(opts);
+    
+    while (burn_src->read_xt(burn_src, buf, 2048) == 2048) {
+        fwrite(buf, 1, 2048, fp);
+    }
+    fclose(fp);
+    burn_src->free_data(burn_src);
+    free(burn_src);
+    
+    iso_image_unref(image);
+    iso_finish();
+    return 0;
+ex:
+    if (fp != NULL)
+        fclose(fp);
+    return 1;
+}
+
+
+/* ------------------------- from demo/iso_ms.c ----------------------- */
+
+
+void iso_ms_usage(char **argv)
+{
+    printf("%s LSS NWA DISC DIRECTORY OUTPUT\n", argv[0]);
+}
+
+int gesture_iso_ms(int argc, char **argv)
+{
+    int result;
+    IsoImage *image;
+    IsoDataSource *src;
+    struct burn_source *burn_src;
+    unsigned char buf[2048];
+    FILE *fp = NULL;
+    IsoWriteOpts *opts;
+    IsoReadOpts *ropts;
+    uint32_t ms_block;
+	
+    if (argc < 6) {
+        iso_ms_usage(argv);
+        goto ex;
+    }
+
+    if (strcmp(argv[3], argv[5]) == 0) {
+        fprintf(stderr,
+                "image_file and output_file must not be the same file.\n");
+        goto ex;
+    }
+
+    fp = fopen(argv[5], "w");
+    if (!fp) {
+        err(1, "error opening output file");
+        goto ex;
+    }
+
+    iso_init();
+    iso_set_msgs_severities("NEVER", "ALL", "");
+    
+    /* create the data source to accesss previous image */
+    result = iso_data_source_new_from_file(argv[3], &src);
+    if (result < 0) {
+        printf ("Error creating data source\n");
+        goto ex;
+    }
+    
+    /* create the image context */
+    result = iso_image_new("volume_id", &image);
+    if (result < 0) {
+        printf ("Error creating image\n");
+        goto ex;
+    }
+    iso_tree_set_follow_symlinks(image, 0);
+    iso_tree_set_ignore_hidden(image, 0);
+    
+    /* import previous image */
+    result = iso_read_opts_new(&ropts, 0);
+    if (result < 0) {
+        fprintf(stderr, "Error creating read options\n");
+        goto ex;
+    }
+    iso_read_opts_set_start_block(ropts, atoi(argv[1]));
+    result = iso_image_import(image, src, ropts, NULL);
+    iso_read_opts_free(ropts);
+    iso_data_source_unref(src);
+    if (result < 0) {
+        printf ("Error importing previous session %d\n", result);
+        goto ex;
+    }
+    
+    /* add new dir */
+    result = iso_tree_add_dir_rec(image, iso_image_get_root(image), argv[4]);
+    if (result < 0) {
+        printf ("Error adding directory %d\n", result);
+        goto ex;
+    }
+    
+    /* generate a multisession image with new contents */
+    result = iso_write_opts_new(&opts, 1);
+    if (result < 0) {
+        printf("Cant create write opts, error %d\n", result);
+        goto ex;
+    }
+    
+    /* round up to 32kb aligment = 16 block */
+    ms_block =  atoi(argv[2]);
+    iso_write_opts_set_ms_block(opts, ms_block);
+    iso_write_opts_set_appendable(opts, 1);
+
+    result = iso_image_create_burn_source(image, opts, &burn_src);
+    if (result < 0) {
+        printf ("Cant create image, error %d\n", result);
+        goto ex;
+    }
+    iso_write_opts_free(opts);
+    
+    while (burn_src->read_xt(burn_src, buf, 2048) == 2048) {
+        fwrite(buf, 1, 2048, fp);
+    }
+    fclose(fp);
+    burn_src->free_data(burn_src);
+    free(burn_src);
+    
+    iso_image_unref(image);
+    iso_finish();
+    return 0;
+ex:;
+    if (fp != NULL)
+        fclose(fp);
+    return 1;
+}
+
+
+/* ------------------------- switcher ----------------------- */
+
+
+int main(int argc, char **argv)
+{
+    char *gesture;
+    int i;
+
+    if (argc < 2) {
+usage:;
+        fprintf(stderr, "usage: %s gesture [gesture_options]\n", argv[0]);
+        for (i = 0; helptext[i][0] != '@'; i++)
+             fprintf(stderr, "%s\n", helptext[i]);
+        exit(1);
+    }
+    for (gesture = argv[1]; *gesture == '-'; gesture++);
+    if (strcmp(gesture, "tree") == 0) {
+        gesture_tree(argc - 1, &(argv[1]));
+    } else if(strcmp(gesture, "find") == 0) {
+        gesture_find(argc - 1, &(argv[1]));
+    } else if(strcmp(gesture, "iso") == 0) {
+        gesture_iso(argc - 1, &(argv[1]));
+    } else if(strcmp(gesture, "iso_read") == 0) {
+        gesture_iso_read(argc - 1, &(argv[1]));
+    } else if(strcmp(gesture, "iso_cat") == 0) {
+        gesture_iso_cat(argc - 1, &(argv[1]));
+    } else if(strcmp(gesture, "iso_modify") == 0) {
+        gesture_iso_modify(argc - 1, &(argv[1]));
+    } else if(strcmp(gesture, "iso_ms") == 0) {
+        gesture_iso_ms(argc - 1, &(argv[1]));
+    } else {
+        goto usage;
+    }
+    exit(0);
+}

demo/iso_cat.c

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 

demo/iso_read.c

@@ -10,12 +10,18 @@
 
 #include "libisofs.h"
 
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
 static void
 print_permissions(mode_t mode)
 {
     char perm[10];
 
-    //TODO suid, sticky...
+    /* TODO suid, sticky... */
 
     perm[9] = '\0';
     perm[8] = mode & S_IXOTH ? 'x' : '-';
@@ -51,8 +57,8 @@ print_file_src(IsoFileSource *file)
     iso_file_source_lstat(file, &info);
     print_type(info.st_mode);
     print_permissions(info.st_mode);
-    printf(" %10llu ", info.st_size);
+    printf(" %10.f ", (double) info.st_size);
-    //printf(" {%ld,%ld} ", (long)info.st_dev, (long)info.st_ino);
+    /* printf(" {%ld,%ld} ", (long)info.st_dev, (long)info.st_ino); */
     name = iso_file_source_get_name(file);
     printf(" %s", name);
     free(name);
@@ -156,7 +162,7 @@ int main(int argc, char **argv)
         printf ("Can't get root %d\n", result);
         return 1;
     }
-    //print_file_src(root);
+    /* print_file_src(root); */
     print_dir(root, 0);
     iso_file_source_unref(root);
 

demo/lsl.c

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #include "libisofs.h"
@@ -13,6 +14,12 @@
 #include <stdlib.h>
 #include <limits.h>
 
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
 /*
  * Little test program to test filesystem implementations.
  * 

demo/tree.c

@@ -15,7 +15,7 @@ print_permissions(mode_t mode)
 {
 	char perm[10];
 	
-	//TODO suid, sticky...
+	/* TODO suid, sticky... */
 	
 	perm[9] = '\0';
 	perm[8] = mode & S_IXOTH ? 'x' : '-';

doc/boot_sectors.txt

@@ -0,0 +1,727 @@
+
+
+
+            Collection of Boot Sector Formats for ISO 9660 Images
+
+
+            by Thomas Schmitt    - mailto:scdbackup@gmx.net
+            Libburnia project    - mailto:libburn-hackers@pykix.org
+
+This information is collected from various sources. Some is backed by
+specifications, some is just rumor which happens to work (maybe not even that).
+
+
+                                  Content
+
+EL Torito CD booting, for PC-BIOS x86, PowerPC, (old) Mac, EFI.
+
+MBR, for PC-BIOS x86 from (pseudo-) hard disk
+- SYSLINUX isohybrid MBR
+- GRUB2 grub-mkrescue MBR.
+
+MIPS Volume Header, for MIPS Big Endian, e.g. SGI Indigo2.
+
+DEC Boot Block, for MIPS Little Endian , e.g. DECstation.
+
+SUN Disk Label and boot images, for SUN SPARC
+
+
+------------------------------------------------------------------------------
+
+
+                         EL Torito CD booting
+                 for PC-BIOS x86, PowerPC, (old) Mac, EFI
+
+Sources:
+    El Torito, Bootable CD-ROM Format Specification, Version 1.0, 1995
+    which refers to ECMA-119, the standard for ISO 9660 filesystems.
+    libisofs/eltorito.[ch] by Vreixo Formoso.
+    man mkisofs by Joerg Schilling.
+
+
+ECMA-119 prescribes that the first 32 kB of an ISO 9660 image are System Area
+with arbitrary content. This prescription is obeyed by PC-BIOS systems only
+if the ISO 9660 image is presented on CD, DVD or BD media.
+In this case the El Torito Boot record is the starting point of booting.
+
+The Boot Record is a ECMA-119 Volume Descriptor which is eventually located
+at 2 kB block number 17 (decimal). Its content points to the location of the
+Boot Catalog.
+The format is described in part by ECMA-119 8.2 "Boot Record" and further
+specified by El Torito figure 7.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   0 |          0 | Volume Descriptor Type. 0= Boot record
+   1 -   5 |    "CD001" | Standard Identifier
+   6 -   6 |          1 | Volume Descriptor Version
+   7 -  38 |  el_torito | Boot System Identifier
+  39 -  70 |          0 | Boot Identifier
+           |            |
+  71 -2047 | ========== | Boot System Use
+           |            |
+  71 - 74  | cataloglba | The 2 kB block number of the Boot Catalog
+           |            | as little-endian 32 bit number.
+           |            |
+  75 -2047 |          0 | Unused
+---------- | ---------- | ----------------------------------------------------
+
+el_torito is the constant string "EL TORITO SPECIFICATION" padded by 9 zeros.
+cataloglba has to be provided by the file system generator.
+
+
+The Boot Catalog lists the available boot images which may be prepared for
+multiple system architectures, called "platforms".
+It consists of one or more 2 kB blocks. The content is a sequence of fixed
+format entries, 32 bytes each.
+The entries are grouped in sections, which assign the entries to a particular
+system architecture. The booting system will then choose an entry from an
+appropriate section.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -  31 | ========== | Validation Entry
+           |            | begins the first section, specifies an architecture
+  32 -  63 | ========== | Initial/Default Entry
+           |            | points to a boot image for given architecture
+---------- | ---------- | ----------------------------------------------------
+Optional: 
+---------- | ---------- | ----------------------------------------------------
+  64 -  95 | ========== | Section Header entry
+           |            | begins new section, specifies an architecture
+  96 - 127 | ========== | Section Entry
+           |            | points to a boot image for given architecture
+    ...    | .......... | Optional more Section Entries
+    ...    | .......... | Optional more Section Headers and their Section
+           |            | Entries
+---------- | ---------- | ----------------------------------------------------
+
+An architecture is refered by a Platform Id number.
+Defined by El Torito are:
+    0 = "80x86" which is used for standard PCs with Intel x86 or compatible CPU
+    1 = "PowerPC" (possibly for IBM machines with PowerPC CPU)
+    2 = "Mac" (possibly for Apple computers with MC68000 or PowerPC CPU)
+Further in use by GRUB2 is:
+ 0xef = EFI, a competitor resp. successor to PC-BIOS, possibly in use with
+             Intel ia64 Itanium and possibly with newer Apple machines.
+
+Words resp. numbers are represented are little-endian. 
+
+Validation Entry:
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   0 |          1 | Header Id
+           |            |
+   1 -   1 | platform_id| Platform Id. One of: 0, 1, 2, 0xef. See above.
+           |            |
+   2 -   3 |          0 | Reserved
+   4 -  27 |  manuf_dev | ID string identifies the manufacturer/developer
+           |            | (no non-zero examples known yet)
+           |            |
+  28 -  29 |   checksum | Checksum Word for the Validation Entry. 
+           |            | The sum of all words in the entry has to be 0.
+           |            |
+  30 -  30 |       0x55 | 
+  31 -  31 |       0xaa | 
+---------- | ---------- | ----------------------------------------------------
+
+
+Initial/Default Entry:
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   0 | boot_indct | Boot Indicator: 0x88 = bootable, 0x00 = not bootable
+           |            |
+   1 -   1 | boot_media | Boot Media Type (i.e. media emulated by boot image):
+           |            | 0= no emulation , 1= 1.2 MB diskette, 2=1.44 MB,
+           |            | 3= 2.88 MB , 4= hard disk
+           |            | (About everybody uses 0 = no emulation)
+           |            |
+   2 -   3 |   load_seg | Load Segment. (meaning unclear)
+           |            | "If this value is 0 the system will use the
+           |            |  traditional segment of 7C0."
+           |            | libisofs default is 0
+           |            |
+   4 -   4 |   sys_type | System Type.
+           |            | "Must be a copy of byte 5 from the partition table
+           |            |  found in the boot image."
+           |            | libisofs reads the start the boot image as MBR
+           |            | if  boot_media == 4. This emulated MBR has a
+           |            | partition table from where a byte gets copied.
+           |            | Else this byte is 0.
+           |            |
+   5 -   5 |          0 | Unused
+           |            |
+   6 -   7 |  sec_count | Sector Count.
+           |            | "the number of virtual/emulated sectors the system
+           |            |  will store at Load Segment during the initial boot
+           |            |  procedure."
+           |            | libisofs stores 1 for emulated boot_media and a
+           |            | user defined value for boot_media == 0. Often: 4.
+           |            |
+   8 -  11 |  load_rba  | Load RBA. The 2 kB block address where the boot
+           |            | image file content is located in the ISO 9660 image.
+           |            |
+  12 -  31 |          0 | Unused
+---------- | ---------- | ----------------------------------------------------
+
+
+Section Header Entry:
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   0 |   head_ind | Header Indicator: 0x90 = more headers follow
+           |            |                   0x91 = final header, last section
+           |            |
+   1 -   1 | platform_id| Platform Id. One of: 0, 1, 2, 0xef. See above.
+           |            |
+   2 -   3 | num_entries| Number of entries to follow in this section
+           |            |
+   4 -  31 |            | ID string identifies the manufacturer/developer
+---------- | ---------- | ----------------------------------------------------
+
+
+Section Entry:
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   0 | boot_indct | Boot Indicator: 0x88 = bootable, 0x00 = not bootable
+           |            |
+   1 -   1 | boot_media | Boot Media Type (i.e. media emulated by boot image):
+           |            | Bit 0 to 3 govern emulation
+           |            |   0= no emulation , 1= 1.2 MB diskette, 2=1.44 MB,
+           |            |   3= 2.88 MB , 4= hard disk
+           |            |   (About everybody uses 0 = no emulation)
+           |            | Bit 4 is reserved and must be 0
+           |            | Bit 5 "Continuation entry follows" (meaning unclear)
+           |            |       Might be the indicator for Extension Entries,
+           |            |       which are not described here.
+           |            | Bit 6 "Image contains an ATAPI driver"
+           |            | Bit 7 "Image contains SCSI drivers"
+           |            |
+   2 -   3 |   load_seg | Load Segment. (meaning unclear)
+           |            | See above Initial/Default Entry
+           |            | libisofs default is 0.
+   4 -   4 |   sys_type | System Type.
+           |            | See above Initial/Default Entry
+           |            | 0 if not emulation == 4.
+   5 -   5 |          0 | Unused
+           |            |
+   6 -   7 |  sec_count | Sector Count.
+           |            | See above Initial/Default Entry
+           |            | libisofs stores 1 for emulated boot_media and a
+           |            | user defined value for boot_media == 0. Often: 4.
+           |            |
+   8 -  11 |  load_rba  | Load RBA. The 2 kB block address where the boot
+           |            | image file content is located in the ISO 9660 image.
+           |            |
+  12 -  31 |   sel_crit | "Vendor unique selection criteria."
+---------- | ---------- | ----------------------------------------------------
+
+
+The boot image file content is mostly opaque to the ISO 9660 image generator.
+Nevertheless there is a tradition named "Boot Info Table" which prescribes
+to write information into byte fields of the boot image file content.
+There are no general means known how a producer of ISO 9660 images could
+detect the need for Boot Info Table production.
+It rather needs a hint from the user who has to know whether the boot image
+expects a Boot Info Table.
+The Boot Info Table begins at byte 8 of the boot image content.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   8 -  11 |    pvd_lba | Block address of the Primary Volume Descriptor
+           |            | This is the session start LBA + 16.
+           |            |
+  12 -  15 |   file_lba | Block address of the start of the boot image file
+           |            | content.
+           |            |
+  16 -  19 |   file_len | Number of bytes in boot image file content.
+           |            |
+  20 -  23 |   checksum | Little-endian: The sum of all 32-bit words of the
+           |            | file content from byte 64 to file end.
+           |            |
+  24 -  63 |          0 | Reserved
+---------- | ---------- | ----------------------------------------------------
+
+------------------------------------------------------------------------------
+
+
+                                   MBR
+                  for PC-BIOS x86 from (pseudo-) hard disk
+
+Sources:
+    http://en.wikipedia.org/wiki/Master_boot_record
+    Mailing list conversations with H. Peter Anvin and Vladimir Serbinenko.
+
+
+The candidates for MBR booting will normally use El Torito rather than MBR
+if the ISO image is presented on CD, DVD, or BD media.
+The eventual MBR comes into effect if the image is on a media that is
+interpreted by the BIOS as some kind of hard disk. Usually real hard disks,
+floppy disks, USB sticks, memory cards.
+
+An important part of an MBR is the DOS style partition table. It describes up
+to four primary partitions. There are two formats used for block address:
+Cylinder/Head/Sector (C/H/S) and Logical Block Address (LBA). Both are based
+on units of 512 bytes. So MBR_LBA = ISO_LBA * 4.
+
+For C/H/S, the sector address is broken up into whole cylinders, remaining
+heads, and remaining sectors + 1. The nomenclature seems to stem from antique
+drum storage.
+There are two parameters, sectors_per_head and heads_per_cylinder which are not
+stored in the MBR. So it is more or less arbitray how to convert a LBA into
+a C/H/S address and vice versa. For maximum range of C/H/S addresses one
+may use sectors_per_head = 63 , heads_per_cylinder = 255.
+
+
+Words are composed little-endian style.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 - 439 | = opaque = | Code Area filled with bytes for some boot system,
+           |            | typically machine code.
+           |            |
+ 440 - 443 | disk_sgntr | Disc signature, an individual disk id of obscure
+           |            | usability.
+           |            | (The Code Area might extend up to this field.)
+           |            |
+ 444 - 445 |          0 | "usually nulls"
+           |            | (The Code Area might extend up to this field.)
+           |            |
+ 446 - 461 | ========== | Partition Table Entry for partition 1
+           |            |
+ 446 - 446 |     status | Governs bootability:
+           |            | 0x80 = bootable/active , 0x00 non-bootable/inactive
+           |            |
+ 447 - 449 | ========== | C/H/S address of partition start
+ 447 - 447 | start_head | Heads part of start address.
+ 448 - 448 |  start_c_s | Bits 0 to 5 : Sectors part of start address.
+           |            | Bits 6 to 7 : Bits 8 to 9 of cylinders part.
+ 449 - 449 |  start_cyl | Lower 8 bits of cylinders part of start address
+           |            |
+ 450 - 450 |  part_type | Partition type indicates the purpose or kind of
+           |            | filesystem in the partition.
+           |            |
+ 451 - 453 | ========== | C/H/S address of last absolute sector in partition
+ 451 - 451 |   end_head | Heads part of end address.
+ 452 - 452 |    end_c_s | Bits 0 to 5 : Sectors part of end address.
+                        |               Values: 1 to 63, not 0.
+           |            | Bits 6 to 7 : Bits 8 to 9 of cylinders part.
+ 453 - 453 |    end_cyl | Lower 8 bits of cylinders part of end address
+           |            |
+ 454 - 457 |  start_lba | LBA of first absolute sector in partiton. 
+           |            | Block size is 512. Counting starts at 0.
+           |            |
+ 458 - 461 | num_blocks | Number of sectors in partition.
+           |            |
+ 462 - 477 | ========== | Partition Table Entry for partition 2
+           | part_entr2 | 16 bytes. Format as with partition 1.
+           |            | All 0 means that partition is unused/undefined.
+           |            |
+ 478 - 493 | ========== | Partition Table Entry for partition 3
+           | part_entr3 | 16 bytes. See above.
+           |            |
+ 494 - 509 | ========== | Partition Table Entry for partition 4
+           | part_entr4 | 16 bytes. See above.
+           |            |
+ 510 - 510 |       0x55 | MBR signature
+ 511 - 511 |       0xaa | MBR signature
+           |            |
+---------- | ---------- | ----------------------------------------------------
+
+
+By tradition the MBR itself and possibly more blocks are not claimed by any
+partition. But starting the first partition at a non-zero block address causes
+on Linux a partition device file (e.g. /dev/sdb1) which cannot be used to mount
+the ISO filesystem.
+
+libisofs is able to produce a second set of trees and meta data which is
+suitable for being mounted at start block 16 (ISO) resp. 64 (MBR).
+See <libisofs/libisofs.h> for call iso_write_opts_set_part_offset()
+and http://libburnia-project.org/wiki/PartitionOffset for examples with
+program xorriso.
+
+
+------------------------------------------------------------------------------
+
+
+                             SYSLINUX Isohybrid MBR
+
+Sources:
+    syslinux-3.72/utils/isohybrid , a perl script by H. Peter Anvin = hpa.
+    Mailing list conversations with hpa.
+
+
+An isohybrid MBR directs the booting BIOS to an ISOLINUX boot image which
+is also the target of an El Torito boot catalog entry.
+For that purpose one has to take an MBR template and has to set a few bytes
+to values which sufficiently describe the ISO image and the boot image file.
+
+Words are composed little-endian style.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 - 431 | = opaque = | Syslinux machine code provided by MBR template
+           |            |
+ 432 - 439 | hd_bootlba | Address of the ISOLINUX boot image file in the
+           |            | ISO image. Counted in 512 byte blocks.
+           |            |
+ 440 - 443 |     mbr_id | Random number
+ 444 - 445 |          0 | Padding
+           |            |
+ 446 - 509 | ========== | Partition table
+           |            |
+ 446 - 461 | part_entry | Partition table entry 1 describing ISO image size
+           |            | starting at LBA 0. I.e. contrary to tradition.
+           |            | See above for partition table entry format.
+           |            |
+ 462 - 509 |          0 | Unused partition entries 2 to 4
+ 510 - 511 |     0xaa55 | MBR signature
+---------- | ---------- | ----------------------------------------------------
+
+
+hpa about MBR templates and partition table filesystem types:
+
+"[MBR templates] are available in the Syslinux build tree under the names:
+   mbr/isohdp[fp]x*.bin
+ The default probably should be mbr/isohdppx.bin, but it's ultimately up
+ to the user.
+ [...]
+ Note: the filesystem type is largely arbitrary, in theory it can be any
+ value other than 0x00, 0x05, 0x0f, 0x85, 0xee, or 0xef.  0x17 ("Windows
+ IFS Hidden") seems safeish, some people believe 0x83 (Linux) is better.
+"
+
+------------------------------------------------------------------------------
+
+
+                          GRUB2 grub-mkrescue MBR
+
+Sources:
+    Mailing list conversations with Vladimir Serbinenko.
+
+
+The MBR file that is used with GRUB2 script grub-mkrescue needs only a
+partition table entry which describes the image size.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 - 445 | = opaque = | GRUB2 machine code provided by MBR template
+           |            |
+ 446 - 509 | ========== | Partition table
+           |            |
+ 446 - 461 | part_entry | Partition table entry 1 describing ISO image size
+           |            | Peculiar is the start offset of 1 block.
+           |            | This prevents mounting of the partition.
+           |            | See above for partition table entry format.
+           |            |
+ 462 - 509 |          0 | Unused partition entries 2 to 4
+ 510 - 511 |     0xaa55 | MBR signature
+---------- | ---------- | ----------------------------------------------------
+
+Vladimir Serbinenko about the partition table entry:
+"Currently we use first and not last entry. You need to:
+ 1) Zero-fill 446-510
+ 2) Put 0x55, 0xAA into 510-512
+ 3) Put 0x80 (for bootable partition), 0, 2, 0 (C/H/S of the start), 0xcd
+   (partition type), [3 bytes of C/H/S end], 0x01, 0x00, 0x00, 0x00 (LBA
+   start in little endian), [LBA end in little endian] at 446-462
+"
+
+------------------------------------------------------------------------------
+
+
+                               MIPS Volume Header
+                       for MIPS Big Endian, e.g. SGI Indigo2
+
+Sources:
+   cdrkit-1.1.10/genisoimage/boot-mips.c
+   by Steve McIntyre <steve@einval.com>
+   which refers to
+      genisovh by Florian Lohoff <flo@rfc822.org>
+              and Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
+      who seem to have learned parameter settings from IRIX CD media
+      There are traces in the web which relate this to specs by
+               MIPS Computer Systems, Inc. , 1985
+               Silicon Graphics Computer Systems, Inc. , 2000
+
+
+The first 512 bytes of the media constitute the Volume Header.
+Words are composed big-endian style.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   3 | 0x0be5a941 | Magic number
+   4 -   5 |          0 | Root partition number
+   6 -   7 |          0 | Swap partition number
+   8 -  23 |          0 | Name of file to boot (unclear what this means)
+           |            |
+  24 -  71 | ========== | Device Parameters
+           |            |
+  24 -  24 |          0 | Spiral addressing skew (unclear what this means)
+  25 -  25 |          0 | Words of 0 before header
+  26 -  26 |          0 | Words of 0 between hdr and data
+  27 -  27 |          0 | Spare sectors per cylinder
+  28 -  29 |  num_cyl_l | Number of usable cylinder, lower two bytes
+           |            |  ((iso_size + BYTES_PER_SECTOR - 1) /
+           |            |   (SECTORS_PER_TRACK * BYTES_PER_SECTOR)) & 0xffff
+  30 -  31 |          0 | Starting head of volume 0 
+  32 -  33 |          1 | Number of tracks per cylinder
+  34 -  34 |          0 | Depth of CTQ queue (unclear what this means)
+  35 -  35 |  num_cyl_h | Number of usable cylinders, high byte 
+           |            |  ((iso_size + BYTES_PER_SECTOR - 1) /
+           |            |   (SECTORS_PER_TRACK * BYTES_PER_SECTOR)) >> 16
+  36 -  37 |          0 | unused
+  38 -  39 |         32 | SECTORS_PER_TRACK
+  40 -  41 |        512 | BYTES_PER_SECTOR
+  42 -  43 |          0 | Sector interleave (unclear what this means)
+  44 -  47 | 0x00000034 | Controller characteristics composed from
+           |            | DP_RESEEK 0x00000020 /* recalibrate as last resort */
+           |            | DP_IGNOREERRORS 0x00000010
+           |            |              /* transfer data regardless of errors */
+           |            | DP_TRKFWD 0x00000004
+           |            |                    /* forward to replacement track */
+  48 -  51 |          0 | Bytes/sec for kernel stats
+  52 -  55 |          0 | Max num retries on data error
+  56 -  59 |          0 | ms per word to xfer, for iostat
+  60 -  71 |          0 | 6 parameter words for xylogics controllers
+           |            |
+  72 - 311 | ========== | Volume Directory with 15 entries of 16 bytes each
+           |            |
+  72 -  87 | ========== | Volume Directory Entry 1
+  72 -  79 |  boot_name | Boot file basename, eventually padded by 0 to lenght 8
+  80 -  83 | boot_block | ISO 9660 LBA of boot file * 4, i.e. in blocks of 512
+  84 -  87 | boot_bytes | File length in bytes
+           |            |
+  88 - 311 | see above  | Volume Directory Entries 2 to 15
+           |            |
+ 312 - 504 | ========== | Partition Table with 16 entries of 12 bytes each
+           |            |
+ 312 - 407 |          0 | Unused partition entries 1 to 8
+           |            |
+ 408 - 419 | ========== | Partition Table Entry 9 for Volume Header
+ 408 - 411 |  part_blks | Number of 512 byte blocks in partition
+           |            |(iso_size + (BYTES_PER_SECTOR - 1)) / BYTES_PER_SECTOR
+ 412 - 415 |          0 | Start block of partition
+ 416 - 419 |          0 | PTYPE_VOLHDR = Partition is volume header
+           |            |
+ 420 - 431 |          0 | Unused partition entry 10
+           |            |
+ 432 - 443 | ========== | Partition Table Entry 11 for Volume
+ 432 - 435 |  part_blks | Number of 512 byte blocks in partition
+           |            |(iso_size + (BYTES_PER_SECTOR - 1)) / BYTES_PER_SECTOR
+ 436 - 439 |          0 | Start block of partition
+ 440 - 443 |          6 | PTYPE_VOLUME = Partition is entire volume
+           |            |
+ 444 - 503 |          0 | Unused partition entries 12 to 16
+           |            |
+ 504 - 507 |   head_chk | Volume header checksum
+           |            | The two's complement of bytes 0 to 503 read as big
+           |            | endian unsigned 32 bit: sum(words) + head_chk == 0
+           |            |
+ 508 - 511 |          0 | Volume header end padding
+           |            |
+up to 2048 |          0 | ISO 9660 Block end padding
+---------- | ---------- | ----------------------------------------------------
+
+
+------------------------------------------------------------------------------
+
+
+                               DEC Boot Block
+                    for MIPS Little Endian , e.g. DECstation
+
+Sources:
+   cdrkit-1.1.10/genisoimage/boot-mipsel.c
+   by Steve McIntyre <steve@einval.com>
+   which refers to
+      delo   by Florian Lohoff <flo@rfc822.org>
+            and Thiemo Seufer <seufer@csv.ica.uni-stuttgart.de>
+   cdrkit-1.1.10/include/glibc_elf.h
+   by Steve McIntyre
+   which is based on
+      <elf.h> from GNUC C Library by Free Software Foundation, Inc. 
+
+
+There seems to be only one boot file possible.
+Some information needs to be read out of the ELF headers of this boot file.
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   7 |          0 | Padding
+           |            |
+   8 -  11 | 0x0002757a | Magic number
+           |            |
+  12 -  15 |          1 | Mode /* 0: Single extent, 1: Multi extent boot */
+           |            |
+  16 -  19 |   load_adr | Load address /* Load below kernel */
+           |            | Stems from ELF header of boot file.
+           |            | See below Elf32_Phdr field p_vaddr.
+           |            |
+  20 -  23 |   exec_adr | Execution address /* And exec there */
+           |            | Stems from ELF header of boot file.
+           |            | See below Elf32_Ehdr field e_entry.
+           |            |
+  24 -  31 | ========== | Boot Map Entry 1
+           |            |
+  24 -  27 |   seg_size | Segment size in file. Blocks of 512 bytes.
+           |            | Stems from ELF header of boot file.
+           |            | (Elf32_Phdr field p_filesz + 511) / 512;
+           |            |
+  28 -  31 |  seg_start | Segment file offset. Blocks 512 bytes.
+           |            | ISO 9660 LBA of boot file * 4 plus offset
+           |            | + offset which stems from ELF header of boot file:
+           |            | (Elf32_Phdr field p_offset + 511) / 512;
+           |            |
+  32 - 431 | ========== | Boot Map Entries 2 to 51
+           |          0 |
+           |            |
+---------- | ---------- | ----------------------------------------------------
+
+
+Elf32_Ehdr gets loaded from boot file byte address 0:
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -  23 |            | ( Magic number, file information )
+           |            | 
+  24 -  27 |    e_entry | /* Entry point virtual address */
+           | = exec_adr | Needed for exec_adr
+           |            |
+  28 -  31 |    e_phoff | /* Program header table file offset */
+           |            | Byte address of Elf32_Phdr
+           |            |
+
+Elf32_Phdr gets loaded from boot file byte_address Elf32_Ehdr.e_phoff :
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 -   3 |            | ( Segment type )
+           |            |
+   4 -   7 |   p_offset | /* Segment file offset */
+           |-> seg_start| Needed for seg_start
+           |            |
+   8 -  11 |    p_vaddr | /* Segment virtual address */
+           |  =load_adr | Needed for load_adr
+           |            |
+  12 -  15 |            | (Segment physical address)
+           |            |
+  16 -  19 |   p_filesz | /* Segment size in file */
+           |-> seg_size | Needed for seg_size
+           |            |
+---------- | ---------- | ----------------------------------------------------
+
+
+------------------------------------------------------------------------------
+
+
+                          SUN Disk Label and boot images
+                                 for SUN SPARC
+
+Sources:
+   cdrtools-2.01.01a77/mkisofs/sunlabel.h
+   cdrtools-2.01.01a77/mkisofs/mkisofs.8
+   by Joerg Schilling
+
+
+The Disk Label is written to the first 512 bytes of the image. It can mark
+8 partitions (slices ) of which the first contains the ISO image. The other
+7 may contain boot images.
+Words are composed big-endian style.
+
+Boot images are provided externally. mkisofs arranges them after the end of
+the ISO image so that each starts at a cylinder boundary (320 kB).
+
+There is a mechanism in mkisofs which fills unused partitions by copies of
+their predecessor in the partition table:
+"If the special filename ... is used, the actual and all following
+ boot partitions are mapped to  the  previous  partition.
+ If mkisofs is called with -G image -B ... all boot partitions are
+ mapped to the partition that contains the ISO9660 filesystem."
+
+
+Disk Label components:
+
+Byte Range | Value      | Meaning
+---------- | ---------- | ----------------------------------------------------
+   0 - 127 |      label | ASCII Label
+           |            | "CD-ROM Disc with Sun sparc boot created by ..."
+           |            | mkisofs option -sparc-label
+           |            |
+ 128 - 263 | ========== | /* vtoc inclusions from AT&T SVr4 */
+           |            |
+ 128 - 131 |          1 | Layout version
+ 132 - 139 |          0 | /* volume name           */
+ 140 - 141 |          8 | Number of partitions
+           |            |
+ 142 - 173 | ========== | 8 partition entries of 4 bytes
+           |            |
+ 142 - 145 | ========== | Entry for partition 1
+ 142 - 143 |          4 | ID tag of partition: 4 = User partition
+ 144 - 145 |       0x10 | Permissions: 0x10 = read-only
+           |            |
+ 146 - 149 | ========== | Entry for partition 2 
+ 146 - 147 |    id_tag2 | ID tag of partition:
+           |            | 0 = unused
+           |            | 2 = Root partition with boot image
+ 148 - 149 |      perm2 | Permissions:
+           |            | 0 = unused
+           |            | 0x10 = read-only (if used)
+           |            |
+ 150 - 173 | ========== | Entries for partition 3 to 8.
+           |            | See above: Entry for partition 2
+           |            |
+ 174 - 175 |          0 | Padding
+           |            |
+ 176 - 187 |          0 | /* info for mboot       */
+           |            |
+ 188 - 191 | 0x600ddeee | /* to verify vtoc sanity */
+           |            |
+ 192 - 231 |          0 | Reserved
+           |            |
+ 232 - 263 |          0 | 8 Timestamps of yet unknown format
+           |            |
+ 264 - 419 |          0 | Padding
+           |            |
+ 420 - 443 | ========== | Disk properties
+           |            |
+ 420 - 421 |        350 | Rotations per minute
+ 422 - 423 |       2048 | Number of physical cylinders (fixely 640 MB)
+ 424 - 425 |          0 | /* alternates per cylinder */
+ 426 - 429 |          0 | /* obsolete */
+ 430 - 431 |          1 | /* interleave factor */
+ 432 - 433 |       2048 | Number of data cylinders (fixely 640 MB)
+ 434 - 435 |          0 | /* # of alternate cylinders */
+ 436 - 437 |          1 | Number of heads per cylinder (i.e. 1 cyl = 320 kB)
+ 438 - 439 |        640 | Number of sectors per head (i.e. 1 head = 320 kB)
+ 440 - 443 |          0 | /* obsolete */
+           |            |
+ 444 - 507 | ========== | Partition table
+           |            |
+ 444 - 451 | ========== | Partition table entry #1
+           |            |
+ 444 - 447 |  start_cyl | Start cylinder
+           |            |
+ 448 - 451 | num_blocks | Number of blocks in partition
+           |            |
+ 452 - 507 | ========== | Partition table entries #2 to #8
+           |        ... | See above Partition table entry #1
+           |            |
+ 508 - 509 |     0xdabe | Magic Number
+           |            |
+ 510 - 511 |   checksum | The result of exoring 2-byte words 0 to 254
+           |            |
+---------- | ---------- | ----------------------------------------------------
+
+
+------------------------------------------------------------------------------
+
+>>> ??? HP-PA
+
+------------------------------------------------------------------------------
+
+>>> ??? DEC Alpha
+
+------------------------------------------------------------------------------
+

doc/checksums.txt

@@ -0,0 +1,320 @@
+
+                    Description of libisofs MD5 checksumming
+
+               by Thomas Schmitt    - mailto:scdbackup@gmx.net
+               Libburnia project    - mailto:libburn-hackers@pykix.org
+                                 26 Aug 2009
+
+
+MD5 is a 128 bit message digest with a very low probability to be the same for
+any pair of differing data files. It is described in RFC 1321. and can be
+computed e.g. by program md5sum.
+
+libisofs can equip its images with MD5 checksums for superblock, directory
+tree, the whole session, and for each single data file.
+See libisofs.h, iso_write_opts_set_record_md5().
+
+The data file checksums get loaded together with the directory tree if this
+is enabled by iso_read_opts_set_no_md5(). Loaded checksums can be inquired by
+iso_image_get_session_md5() and iso_file_get_md5().
+
+Stream recognizable checksum tags occupy exactly one block each. They can
+be detected by submitting a block to iso_util_decode_md5_tag(). 
+
+libisofs has own MD5 computation functions:
+iso_md5_start(), iso_md5_compute(), iso_md5_clone(), iso_md5_end(),
+iso_md5_match()
+
+
+                          Representation in the Image
+
+There may be several stream recognizable checksum tags and a compact array
+of MD5 items at the end of the session. The latter allows to quickly load many
+file checksums from media with slow random access.
+
+
+                              The Checksum Array
+
+Location and layout of the checksum array is recorded as AAIP attribute
+"isofs.ca" of the root node.
+See doc/susp_aaip_2_0.txt for a general description of AAIP and
+doc/susp_aaip_isofs_names.txt for the layout of "isofs.ca".
+
+The single data files hold an index to their MD5 checksum in individual AAIP
+attributes "isofs.cx". Index I means: array base address + 16 * I. 
+
+If there are N checksummed data files then the array consists of N + 2 entries
+with 16 bytes each.
+
+Entry number 0 holds a session checksum which covers the range from the session
+start block up to (but not including) the start block of the checksum area.
+This range is described by attribute "isofs.ca" of the root node.
+
+Entries 1 to N hold the checksums of individual data files.
+
+Entry number N + 1 holds the MD5 checksum of entries 0 to N.
+
+
+                             The Checksum Tags
+
+Because the inquiry of AAIP attributes demands loading of the image tree,
+there are also checksum tags which can be detected on the fly when reading
+and checksumming the session from its start point as learned from a media
+table-of-content. 
+
+The superblock checksum tag is written after the ECMA-119 volume descriptors.
+The tree checksum tag is written after the  ECMA-119 directory entries.
+The session checksum tag is written after all payload including the checksum
+array. (Then follows eventual padding.)
+
+The tags are single lines of printable text at the very beginning of a block
+of 2048 bytes. They have the following format:
+
+ Tag_id pos=# range_start=# range_size=# [session_start|next=#] md5=# self=#\n
+
+Tag_id distinguishes the following tag types
+  "libisofs_rlsb32_checksum_tag_v1"     Relocated 64 kB superblock tag
+  "libisofs_sb_checksum_tag_v1"         Superblock tag
+  "libisofs_tree_checksum_tag_v1"       Directory tree tag
+  "libisofs_checksum_tag_v1"            Session tag
+
+A relocated superblock may appear at LBA 0 of an image which was produced for
+being stored in a disk file or on overwriteable media (e.g. DVD+RW, BD-RE).
+Typically there is a first session recorded with a superblock at LBA 32 and
+the next session may follow shortly after its session tag. (Typically at the
+next block address which is divisible by 32.) Normally no session starts after 
+the address given by parameter session_start=.
+
+Session oriented media like CD-R[W], DVD+R, BD-R will have no relocated
+superblock but rather bear a table-of-content on media level (to be inquired
+by MMC commands).
+
+
+Example:
+A relocated superblock which points to the last session. Then the first session
+which starts at Logical Block Address 32. The following sessions have the same
+structure as the first one.
+
+LBA 0:
+   <... ECMA-119 System Area and Volume Descriptors ...>
+LBA 18:
+   libisofs_rlsb32_checksum_tag_v1 pos=18 range_start=0 range_size=18 session_start=311936 md5=6fd252d5b1db52b3c5193447081820e4 self=526f7a3c7fefce09754275c6b924b6d9
+   <... padding up to LBA 32 ...>
+LBA 32:
+   <... First Session: ECMA-119 System Area and Volume Descriptors ...>
+   libisofs_sb_checksum_tag_v1 pos=50 range_start=32 range_size=18 md5=17471035f1360a69eedbd1d0c67a6aa2 self=52d602210883eeababfc9cd287e28682
+   <... ECMA-119 Directory Entries (the tree of file names) ...>
+LBA 334:
+   libisofs_tree_checksum_tag_v1 pos=334 range_start=32 range_size=302 md5=41acd50285339be5318decce39834a45 self=fe100c338c8f9a494a5432b5bfe6bf3c
+   <... Data file payload and checksum array ...>
+LBA 81554:
+   libisofs_checksum_tag_v1 pos=81554 range_start=32 range_size=81522 md5=8adb404bdf7f5c0a078873bb129ee5b9 self=57c2c2192822b658240d62cbc88270cb
+
+   <... more sessions ...>
+
+LBA 311936:
+   <... Last Session: ECMA-119 System Area and Volume Descriptors ...>
+LBA 311954:
+   libisofs_sb_checksum_tag_v1 pos=311954 range_start=311936 range_size=18 next=312286 md5=7f1586e02ac962432dc859a4ae166027 self=2c5fce263cd0ca6984699060f6253e62
+   <... Last Session: tree, tree checksum tag, data payload, session tag ...>
+
+
+There are several tag parameters. Addresses are given as decimal numbers, MD5
+checksums as strings of 32 hex digits.
+
+  pos=
+  gives the block address where the tag supposes itself to be stored.
+  If this does not match the block address where the tag is found then this
+  either indicates that the tag is payload of the image or that the image has
+  been relocated. (The latter makes the image unusable.)
+
+  range_start=
+  The block address where the session is supposed to start. If this does not
+  match the session start on media then the volume descriptors of the
+  image have been relocated. (This can happen with overwriteable media. If
+  checksumming started at LBA 0 and finds range_start=32, then one has to
+  restart checksumming at LBA 32. See libburn/doc/cookbook.txt
+  "ISO 9660 multi-session emulation on overwriteable media" for background
+  information.)
+
+  range_size=
+  The number of blocks beginning at range_start which are covered by the
+  checksum of the tag.  
+
+  Only with superblock tag and tree tag:
+  next=
+  The block address where the next tag is supposed to be found. This is
+  to avoid the small possibility that a checksum tag with matching position
+  is part of a directory entry or data file. The superblock tag is quite
+  uniquely placed directly after the ECMA-119 Volume Descriptor Set Terminator
+  where no such cleartext is supposed to reside by accident.
+
+  Only with relocated 64 kB superblock tag:
+  session_start=
+  The start block address (System Area) of the session to which the relocated
+  superblock points.  
+
+  md5=
+  The checksum payload of the tag as lower case hex digits.
+
+  self=
+  The MD5 checksum of the tag itself up to and including the last hex digit of
+  parameter "md5=".
+  
+The newline character at the end is mandatory. After that newline there may
+follow more lines. Their meaning is not necessarily described in this document.
+
+One such line type is the scdbackup checksum tag, an ancestor of libisofs tags
+which is suitable only for single session images which begin at LBA 0. It bears
+a checksum record which by its MD5 covers all bytes from LBA 0 up to the
+newline character preceding the scdbackup tag. See scdbackup/README appendix
+VERIFY for details.
+
+-------------------------------------------------------------------------------
+
+                              Usage at Read Time
+
+                     Checking Before Image Tree Loading
+
+In order to check for a trustworthy loadable image tree, read the first 32
+blocks from to the session start and look in block 16 to 32 for a superblock
+checksum tag by
+  iso_util_decode_md5_tag(block, &tag_type, &pos,
+                          &range_start, &range_size, &next_tag, md5, 0);
+
+If a tag of type 2 or 4 appears and has plausible parameters, then check
+whether its MD5 matches the MD5 of the data blocks which were read before.
+
+With tag type 2:
+
+Keep the original MD5 context of the data blocks and clone one for obtaining
+the MD5 bytes.
+If the MD5s match, then compute the checksum block and all folowing ones into
+the kept MD5 context and go on with reading and computing for the tree checksum
+tag. This will be found at block address next_tag, verified and parsed by:
+  iso_util_decode_md5_tag(block, &tag_type, &pos,
+                          &range_start, &range_size, &next_tag, md5, 3);
+
+Again, if the parameters match the reading state, the MD5 must match the
+MD5 computed from the data blocks which were before.
+If so, then the tree is ok and safe to be loaded by iso_image_import().
+
+With tag type 4:
+
+End the MD5 context and start a new context for the session which you will
+read next.
+
+Then look for the actual session by starting to read at the address given by
+parameter session_start= which is returned by iso_util_decode_md5_tag() as
+next_tag. Go on by looking for tag type 2 and follow above prescription.
+
+
+                      Checking the Data Part of the Session
+
+In order to check the trustworthyness of a whole session, continue reading
+and checksumming after the tree was verified. 
+
+Read and checksum the blocks. When reaching block address next_tag (from the
+tree tag) submit this block to
+
+  iso_util_decode_md5_tag(block, &tag_type, &pos,
+                          &range_start, &range_size, &next_tag, md5, 1);
+
+If this returns 1, then check whether the returned parameters pos, range_start,
+and range_size match the state of block reading, and whether the returned
+bytes in parameter md5 match the MD5 computed from the data blocks which were
+read before the tag block.
+
+
+                           Checking All Sessions
+
+If the media is sequentially recordable, obtain a table of content and check
+the first track of each session as prescribed above in Checking Before Image
+Tree Loading and in Checking the Data Part of the Session.
+
+With disk files or overwriteable media, look for a relocated superblock tag
+but do not hop to address next_tag (given by session_start=). Instead look at
+LBA 32 for the first session and check it as prescribed above.
+After reaching its end, round up the read address to the next multiple of 32
+and check whether it is smaller than session_start= from the super block.
+If so, expect another session to start there.
+
+
+                   Checking Single Files in a Loaded Image
+
+An image may consist of many sessions wherein many data blocks may not belong
+to files in the directory tree of the most recent session. Checking this
+tree and all its data files can ensure that all actually valid data in the
+image are trustworthy. This will leave out the trees of the older sessions
+and the obsolete data blocks of overwritten or deleted files.
+
+Once the image has been loaded, you can obtain MD5 sums from IsoNode objects
+which fulfill
+  iso_node_get_type(node) == LIBISO_FILE
+
+The recorded checksum can be obtained by
+  iso_file_get_md5(image, (IsoFile *) node, md5, 0);
+
+For accessing the file data in the loaded image use 
+  iso_file_get_stream((IsoFile *) node);
+to get the data stream of the object.
+The checksums cover the data content as it was actually written into the ISO
+image stream, not necessarily as it was on hard disk before or afterwards.
+This implies that content filtered files bear the MD5 of the filtered data
+and not of the original files on disk. When checkreading, one has to avoid
+any reverse filtering. Dig out the stream which directly reads image data
+by calling iso_stream_get_input_stream() until it returns NULL and use
+iso_stream_get_size() rather than iso_file_get_size().
+
+Now you may call iso_stream_open(), iso_stream_read(), iso_stream_close()
+for reading file content from the loaded image.
+
+
+                        Session Check in a Loaded Image
+
+iso_image_get_session_md5() gives start LBA and session payload size as of
+"isofs.ca" and the session checksum as of the checksum array.
+
+For reading you may use the IsoDataSource object which you submitted
+to iso_image_import() when reading the image. If this source is associated
+to a libburn drive, then libburn function burn_read_data() can read directly
+from it.
+
+-------------------------------------------------------------------------------
+
+                            scdbackup Checksum Tags
+
+The session checksum tag does not occupy its whole block. So there is room to
+store a scdbackup stream checksum tag, which is an ancestor format of the tags
+described here. This feature allows scdbackup to omit its own checksum filter
+if using xorriso as ISO 9660 formatter program.
+Such a tag makes only sense if the session begins at LBA 0.
+
+See scdbackup-*/README, appendix VERIFY for a specification.
+
+Example of a scdbackup checksum tag:
+scdbackup_checksum_tag_v0.1 2456606865 61 2_2 B00109.143415 2456606865 485bbef110870c45754d7adcc844a72c c2355d5ea3c94d792ff5893dfe0d6d7b
+
+The tag is located at byte position 2456606865, contains 61 bytes of scdbackup
+checksum record (the next four words):
+Name of the backup volume is "2_2".
+Written in year B0 = 2010 (A9 = 2009, B1 = 2011), January (01), 9th (09),
+14:34:15 local time. 
+The size of the volume is 2456606865 bytes, which have a MD5 sum of
+485bbef110870c45754d7adcc844a72c.
+The checksum of "2_2 B00109.143415 2456606865 485bbef110870c45754d7adcc844a72c"
+is c2355d5ea3c94d792ff5893dfe0d6d7b.
+
+-------------------------------------------------------------------------------
+
+This text is under
+Copyright (c) 2009 - 2010 Thomas Schmitt <scdbackup@gmx.net>
+It shall only be modified in sync with libisofs and other software which
+makes use of libisofs checksums. Please mail change requests to mailing list
+<libburn-hackers@pykix.org> or to the copyright holder in private.
+Only if you cannot reach the copyright holder for at least one month it is
+permissible to modify this text under the same license as the affected
+copy of libisofs.
+If you do so, you commit yourself to taking reasonable effort to stay in 
+sync with the other interested users of this text.
+

doc/doxygen.conf.in

@@ -154,13 +154,6 @@ QT_AUTOBRIEF           = NO
 
 MULTILINE_CPP_IS_BRIEF = YES
 
-# If the DETAILS_AT_TOP tag is set to YES then Doxygen 
-# will output the detailed description near the top, like JavaDoc.
-# If set to NO, the detailed description appears after the member 
-# documentation.
-
-DETAILS_AT_TOP         = YES
-
 # If the INHERIT_DOCS tag is set to YES (the default) then an undocumented 
 # member inherits the documentation from any documented member that it 
 # re-implements.

doc/susp_aaip_2_0.txt

@@ -443,3 +443,14 @@ Program mkisofs emits entry XA
 
 -------------------------------------------------------------------------------
 
+This text is under
+Copyright (c) 2009 - 2010 Thomas Schmitt <scdbackup@gmx.net>
+It shall only be modified in sync with libisofs and other software which
+makes use of AAIP. Please mail change requests to mailing list
+<libburn-hackers@pykix.org> or to the copyright holder in private.
+Only if you cannot reach the copyright holder for at least one month it is
+permissible to modify this text under the same license as the affected
+copy of libisofs.
+If you do so, you commit yourself to taking reasonable effort to stay in 
+sync with the other interested users of this text.
+

doc/susp_aaip_isofs_names.txt

@@ -12,6 +12,95 @@ specification of AAIP :
 
 -------------------------------------------------------------------------------
 
+Name:
+      isofs.ca
+
+Purpose:
+      Records the range of checksummed image data (START, END), the number
+      of checksum items (COUNT), the number of bytes in a single checksum item
+      (SIZE), and the name of the checksum algorithm (CHECKSUM_TYPE).
+      END is also the block address of the start of the checksum recording
+      area in the image.
+      See also isofs.cx .
+      This attribute shall eventually be attached to the root directory entry
+      and be global for the whole image.
+
+Format of Value:
+      START_LEN | START_BYTES | END_LEN | END_BYTES |
+      COUNT_LEN | COUNT_BYTES | SIZE_LEN | SIZE_BYTES | CHECKSUM_TYPE
+      Each number is encoded as _LEN byte and _BYTES value string.
+      The _LEN fields comply to ISO 9660 Format section 7.1.1.
+      The byte strings START_BYTES, END_BYTES, COUNT_BYTES, SIZE_BYTES begin
+      with the most significant byte. Leading zero bytes are allowed.
+      CHECKSUM_TYPE consists of the bytes after
+      START_LEN + END_LEN + COUNT_LEN + SIZE_LEN + 4.
+      It shall be a string of printable characters without terminating 0-byte.
+      Type names shall be registered here.
+      For now there is:
+       "MD5"   128 bit message digest, see RFC 1321, see man md5sum 
+
+Example:
+      LBA range 32 to 1000000 , 520 checksums recorded, MD5 
+      {   1,  32,
+          3,  15,  66,  64,
+          2,   2,   8,
+          1,  16,
+        'M', 'D', '5' }
+      or
+      {   4,   0,   0,   0,  32,
+          4,   0,  15,  66,  64,
+          4,   0,   0,   2,   8,
+          1,  16,
+         'M', 'D', '5' }
+
+Registered:
+      16 Jul 2009 by Thomas Schmitt for libisofs.
+
+-------------------------------------------------------------------------------
+
+Name:
+      isofs.cs
+
+Purpose:
+      Records the name of the character set that was used as output character
+      set when writing the RRIP name tree of the ISO 9660 image. It shall be
+      suitable as parameter for function iconv_open(3).
+      This attribute shall eventually be attached to the root directory entry
+      and be global for the whole image.
+
+Format of Value:
+      Shall hold the character set name without terminating 0-byte.
+
+Example:
+      { 'I', 'S', 'O', '-', '8', '8', '5', '9' , '-', '1' }
+
+Registered:
+      18 Mar 2009 by Thomas Schmitt for libisofs.
+
+-------------------------------------------------------------------------------
+
+Name:
+      isofs.cx
+
+Purpose:
+      Records the index of the file's checksum in the checksum area at the
+      end of the image. The byte address of the checksum is 
+       checksum_area_lba * 2048 + isofs.cx * checksum_size
+      Default checksum algorithm is MD5 with a size of 16 byte.
+      See also isofs.ca .
+
+Format of Value:
+      A byte string which begins with the most significant byte.
+
+Example:
+      Index 123456
+      {   1, 226,  64 }
+
+Registered:
+      16 Jul 2009 by Thomas Schmitt for libisofs.
+
+-------------------------------------------------------------------------------
+
 Name:
       isofs.di
 
@@ -31,28 +120,7 @@ Example:
 
 Registered:
       17 Feb 2009 by Thomas Schmitt for xorriso.
-                 
--------------------------------------------------------------------------------
 
-Name:
-      isofs.cs
-
-Purpose:
-      Records the name of the character set that was used as output character
-      set when writing the RRIP name tree of the ISO 9660 image. It shall be
-      suitable as parameter for function iconv_open(3).
-      This attribute shall eventually be attached to the root directory entry
-      and be global for the whole image.
-
-Format of Value:
-      Shall hold the character set name without terminating 0-byte.
-
-Example:
-      { 'I', 'S', 'O', '-', '8', '8', '5', '9' , '-', '1' }
-
-Registered:
-      18 Mar 2009 by Thomas Schmitt for libisofs.
-                 
 -------------------------------------------------------------------------------
 
 Name:
@@ -78,8 +146,19 @@ Example:
 
 Registered:
       03 Apr 2009 by Thomas Schmitt for xorriso.
-                 
+
 -------------------------------------------------------------------------------
 -------------------------------------------------------------------------------
 -------------------------------------------------------------------------------
 
+This text is under
+Copyright (c) 2009 - 2011 Thomas Schmitt <scdbackup@gmx.net>
+It shall only be modified in sync with libisofs and other software which
+makes use of AAIP. Please mail change requests to mailing list
+<libburn-hackers@pykix.org> or to the copyright holder in private.
+Only if you cannot reach the copyright holder for at least one month it is
+permissible to modify this text under the same license as the affected
+copy of libisofs.
+If you do so, you commit yourself to taking reasonable effort to stay in
+sync with the other interested users of this text.
+

doc/zisofs_format.txt

@@ -141,3 +141,14 @@ SUSP 1.12
 RRIP 1.12
   ftp://ftp.ymi.com/pub/rockridge/rrip112.ps
 
+-------------------------------------------------------------------------------
+
+This text is under
+Copyright (c) 2009 - 2010 Thomas Schmitt <scdbackup@gmx.net>
+It shall reflect the effective technical specifications as implemented in
+zisofs-tools and the Linux kernel. So please contact mailing list
+<libburn-hackers@pykix.org> or to the copyright holder in private, if you
+want to make changes.
+Only if you cannot reach the copyright holder for at least one month it is
+permissible to modify and distribute this text under the license "BSD revised".
+

libisofs/aaip-os-dummy.c

@@ -11,10 +11,14 @@
 
  To be included by aaip_0_2.c
 
- Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2
+ Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2+
 
 */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <ctype.h>
 #include <sys/types.h>
 #include <unistd.h>

libisofs/aaip-os-freebsd.c

@@ -3,7 +3,7 @@
 
  aaip-os-freebsd.c
  Arbitrary Attribute Interchange Protocol , system adapter for getting and
- setting of ACLs and ixattr.
+ setting of ACLs and xattr.
 
  To be included by aaip_0_2.c
 
@@ -11,6 +11,10 @@
 
 */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <ctype.h>
 #include <sys/types.h>
 #include <unistd.h>
@@ -18,6 +22,7 @@
 #include <string.h>
 #include <stdio.h>
 #include <sys/stat.h>
+#include <errno.h>
 
 #ifndef Libisofs_with_aaip_acL
 /* It seems ACL is fixely integrated in FreeBSD libc. There is no libacl. */
@@ -37,13 +42,14 @@
                         finally has to be freed by a call to this function
                         with bit15 of flag.
    @param flag          Bitfield for control purposes
-                        bit0=  obtain default ACL rather than access ACL
+                        (bit0=  obtain default ACL rather than access ACL)
                         bit4=  set *text = NULL and return 2
                                if the ACL matches st_mode permissions.
                         bit5=  in case of symbolic link: inquire link target
                         bit15= free text and return 1
    @return              > 0 ok
                           0 ACL support not enabled at compile time
+                            or filesystem does not support ACL
                          -1 failure of system ACL service (see errno)
                          -2 attempt to inquire ACL of a symbolic
                             link without bit4 or bit5
@@ -88,8 +94,18 @@ int aaip_get_acl_text(char *path, char **text, int flag)
 
  acl= acl_get_file(path, ACL_TYPE_ACCESS);
 
- if(acl == NULL)
+ if(acl == NULL) {
+   if(errno == EOPNOTSUPP) {
+     /* filesystem does not support ACL */
+     if(flag & 16)
+       return(2);
+
+     /* >>> ??? fake ACL from POSIX permissions ? */;
+
+     return(0);
+   }
    return(-1);
+ }
  *text= acl_to_text(acl, NULL);
  acl_free(acl);
 
@@ -181,10 +197,8 @@ int aaip_get_attr_list(char *path, size_t *num_attrs, char ***names,
 
  if(flag & 1) { /* Obtain ACL */
    /* access-ACL */
-   ret= aaip_get_acl_text(path, &acl_text, flag & (16 | 32));
+   aaip_get_acl_text(path, &acl_text, flag & (16 | 32));
-   if(ret <= 0)
+   if(acl_text == NULL)
-     goto ex;
-   if(ret == 2)
      {ret= 1; goto ex;} /* empty ACL / only st_mode info was found in ACL */
    ret= aaip_encode_acl(acl_text, (mode_t) 0, &a_acl_len, &a_acl, flag & 2);
    if(ret <= 0)

libisofs/aaip-os-linux.c

@@ -7,10 +7,14 @@
 
  To be included by aaip_0_2.c
 
- Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2
+ Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2+
 
 */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <ctype.h>
 #include <sys/types.h>
 #include <unistd.h>
@@ -18,6 +22,8 @@
 #include <string.h>
 #include <stdio.h>
 #include <sys/stat.h>
+#include <errno.h>
+
 
 #ifdef Libisofs_with_aaip_acL
 #include <sys/acl.h>
@@ -46,6 +52,7 @@
                           2 only st_mode permissions exist and bit 4 is set
                             or empty ACL and bit0 is set
                           0 ACL support not enabled at compile time
+                            or filesystem does not support ACL
                          -1 failure of system ACL service (see errno)
                          -2 attempt to inquire ACL of a symbolic link without
                             bit4 or bit5 resp. with no suitable link target
@@ -79,8 +86,18 @@ int aaip_get_acl_text(char *path, char **text, int flag)
  }
  
  acl= acl_get_file(path, (flag & 1) ? ACL_TYPE_DEFAULT : ACL_TYPE_ACCESS);
- if(acl == NULL)
+ if(acl == NULL) {
+   if(errno == ENOTSUP) {
+     /* filesystem does not support ACL */
+     if(flag & 16)
+       return(2);
+   
+     /* >>> ??? fake ACL from POSIX permissions ? */;
+
+     return(0);   
+   }
    return(-1);
+ }
  *text= acl_to_text(acl, NULL);
  acl_free(acl);
 
@@ -131,13 +148,19 @@ int aaip_get_acl_text(char *path, char **text, int flag)
 int aaip_get_attr_list(char *path, size_t *num_attrs, char ***names,
                        size_t **value_lengths, char ***values, int flag)
 {
- int ret, retry= 0;
+ int ret;
  char *list= NULL;
- ssize_t list_size= 0, i, num_names= 0, value_ret;
+ ssize_t list_size= 0, i, num_names= 0;
- size_t acl_len= 0;
  unsigned char *acl= NULL;
  char *a_acl_text= NULL, *d_acl_text= NULL;
 
+#ifdef Libisofs_with_aaip_acL
+ size_t acl_len= 0;
+#endif
+#ifdef Libisofs_with_aaip_xattR
+ ssize_t value_ret, retry= 0;
+#endif
+
  if(flag & (1 << 15)) { /* Free memory */
    {ret= 1; goto ex;}
  }
@@ -272,9 +295,9 @@ ex:;
    aaip_get_acl_text("", &a_acl_text, 1 << 15); /* free */
  if(d_acl_text != NULL)
    aaip_get_acl_text("", &d_acl_text, 1 << 15); /* free */
+ if(list != NULL)
+   free(list);
  if(ret <= 0 || (flag & (1 << 15))) {
-   if(list != NULL)
-     free(list);
    if(*names != NULL) {
      for(i= 0; i < *num_attrs; i++)
        free((*names)[i]);
@@ -375,8 +398,11 @@ int aaip_set_attr_list(char *path, size_t num_attrs, char **names,
                        size_t *value_lengths, char **values, int flag)
 {
  int ret, has_default_acl= 0;
- size_t i, consumed, acl_text_fill, list_size= 0, acl_idx= 0, h_consumed;
+ size_t i, consumed, acl_text_fill, acl_idx= 0, h_consumed;
  char *acl_text= NULL, *list= NULL;
+#ifdef Libisofs_with_aaip_xattR
+ size_t list_size= 0;
+#endif
 
 #ifdef Libisofs_with_aaip_xattR
 

libisofs/aaip_0_2.c

@@ -7,10 +7,14 @@
  See test/aaip_0_2.h
      http://libburnia-project.org/wiki/AAIP
 
- Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2
+ Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2+
 
 */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <ctype.h>
 #include <sys/types.h>
 #include <unistd.h>
@@ -23,9 +27,9 @@
 
 #include "libisofs.h"
 
-/* <<<
+/*
-*/
 #define Aaip_encode_debuG 1
+*/
 
 #include "aaip_0_2.h"
 
@@ -128,13 +132,7 @@ size_t aaip_encode(size_t num_attrs, char **names,
  /* write the field headers */
  for(i= 0; i < number_of_fields; i++) {
    (*result)[i * 255 + 0]= 'A';
-
-#ifdef Libisofs_aaip_2_0
    (*result)[i * 255 + 1]= 'L';
-#else /* Libisofs_aaip_2_0 */
-   (*result)[i * 255 + 1]= 'A';
-#endif /* ! Libisofs_aaip_2_0 */
-
    if(i < number_of_fields - 1 || (mem_size % 255) == 0)
      (*result)[i * 255 + 2]= 255;
    else 
@@ -573,7 +571,7 @@ ex:;
 }
 
 
-/* Linux man 5 acl says:
+/* GNU/Linux man 5 acl says:
      The permissions defined by ACLs are a superset of the permissions speci-
      fied by the file permission bits. The permissions defined for the file
      owner correspond to the permissions of the ACL_USER_OBJ entry.  The per-
@@ -1186,16 +1184,9 @@ static int aaip_consume_aa_head(struct aaip_state *aaip,
  aaip->aa_head_missing-= todo;
  if(aaip->aa_head_missing == 0) {
    aaip_read_from_recs(aaip, aaip->recs_fill - 5, aa_head, 5, 0);
-
-#ifdef Libisofs_aaip_2_0
    if(aa_head[0] != 'A' || (aa_head[1] != 'L' && aa_head[1] != 'A') ||
       aa_head[3] != 1)
      return(-1);
-#else /* Libisofs_aaip_2_0 */
-   if(aa_head[0] != 'A' || aa_head[1] != 'A' || aa_head[3] != 1)
-     return(-1);
-#endif /* ! Libisofs_aaip_2_0 */
-
    aaip->aa_missing= aa_head[2];
    aaip->aa_ends= !(aa_head[4] & 1);
    aaip->recs_fill-= 5; /* AAIP field heads do not get delivered */

libisofs/aaip_0_2.h

@@ -9,7 +9,7 @@
 
  test/aaip_0_2.h - Public declarations
 
- Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2
+ Copyright (c) 2009 Thomas Schmitt, libburnia project, GPLv2+
 
 */
 

libisofs/buffer.c

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /*
@@ -16,6 +17,10 @@
  *
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 /*
    Use the copy of the struct burn_source definition in libisofs.h
 */
@@ -288,7 +293,51 @@ unsigned int iso_ring_buffer_get_times_empty(IsoRingBuffer *buf)
 }
 
 
-/**
+/** Internal via buffer.h
+ *
+ * Get the status of a ring buffer.
+ *
+ * @param buf
+ *      The ring buffer object to inquire
+ * @param size
+ *      Will be filled with the total size of the buffer, in bytes
+ * @param free_bytes
+ *      Will be filled with the bytes currently available in buffer
+ * @return
+ *      < 0 error, > 0 state:
+ *           1="active"    : input and consumption are active
+ *           2="ending"    : input has ended without error
+ *           3="failing"   : input had error and ended,
+ *           5="abandoned" : consumption has ended prematurely
+ *           6="ended"     : consumption has ended without input error
+ *           7="aborted"   : consumption has ended after input error
+ */
+int iso_ring_buffer_get_buf_status(IsoRingBuffer *buf, size_t *size,
+                                   size_t *free_bytes)
+{
+    int ret;
+
+    if (buf == NULL) {
+        return ISO_NULL_POINTER;
+    }
+
+    /* get mutex */
+    pthread_mutex_lock(&buf->mutex);
+    if (size) {
+        *size = buf->cap;
+    }
+    if (free_bytes) {
+        *free_bytes = buf->cap - buf->size;
+    }
+
+    ret = (buf->rend ? 4 : 0) + (buf->wend + 1);
+
+    pthread_mutex_unlock(&buf->mutex);
+    return ret;
+}
+
+/** API via libisofs.h
+ *
  * Get the status of the buffer used by a burn_source.
  *
  * @param b
@@ -316,18 +365,7 @@ int iso_ring_buffer_get_status(struct burn_source *b, size_t *size,
         return ISO_NULL_POINTER;
     }
     buf = ((Ecma119Image*)(b->data))->buffer;
-
+    ret = iso_ring_buffer_get_buf_status(buf, size, free_bytes);
-    /* get mutex */
-    pthread_mutex_lock(&buf->mutex);
-    if (size) {
-        *size = buf->cap;
-    }
-    if (free_bytes) {
-        *free_bytes = buf->cap - buf->size;
-    }
-
-    ret = (buf->rend ? 4 : 0) + (buf->wend + 1);
-
-    pthread_mutex_unlock(&buf->mutex);
     return ret;
 }
+

libisofs/buffer.h

@@ -2,15 +2,23 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #ifndef LIBISO_BUFFER_H_
 #define LIBISO_BUFFER_H_
 
 #include <stdlib.h>
+
+#ifdef HAVE_STDINT_H
 #include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
 
 #define BLOCK_SIZE          2048
 
@@ -62,6 +70,28 @@ int iso_ring_buffer_write(IsoRingBuffer *buf, uint8_t *data, size_t count);
  */
 int iso_ring_buffer_read(IsoRingBuffer *buf, uint8_t *dest, size_t count);
 
+/** Backend of API call iso_ring_buffer_get_status()
+ *
+ * Get the status of a ring buffer.
+ *
+ * @param buf
+ *      The ring buffer object to inquire
+ * @param size
+ *      Will be filled with the total size of the buffer, in bytes
+ * @param free_bytes
+ *      Will be filled with the bytes currently available in buffer
+ * @return
+ *      < 0 error, > 0 state:
+ *           1="active"    : input and consumption are active
+ *           2="ending"    : input has ended without error
+ *           3="failing"   : input had error and ended,
+ *           5="abandoned" : consumption has ended prematurely
+ *           6="ended"     : consumption has ended without input error
+ *           7="aborted"   : consumption has ended after input error
+ */
+int iso_ring_buffer_get_buf_status(IsoRingBuffer *buf, size_t *size,
+                                   size_t *free_bytes);
+
 /** 
  * Close the buffer (to be called by the writer).
  * You have to explicity close the buffer when you don't have more data to

libisofs/builder.c

@@ -3,10 +3,15 @@
  * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 /* libisofs.h defines aaip_xinfo_func */
 #include "libisofs.h"
 
@@ -19,6 +24,11 @@
 #include <stdlib.h>
 #include <string.h>
 #include <limits.h>
+#include <stdio.h>
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
 
 
 
@@ -92,8 +102,9 @@ int default_create_node(IsoNodeBuilder *builder, IsoImage *image,
     int ret;
     struct stat info;
     IsoNode *new;
+    IsoFilesystem *fs;
     char *name;
-    unsigned char *aa_string;
+    unsigned char *aa_string = NULL;
     char *a_text = NULL, *d_text = NULL;
 
     if (builder == NULL || src == NULL || node == NULL) {
@@ -111,6 +122,7 @@ int default_create_node(IsoNodeBuilder *builder, IsoImage *image,
     }
 
     name = iso_file_source_get_name(src);
+    fs = iso_file_source_get_filesystem(src);
     new = NULL;
 
     switch (info.st_mode & S_IFMT) {
@@ -154,6 +166,13 @@ int default_create_node(IsoNodeBuilder *builder, IsoImage *image,
             }
             ret = iso_node_new_symlink(name, strdup(dest), &link);
             new = (IsoNode*) link;
+            if (fs != NULL) {
+                link->fs_id = fs->get_id(fs);
+                if (link->fs_id != 0) {
+                    link->st_ino = info.st_ino;
+                    link->st_dev = info.st_dev;
+                }
+            }
         }
         break;
     case S_IFSOCK:
@@ -166,6 +185,13 @@ int default_create_node(IsoNodeBuilder *builder, IsoImage *image,
             ret = iso_node_new_special(name, info.st_mode, info.st_rdev, 
                                        &special);
             new = (IsoNode*) special;
+            if (fs != NULL) {
+                special->fs_id = fs->get_id(fs);
+                if (special->fs_id != 0) {
+                    special->st_ino = info.st_ino;
+                    special->st_dev = info.st_dev;
+                }
+            }
         }
         break;
     }
@@ -195,6 +221,9 @@ int default_create_node(IsoNodeBuilder *builder, IsoImage *image,
         }
         iso_aa_get_acl_text(aa_string, info.st_mode, &a_text, &d_text,
                             1 << 15); /* free ACL texts */
+        if(aa_string != NULL)
+            free(aa_string);
+        aa_string = NULL;
     }
 
     /* Obtain ownership of eventual AAIP string */
@@ -205,6 +234,8 @@ int default_create_node(IsoNodeBuilder *builder, IsoImage *image,
         ret = iso_node_add_xinfo(new, aaip_xinfo_func, aa_string);
         if (ret < 0)
             return ret;
+    } else if(aa_string != NULL) {
+        free(aa_string);
     }
 
     *node = new;

libisofs/builder.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #ifndef LIBISO_BUILDER_H_

libisofs/data_source.c

@@ -2,10 +2,15 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "util.h"
 

libisofs/ecma119.h

@@ -1,10 +1,11 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
- * Copyright (c) 2009 Thomas Schmitt
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #ifndef LIBISO_ECMA119_H_
@@ -14,7 +15,14 @@
 #include "util.h"
 #include "buffer.h"
 
+#ifdef HAVE_STDINT_H
 #include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
+
 #include <pthread.h>
 
 #define BLOCK_SIZE      2048
@@ -31,11 +39,44 @@
  */
 #define ISO_EXTENT_SIZE             0xFFFFF800
 
+/*
+ * The maximum number of partition images that can be registered. Depending
+ * on the system area type, the effectively usable number may be smaller or
+ * even 0.
+ */
+#define ISO_MAX_PARTITIONS 8
+
+/*
+ * The cylindersize with SUN Disk Label
+ * (512 bytes/sector, 640 sectors/head, 1 head/cyl = 320 KiB).
+ * Expressed in ECMA-119 blocks of 2048 bytes/block.
+ */
+#define ISO_SUN_CYL_SIZE 160
+
+/*
+ * Maximum length of a disc label text plus 1.
+ */
+#define ISO_DISC_LABEL_SIZE 129
+
+
+/* The maximum lenght of an specs violating ECMA-119 file identifier.
+   The theoretical limit is  254 - 34 - 28 (len of SUSP CE entry) = 192
+   Currently the practical limit is 254 - 34 - 96 (non-CE RR entries) - 28 (CE)
+*/
+#ifdef Libisofs_with_rrip_rR
+#define ISO_UNTRANSLATED_NAMES_MAX 92
+#else
+#define ISO_UNTRANSLATED_NAMES_MAX 96
+#endif
+
+
 /**
  * Holds the options for the image generation.
  */
 struct iso_write_opts {
 
+    int will_cancel;
+
     int level; /**< ISO level to write at. (ECMA-119, 10) */
 
     /** Which extensions to support. */
@@ -53,11 +94,21 @@ struct iso_write_opts {
      * but it is supposed to work on most moderns systems. Use with caution.
      */
 
+    /**
+     * Convert directory names for ECMA-119 the same way as other file names
+     * but do not force dots or add version numbers.
+     * This violates ECMA-119 by allowing one "." and especially ISO level 1
+     * by allowing DOS style 8.3 names rather than only 8 characters.
+     */
+    unsigned int allow_dir_id_ext :1;
+
     /**
      * Omit the version number (";1") at the end of the ISO-9660 identifiers.
      * Version numbers are usually not used.
+     * bit0= ECMA-119 and Joliet (for historical reasons)
+     * bit1= Joliet
      */
-    unsigned int omit_version_numbers :1;
+    unsigned int omit_version_numbers :2;
 
     /**
      * Allow ISO-9660 directory hierarchy to be deeper than 8 levels.
@@ -81,8 +132,10 @@ struct iso_write_opts {
      * ISO-9660 forces filenames to have a ".", that separates file name from
      * extension. libisofs adds it if original filename doesn't has one. Set
      * this to 1 to prevent this behavior
+     * bit0= ECMA-119
+     * bit1= Joliet
      */
-    unsigned int no_force_dots :1;
+    unsigned int no_force_dots :2;
 
     /**
      * Allow lowercase characters in ISO-9660 filenames. By default, only
@@ -115,6 +168,16 @@ struct iso_write_opts {
      */
     unsigned int rrip_version_1_10 :1;
 
+    /**
+     * Write field PX with file serial number even with RRIP-1.10
+     */
+    unsigned int rrip_1_10_px_ino :1;
+
+    /**
+     * See iso_write_opts_set_hardlinks()
+     */
+    unsigned int hardlinks:1;
+
     /**
      * Write AAIP as extension according to SUSP 1.10 rather than SUSP 1.12.
      * I.e. without announcing it by an ER field and thus without the need
@@ -137,6 +200,26 @@ struct iso_write_opts {
      */
     unsigned int dir_rec_mtime :1;
 
+    /**
+     * Compute MD5 checksum for the whole session and record it as index 0 of
+     * the checksum blocks after the data area of the session. The layout and
+     * position of these blocks will be recorded in xattr "isofs.ca" of the
+     * root node. See see also API call iso_image_get_session_md5().
+     */
+    unsigned int md5_session_checksum :1;
+
+    /**
+     * Compute MD5 checksums for IsoFile objects and write them to blocks
+     * after the data area of the session. The layout and position of these
+     * blocks will be recorded in xattr "isofs.ca" of the root node.
+     * The indice of the MD5 sums will be recorded with the IsoFile directory
+     * entries as xattr "isofs.cx". See also API call iso_file_get_md5().
+     * bit0= compute individual checksums
+     * bit1= pre-compute checksum and compare it with actual one.
+     *       Raise MISHAP if mismatch.
+     */
+    unsigned int md5_file_checksums :2;
+
     /** If files should be sorted based on their weight. */
     unsigned int sort_files :1;
 
@@ -161,6 +244,26 @@ struct iso_write_opts {
     uid_t uid; /** uid to use when replace_uid == 2. */
     gid_t gid; /** gid to use when replace_gid == 2. */
 
+    /**
+     * See API call iso_write_opts_set_old_empty().
+     */ 
+    unsigned int old_empty :1;
+
+    /**
+     * Extra Caution: This option breaks any assumptions about names that
+     *                are supported by ECMA-119 specifications. 
+     * Omit any translation which would make a file name compliant to the
+     * ECMA-119 rules. This includes and exceeds omit_version_numbers,
+     * max_37_char_filenames, no_force_dots bit0, allow_lowercase.
+     * The maximum name length is given by this variable.
+     * There is a length limit of ISO_UNTRANSLATED_NAMES_MAX characters,
+     * because ECMA-119 allows 254 byte in a directory record, some
+     * of them are occupied by ECMA-119, some more are needed for SUSP CE,
+     * and some are fixely occupied by libisofs Rock Ridge code.
+     * The default value 0 disables this feature.
+     */
+    unsigned int untranslated_name_len;
+
     /**
      * 0 to use IsoNode timestamps, 1 to use recording time, 2 to use
      * values from timestamp field. This has only meaning if RR extensions
@@ -247,6 +350,67 @@ struct iso_write_opts {
      */
     uint32_t data_start_lba;
 
+    /**
+     * If not empty: A text holding parameters "name" and "timestamp" for
+     * a scdbackup stream checksum tag. See scdbackup/README appendix VERIFY.
+     * It makes sense only for single session images which start at LBA 0.
+     * Such a tag may be part of a libisofs checksum tag block after the
+     * session tag line. It then covers the whole session up to its own start
+     * position.
+     */
+    char scdbackup_tag_parm[100];
+
+    /* If not NULL: A pointer to an application provided array with
+       at least 512 characters. The effectively written scdbackup tag
+       will be copied to this memory location.
+     */
+    char *scdbackup_tag_written;
+
+    /*
+     * See ecma119_image : System Area related information
+     */
+    char *system_area_data;
+    int system_area_options;
+
+    /* User settable PVD time stamps */
+    time_t vol_creation_time;
+    time_t vol_modification_time;
+    time_t vol_expiration_time;
+    time_t vol_effective_time;
+    /* To eventually override vol_creation_time and vol_modification_time
+     * by unconverted string with timezone 0
+     */
+    char vol_uuid[17];
+
+    /* The number of unclaimed 2K blocks before start of partition 1 as of
+       the MBR in system area.
+       Must be 0 or >= 16. (Actually >= number of voldescr + checksum tag)
+    */
+    uint32_t partition_offset;
+    /* Partition table parameter: 1 to 63, 0= disabled/default */
+    int partition_secs_per_head;
+    /* 1 to 255, 0= disabled/default */
+    int partition_heads_per_cyl;
+
+#ifdef Libisofs_with_libjtE
+    /* Parameters and state of Jigdo Template Export environment.
+    */
+    struct libjte_env *libjte_handle;
+#endif /* Libisofs_with_libjtE */
+
+    /* A trailing padding of zero bytes which belongs to the image
+    */
+    uint32_t tail_blocks;
+
+    /* Eventual disk file paths of prepared images which shall be appended
+       after the ISO image and described by partiton table entries in a MBR
+    */
+    char *appended_partitions[ISO_MAX_PARTITIONS]; 
+    uint8_t appended_part_types[ISO_MAX_PARTITIONS];
+
+    /* Eventual name of the non-ISO aspect of the image. E.g. SUN ASCII label.
+     */
+    char ascii_disc_label[ISO_DISC_LABEL_SIZE];
 };
 
 typedef struct ecma119_image Ecma119Image;
@@ -258,9 +422,13 @@ typedef struct Iso_Image_Writer IsoImageWriter;
 
 struct ecma119_image
 {
+    int refcount;
+
     IsoImage *image;
     Ecma119Node *root;
 
+    int will_cancel :1;
+
     unsigned int iso_level :2;
 
     /* extensions */
@@ -269,17 +437,20 @@ struct ecma119_image
     unsigned int eltorito :1;
     unsigned int iso1999 :1;
 
-    unsigned int aaip :1;  /* whether to write eventual ACLs and EAs */
+    unsigned int hardlinks:1; /* see iso_write_opts_set_hardlinks() */
+
+    unsigned int aaip :1;     /* see iso_write_opts_set_aaip() */
 
     /* allways write timestamps in GMT */
     unsigned int always_gmt :1;
 
     /* relaxed constraints */
-    unsigned int omit_version_numbers :1;
+    unsigned int allow_dir_id_ext :1;
+    unsigned int omit_version_numbers :2;
     unsigned int allow_deep_paths :1;
     unsigned int allow_longer_paths :1;
     unsigned int max_37_char_filenames :1;
-    unsigned int no_force_dots :1;
+    unsigned int no_force_dots :2;
     unsigned int allow_lowercase :1;
     unsigned int allow_full_ascii :1;
 
@@ -291,12 +462,18 @@ struct ecma119_image
     /** Write old fashioned RRIP-1.10 rather than RRIP-1.12 */
     unsigned int rrip_version_1_10 :1;
 
+    /** Write field PX with file serial number even with RRIP-1.10 */
+    unsigned int rrip_1_10_px_ino :1;
+
     /* Write AAIP as extension according to SUSP 1.10 rather than SUSP 1.12. */
     unsigned int aaip_susp_1_10 :1;
 
     /* Store in ECMA-119 timestamp mtime of source */
     unsigned int dir_rec_mtime :1;
 
+    unsigned int md5_session_checksum :1;
+    unsigned int md5_file_checksums :2;
+
     /*
      * Mode replace. If one of these flags is set, the correspodent values are
      * replaced with values below.
@@ -313,17 +490,14 @@ struct ecma119_image
     mode_t dir_mode;
     time_t timestamp;
 
+    unsigned int old_empty :1;
+    unsigned int untranslated_name_len;
+
     /**
      *  if sort files or not. Sorting is based of the weight of each file
      */
     int sort_files;
 
-    /**
-     * In the CD, each file must have an unique inode number. So each
-     * time we add a new file, this is incremented.
-     */
-    ino_t ino;
-
     char *input_charset;
     char *output_charset;
 
@@ -335,8 +509,9 @@ struct ecma119_image
     off_t total_size;
     uint32_t vol_space_size;
 
-    /* Bytes already written, just for progress notification */
+    /* Bytes already written to image output */
     off_t bytes_written;
+    /* just for progress notification */
     int percent_written;
 
     /*
@@ -345,6 +520,17 @@ struct ecma119_image
      */
     uint32_t curblock;
 
+    /*
+     * The address to be used for the content pointer of empty data files.
+     */
+    uint32_t empty_file_block;
+
+    /*
+     * The calculated block address after ECMA-119 tree and eventual
+     * tree checksum tag.
+     */
+    uint32_t tree_end_block;
+
     /*
      * number of dirs in ECMA-119 tree, computed together with dir position,
      * and needed for path table computation in a efficient way
@@ -377,7 +563,33 @@ struct ecma119_image
      */
     struct el_torito_boot_catalog *catalog;
     IsoFileSrc *cat; /**< location of the boot catalog in the new image */
-    IsoFileSrc *bootimg; /**< location of the boot image in the new image */
+
+    int num_bootsrc;
+    IsoFileSrc **bootsrc; /* location of the boot images in the new image */
+
+    /*
+     * System Area related information
+     */
+    /* Content of an embedded boot image. Valid if not NULL.
+     * In that case it must point to a memory buffer at least 32 kB.
+     */
+    char *system_area_data;
+    /*
+     * bit0= Only with PC-BIOS DOS MBR
+     *       Make bytes 446 - 512 of the system area a partition
+     *       table which reserves partition 1 from byte 63*512 to the
+     *       end of the ISO image. Assumed are 63 secs/hed, 255 head/cyl.
+     *       (GRUB protective msdos label.)
+     *       This works with and without system_area_data.
+     * bit1= Only with PC-BIOS DOS MBR
+     *       Apply isohybrid MBR patching to the system area.
+     *       This works only with system_area_data plus ISOLINUX boot image
+     *       and only if not bit0 is set.
+     * bit2-7= System area type
+     *       0= DOS MBR
+     *       1= MIPS Big Endian Volume Header
+     */
+    int system_area_options;
 
     /*
      * Number of pad blocks that we need to write. Padding blocks are blocks
@@ -392,7 +604,7 @@ struct ecma119_image
      * file data is written in the first 64 KiB, that are the bytes we usually
      * overwrite.
      */
-    uint32_t pad_blocks;
+    uint32_t mspad_blocks;
 
     size_t nwriters;
     IsoImageWriter **writers;
@@ -400,12 +612,94 @@ struct ecma119_image
     /* tree of files sources */
     IsoRBTree *files;
 
+    unsigned int checksum_idx_counter;
+    void *checksum_ctx;
+    off_t checksum_counter;
+    uint32_t checksum_rlsb_tag_pos;
+    uint32_t checksum_sb_tag_pos;
+    uint32_t checksum_tree_tag_pos;
+    uint32_t checksum_tag_pos;
+    char image_md5[16];
+    char *checksum_buffer;
+    uint32_t checksum_array_pos;
+    uint32_t checksum_range_start;
+    uint32_t checksum_range_size;
+
+    char *opts_overwrite; /* Points to IsoWriteOpts->overwrite.
+                             Use only underneath ecma119_image_new()
+                             and if not NULL*/
+
+    /* ??? Is there a reason why we copy lots of items from IsoWriteOpts
+           rather than taking ownership of the IsoWriteOpts object which
+           is submitted with ecma119_image_new() ?
+     */
+
+    char scdbackup_tag_parm[100];
+    char *scdbackup_tag_written;
+
     /* Buffer for communication between burn_source and writer thread */
     IsoRingBuffer *buffer;
 
     /* writer thread descriptor */
     pthread_t wthread;
+    int wthread_is_running;
     pthread_attr_t th_attr;
+
+    /* User settable PVD time stamps */
+    time_t vol_creation_time;
+    time_t vol_modification_time;
+    time_t vol_expiration_time;
+    time_t vol_effective_time;
+    /* To eventually override vol_creation_time and vol_modification_time
+     * by unconverted string with timezone 0
+     */
+    char vol_uuid[17];
+
+    /* The number of unclaimed 2K blocks before
+       start of partition 1 as of the MBR in system area. */
+    uint32_t partition_offset;
+    /* Partition table parameter: 1 to 63, 0= disabled/default */
+    int partition_secs_per_head;
+    /* 1 to 255, 0= disabled/default */
+    int partition_heads_per_cyl;
+
+    /* The currently applicable LBA offset. To be subtracted from any LBA
+     *  that is mentioned in volume descriptors, trees, path tables,
+     *  Either 0 or .partition_offset
+     */
+    uint32_t eff_partition_offset;
+
+    /* The second ECMA-119 directory tree and path tables */
+    Ecma119Node *partition_root;
+    uint32_t partition_l_table_pos;
+    uint32_t partition_m_table_pos;
+
+    /* The second Joliet directory tree and path tables */
+    JolietNode *j_part_root;
+    uint32_t j_part_l_path_table_pos;
+    uint32_t j_part_m_path_table_pos;
+
+#ifdef Libisofs_with_libjtE
+    struct libjte_env *libjte_handle;
+#endif /* Libisofs_with_libjtE */
+
+    uint32_t tail_blocks;
+
+    /* Memorized ELF parameters from MIPS Little Endian boot file */
+    uint32_t mipsel_e_entry;
+    uint32_t mipsel_p_offset;
+    uint32_t mipsel_p_vaddr;
+    uint32_t mipsel_p_filesz;
+
+    char *appended_partitions[ISO_MAX_PARTITIONS]; 
+    uint8_t appended_part_types[ISO_MAX_PARTITIONS];
+    /* Counted in blocks of 2048 */
+    uint32_t appended_part_prepad[ISO_MAX_PARTITIONS];
+    uint32_t appended_part_start[ISO_MAX_PARTITIONS];
+    uint32_t appended_part_size[ISO_MAX_PARTITIONS];
+
+    char ascii_disc_label[ISO_DISC_LABEL_SIZE];
+
 };
 
 #define BP(a,b) [(b) - (a) + 1]

libisofs/ecma119_tree.c

@@ -1,11 +1,20 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
+/* Must be before ecma119.h because of eventual Libisofs_with_rrip_rR */
+#include "libisofs.h"
+
 #include "ecma119_tree.h"
 #include "ecma119.h"
 #include "node.h"
@@ -23,7 +32,7 @@
 static
 int get_iso_name(Ecma119Image *img, IsoNode *iso, char **name)
 {
-    int ret, relaxed;
+    int ret, relaxed, free_ascii_name= 0, force_dots = 0, max_len;
     char *ascii_name;
     char *isoname= NULL;
 
@@ -32,9 +41,16 @@ int get_iso_name(Ecma119Image *img, IsoNode *iso, char **name)
         return ISO_SUCCESS;
     }
 
-    ret = str2ascii(img->input_charset, iso->name, &ascii_name);
+    if (img->untranslated_name_len > 0) {
+        ascii_name = iso->name;
+        ret = 1;
+    } else {
+        ret = str2ascii(img->input_charset, iso->name, &ascii_name);
+        free_ascii_name = 1;
+    }
     if (ret < 0) {
-        iso_msg_submit(img->image->id, ret, 0, "Can't convert %s", iso->name);
+        iso_msg_submit(img->image->id, ret, 0,
+                       "Cannot convert name '%s' to ASCII", iso->name);
         return ret;
     }
 
@@ -43,8 +59,18 @@ int get_iso_name(Ecma119Image *img, IsoNode *iso, char **name)
     } else {
         relaxed = (int)img->allow_lowercase;
     }
-    if (iso->type == LIBISO_DIR) {
+    if (iso->type == LIBISO_DIR && !(img->allow_dir_id_ext)) {
-        if (img->max_37_char_filenames) {
+        if (img->untranslated_name_len > 0) {
+            if (strlen(ascii_name) > img->untranslated_name_len) {
+needs_transl:;
+                iso_msg_submit(img->image->id, ISO_NAME_NEEDS_TRANSL, 0,
+              "File name too long (%d > %d) for untranslated recording:  '%s'",
+                               strlen(ascii_name), img->untranslated_name_len,
+                               ascii_name);
+                return ISO_NAME_NEEDS_TRANSL;
+            }
+            isoname = strdup(ascii_name);
+        } else if (img->max_37_char_filenames) {
             isoname = iso_r_dirid(ascii_name, 37, relaxed);
         } else if (img->iso_level == 1) {
             if (relaxed) {
@@ -54,32 +80,40 @@ int get_iso_name(Ecma119Image *img, IsoNode *iso, char **name)
             }
         } else {
             if (relaxed) {
-                isoname = iso_r_dirid(ascii_name, 8, relaxed);
+                isoname = iso_r_dirid(ascii_name, 31, relaxed);
             } else {
                 isoname = iso_2_dirid(ascii_name);
             }
         }
     } else {
-        if (img->max_37_char_filenames) {
+        force_dots = !((img->no_force_dots & 1) || iso->type == LIBISO_DIR);
-            isoname = iso_r_fileid(ascii_name, 36, relaxed,
+        if (img->untranslated_name_len > 0) {
-                                   img->no_force_dots ? 0 : 1);
+            if (strlen(ascii_name) > img->untranslated_name_len)
+                goto needs_transl;
+            isoname = strdup(ascii_name);
+        } else if (img->max_37_char_filenames) {
+            isoname = iso_r_fileid(ascii_name, 36, relaxed, force_dots);
         } else if (img->iso_level == 1) {
-            if (relaxed) {
+            if (relaxed || !force_dots) {
-                isoname = iso_r_fileid(ascii_name, 11, relaxed,
+                if (strchr(ascii_name, '.') == NULL)
-                                       img->no_force_dots ? 0 : 1);
+                    max_len = 8;
+                else
+                    max_len = 11;
+                isoname = iso_r_fileid(ascii_name, max_len, relaxed,
+                                       force_dots);
             } else {
                 isoname = iso_1_fileid(ascii_name);
             }
         } else {
-            if (relaxed) {
+            if (relaxed || !force_dots) {
-                isoname = iso_r_fileid(ascii_name, 30, relaxed,
+                isoname = iso_r_fileid(ascii_name, 30, relaxed, force_dots);
-                                       img->no_force_dots ? 0 : 1);
             } else {
                 isoname = iso_2_fileid(ascii_name);
             }
         }
     }
-    free(ascii_name);
+    if (free_ascii_name)
+        free(ascii_name);
     if (isoname != NULL) {
         *name = isoname;
         return ISO_SUCCESS;
@@ -102,14 +136,9 @@ int create_ecma119_node(Ecma119Image *img, IsoNode *iso, Ecma119Node **node)
         return ISO_OUT_OF_MEM;
     }
 
-    /* take a ref to the IsoNode */
     ecma->node = iso;
     iso_node_ref(iso);
-
-    /* TODO #00009 : add true support for harlinks and inode numbers */
     ecma->nlink = 1;
-    ecma->ino = ++img->ino;
-
     *node = ecma;
     return ISO_SUCCESS;
 }
@@ -149,15 +178,11 @@ int create_dir(Ecma119Image *img, IsoDir *iso, Ecma119Node **node)
     return ISO_SUCCESS;
 }
 
-/**
+
- * Create a new ECMA-119 node representing a regular file from a iso file
- * node.
- */
 static
-int create_file(Ecma119Image *img, IsoFile *iso, Ecma119Node **node)
+int create_file_src(Ecma119Image *img, IsoFile *iso, IsoFileSrc **src)
 {
     int ret;
-    IsoFileSrc *src;
     off_t size;
 
     size = iso_stream_get_size(iso->stream);
@@ -169,8 +194,25 @@ int create_file(Ecma119Image *img, IsoFile *iso, Ecma119Node **node)
         free(ipath);
         return ret;
     }
+    ret = iso_file_src_create(img, iso, src);
+    if (ret < 0) {
+        return ret;
+    }
+    return 0;
+}
 
-    ret = iso_file_src_create(img, iso, &src);
+
+/**
+ * Create a new ECMA-119 node representing a regular file from a iso file
+ * node.
+ */
+static
+int create_file(Ecma119Image *img, IsoFile *iso, Ecma119Node **node)
+{
+    int ret;
+    IsoFileSrc *src;
+
+    ret = create_file_src(img, iso, &src);
     if (ret < 0) {
         return ret;
     }
@@ -270,58 +312,78 @@ void ecma119_node_free(Ecma119Node *node)
 }
 
 /**
- *
+ * @param flag
+ *      bit0= iso is in a hidden directory. Thus hide it.
  * @return
  *      1 success, 0 node ignored,  < 0 error
  *
  */
 static
 int create_tree(Ecma119Image *image, IsoNode *iso, Ecma119Node **tree,
-                int depth, int pathlen)
+                int depth, int pathlen, int flag)
 {
-    int ret;
+    int ret, hidden;
-    Ecma119Node *node;
+    Ecma119Node *node = NULL;
     int max_path;
-    char *iso_name= NULL;
+    char *iso_name= NULL, *ipath = NULL;
+    IsoFileSrc *src = NULL;
 
     if (image == NULL || iso == NULL || tree == NULL) {
         return ISO_NULL_POINTER;
     }
+    *tree = NULL;
 
+    hidden = flag & 1;
     if (iso->hidden & LIBISO_HIDE_ON_RR) {
-        /* file will be ignored */
+        hidden = 1;
-        return 0;
+        if (!((iso->hidden & LIBISO_HIDE_BUT_WRITE) ||
+              iso->type == LIBISO_BOOT)) {
+            return 0; /* file will be ignored */
+        }
     }
-    ret = get_iso_name(image, iso, &iso_name);
+
-    if (ret < 0) {
+    if (hidden) {
-        return ret;
+        max_path= pathlen;
-    }
+    } else {
-    max_path = pathlen + 1 + (iso_name ? strlen(iso_name) : 0);
+        ret = get_iso_name(image, iso, &iso_name);
-    if (!image->rockridge) {
+        if (ret < 0) {
-        if ((iso->type == LIBISO_DIR && depth > 8) && !image->allow_deep_paths) {
+            iso_name = NULL; /* invalid, do not free */
-            char *ipath = iso_tree_get_node_path(iso);
+            goto ex;
-            return iso_msg_submit(image->image->id, ISO_FILE_IMGPATH_WRONG, 0,
+        }
-                         "File \"%s\" can't be added, because directory depth "
+        max_path = pathlen + 1 + (iso_name ? strlen(iso_name) : 0);
-                         "is greater than 8.", ipath);
+        if (!image->rockridge) {
-            free(iso_name);
+            if ((iso->type == LIBISO_DIR && depth > 8) &&
-            free(ipath);
+                !image->allow_deep_paths) {
-            return ret;
+	        ipath = iso_tree_get_node_path(iso);
-        } else if (max_path > 255 && !image->allow_longer_paths) {
+                ret = iso_msg_submit(image->image->id, ISO_FILE_IMGPATH_WRONG,
-            char *ipath = iso_tree_get_node_path(iso);
+                                     0, "File \"%s\" can't be added, "
-            ret = iso_msg_submit(image->image->id, ISO_FILE_IMGPATH_WRONG, 0,
+                                     "because directory depth "
-                         "File \"%s\" can't be added, because path length "
+                                     "is greater than 8.", ipath);
-                         "is greater than 255 characters", ipath);
+                goto ex;
-            free(iso_name);
+            } else if (max_path > 255 && !image->allow_longer_paths) {
-            free(ipath);
+                ipath = iso_tree_get_node_path(iso);
-            return ret;
+                ret = iso_msg_submit(image->image->id, ISO_FILE_IMGPATH_WRONG,
+                                     0, "File \"%s\" can't be added, "
+                                     "because path length "
+                                     "is greater than 255 characters", ipath);
+                goto ex;
+            }
         }
     }
 
     switch (iso->type) {
     case LIBISO_FILE:
-        ret = create_file(image, (IsoFile*)iso, &node);
+        if (hidden) {
+            ret = create_file_src(image, (IsoFile *) iso, &src);
+        } else {
+            ret = create_file(image, (IsoFile*)iso, &node);
+        }
         break;
     case LIBISO_SYMLINK:
+        if (hidden) {
+            ret = 0; /* Hidden means non-existing */
+            goto ex;
+        }
         if (image->rockridge) {
             ret = create_symlink(image, (IsoSymlink*)iso, &node);
         } else {
@@ -334,6 +396,10 @@ int create_tree(Ecma119Image *image, IsoNode *iso, Ecma119Node **tree,
         }
         break;
     case LIBISO_SPECIAL:
+        if (hidden) {
+            ret = 0; /* Hidden means non-existing */
+            goto ex;
+        }
         if (image->rockridge) {
             ret = create_special(image, (IsoSpecial*)iso, &node);
         } else {
@@ -347,7 +413,11 @@ int create_tree(Ecma119Image *image, IsoNode *iso, Ecma119Node **tree,
         break;
     case LIBISO_BOOT:
         if (image->eltorito) {
-            ret = create_boot_cat(image, (IsoBoot*)iso, &node);
+            if (hidden) {
+                ret = el_torito_catalog_file_src_create(image, &src);
+            } else {
+                ret = create_boot_cat(image, (IsoBoot*)iso, &node);
+            }
         } else {
             /* log and ignore */
             ret = iso_msg_submit(image->image->id, ISO_FILE_IGNORED, 0,
@@ -358,21 +428,25 @@ int create_tree(Ecma119Image *image, IsoNode *iso, Ecma119Node **tree,
         {
             IsoNode *pos;
             IsoDir *dir = (IsoDir*)iso;
-            ret = create_dir(image, dir, &node);
+
-            if (ret < 0) {
+            if (!hidden) {
-                return ret;
+                ret = create_dir(image, dir, &node);
+                if (ret < 0) {
+                    goto ex;
+                }
             }
+            ret = ISO_SUCCESS;
             pos = dir->children;
             while (pos) {
                 int cret;
                 Ecma119Node *child;
-                cret = create_tree(image, pos, &child, depth + 1, max_path);
+                cret = create_tree(image, pos, &child, depth + 1, max_path,
+                                   !!hidden);
                 if (cret < 0) {
                     /* error */
-                    ecma119_node_free(node);
                     ret = cret;
                     break;
-                } else if (cret == ISO_SUCCESS) {
+                } else if (cret == ISO_SUCCESS && !hidden) {
                     /* add child to this node */
                     int nchildren = node->info.dir->nchildren++;
                     node->info.dir->children[nchildren] = child;
@@ -384,15 +458,30 @@ int create_tree(Ecma119Image *image, IsoNode *iso, Ecma119Node **tree,
         break;
     default:
         /* should never happen */
-        return ISO_ASSERT_FAILURE;
+        ret = ISO_ASSERT_FAILURE;
+        goto ex;
     }
     if (ret <= 0) {
-        free(iso_name);
+        goto ex;
-        return ret;
     }
-    node->iso_name = iso_name;
+    if (!hidden) {
-    *tree = node;
+        node->iso_name = iso_name;
-    return ISO_SUCCESS;
+        iso_name = NULL; /* now owned by node, do not free */
+        *tree = node;
+        node = NULL;     /* now owned by caller, do not free */
+    }
+    ret = ISO_SUCCESS;
+ex:
+    if (iso_name != NULL)
+        free(iso_name);
+    if (ipath != NULL)
+        free(ipath);
+    if (node != NULL)
+        ecma119_node_free(node);
+    if (hidden && ret == ISO_SUCCESS)
+        ret = 0;
+    /* The sources of hidden files are now owned by the rb-tree */
+    return ret;
 }
 
 /**
@@ -474,6 +563,18 @@ int mangle_single_dir(Ecma119Image *img, Ecma119Node *dir, int max_file_len,
             continue;
         }
 
+        if (img->untranslated_name_len) {
+            /* This should not happen because no two IsoNode names should be
+               identical and only unaltered IsoNode names should be seen here.
+               Thus the Ema119Node names should be unique.
+            */
+            iso_msg_submit(img->image->id, ISO_NAME_NEEDS_TRANSL, 0,
+                           "ECMA-119 file name collision: '%s'",
+                           children[i]->iso_name);
+            ret = ISO_NAME_NEEDS_TRANSL;
+            goto mangle_cleanup;
+        }
+
         /*
          * A max of 7 characters is good enought, it allows handling up to
          * 9,999,999 files with same name. We can increment this to
@@ -490,10 +591,11 @@ int mangle_single_dir(Ecma119Image *img, Ecma119Node *dir, int max_file_len,
 
             /* compute name and extension */
             dot = strrchr(full_name, '.');
-            if (dot != NULL && children[i]->type != ECMA119_DIR) {
+            if (dot != NULL &&
+                (children[i]->type != ECMA119_DIR || img->allow_dir_id_ext)) {
 
                 /*
-                 * File (not dir) with extension
+                 * File (normally not dir) with extension
                  * Note that we don't need to check for placeholders, as
                  * tree reparent happens later, so no placeholders can be
                  * here at this time.
@@ -533,10 +635,10 @@ int mangle_single_dir(Ecma119Image *img, Ecma119Node *dir, int max_file_len,
                     name[max] = '\0';
                 }
             } else {
-                /* Directory, or file without extension */
+                /* Directory (normally), or file without extension */
                 if (children[i]->type == ECMA119_DIR) {
                     max = max_dir_len - digits;
-                    dot = NULL; /* dots have no meaning in dirs */
+                    dot = NULL; /* dots (normally) have no meaning in dirs */
                 } else {
                     max = max_file_len - digits;
                 }
@@ -651,8 +753,11 @@ static
 int mangle_tree(Ecma119Image *img, int recurse)
 {
     int max_file, max_dir;
+    Ecma119Node *root;
 
-    if (img->max_37_char_filenames) {
+    if (img->untranslated_name_len > 0) {
+        max_file = max_dir = img->untranslated_name_len;
+    } else if (img->max_37_char_filenames) {
         max_file = max_dir = 37;
     } else if (img->iso_level == 1) {
         max_file = 12; /* 8 + 3 + 1 */
@@ -660,10 +765,15 @@ int mangle_tree(Ecma119Image *img, int recurse)
     } else {
         max_file = max_dir = 31;
     }
-    if (recurse) {
+    if (img->eff_partition_offset > 0) {
-        return mangle_dir(img, img->root, max_file, max_dir);
+        root = img->partition_root;
     } else {
-        return mangle_single_dir(img, img->root, max_file, max_dir);
+        root = img->root;
+    }
+    if (recurse) {
+        return mangle_dir(img, root, max_file, max_dir);
+    } else {
+        return mangle_single_dir(img, root, max_file, max_dir);
     }
 }
 
@@ -781,11 +891,17 @@ int reorder_tree(Ecma119Image *img, Ecma119Node *dir, int level, int pathlen)
 {
     int ret;
     size_t max_path;
+    Ecma119Node *root;
 
     max_path = pathlen + 1 + max_child_name_len(dir);
 
     if (level > 8 || max_path > 255) {
-        ret = reparent(dir, img->root);
+        if (img->eff_partition_offset > 0) {
+            root = img->partition_root;
+        } else {
+            root = img->root;
+        }
+        ret = reparent(dir, root);
         if (ret < 0) {
             return ret;
         }
@@ -811,12 +927,181 @@ int reorder_tree(Ecma119Image *img, Ecma119Node *dir, int level, int pathlen)
     return ISO_SUCCESS;
 }
 
+/*
+ * @param flag
+ *     bit0= recursion
+ *     bit1= count nodes rather than fill them into *nodes
+ * @return
+ *     <0 error
+ *     bit0= saw ino == 0
+ *     bit1= saw ino != 0
+ */
+static
+int make_node_array(Ecma119Image *img, Ecma119Node *dir,
+                    Ecma119Node **nodes, size_t nodes_size, size_t *node_count,
+                    int flag)
+{
+    int ret, result = 0;
+    size_t i;
+    Ecma119Node *child;
+
+    if (!(flag & 1)) {
+        *node_count = 0;
+        if (!(flag & 2)) {
+            /* Register the tree root node */
+            if (*node_count >= nodes_size) {
+                iso_msg_submit(img->image->id, ISO_ASSERT_FAILURE, 0,
+                         "Programming error: Overflow of hardlink sort array");
+                return ISO_ASSERT_FAILURE;
+            }
+            nodes[*node_count] = dir;
+        }
+        result|= (dir->ino == 0 ? 1 : 2);
+        (*node_count)++;
+    }
+        
+    for (i = 0; i < dir->info.dir->nchildren; i++) {
+        child = dir->info.dir->children[i];
+        if (!(flag & 2)) {
+            if (*node_count >= nodes_size) {
+                iso_msg_submit(img->image->id, ISO_ASSERT_FAILURE, 0,
+                         "Programming error: Overflow of hardlink sort array");
+                return ISO_ASSERT_FAILURE;
+            }
+            nodes[*node_count] = child;
+        }
+        result|= (child->ino == 0 ? 1 : 2);
+        (*node_count)++;
+
+        if (child->type == ECMA119_DIR) {
+            ret = make_node_array(img, child,
+                                  nodes, nodes_size, node_count, flag | 1);
+            if (ret < 0)
+                return ret;
+        }
+    }
+    return result;
+}
+
+/*
+ * @param flag
+ *     bit0= compare stat properties and attributes 
+ *     bit1= treat all nodes with image ino == 0 as unique
+ */
+static
+int ecma119_node_cmp_flag(const void *v1, const void *v2, int flag)
+{
+    int ret;
+    Ecma119Node *n1, *n2;
+
+    n1 = *((Ecma119Node **) v1);
+    n2 = *((Ecma119Node **) v2);
+    if (n1 == n2)
+        return 0;
+
+    ret = iso_node_cmp_flag(n1->node, n2->node, flag & (1 | 2));
+    return ret;
+}
+
+static 
+int ecma119_node_cmp_hard(const void *v1, const void *v2)
+{
+    return ecma119_node_cmp_flag(v1, v2, 1);
+}   
+
+static 
+int ecma119_node_cmp_nohard(const void *v1, const void *v2)
+{
+    return ecma119_node_cmp_flag(v1, v2, 1 | 2);
+}   
+
+static
+int family_set_ino(Ecma119Image *img, Ecma119Node **nodes, size_t family_start,
+                   size_t next_family, ino_t img_ino, ino_t prev_ino, int flag)
+{
+    size_t i;
+
+    if (img_ino != 0) {
+        /* Check whether this is the same img_ino as in the previous
+           family (e.g. by property divergence of imported hardlink).
+        */
+        if (img_ino == prev_ino)
+            img_ino = 0;
+    }
+    if (img_ino == 0) {
+        img_ino = img_give_ino_number(img->image, 0);
+    }
+    for (i = family_start; i < next_family; i++) {
+        nodes[i]->ino = img_ino;
+        nodes[i]->nlink = next_family - family_start;
+    }
+    return 1;
+}
+
+static
+int match_hardlinks(Ecma119Image *img, Ecma119Node *dir, int flag)
+{
+    int ret;
+    size_t nodes_size = 0, node_count = 0, i, family_start;
+    Ecma119Node **nodes = NULL;
+    unsigned int fs_id;
+    dev_t dev_id;
+    ino_t img_ino = 0, prev_ino = 0;
+
+    ret = make_node_array(img, dir, nodes, nodes_size, &node_count, 2);
+    if (ret < 0)
+        return ret;
+    nodes_size = node_count;
+    nodes = (Ecma119Node **) calloc(sizeof(Ecma119Node *), nodes_size);
+    if (nodes == NULL)
+        return ISO_OUT_OF_MEM;
+    ret = make_node_array(img, dir, nodes, nodes_size, &node_count, 0);
+    if (ret < 0)
+        goto ex;
+
+    /* Sort according to id tuples, IsoFileSrc identity, properties, xattr. */
+    if (img->hardlinks)
+        qsort(nodes, node_count, sizeof(Ecma119Node *), ecma119_node_cmp_hard);
+    else
+        qsort(nodes, node_count, sizeof(Ecma119Node *),
+              ecma119_node_cmp_nohard);
+
+    /* Hand out image inode numbers to all Ecma119Node.ino == 0 .
+       Same sorting rank gets same inode number.
+       Split those image inode number families where the sort criterion
+       differs.
+    */
+    iso_node_get_id(nodes[0]->node, &fs_id, &dev_id, &img_ino, 1);
+    family_start = 0;
+    for (i = 1; i < node_count; i++) {
+        if (nodes[i]->type != ECMA119_DIR &&
+            ecma119_node_cmp_hard(nodes + (i - 1), nodes + i) == 0) {
+            /* Still in same ino family */
+            if (img_ino == 0) { /* Just in case any member knows its img_ino */
+                iso_node_get_id(nodes[0]->node, &fs_id, &dev_id, &img_ino, 1);
+            }
+    continue;
+        }
+        family_set_ino(img, nodes, family_start, i, img_ino, prev_ino, 0);
+        prev_ino = img_ino;
+        iso_node_get_id(nodes[i]->node, &fs_id, &dev_id, &img_ino, 1);
+        family_start = i;
+    }
+    family_set_ino(img, nodes, family_start, i, img_ino, prev_ino, 0);
+
+    ret = ISO_SUCCESS;
+ex:;
+    if (nodes != NULL)
+        free((char *) nodes);
+    return ret;
+}
+
 int ecma119_tree_create(Ecma119Image *img)
 {
     int ret;
     Ecma119Node *root;
 
-    ret = create_tree(img, (IsoNode*)img->image->root, &root, 1, 0);
+    ret = create_tree(img, (IsoNode*)img->image->root, &root, 1, 0, 0);
     if (ret <= 0) {
         if (ret == 0) {
             /* unexpected error, root ignored!! This can't happen */
@@ -824,7 +1109,17 @@ int ecma119_tree_create(Ecma119Image *img)
         }
         return ret;
     }
-    img->root = root;
+    if (img->eff_partition_offset > 0) {
+        img->partition_root = root;
+    } else {
+        img->root = root;
+    }
+
+    iso_msg_debug(img->image->id, "Matching hardlinks...");
+    ret = match_hardlinks(img, root, 0);
+    if (ret < 0) {
+        return ret;
+    }
 
     iso_msg_debug(img->image->id, "Sorting the low level tree...");
     sort_tree(root);
@@ -838,7 +1133,7 @@ int ecma119_tree_create(Ecma119Image *img)
     if (img->rockridge && !img->allow_deep_paths) {
 
         /* reorder the tree, acording to RRIP, 4.1.5 */
-        ret = reorder_tree(img, img->root, 1, 0);
+        ret = reorder_tree(img, root, 1, 0);
         if (ret < 0) {
             return ret;
         }
@@ -856,3 +1151,35 @@ int ecma119_tree_create(Ecma119Image *img)
 
     return ISO_SUCCESS;
 }
+
+/**
+ * Search the tree for a certain IsoNode and return its owning Ecma119Node
+ * or NULL.
+ */
+static
+Ecma119Node *search_iso_node(Ecma119Node *root, IsoNode *node)
+{
+    size_t i;
+    Ecma119Node *res = NULL;
+
+    if (root->node == node)
+        return root;
+    for (i = 0; i < root->info.dir->nchildren && res == NULL; i++) {
+        if (root->info.dir->children[i]->type == ECMA119_DIR)
+            res = search_iso_node(root->info.dir->children[i], node);
+        else if (root->info.dir->children[i]->node == node)
+            res = root->info.dir->children[i];
+    }
+    return res;
+}
+
+
+Ecma119Node *ecma119_search_iso_node(Ecma119Image *img, IsoNode *node)
+{
+    Ecma119Node *res = NULL;
+
+    if (img->root != NULL)
+        res = search_iso_node(img->root, node);
+    return res;
+}
+

libisofs/ecma119_tree.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #ifndef LIBISO_ECMA119_TREE_H_
@@ -62,8 +63,10 @@ struct ecma119_node
 
     IsoNode *node; /*< reference to the iso node */
 
-    /* TODO #00009 : add true support for harlinks and inode numbers */
+    /* >>> ts A90501 : Shouldn't this be uint32_t
+                       as this is what PX will take ? */
     ino_t ino;
+
     nlink_t nlink;
 
     /**< file, symlink, special, directory or placeholder */
@@ -87,4 +90,11 @@ int ecma119_tree_create(Ecma119Image *img);
  */
 void ecma119_node_free(Ecma119Node *node);
 
+/**
+ * Search the tree for a certain IsoNode and return its owning Ecma119Node
+ * or NULL.
+ */
+Ecma119Node *ecma119_search_iso_node(Ecma119Image *img, IsoNode *node);
+
+
 #endif /*LIBISO_ECMA119_TREE_H_*/

libisofs/eltorito.c

@@ -1,13 +1,19 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2010 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
+#include "libisofs.h"
 #include "eltorito.h"
-#include "stream.h"
 #include "fsource.h"
 #include "filesrc.h"
 #include "image.h"
@@ -16,9 +22,10 @@
 
 #include <stdlib.h>
 #include <string.h>
+#include <stdio.h>
 
 /**
- * This table should be written with accuracy values at offset
+ * This table should be written with the actual values at offset
  * 8 of boot image, when used ISOLINUX boot loader
  */
 struct boot_info_table {
@@ -54,6 +61,19 @@ struct hard_disc_mbr {
     uint8_t sign2;
 };
 
+/* API */
+int el_torito_set_boot_platform_id(ElToritoBootImage *bootimg, uint8_t id)
+{
+    bootimg->platform_id = id;
+    return 1;
+}
+
+/* API */
+int el_torito_get_boot_platform_id(ElToritoBootImage *bootimg)
+{
+    return bootimg->platform_id;
+}
+
 /**
  * Sets the load segment for the initial boot image. This is only for
  * no emulation boot images, and is a NOP for other image types.
@@ -65,6 +85,14 @@ void el_torito_set_load_seg(ElToritoBootImage *bootimg, short segment)
     bootimg->load_seg = segment;
 }
 
+/* API */
+int el_torito_get_load_seg(ElToritoBootImage *bootimg)
+{
+   if (bootimg->load_seg < 0)
+       return 0xffff - bootimg->load_seg;
+   return bootimg->load_seg;
+}
+ 
 /**
  * Sets the number of sectors (512b) to be load at load segment during
  * the initial boot procedure. This is only for no emulation boot images,
@@ -77,6 +105,14 @@ void el_torito_set_load_size(ElToritoBootImage *bootimg, short sectors)
     bootimg->load_size = sectors;
 }
 
+/* API */
+int el_torito_get_load_size(ElToritoBootImage *bootimg)
+{
+   if (bootimg->load_size < 0)
+       return 0xffff - bootimg->load_size;
+   return bootimg->load_size;
+}
+
 /**
  * Marks the specified boot image as not bootable
  */
@@ -85,8 +121,50 @@ void el_torito_set_no_bootable(ElToritoBootImage *bootimg)
     bootimg->bootable = 0;
 }
 
+/* API */
+int el_torito_get_bootable(ElToritoBootImage *bootimg)
+{   
+    return !!bootimg->bootable;
+}
+
+/* API */
+int el_torito_set_id_string(ElToritoBootImage *bootimg, uint8_t id_string[28])
+{
+    memcpy(bootimg->id_string, id_string, 28);
+    return 1;
+}
+
+/* API */
+int el_torito_get_id_string(ElToritoBootImage *bootimg, uint8_t id_string[28])
+{
+    
+    memcpy(id_string, bootimg->id_string, 28);
+    return 1;
+}
+
+/* API */
+int el_torito_set_selection_crit(ElToritoBootImage *bootimg, uint8_t crit[20])
+{
+    memcpy(bootimg->selection_crit, crit, 20);
+    return 1;
+}
+
+/* API */
+int el_torito_get_selection_crit(ElToritoBootImage *bootimg, uint8_t crit[20])
+{
+    
+    memcpy(crit, bootimg->selection_crit, 20);
+    return 1;
+}
+
+/* API */
+int el_torito_seems_boot_info_table(ElToritoBootImage *bootimg, int flag)
+{
+    return bootimg->seems_boot_info_table;
+}
+
 /**
- * Specifies that this image needs to be patched. This involves the writting
+ * Specifies that this image needs to be patched. This involves the writing
  * of a 56 bytes boot information table at offset 8 of the boot image file.
  * The original boot image file won't be modified.
  * This is needed for isolinux boot images.
@@ -105,7 +183,7 @@ void el_torito_patch_isolinux_image(ElToritoBootImage *bootimg)
  *        bitmask style flag. The following values are defined:
  *
  *        bit 0 -> 1 to path the image, 0 to not
- *                 Patching the image involves the writting of a 56 bytes
+ *                 Patching the image involves the writing of a 56 bytes
  *                 boot information table at offset 8 of the boot image file.
  *                 The original boot image file won't be modified. This is needed
  *                 to allow isolinux images to be bootable.
@@ -123,31 +201,36 @@ int el_torito_set_isolinux_options(ElToritoBootImage *bootimg, int options, int
     return ISO_SUCCESS;
 }
 
-/* TODO getter for boot image properties should be exposed
+/* API */
- * useful when reading discs */
+int el_torito_get_isolinux_options(ElToritoBootImage *bootimg, int flag)
-int el_torito_get_boot_media_type(const ElToritoBootImage *bootimg)
+{
+    return bootimg->isolinux_options & 0x03;
+}
+
+/* API */
+int el_torito_get_boot_media_type(ElToritoBootImage *bootimg,
+                                  enum eltorito_boot_media_type *media_type)
 {
     if (bootimg) {
         switch (bootimg->type) {
         case 1:
         case 2:
         case 3:
-            return ELTORITO_FLOPPY_EMUL;
+            *media_type = ELTORITO_FLOPPY_EMUL;
-            break;
+            return 1;
         case 4:
-            return ELTORITO_HARD_DISC_EMUL;
+            *media_type = ELTORITO_HARD_DISC_EMUL;
-            break;
+            return 1;
         case 0:
-            return ELTORITO_NO_EMUL;
+            *media_type = ELTORITO_NO_EMUL;
-            break;
+            return 1;
         default:
             /* should never happen */
             return ISO_ASSERT_FAILURE;
             break;
         }
-    } else {
-        return ISO_WRONG_ARG_VALUE;
     }
+    return ISO_WRONG_ARG_VALUE;
 }
 
 static
@@ -226,6 +309,7 @@ int create_image(IsoImage *image, const char *image_path,
     int boot_media_type = 0;
     int load_sectors = 0; /* number of sector to load */
     unsigned char partition_type = 0;
+    off_t size;
     IsoNode *imgfile;
     IsoStream *stream;
 
@@ -234,6 +318,9 @@ int create_image(IsoImage *image, const char *image_path,
         return ret;
     }
     if (ret == 0) {
+        iso_msg_submit(image->id, ISO_NODE_DOESNT_EXIST, 0,
+                       "El Torito boot image file missing in ISO image: '%s'",
+                       image_path);
         return ISO_NODE_DOESNT_EXIST;
     }
 
@@ -248,9 +335,16 @@ int create_image(IsoImage *image, const char *image_path,
         return ISO_BOOT_IMAGE_NOT_VALID;
     }
 
+    size = iso_stream_get_size(stream);
+    if (size <= 0) {
+        iso_msg_submit(image->id, ISO_BOOT_IMAGE_NOT_VALID, 0,
+                       "Boot image file is empty");
+        return ISO_BOOT_IMAGE_NOT_VALID;
+    }
+
     switch (type) {
     case ELTORITO_FLOPPY_EMUL:
-        switch (iso_stream_get_size(stream)) {
+        switch (size) {
         case 1200 * 1024:
             boot_media_type = 1; /* 1.2 meg diskette */
             break;
@@ -332,10 +426,15 @@ int create_image(IsoImage *image, const char *image_path,
     boot->image = (IsoFile*)imgfile;
     iso_node_ref(imgfile); /* get our ref */
     boot->bootable = 1;
+    boot->seems_boot_info_table = 0;
+    boot->isolinux_options = 0;
     boot->type = boot_media_type;
-    boot->load_size = load_sectors;
     boot->partition_type = partition_type;
-
+    boot->load_seg = 0;
+    boot->load_size = load_sectors;
+    boot->platform_id = 0; /* 80x86 */
+    memset(boot->id_string, 0, sizeof(boot->id_string));
+    memset(boot->selection_crit, 0, sizeof(boot->selection_crit));
     if (bootimg) {
         *bootimg = boot;
     }
@@ -348,7 +447,7 @@ int iso_image_set_boot_image(IsoImage *image, const char *image_path,
                              const char *catalog_path,
                              ElToritoBootImage **boot)
 {
-    int ret;
+    int ret, i;
     struct el_torito_boot_catalog *catalog;
     ElToritoBootImage *boot_image= NULL;
     IsoBoot *cat_node= NULL;
@@ -383,6 +482,9 @@ int iso_image_set_boot_image(IsoImage *image, const char *image_path,
             catname[0] = '\0';
             ret = iso_tree_path_to_node(image, catdir, &p);
             if (ret <= 0) {
+                iso_msg_submit(image->id, ISO_NODE_DOESNT_EXIST, 0,
+         "Cannot find directory for El Torito boot catalog in ISO image: '%s'",
+                               catdir);
                 free(catdir);
                 return ret < 0 ? ret : ISO_NODE_DOESNT_EXIST;
             }
@@ -407,13 +509,17 @@ int iso_image_set_boot_image(IsoImage *image, const char *image_path,
     }
 
     /* creates the catalog with the given image */
-    catalog = malloc(sizeof(struct el_torito_boot_catalog));
+    catalog = calloc(1, sizeof(struct el_torito_boot_catalog));
     if (catalog == NULL) {
         ret = ISO_OUT_OF_MEM;
         goto boot_image_cleanup;
     }
-    catalog->image = boot_image;
+    catalog->num_bootimages = 1;
+    catalog->bootimages[0] = boot_image;
+    for (i = 1; i < Libisofs_max_boot_imageS; i++)
+        catalog->bootimages[i] = NULL;
     catalog->node = cat_node;
+    catalog->sort_weight = 1000;                            /* slightly high */
     iso_node_ref((IsoNode*)cat_node);
     image->bootcat = catalog;
 
@@ -436,7 +542,7 @@ boot_image_cleanup:;
 }
 
 /**
- * Get El-Torito boot image of an ISO image, if any.
+ * Get the boot catalog and the El-Torito default boot image of an ISO image.
  *
  * This can be useful, for example, to check if a volume read from a previous
  * session or an existing image is bootable. It can also be useful to get
@@ -480,10 +586,10 @@ int iso_image_get_boot_image(IsoImage *image, ElToritoBootImage **boot,
 
     /* ok, image is bootable */
     if (boot) {
-        *boot = image->bootcat->image;
+        *boot = image->bootcat->bootimages[0];
     }
     if (imgnode) {
-        *imgnode = image->bootcat->image->image;
+        *imgnode = image->bootcat->bootimages[0]->image;
     }
     if (catnode) {
         *catnode = image->bootcat->node;
@@ -491,6 +597,40 @@ int iso_image_get_boot_image(IsoImage *image, ElToritoBootImage **boot,
     return ISO_SUCCESS;
 }
 
+int iso_image_get_all_boot_imgs(IsoImage *image, int *num_boots,
+               ElToritoBootImage ***boots, IsoFile ***bootnodes, int flag)
+{
+    int i;
+    struct el_torito_boot_catalog *cat;
+
+    if (image == NULL)
+        return ISO_NULL_POINTER;
+    if (image->bootcat == NULL)
+        return 0;
+    cat = image->bootcat;
+    *num_boots = cat->num_bootimages;
+    *boots = NULL;
+    *bootnodes = NULL;
+    if (*num_boots <= 0)
+        return 0;
+    *boots = calloc(*num_boots, sizeof(ElToritoBootImage *));
+    *bootnodes = calloc(*num_boots, sizeof(IsoFile *));
+    if(*boots == NULL || *bootnodes == NULL) {
+        if (*boots != NULL)
+            free(*boots);
+        if (*bootnodes != NULL)
+            free(*bootnodes);
+        *boots = NULL;
+        *bootnodes = NULL;
+        return ISO_OUT_OF_MEM;
+    }
+    for (i = 0; i < *num_boots; i++) {
+        (*boots)[i] = cat->bootimages[i];
+        (*bootnodes)[i] = image->bootcat->bootimages[i]->image;
+    }
+    return 1;
+}
+
 /**
  * Removes the El-Torito bootable image.
  *
@@ -505,28 +645,82 @@ void iso_image_remove_boot_image(IsoImage *image)
         return;
 
     /*
-     * remove catalog node from its parent
+     * remove catalog node from its parent and dispose it
-     * (the reference will be disposed next)
+     * (another reference is with the catalog)
      */
-    iso_node_take((IsoNode*)image->bootcat->node);
+    if (iso_node_get_parent((IsoNode*) image->bootcat->node) != NULL) {
+        iso_node_take((IsoNode*) image->bootcat->node);
+        iso_node_unref((IsoNode*) image->bootcat->node);
+    }
 
     /* free boot catalog and image, including references to nodes */
     el_torito_boot_catalog_free(image->bootcat);
     image->bootcat = NULL;
 }
 
+/* API */
+int iso_image_add_boot_image(IsoImage *image, const char *image_path,
+                             enum eltorito_boot_media_type type, int flag,
+                             ElToritoBootImage **boot)
+{
+    int ret;
+    struct el_torito_boot_catalog *catalog = image->bootcat;
+    ElToritoBootImage *boot_img;
+
+    if(catalog == NULL)
+      return ISO_BOOT_NO_CATALOG;
+    if (catalog->num_bootimages >= Libisofs_max_boot_imageS)
+        return ISO_BOOT_IMAGE_OVERFLOW;
+    ret = create_image(image, image_path, type, &boot_img);
+    if (ret < 0) 
+        return ret;
+    catalog->bootimages[catalog->num_bootimages] = boot_img;
+    catalog->num_bootimages++;
+    if (boot != NULL)
+        *boot = boot_img;
+    return 1;
+}
+
+/* API */
+int iso_image_set_boot_catalog_weight(IsoImage *image, int sort_weight)
+{
+    if (image->bootcat == NULL)
+        return 0;
+    image->bootcat->sort_weight = sort_weight;
+    return 1;
+}
+
+/* API */
+int iso_image_set_boot_catalog_hidden(IsoImage *image, int hide_attrs)
+{
+    if (image->bootcat == NULL)
+        return 0;
+    if (image->bootcat->node == NULL)
+        return 0;
+    iso_node_set_hidden((IsoNode *) image->bootcat->node, hide_attrs);
+    return 1;
+}
+
+
 void el_torito_boot_catalog_free(struct el_torito_boot_catalog *cat)
 {
     struct el_torito_boot_image *image;
+    int i;
 
     if (cat == NULL) {
         return;
     }
 
-    image = cat->image;
+    for (i = 0; i < Libisofs_max_boot_imageS; i++) {
-    iso_node_unref((IsoNode*)image->image);
+        image = cat->bootimages[i];
-    free(image);
+        if (image == NULL)
-    iso_node_unref((IsoNode*)cat->node);
+    continue;
+        if ((IsoNode*)image->image != NULL)
+            iso_node_unref((IsoNode*)image->image);
+        free(image);
+    }
+    if ((IsoNode*)cat->node != NULL)
+        iso_node_unref((IsoNode*)cat->node);
     free(cat);
 }
 
@@ -537,11 +731,12 @@ struct catalog_stream
 {
     Ecma119Image *target;
     uint8_t buffer[BLOCK_SIZE];
-    int offset; /* -1 if stream is not openned */
+    int offset; /* -1 if stream is not opened */
 };
 
 static void
-write_validation_entry(uint8_t *buf)
+write_validation_entry(uint8_t *buf, uint8_t platform_id,
+                       uint8_t id_string[24])
 {
     size_t i;
     int checksum;
@@ -549,10 +744,10 @@ write_validation_entry(uint8_t *buf)
     struct el_torito_validation_entry *ve =
         (struct el_torito_validation_entry*)buf;
     ve->header_id[0] = 1;
-    ve->platform_id[0] = 0; /* 0: 80x86, 1: PowerPC, 2: Mac */
+    ve->platform_id[0] = platform_id;
+    memcpy(ve->id_string, id_string, sizeof(ve->id_string));
     ve->key_byte1[0] = 0x55;
     ve->key_byte2[0] = 0xAA;
-
     /* calculate the checksum, to ensure sum of all words is 0 */
     checksum = 0;
     for (i = 0; i < sizeof(struct el_torito_validation_entry); i += 2) {
@@ -561,35 +756,64 @@ write_validation_entry(uint8_t *buf)
     iso_lsb(ve->checksum, checksum, 2);
 }
 
+static void
+write_section_header(uint8_t *buf, Ecma119Image *t, int idx, int num_entries)
+{
+    int pi;
+    char *id_string;
+
+    struct el_torito_section_header *e =
+        (struct el_torito_section_header *) buf;
+
+    /* 0x90 = more section headers follow , 0x91 = final section */
+    e->header_indicator[0] = 0x90 + (idx == t->catalog->num_bootimages - 1);
+    pi= e->platform_id[0] = t->catalog->bootimages[idx]->platform_id;
+    e->num_entries[0] = num_entries & 0xff;
+    e->num_entries[1] = (num_entries >> 8) & 0xff;;
+    id_string = (char *) e->id_string;
+    memcpy(id_string,  t->catalog->bootimages[idx]->id_string,
+           sizeof(e->id_string));
+}
+
 /**
  * Write one section entry.
- * Currently this is used only for default image (the only supported just now)
+ * Usable for the Default Entry
+ * and for Section Entries with Selection criteria type == 0
  */
 static void
-write_section_entry(uint8_t *buf, Ecma119Image *t)
+write_section_entry(uint8_t *buf, Ecma119Image *t, int idx)
 {
     struct el_torito_boot_image *img;
     struct el_torito_section_entry *se =
         (struct el_torito_section_entry*)buf;
 
-    img = t->catalog->image;
+    img = t->catalog->bootimages[idx];
 
     se->boot_indicator[0] = img->bootable ? 0x88 : 0x00;
     se->boot_media_type[0] = img->type;
     iso_lsb(se->load_seg, img->load_seg, 2);
     se->system_type[0] = img->partition_type;
     iso_lsb(se->sec_count, img->load_size, 2);
-    iso_lsb(se->block, t->bootimg->sections[0].block, 4);
+    iso_lsb(se->block, t->bootsrc[idx]->sections[0].block, 4);
+    se->selec_criteria[0] = img->selection_crit[0];
+    memcpy(se->vendor_sc, img->selection_crit + 1, 19);
 }
 
 static
 int catalog_open(IsoStream *stream)
 {
+    int i, j, k, num_entries;
     struct catalog_stream *data;
+    uint8_t *wpt;
+    struct el_torito_boot_catalog *cat;
+    struct el_torito_boot_image **boots;
+
     if (stream == NULL) {
         return ISO_NULL_POINTER;
     }
     data = stream->data;
+    cat = data->target->catalog;
+    boots = cat->bootimages;
 
     if (data->offset != -1) {
         return ISO_FILE_ALREADY_OPENED;
@@ -598,11 +822,36 @@ int catalog_open(IsoStream *stream)
     memset(data->buffer, 0, BLOCK_SIZE);
 
     /* fill the buffer with the catalog contents */
-    write_validation_entry(data->buffer);
+    write_validation_entry(data->buffer,
+                           boots[0]->platform_id, boots[0]->id_string);
 
-    /* write default entry */
+    /* write default entry = first boot image */
-    write_section_entry(data->buffer + 32, data->target);
+    write_section_entry(data->buffer + 32, data->target, 0);
 
+    /* IMPORTANT: The maximum number of boot images must fit into BLOCK_SIZE */
+    wpt = data->buffer + 64;
+    for (i = 1; i < cat->num_bootimages; ) {
+        /* Look ahead and put images of same platform_id and id_string
+           into the same section */
+        for (j = i + 1; j < cat->num_bootimages; j++) {
+             if (boots[i]->platform_id != boots[j]->platform_id)
+        break;
+             for (k = 0; k < sizeof(boots[i]->id_string); k++)
+                 if (boots[i]->id_string[k] != boots[j]->id_string[k])
+             break;
+             if (k < sizeof(boots[i]->id_string))
+        break;
+        }
+        num_entries = j - i;
+
+        write_section_header(wpt, data->target, i, num_entries);
+        wpt += 32;
+        for (j = 0; j < num_entries; j++) {
+            write_section_entry(wpt,  data->target, i);
+            wpt += 32;
+            i++;
+        }
+    }
     data->offset = 0;
     return ISO_SUCCESS;
 }
@@ -702,12 +951,12 @@ int catalog_stream_new(Ecma119Image *target, IsoStream **stream)
         return ISO_NULL_POINTER;
     }
 
-    str = malloc(sizeof(IsoStream));
+    str = calloc(1, sizeof(IsoStream));
     if (str == NULL) {
         return ISO_OUT_OF_MEM;
     }
-    data = malloc(sizeof(struct catalog_stream));
+    data = calloc(1, sizeof(struct catalog_stream));
-    if (str == NULL) {
+    if (data == NULL) {
         free(str);
         return ISO_OUT_OF_MEM;
     }
@@ -740,7 +989,7 @@ int el_torito_catalog_file_src_create(Ecma119Image *target, IsoFileSrc **src)
         return ISO_SUCCESS;
     }
 
-    file = malloc(sizeof(IsoFileSrc));
+    file = calloc(1, sizeof(IsoFileSrc));
     if (file == NULL) {
         return ISO_OUT_OF_MEM;
     }
@@ -753,9 +1002,10 @@ int el_torito_catalog_file_src_create(Ecma119Image *target, IsoFileSrc **src)
 
     /* fill fields */
     file->prev_img = 0; /* TODO allow copy of old img catalog???? */
+    file->checksum_index = 0;
     file->nsections = 1;
     file->sections = calloc(1, sizeof(struct iso_file_section));
-    file->sort_weight = 1000; /* slightly high */
+    file->sort_weight = target->catalog->sort_weight;
     file->stream = stream;
 
     ret = iso_file_src_add(target, file, src);
@@ -771,27 +1021,26 @@ int el_torito_catalog_file_src_create(Ecma119Image *target, IsoFileSrc **src)
 /******************* EL-TORITO WRITER *******************************/
 
 /**
- * Patch an isolinux boot image.
+ * Insert boot info table content into buf.
  *
  * @return
  *      1 on success, 0 error (but continue), < 0 error
  */
-static
+int make_boot_info_table(uint8_t *buf, uint32_t pvd_lba,
-int patch_boot_image(uint8_t *buf, Ecma119Image *t, size_t imgsize)
+                         uint32_t boot_lba, uint32_t imgsize)
 {
     struct boot_info_table *info;
     uint32_t checksum;
-    size_t offset;
+    uint32_t offset;
 
-    if (imgsize < 64) {
+    info = (struct boot_info_table *) (buf + 8);
-        return iso_msg_submit(t->image->id, ISO_ISOLINUX_CANT_PATCH, 0,
+    if (imgsize < 64)
-            "Isolinux image too small. We won't patch it.");
+        return ISO_ISOLINUX_CANT_PATCH;
-    }
 
     /* compute checksum, as the the sum of all 32 bit words in boot image
      * from offset 64 */
     checksum = 0;
-    offset = (size_t) 64;
+    offset = 64;
 
     while (offset <= imgsize - 4) {
         checksum += iso_read_lsb(buf + offset, 4);
@@ -806,16 +1055,36 @@ int patch_boot_image(uint8_t *buf, Ecma119Image *t, size_t imgsize)
         checksum += iso_read_lsb(buf + offset, imgsize - offset);
     }
 
-    /* patch boot info table */
-    info = (struct boot_info_table*)(buf + 8);
     /*memset(info, 0, sizeof(struct boot_info_table));*/
-    iso_lsb(info->bi_pvd, t->ms_block + 16, 4);
+    iso_lsb(info->bi_pvd, pvd_lba, 4);
-    iso_lsb(info->bi_file, t->bootimg->sections[0].block, 4);
+    iso_lsb(info->bi_file, boot_lba, 4);
     iso_lsb(info->bi_length, imgsize, 4);
     iso_lsb(info->bi_csum, checksum, 4);
+    memset(buf + 24, 0, 40);
     return ISO_SUCCESS;
 }
 
+/**
+ * Patch an isolinux boot image.
+ *
+ * @return
+ *      1 on success, 0 error (but continue), < 0 error
+ */
+static
+int patch_boot_image(uint8_t *buf, Ecma119Image *t, size_t imgsize, int idx)
+{
+    int ret;
+
+    if (imgsize < 64) {
+        return iso_msg_submit(t->image->id, ISO_ISOLINUX_CANT_PATCH, 0,
+            "Isolinux image too small. We won't patch it.");
+    }
+    ret = make_boot_info_table(buf, t->ms_block + (uint32_t) 16,
+                               t->bootsrc[idx]->sections[0].block,
+                               (uint32_t) imgsize);
+    return ret;
+}
+
 static
 int eltorito_writer_compute_data_blocks(IsoImageWriter *writer)
 {
@@ -824,7 +1093,11 @@ int eltorito_writer_compute_data_blocks(IsoImageWriter *writer)
      * this is a good place to do so.
      */
     Ecma119Image *t;
-    int ret;
+    int ret, idx;
+    size_t size;
+    uint8_t *buf;
+    IsoStream *new = NULL;
+    IsoStream *original = NULL;
 
     if (writer == NULL) {
         return ISO_NULL_POINTER;
@@ -832,19 +1105,23 @@ int eltorito_writer_compute_data_blocks(IsoImageWriter *writer)
 
     t = writer->target;
 
-    if (t->catalog->image->isolinux_options & 0x01) {
+    /* Patch the boot image info tables if indicated */
-        /* we need to patch the image */
+    for (idx = 0; idx < t->catalog->num_bootimages; idx++) {
-        size_t size;
+        if (!(t->catalog->bootimages[idx]->isolinux_options & 0x01))
-        uint8_t *buf;
+    continue;
-        IsoStream *new = NULL;
+        original = t->bootsrc[idx]->stream;
-        IsoStream *original = t->bootimg->stream;
         size = (size_t) iso_stream_get_size(original);
-        buf = malloc(size);
+
+        /* >>> BOOT ts B00428 :
+               check whether size is not too large for buffering */;
+
+        buf = calloc(1, size);
         if (buf == NULL) {
             return ISO_OUT_OF_MEM;
         }
         ret = iso_stream_open(original);
         if (ret < 0) {
+            free(buf);
             return ret;
         }
         ret = iso_stream_read(original, buf, size);
@@ -854,7 +1131,7 @@ int eltorito_writer_compute_data_blocks(IsoImageWriter *writer)
         }
 
         /* ok, patch the read buffer */
-        ret = patch_boot_image(buf, t, size);
+        ret = patch_boot_image(buf, t, size, idx);
         if (ret < 0) {
             return ret;
         }
@@ -865,7 +1142,7 @@ int eltorito_writer_compute_data_blocks(IsoImageWriter *writer)
         if (ret < 0) {
             return ret;
         }
-        t->bootimg->stream = new;
+        t->bootsrc[idx]->stream = new;
         iso_stream_unref(original);
     }
     return ISO_SUCCESS;
@@ -887,7 +1164,6 @@ int eltorito_writer_write_vol_desc(IsoImageWriter *writer)
 
     t = writer->target;
     cat = t->catalog;
-
     iso_msg_debug(t->image->id, "Write El-Torito boot record");
 
     memset(&vol, 0, sizeof(struct ecma119_boot_rec_vol_desc));
@@ -895,8 +1171,8 @@ int eltorito_writer_write_vol_desc(IsoImageWriter *writer)
     memcpy(vol.std_identifier, "CD001", 5);
     vol.vol_desc_version[0] = 1;
     memcpy(vol.boot_sys_id, "EL TORITO SPECIFICATION", 23);
-    iso_lsb(vol.boot_catalog, t->cat->sections[0].block, 4);
+    iso_lsb(vol.boot_catalog,
-
+            t->cat->sections[0].block - t->eff_partition_offset, 4);
     return iso_write(t, &vol, sizeof(struct ecma119_boot_rec_vol_desc));
 }
 
@@ -916,12 +1192,12 @@ int eltorito_writer_free_data(IsoImageWriter *writer)
 
 int eltorito_writer_create(Ecma119Image *target)
 {
-    int ret;
+    int ret, idx;
     IsoImageWriter *writer;
     IsoFile *bootimg;
     IsoFileSrc *src;
 
-    writer = malloc(sizeof(IsoImageWriter));
+    writer = calloc(1, sizeof(IsoImageWriter));
     if (writer == NULL) {
         return ISO_OUT_OF_MEM;
     }
@@ -947,16 +1223,19 @@ int eltorito_writer_create(Ecma119Image *target)
             return ret;
         }
     }
-    bootimg = target->catalog->image->image;
-    ret = iso_file_src_create(target, bootimg, &src);
-    if (ret < 0) {
-        return ret;
-    }
-    target->bootimg = src;
 
-    /* if we have selected to patch the image, it needs to be copied always */
+    for (idx = 0; idx < target->catalog->num_bootimages; idx++) {
-    if (target->catalog->image->isolinux_options & 0x01) {
+        bootimg = target->catalog->bootimages[idx]->image;
-        src->prev_img = 0;
+        ret = iso_file_src_create(target, bootimg, &src);
+        if (ret < 0) {
+            return ret;
+        }
+        target->bootsrc[idx] = src;
+
+        /* For patching an image, it needs to be copied always */
+        if (target->catalog->bootimages[idx]->isolinux_options & 0x01) {
+            src->prev_img = 0;
+        }
     }
 
     /* we need the bootable volume descriptor */

libisofs/eltorito.h

@@ -1,9 +1,11 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2010 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /**
@@ -26,25 +28,45 @@ struct Iso_Boot
     IsoNode node;
 };
 
+/* Not more than 32 so that all entries fit into 2048 bytes */
+#define Libisofs_max_boot_imageS 32
+
 struct el_torito_boot_catalog {
     IsoBoot *node; /* node of the catalog */
-    struct el_torito_boot_image *image; /* default boot image */
+
+    int num_bootimages;
+    struct el_torito_boot_image *bootimages[Libisofs_max_boot_imageS];
+                                                  /* [0]= default boot image */
+
+    /* Weight value for image sorting */
+    int sort_weight;
 };
 
 struct el_torito_boot_image {
     IsoFile *image;
 
     unsigned int bootable:1; /**< If the entry is bootable. */
+    /**
+     * Whether the boot image seems to contain a boot_info_table
+     */
+    unsigned int seems_boot_info_table:1;
     /**
      * isolinux options
      * bit 0 -> whether to patch image
-     * bit 1 -> whether to create isolinux image
+     * bit 1 -> whether to put built-in isolinux 3.72 isohybrid-MBR into image
+     *          System Area (deprecated)
      */
     unsigned int isolinux_options:2;
     unsigned char type; /**< The type of image */
     unsigned char partition_type; /**< type of partition for HD-emul images */
     short load_seg; /**< Load segment for the initial boot image. */
     short load_size; /**< Number of sectors to load. */
+
+    /* Byte 1 of Validation Entry or Section Header Entry:
+       0= 80x86, 1= PowerPC, 2= Mac, 0xef= EFI */
+    uint8_t platform_id;
+    uint8_t id_string[28];
+    uint8_t selection_crit[20];
 };
 
 /** El-Torito, 2.1 */
@@ -74,8 +96,8 @@ struct el_torito_default_entry {
 struct el_torito_section_header {
     uint8_t header_indicator    BP(1, 1);
     uint8_t platform_id         BP(2, 2);
-    uint8_t number              BP(3, 4);
+    uint8_t num_entries         BP(3, 4);
-    uint8_t character           BP(5, 32);
+    uint8_t id_string           BP(5, 32);
 };
 
 /** El-Torito, 2.4 */
@@ -105,4 +127,14 @@ int el_torito_catalog_file_src_create(Ecma119Image *target, IsoFileSrc **src);
  */
 int eltorito_writer_create(Ecma119Image *target);
 
+/**
+ * Insert boot info table content into buf.
+ *
+ * @return
+ *      1 on success, 0 error (but continue), < 0 error
+ */
+int make_boot_info_table(uint8_t *buf, uint32_t pvd_lba,
+                         uint32_t boot_lba, uint32_t imgsize);
+
+
 #endif /* LIBISO_ELTORITO_H */

libisofs/filesrc.c

@@ -1,60 +1,67 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ *               2010 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
+#include "libisofs.h"
 #include "filesrc.h"
 #include "node.h"
 #include "util.h"
 #include "writer.h"
 #include "messages.h"
 #include "image.h"
+#include "stream.h"
+#include "md5.h"
 
 #include <stdlib.h>
 #include <string.h>
 #include <limits.h>
 
+
+#ifdef Xorriso_standalonE
+
+#ifdef Xorriso_with_libjtE
+#include "../libjte/libjte.h"
+#endif
+
+#else
+
+#ifdef Libisofs_with_libjtE
+#include <libjte/libjte.h>
+#endif
+
+#endif /* ! Xorriso_standalonE */
+
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
 int iso_file_src_cmp(const void *n1, const void *n2)
 {
+    int ret;
     const IsoFileSrc *f1, *f2;
-    unsigned int fs_id1, fs_id2;
+
-    dev_t dev_id1, dev_id2;
+    if (n1 == n2) {
-    ino_t ino_id1, ino_id2;
+        return 0; /* Normally just a shortcut.
-    off_t size1, size2;
+                     But important if Libisofs_file_src_cmp_non_zerO */
+    }
+
     f1 = (const IsoFileSrc *)n1;
     f2 = (const IsoFileSrc *)n2;
 
-    iso_stream_get_id(f1->stream, &fs_id1, &dev_id1, &ino_id1);
+    ret = iso_stream_cmp_ino(f1->stream, f2->stream, 0);
-    iso_stream_get_id(f2->stream, &fs_id2, &dev_id2, &ino_id2);
+    return ret;
-
-    if (fs_id1 < fs_id2) {
-        return -1;
-    } else if (fs_id1 > fs_id2) {
-        return 1;
-    } else {
-        /* files belong to the same fs */
-        if (dev_id1 > dev_id2) {
-            return -1;
-        } else if (dev_id1 < dev_id2) {
-            return 1;
-        } else if (ino_id1 < ino_id2) {
-            return -1;
-        } else if (ino_id1 > ino_id2) {
-            return 1;
-        } else {
-            size1 = iso_stream_get_size(f1->stream);
-            size2 = iso_stream_get_size(f2->stream);
-            if (size1 < size2) {
-                return -1;
-            } else if (size1 > size2) {
-                return 1;
-            }
-            return 0;
-        }
-    }
 }
 
 int iso_file_src_create(Ecma119Image *img, IsoFile *file, IsoFileSrc **src)
@@ -64,6 +71,8 @@ int iso_file_src_create(Ecma119Image *img, IsoFile *file, IsoFileSrc **src)
     unsigned int fs_id;
     dev_t dev_id;
     ino_t ino_id;
+    int cret, no_md5= 0;
+    void *xipt = NULL;
 
     if (img == NULL || file == NULL || src == NULL) {
         return ISO_NULL_POINTER;
@@ -109,11 +118,44 @@ int iso_file_src_create(Ecma119Image *img, IsoFile *file, IsoFileSrc **src)
     /* insert the filesrc in the tree */
     ret = iso_rbtree_insert(img->files, fsrc, (void**)src);
     if (ret <= 0) {
+        if (ret == 0 && (*src)->checksum_index > 0) {
+            /* Duplicate file source was mapped to previously registered source
+            */
+            cret = iso_file_set_isofscx(file, (*src)->checksum_index, 0);
+            if (cret < 0)
+                ret = cret;
+        }
         free(fsrc->sections);
         free(fsrc);
         return ret;
     }
     iso_stream_ref(fsrc->stream);
+
+    if ((img->md5_file_checksums & 1) &&
+        file->from_old_session && img->appendable) {
+        ret = iso_node_get_xinfo((IsoNode *) file, checksum_md5_xinfo_func,
+                                  &xipt);
+        if (ret <= 0)
+            ret = iso_node_get_xinfo((IsoNode *) file, checksum_cx_xinfo_func,
+                                      &xipt);
+        if (ret <= 0)
+            /* Omit MD5 indexing with old image nodes which have no MD5 */
+            no_md5 = 1;
+    }
+
+    if ((img->md5_file_checksums & 1) && !no_md5) {
+        img->checksum_idx_counter++;
+        if (img->checksum_idx_counter < 0x7fffffff) {
+            fsrc->checksum_index = img->checksum_idx_counter;
+        } else {
+            fsrc->checksum_index= 0;
+            img->checksum_idx_counter= 0x7fffffff; /* keep from rolling over */
+        }
+        cret = iso_file_set_isofscx(file, (*src)->checksum_index, 0);
+        if (cret < 0)
+            return cret;
+    }
+
     return ISO_SUCCESS;
 }
 
@@ -190,6 +232,12 @@ int filesrc_writer_compute_data_blocks(IsoImageWriter *writer)
 
     t = writer->target;
 
+    /* Normally reserve a single zeroed block for all files which have
+       no block address: symbolic links, device files, empty data files.
+    */
+    if (! t->old_empty)
+        t->curblock++;
+
     /* on appendable images, ms files shouldn't be included */
     if (t->appendable) {
         inc_item = is_ms_file;
@@ -224,7 +272,12 @@ int filesrc_writer_compute_data_blocks(IsoImageWriter *writer)
         /*
          * final section
          */
-        file->sections[extent].block = t->curblock + extent * (ISO_EXTENT_SIZE / BLOCK_SIZE);
+        if (section_size <= 0) {
+            file->sections[extent].block = t->empty_file_block;
+        } else {
+            file->sections[extent].block =
+                         t->curblock + extent * (ISO_EXTENT_SIZE / BLOCK_SIZE);
+        }
         file->sections[extent].size = (uint32_t)section_size;
 
         t->curblock += DIV_UP(iso_file_src_get_size(file), BLOCK_SIZE);
@@ -262,56 +315,70 @@ int filesrc_close(IsoFileSrc *file)
 static
 int filesrc_read(IsoFileSrc *file, char *buf, size_t count)
 {
-    size_t bytes = 0;
+    size_t got;
 
-    /* loop to ensure the full buffer is filled */
+    return iso_stream_read_buffer(file->stream, buf, count, &got);
-    do {
+}
-        ssize_t result;
-        result = iso_stream_read(file->stream, buf + bytes, count - bytes);
-        if (result < 0) {
-            /* fill buffer with 0s and return */
-            memset(buf + bytes, 0, count - bytes);
-            return result;
-        }
-        if (result == 0)
-            break;
-        bytes += result;
-    } while (bytes < count);
 
-    if (bytes < count) {
+/* @return 1=ok, md5 is valid,
-        /* eof */
+           0= not ok, go on,
-        memset(buf + bytes, 0, count - bytes);
+          <0 fatal error, abort 
-        return 0;
+*/  
-    } else {
+static
-        return 1;
+int filesrc_make_md5(Ecma119Image *t, IsoFileSrc *file, char md5[16], int flag)
-    }
+{
+    return iso_stream_make_md5(file->stream, md5, 0);
 }
 
 static
 int filesrc_writer_write_data(IsoImageWriter *writer)
 {
-    int res;
+    int res, ret, was_error;
     size_t i, b;
     Ecma119Image *t;
     IsoFileSrc *file;
     IsoFileSrc **filelist;
     char name[PATH_MAX];
     char buffer[BLOCK_SIZE];
+    off_t file_size;
+    uint32_t nblocks;
+    void *ctx= NULL;
+    char md5[16], pre_md5[16];
+    int pre_md5_valid = 0;
+#ifdef Libisofs_with_libjtE
+    int jte_begun = 0;
+#endif
 
     if (writer == NULL) {
         return ISO_ASSERT_FAILURE;
     }
 
+    memset(buffer, 0, BLOCK_SIZE);
     t = writer->target;
     filelist = writer->data;
 
     iso_msg_debug(t->image->id, "Writing Files...");
 
+    /* Normally write a single zeroed block as block address target for all
+       files which have no block address:
+       symbolic links, device files, empty data files.
+    */
+    if (! t->old_empty) {
+        ret = iso_write(t, buffer, BLOCK_SIZE);
+        if (ret < 0)
+            goto ex;
+    }
+
     i = 0;
     while ((file = filelist[i++]) != NULL) {
-
+        was_error = 0;
-        uint32_t nblocks = DIV_UP(iso_file_src_get_size(file), BLOCK_SIZE);
+        file_size = iso_file_src_get_size(file);
-
+        nblocks = DIV_UP(file_size, BLOCK_SIZE);
+        pre_md5_valid = 0; 
+        if (file->checksum_index > 0 && (t->md5_file_checksums & 2)) {
+            /* Obtain an MD5 of content by a first read pass */
+            pre_md5_valid = filesrc_make_md5(t, file, pre_md5, 0);
+        }
         res = filesrc_open(file);
         iso_stream_get_file_name(file->stream, name);
         if (res < 0) {
@@ -320,28 +387,33 @@ int filesrc_writer_write_data(IsoImageWriter *writer)
              * 0's to image
              */
             iso_report_errfile(name, ISO_FILE_CANT_WRITE, 0, 0);
+            was_error = 1;
             res = iso_msg_submit(t->image->id, ISO_FILE_CANT_WRITE, res,
                       "File \"%s\" can't be opened. Filling with 0s.", name);
             if (res < 0) {
-                return res; /* aborted due to error severity */
+                ret = res; /* aborted due to error severity */
+                goto ex;
             }
             memset(buffer, 0, BLOCK_SIZE);
             for (b = 0; b < nblocks; ++b) {
                 res = iso_write(t, buffer, BLOCK_SIZE);
                 if (res < 0) {
                     /* ko, writer error, we need to go out! */
-                    return res;
+                    ret = res;
+                    goto ex;
                 }
             }
             continue;
         } else if (res > 1) {
             iso_report_errfile(name, ISO_FILE_CANT_WRITE, 0, 0);
+            was_error = 1;
             res = iso_msg_submit(t->image->id, ISO_FILE_CANT_WRITE, 0,
                       "Size of file \"%s\" has changed. It will be %s", name,
                       (res == 2 ? "truncated" : "padded with 0's"));
             if (res < 0) {
                 filesrc_close(file);
-                return res; /* aborted due to error severity */
+                ret = res; /* aborted due to error severity */
+                goto ex;
             }
         }
 #ifdef LIBISOFS_VERBOSE_DEBUG
@@ -350,6 +422,29 @@ int filesrc_writer_write_data(IsoImageWriter *writer)
         }
 #endif
 
+#ifdef Libisofs_with_libjtE
+        if (t->libjte_handle != NULL) {
+            res = libjte_begin_data_file(t->libjte_handle, name,
+                                         BLOCK_SIZE, file_size);
+            if (res <= 0) {
+                res = iso_libjte_forward_msgs(t->libjte_handle, t->image->id,
+                                        ISO_LIBJTE_FILE_FAILED, 0);
+                if (res < 0) {
+                    filesrc_close(file);
+                    ret = ISO_LIBJTE_FILE_FAILED;
+                    goto ex;
+                }
+            }
+            jte_begun = 1;
+        }
+#endif /* Libisofs_with_libjtE */
+
+        if (file->checksum_index > 0) {
+            /* initialize file checksum */
+            res = iso_md5_start(&ctx);
+            if (res <= 0)
+                file->checksum_index = 0;
+        }
         /* write file contents to image */
         for (b = 0; b < nblocks; ++b) {
             int wres;
@@ -362,7 +457,18 @@ int filesrc_writer_write_data(IsoImageWriter *writer)
             if (wres < 0) {
                 /* ko, writer error, we need to go out! */
                 filesrc_close(file);
-                return wres;
+                ret = wres;
+                goto ex;
+            }
+            if (file->checksum_index > 0) {
+                /* Add to file checksum */
+                if (file_size - b * BLOCK_SIZE > BLOCK_SIZE)
+                    res = BLOCK_SIZE;
+                else
+                    res = file_size - b * BLOCK_SIZE;
+                res = iso_md5_compute(ctx, buffer, res);
+                if (res <= 0)
+                    file->checksum_index = 0;
             }
         }
 
@@ -371,6 +477,7 @@ int filesrc_writer_write_data(IsoImageWriter *writer)
         if (b < nblocks) {
             /* premature end of file, due to error or eof */
             iso_report_errfile(name, ISO_FILE_CANT_WRITE, 0, 0);
+            was_error = 1;
             if (res < 0) {
                 /* error */
                 res = iso_msg_submit(t->image->id, ISO_FILE_CANT_WRITE, res,
@@ -382,7 +489,8 @@ int filesrc_writer_write_data(IsoImageWriter *writer)
             }
 
             if (res < 0) {
-                return res; /* aborted due error severity */
+                ret = res; /* aborted due error severity */
+                goto ex;
             }
 
             /* fill with 0s */
@@ -393,13 +501,75 @@ int filesrc_writer_write_data(IsoImageWriter *writer)
                 res = iso_write(t, buffer, BLOCK_SIZE);
                 if (res < 0) {
                     /* ko, writer error, we need to go out! */
-                    return res;
+                    ret = res;
+                    goto ex;
+                }
+                if (file->checksum_index > 0) {
+                    /* Add to file checksum */
+                    if (file_size - b * BLOCK_SIZE > BLOCK_SIZE)
+                        res = BLOCK_SIZE;
+                    else
+                        res = file_size - b * BLOCK_SIZE;
+                    res = iso_md5_compute(ctx, buffer, res);
+                    if (res <= 0)
+                        file->checksum_index = 0;
                 }
             }
         }
+        if (file->checksum_index > 0 &&
+            file->checksum_index <= t->checksum_idx_counter) {
+            /* Obtain checksum and dispose checksum context */
+            res = iso_md5_end(&ctx, md5);
+            if (res <= 0)
+                file->checksum_index = 0;
+            if ((t->md5_file_checksums & 2) && pre_md5_valid > 0 &&
+                !was_error) {
+                if (! iso_md5_match(md5, pre_md5)) {
+                    /* Issue MISHAP event */
+                    iso_report_errfile(name, ISO_MD5_STREAM_CHANGE, 0, 0);
+                    was_error = 1;
+                    res = iso_msg_submit(t->image->id, ISO_MD5_STREAM_CHANGE,0,
+           "Content of file '%s' changed while it was written into the image.",
+                                         name);
+                    if (res < 0) {
+                        ret = res; /* aborted due to error severity */
+                        goto ex;
+                    }
+                }
+            }
+            /* Write md5 into checksum buffer at file->checksum_index */
+            memcpy(t->checksum_buffer + 16 * file->checksum_index, md5, 16);
+        }
+
+#ifdef Libisofs_with_libjtE
+        if (t->libjte_handle != NULL) {
+            res = libjte_end_data_file(t->libjte_handle);
+            if (res <= 0) {
+                iso_libjte_forward_msgs(t->libjte_handle, t->image->id,
+                                        ISO_LIBJTE_FILE_FAILED, 0);
+                ret = ISO_LIBJTE_FILE_FAILED;
+                goto ex;
+            }
+            jte_begun = 0;
+        }
+#endif /* Libisofs_with_libjtE */
+
     }
 
-    return ISO_SUCCESS;
+    ret = ISO_SUCCESS;
+ex:;
+    if (ctx != NULL) /* avoid any memory leak */
+        iso_md5_end(&ctx, md5);
+
+#ifdef Libisofs_with_libjtE
+    if (jte_begun) {
+        libjte_end_data_file(t->libjte_handle);
+        iso_libjte_forward_msgs(t->libjte_handle, t->image->id,
+                                ISO_LIBJTE_END_FAILED, 0);
+    }
+#endif /* Libisofs_with_libjtE */
+
+    return ret;
 }
 
 static
@@ -414,9 +584,9 @@ int iso_file_src_writer_create(Ecma119Image *target)
 {
     IsoImageWriter *writer;
 
-    writer = malloc(sizeof(IsoImageWriter));
+    writer = calloc(1, sizeof(IsoImageWriter));
     if (writer == NULL) {
-        return ISO_ASSERT_FAILURE;
+        return ISO_OUT_OF_MEM;
     }
 
     writer->compute_data_blocks = filesrc_writer_compute_data_blocks;

libisofs/filesrc.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_FILESRC_H_
 #define LIBISO_FILESRC_H_
@@ -12,11 +13,18 @@
 #include "stream.h"
 #include "ecma119.h"
 
+#ifdef HAVE_STDINT_H
 #include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
 
 struct Iso_File_Src
 {
     unsigned int prev_img :1; /**< if the file comes from a previous image */
+    unsigned int checksum_index :31;
 
     /** File Sections of the file in the image */
     struct iso_file_section *sections;
@@ -43,7 +51,7 @@ int iso_file_src_cmp(const void *n1, const void *n2);
  * @param src
  *      Will be filled with a pointer to the IsoFileSrc
  * @return
- *      1 on success, < 0 on error
+ *      1 if new object was created, 0 if object existed, < 0 on error
  */
 int iso_file_src_create(Ecma119Image *img, IsoFile *file, IsoFileSrc **src);
 

libisofs/filter.c

@@ -3,10 +3,15 @@
  * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "filter.h"
 #include "node.h"

libisofs/filter.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2008 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_FILTER_H_
 #define LIBISO_FILTER_H_

libisofs/filters/external.c

@@ -1,9 +1,10 @@
 /*
- * Copyright (c) 2009 Thomas Schmitt
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  *
  * It implements a filter facility which can pipe a IsoStream into an external
  * process, read its output and forward it as IsoStream output to an IsoFile.
@@ -12,9 +13,14 @@
  * 
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "../libisofs.h"
 #include "../filter.h"
 #include "../fsource.h"
+#include "../stream.h"
 
 #include <sys/types.h>
 #include <sys/time.h>
@@ -35,7 +41,7 @@
  * for classical pipe filtering.
  */
 
-
+/* IMPORTANT: Any change must be reflected by extf_clone_stream() */
 /*
  * Individual runtime properties exist only as long as the stream is opened.
  */
@@ -593,9 +599,44 @@ IsoStream *extf_get_input_stream(IsoStream *stream, int flag)
     return data->orig;
 }
 
+static
+int extf_clone_stream(IsoStream *old_stream, IsoStream **new_stream, int flag)
+{
+    int ret;
+    IsoStream *new_input_stream, *stream;
+    ExternalFilterStreamData *stream_data, *old_stream_data;
+    
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    stream_data = calloc(1, sizeof(ExternalFilterStreamData));
+    if (stream_data == NULL)
+        return ISO_OUT_OF_MEM;
+    ret = iso_stream_clone_filter_common(old_stream, &stream,
+                                         &new_input_stream, 0);
+    if (ret < 0) {
+        free((char *) stream_data);
+        return ret;
+    }
+    old_stream_data = (ExternalFilterStreamData *) old_stream->data;
+    stream_data->id = ++extf_ino_id;
+    stream_data->orig = new_input_stream;
+    stream_data->cmd = old_stream_data->cmd;
+    stream_data->cmd->refcount++;
+    stream_data->size = old_stream_data->size;
+    stream_data->running = NULL;
+    stream->data = stream_data;
+    *new_stream = stream;
+    return ISO_SUCCESS;
+}
+
+static
+int extf_cmp_ino(IsoStream *s1, IsoStream *s2);
+/* Function is defined after definition of extf_stream_class */
+
 
 IsoStreamIface extf_stream_class = {
-    2,
+    4,
     "extf",
     extf_stream_open,
     extf_stream_close,
@@ -605,10 +646,29 @@ IsoStreamIface extf_stream_class = {
     extf_stream_get_id,
     extf_stream_free,
     extf_update_size,
-    extf_get_input_stream
+    extf_get_input_stream,
+    extf_cmp_ino,
+    extf_clone_stream
 };
 
 
+static
+int extf_cmp_ino(IsoStream *s1, IsoStream *s2)
+{
+    ExternalFilterStreamData *data1, *data2;
+
+    if (s1->class != &extf_stream_class || s2->class != &extf_stream_class)
+        return iso_stream_cmp_ino(s1, s2, 1);
+    data1 = (ExternalFilterStreamData*) s1->data;
+    data2 = (ExternalFilterStreamData*) s2->data;
+    if (data1->cmd != data2->cmd)
+        return (data1->cmd < data2->cmd ? -1 : 1);
+    return iso_stream_cmp_ino(data1->orig, data2->orig, 0);
+}
+
+
+/* ------------------------------------------------------------------------- */
+
 static
 void extf_filter_free(FilterContext *filter)
 {

libisofs/filters/gzip.c

@@ -1,9 +1,10 @@
 /*
- * Copyright (c) 2009 Thomas Schmitt
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  *
  * It implements a filter facility which can pipe a IsoStream into gzip
  * compression resp. uncompression, read its output and forward it as IsoStream
@@ -16,10 +17,15 @@
  *
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "../libisofs.h"
 #include "../filter.h"
 #include "../fsource.h"
 #include "../util.h"
+#include "../stream.h"
 
 #include <sys/types.h>
 #include <sys/time.h>
@@ -70,6 +76,7 @@ typedef struct
 
 } GzipFilterRuntime;
 
+#ifdef Libisofs_with_zliB
 
 static
 int gzip_running_destroy(GzipFilterRuntime **running, int flag)
@@ -96,9 +103,7 @@ int gzip_running_new(GzipFilterRuntime **running, int flag)
     if (o == NULL) {
         return ISO_OUT_OF_MEM;
     }
-#ifdef Libisofs_with_zliB
     memset(&(o->strm), 0, sizeof(o->strm));
-#endif
     o->in_buffer = NULL;
     o->out_buffer = NULL;
     o->in_buffer_size = 0;
@@ -106,9 +111,7 @@ int gzip_running_new(GzipFilterRuntime **running, int flag)
     o->rpt = NULL;
     o->in_counter = 0;
     o->out_counter = 0;
-#ifdef Libisofs_with_zliB
     o->do_flush = Z_NO_FLUSH;
-#endif
     o->error_ret = 1;
 
     o->in_buffer_size= 2048;
@@ -123,6 +126,7 @@ failed:
     gzip_running_destroy(running, 0);
     return -1;
 }
+#endif /* Libisofs_with_zliB */
 
 
 /* ---------------------------- GzipFilterStreamData --------------------- */
@@ -150,6 +154,7 @@ static int gzip_compression_level = 6;
 /*
  * The data payload of an individual Gzip Filter IsoStream
  */
+/* IMPORTANT: Any change must be reflected by gzip_clone_stream() */
 typedef struct
 {
     IsoStream *orig;
@@ -164,12 +169,17 @@ typedef struct
 
 
 
+#ifdef Libisofs_with_zliB
+
 /* Each individual GzipFilterStreamData needs a unique id number. */
 /* >>> This is very suboptimal:
        The counter can rollover.
 */
 static ino_t gzip_ino_id = 0;
 
+#endif /* Libisofs_with_zliB */
+
+
 static
 int gzip_stream_uncompress(IsoStream *stream, void *buf, size_t desired);
 
@@ -521,8 +531,41 @@ IsoStream *gzip_get_input_stream(IsoStream *stream, int flag)
 }
 
 
+static
+int gzip_clone_stream(IsoStream *old_stream, IsoStream **new_stream, int flag)
+{
+    int ret;
+    IsoStream *new_input_stream, *stream;
+    GzipFilterStreamData *stream_data, *old_stream_data;
+
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    stream_data = calloc(1, sizeof(GzipFilterStreamData));
+    if (stream_data == NULL)
+        return ISO_OUT_OF_MEM;
+    ret = iso_stream_clone_filter_common(old_stream, &stream,
+                                         &new_input_stream, 0);
+    if (ret < 0) {
+        free((char *) stream_data);
+        return ret;
+    }
+    old_stream_data = (GzipFilterStreamData *) old_stream->data;
+    stream_data->orig = new_input_stream;
+    stream_data->size = old_stream_data->size;
+    stream_data->running = NULL;
+    stream_data->id = ++gzip_ino_id;
+    stream->data = stream_data;
+    *new_stream = stream;
+    return ISO_SUCCESS;
+}
+
+static
+int gzip_cmp_ino(IsoStream *s1, IsoStream *s2);
+
+
 IsoStreamIface gzip_stream_compress_class = {
-    2,
+    4,
     "gzip",
     gzip_stream_open,
     gzip_stream_close,
@@ -532,12 +575,14 @@ IsoStreamIface gzip_stream_compress_class = {
     gzip_stream_get_id,
     gzip_stream_free,
     gzip_update_size,
-    gzip_get_input_stream
+    gzip_get_input_stream,
+    gzip_cmp_ino,
+    gzip_clone_stream
 };
 
 
 IsoStreamIface gzip_stream_uncompress_class = {
-    2,
+    4,
     "pizg",
     gzip_stream_open,
     gzip_stream_close,
@@ -547,9 +592,27 @@ IsoStreamIface gzip_stream_uncompress_class = {
     gzip_stream_get_id,
     gzip_stream_free,
     gzip_update_size,
-    gzip_get_input_stream
+    gzip_get_input_stream,
+    gzip_cmp_ino,
+    gzip_clone_stream
 };
 
+    
+static
+int gzip_cmp_ino(IsoStream *s1, IsoStream *s2)
+{
+    if (s1->class != s2->class || (s1->class != &gzip_stream_compress_class &&
+                                   s2->class != &gzip_stream_compress_class))
+        return iso_stream_cmp_ino(s1, s2, 1);
+    return iso_stream_cmp_ino(iso_stream_get_input_stream(s1, 0),
+                              iso_stream_get_input_stream(s2, 0), 0);
+}   
+
+
+/* ------------------------------------------------------------------------- */
+
+
+#ifdef Libisofs_with_zliB
 
 static
 void gzip_filter_free(FilterContext *filter)
@@ -557,7 +620,6 @@ void gzip_filter_free(FilterContext *filter)
     /* no data are allocated */;
 }
 
-
 /*
  * @param flag bit1= Install a decompression filter
  */
@@ -606,7 +668,6 @@ int gzip_filter_get_filter(FilterContext *filter, IsoStream *original,
     return ISO_SUCCESS;
 }
 
-
 /* To be called by iso_file_add_filter().
  * The FilterContext input parameter is not furtherly needed for the 
  * emerging IsoStream.
@@ -652,6 +713,7 @@ int gzip_create_context(FilterContext **filter, int flag)
     return ISO_SUCCESS;
 }
 
+#endif /* Libisofs_with_zliB */
 
 /*
  * @param flag bit0= if_block_reduction rather than if_reduction

libisofs/filters/xor_encrypt.c

@@ -1,187 +0,0 @@
-/*
- * Copyright (c) 2008 Vreixo Formoso
- * 
- * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
- * published by the Free Software Foundation. See COPYING file for details.
- */
-
-#include "../libisofs.h"
-#include "../filter.h"
-#include "../fsource.h"
-
-/*
- * A simple Filter implementation for example purposes. It encrypts a file
- * by XORing each byte by a given key.
- */
-
-static ino_t xor_ino_id = 0;
-
-typedef struct
-{
-    IsoStream *orig;
-    uint8_t key;
-    ino_t id;
-} XorEncryptStreamData;
-
-static
-int xor_encrypt_stream_open(IsoStream *stream)
-{
-    XorEncryptStreamData *data;
-    if (stream == NULL) {
-        return ISO_NULL_POINTER;
-    }
-    data = (XorEncryptStreamData*)stream->data;
-    return iso_stream_open(data->orig);
-}
-
-static
-int xor_encrypt_stream_close(IsoStream *stream)
-{
-    XorEncryptStreamData *data;
-    if (stream == NULL) {
-        return ISO_NULL_POINTER;
-    }
-    data = stream->data;
-    return iso_stream_close(data->orig);
-}
-
-static
-off_t xor_encrypt_stream_get_size(IsoStream *stream)
-{
-    XorEncryptStreamData *data;
-    if (stream == NULL) {
-        return ISO_NULL_POINTER;
-    }
-    data = stream->data;
-    return iso_stream_get_size(data->orig);
-}
-
-static
-int xor_encrypt_stream_read(IsoStream *stream, void *buf, size_t count)
-{
-    int ret, len;
-    XorEncryptStreamData *data;
-    uint8_t *buffer = buf;
-
-    if (stream == NULL) {
-        return ISO_NULL_POINTER;
-    }
-    data = stream->data;
-    ret = iso_stream_read(data->orig, buf, count);
-    if (ret < 0) {
-        return ret;
-    }
-    
-    /* xor */
-    for (len = 0; len < ret; ++len) {
-        buffer[len] = buffer[len] ^ data->key;
-    }
-    
-    return ret;
-}
-
-static
-int xor_encrypt_stream_is_repeatable(IsoStream *stream)
-{
-    /* the filter can't be created if underlying stream is not repeatable */
-    return 1;
-}
-
-static
-void xor_encrypt_stream_get_id(IsoStream *stream, unsigned int *fs_id, 
-                               dev_t *dev_id, ino_t *ino_id)
-{
-    XorEncryptStreamData *data = stream->data;
-    *fs_id = ISO_FILTER_FS_ID;
-    *dev_id = XOR_ENCRYPT_DEV_ID;
-    *ino_id = data->id;
-}
-
-static
-void xor_encrypt_stream_free(IsoStream *stream)
-{
-    XorEncryptStreamData *data = stream->data;
-    iso_stream_unref(data->orig);
-    free(data);
-}
-
-IsoStreamIface xor_encrypt_stream_class = {
-    0,
-    "xorf",
-    xor_encrypt_stream_open,
-    xor_encrypt_stream_close,
-    xor_encrypt_stream_get_size,
-    xor_encrypt_stream_read,
-    xor_encrypt_stream_is_repeatable,
-    xor_encrypt_stream_get_id,
-    xor_encrypt_stream_free
-};
-
-
-static
-void xor_encrypt_filter_free(FilterContext *filter)
-{
-    free(filter->data);
-}
-
-static
-int xor_encrypt_filter_get_filter(FilterContext *filter, IsoStream *original, 
-                  IsoStream **filtered)
-{
-    IsoStream *str;
-    XorEncryptStreamData *data;
-
-    if (filter == NULL || original == NULL || filtered == NULL) {
-        return ISO_NULL_POINTER;
-    }
-
-    str = malloc(sizeof(IsoStream));
-    if (str == NULL) {
-        return ISO_OUT_OF_MEM;
-    }
-    data = malloc(sizeof(XorEncryptStreamData));
-    if (str == NULL) {
-        free(str);
-        return ISO_OUT_OF_MEM;
-    }
-
-    /* fill data */
-    data->key = *((uint8_t*)filter->data);
-    data->id = xor_ino_id++;
-
-    /* get reference to the source */
-    data->orig = original;
-    iso_stream_ref(original);
-    
-    str->refcount = 1;
-    str->data = data;
-    str->class = &xor_encrypt_stream_class;
-
-    *filtered = str;
-    return ISO_SUCCESS;
-}
-
-int create_xor_encrypt_filter(uint8_t key, FilterContext **filter)
-{
-    FilterContext *f;
-    uint8_t *data;
-    
-    f = calloc(1, sizeof(FilterContext));
-    if (f == NULL) {
-        return ISO_OUT_OF_MEM;
-    }
-    data = malloc(sizeof(uint8_t));
-    if (data == NULL) {
-        free(f);
-        return ISO_OUT_OF_MEM;
-    }
-    f->refcount = 1;
-    f->version = 0;
-    *data = key;
-    f->data = data;
-    f->free = xor_encrypt_filter_free;
-    f->get_filter = xor_encrypt_filter_get_filter;
-    
-    return ISO_SUCCESS;
-}

libisofs/filters/zisofs.c

@@ -1,9 +1,10 @@
 /*
- * Copyright (c) 2009 Thomas Schmitt
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  *
  * It implements a filter facility which can pipe a IsoStream into zisofs
  * compression resp. uncompression, read its output and forward it as IsoStream
@@ -13,10 +14,15 @@
  *
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "../libisofs.h"
 #include "../filter.h"
 #include "../fsource.h"
 #include "../util.h"
+#include "../stream.h"
 
 #include <sys/types.h>
 #include <sys/time.h>
@@ -162,6 +168,7 @@ static int ziso_compression_level = 6;
 
 /*
  * The common data payload of an individual Zisofs Filter IsoStream
+ * IMPORTANT: Any change must be reflected by ziso_clone_stream().
  */
 typedef struct
 {
@@ -178,6 +185,7 @@ typedef struct
 
 /*
  * The data payload of an individual Zisofs Filter Compressor IsoStream
+ * IMPORTANT: Any change must be reflected by ziso_clone_stream().
  */
 typedef struct
 {   
@@ -193,6 +201,7 @@ typedef struct
 
 /*
  * The data payload of an individual Zisofs Filter Uncompressor IsoStream
+ * IMPORTANT: Any change must be reflected by ziso_clone_stream().
  */
 typedef struct
 {
@@ -774,9 +783,63 @@ IsoStream *ziso_get_input_stream(IsoStream *stream, int flag)
     return data->orig;
 }
 
+static
+int ziso_clone_stream(IsoStream *old_stream, IsoStream **new_stream, int flag)
+{
+    int ret;
+    IsoStream *new_input_stream = NULL, *stream = NULL;
+    ZisofsFilterStreamData *stream_data, *old_stream_data;
+    ZisofsUncomprStreamData *uncompr, *old_uncompr;
+    ZisofsComprStreamData *compr, *old_compr;
+
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    ret = iso_stream_clone_filter_common(old_stream, &stream,
+                                         &new_input_stream, 0);
+    if (ret < 0)
+        return ret;
+
+    if (old_stream->class->read == &ziso_stream_uncompress) {
+        uncompr = calloc(1, sizeof(ZisofsUncomprStreamData));
+        if (uncompr == NULL)
+            goto no_mem;
+        stream_data = (ZisofsFilterStreamData *) uncompr;
+        old_uncompr = (ZisofsUncomprStreamData *) old_stream->data;
+        uncompr->header_size_div4 = old_uncompr->header_size_div4;
+        uncompr->block_size_log2 = old_uncompr->block_size_log2;
+    } else {
+        compr = calloc(1, sizeof(ZisofsComprStreamData));
+        if (compr == NULL)
+            goto no_mem;
+        stream_data = (ZisofsFilterStreamData *) compr;
+        old_compr = (ZisofsComprStreamData *) old_stream->data;
+        compr->orig_size = old_compr->orig_size;
+        compr->block_pointers = NULL;
+    }
+    old_stream_data = (ZisofsFilterStreamData *) old_stream->data;
+    stream_data->orig = new_input_stream;
+    stream_data->size = old_stream_data->size;
+    stream_data->running = NULL;
+    stream_data->id = ++ziso_ino_id;
+    stream->data = stream_data;
+    *new_stream = stream;
+    return ISO_SUCCESS;
+no_mem:
+    if (new_input_stream != NULL)
+        iso_stream_unref(new_input_stream);
+    if (stream != NULL)
+        iso_stream_unref(stream);
+    return ISO_OUT_OF_MEM;
+}
+
+
+static
+int ziso_cmp_ino(IsoStream *s1, IsoStream *s2);
+
 
 IsoStreamIface ziso_stream_compress_class = {
-    2,
+    4,
     "ziso",
     ziso_stream_open,
     ziso_stream_close,
@@ -786,12 +849,14 @@ IsoStreamIface ziso_stream_compress_class = {
     ziso_stream_get_id,
     ziso_stream_free,
     ziso_update_size,
-    ziso_get_input_stream
+    ziso_get_input_stream,
+    ziso_cmp_ino,
+    ziso_clone_stream
 };
 
 
 IsoStreamIface ziso_stream_uncompress_class = {
-    2,
+    4,
     "osiz",
     ziso_stream_open,
     ziso_stream_close,
@@ -801,10 +866,29 @@ IsoStreamIface ziso_stream_uncompress_class = {
     ziso_stream_get_id,
     ziso_stream_free,
     ziso_update_size,
-    ziso_get_input_stream
+    ziso_get_input_stream,
+    ziso_cmp_ino,
+    ziso_clone_stream
 };
 
 
+static
+int ziso_cmp_ino(IsoStream *s1, IsoStream *s2)
+{
+    if (s1->class != s2->class || (s1->class != &ziso_stream_compress_class &&
+                                   s2->class != &ziso_stream_uncompress_class))
+        iso_stream_cmp_ino(s1, s2, 1);
+    return iso_stream_cmp_ino(iso_stream_get_input_stream(s1, 0),
+                              iso_stream_get_input_stream(s2, 0), 0);
+}
+
+
+/* ------------------------------------------------------------------------- */
+
+
+
+#ifdef Libisofs_with_zliB
+
 static
 void ziso_filter_free(FilterContext *filter)
 {
@@ -918,6 +1002,7 @@ int ziso_create_context(FilterContext **filter, int flag)
     return ISO_SUCCESS;
 }
 
+#endif /* Libisofs_with_zliB */
 
 /*
  * @param flag bit0= if_block_reduction rather than if_reduction

libisofs/find.c

@@ -2,10 +2,15 @@
  * Copyright (c) 2008 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "node.h"
 

libisofs/fs_local.c

@@ -3,14 +3,19 @@
  * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /*
  * Filesystem/FileSource implementation to access the local filesystem.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "fsource.h"
 #include "util.h"
 #include "aaip_0_2.h"
@@ -34,7 +39,7 @@ int iso_file_source_new_lfs(IsoFileSource *parent, const char *name,
  */
 IsoFilesystem *lfs= NULL;
 
-
+/* IMPORTANT: Any change must be reflected by lfs_clone_src() */
 typedef struct
 {
     /** reference to the parent (if root it points to itself) */
@@ -57,9 +62,19 @@ char* lfs_get_path(IsoFileSource *src)
     if (data->parent == src) {
         return strdup("/");
     } else {
-        char *path = lfs_get_path(data->parent);
+        char *path, *new_path;
-        int pathlen = strlen(path);
+        int pathlen;
-        path = realloc(path, pathlen + strlen(data->name) + 2);
+
+        path = lfs_get_path(data->parent);
+        if (path == NULL)
+            return NULL;
+        pathlen = strlen(path);
+        new_path = realloc(path, pathlen + strlen(data->name) + 2);
+        if (new_path == NULL) {
+            free(path);
+            return NULL;
+        }
+        path= new_path;
         if (pathlen != 1) {
             /* pathlen can only be 1 for root */
             path[pathlen] = '/';
@@ -519,10 +534,56 @@ ex:;
     return ret;
 }
 
+static
+int lfs_clone_src(IsoFileSource *old_source,
+                  IsoFileSource **new_source, int flag)
+{
+    IsoFileSource *src = NULL;
+    char *new_name = NULL;
+    _LocalFsFileSource *old_data, *new_data = NULL;
+
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    old_data = (_LocalFsFileSource *) old_source->data;
+    *new_source = NULL;  
+    src = calloc(1, sizeof(IsoFileSource));
+    if (src == NULL)
+        goto no_mem;
+    new_name = strdup(old_data->name);
+    if (new_name == NULL) 
+        goto no_mem;
+
+    new_data = calloc(1, sizeof(_LocalFsFileSource));
+    if (new_data == NULL)
+        goto no_mem;
+    new_data->openned = 0;
+    new_data->info.fd = -1; /* the value does not matter with (openned == 0) */
+    new_data->name = new_name;
+    new_data->parent = old_data->parent;
+
+    src->class = old_source->class;
+    src->refcount = 1;
+    src->data = new_data;
+    *new_source = src;
+
+    iso_file_source_ref(new_data->parent);
+    iso_filesystem_ref(lfs);
+    return ISO_SUCCESS;
+no_mem:;
+    if (src != NULL)
+        free((char *) src);
+    if (new_data != NULL)
+        free((char *) new_data);
+    if (new_name != NULL)
+        free(new_name);
+    return ISO_OUT_OF_MEM;
+}
+
 
 IsoFileSourceIface lfs_class = { 
 
-    1, /* version */
+    2, /* version */
     lfs_get_path,
     lfs_get_name,
     lfs_lstat,
@@ -536,7 +597,8 @@ IsoFileSourceIface lfs_class = {
     lfs_get_filesystem,
     lfs_free,
     lfs_lseek,
-    lfs_get_aa_string
+    lfs_get_aa_string,
+    lfs_clone_src
 
 };
 

libisofs/fsource.c

@@ -2,10 +2,15 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "fsource.h"
 #include <stdlib.h>
 

libisofs/fsource.h

@@ -1,9 +1,11 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #ifndef LIBISO_FSOURCE_H_
@@ -30,4 +32,17 @@
  */
 int iso_local_filesystem_new(IsoFilesystem **fs);
 
+
+/* Rank two IsoFileSource of ifs_class by their eventual old image LBAs.
+   Other IsoFileSource classes will be ranked only roughly.
+*/
+int iso_ifs_sections_cmp(IsoFileSource *s1, IsoFileSource *s2, int flag);
+
+
+/* Create an independent copy of an ifs_class IsoFileSource.
+*/
+int iso_ifs_source_clone(IsoFileSource *old_source, IsoFileSource **new_source,
+                         int flag);
+
+
 #endif /*LIBISO_FSOURCE_H_*/

libisofs/image.c

@@ -1,11 +1,17 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "image.h"
 #include "node.h"
@@ -14,6 +20,7 @@
 
 #include <stdlib.h>
 #include <string.h>
+#include <stdio.h>
 
 /**
  * Create a new image, empty.
@@ -29,7 +36,7 @@
  */
 int iso_image_new(const char *name, IsoImage **image)
 {
-    int res;
+    int res, i;
     IsoImage *img;
 
     if (image == NULL) {
@@ -71,8 +78,21 @@ int iso_image_new(const char *name, IsoImage **image)
         img->volset_id = strdup(name);
         img->volume_id = strdup(name);
     }
+    img->system_area_data = NULL;
+    img->system_area_options = 0;
+    img->num_mips_boot_files = 0;
+    for (i = 0; i < 15; i++)
+         img->mips_boot_file_paths[i] = NULL;
     img->builder_ignore_acl = 1;
     img->builder_ignore_ea = 1;
+    img->inode_counter = 0;
+    img->used_inodes = NULL;
+    img->used_inodes_start = 0;
+    img->checksum_start_lba = 0;
+    img->checksum_end_lba = 0;
+    img->checksum_idx_count = 0;
+    img->checksum_array = NULL;
+    img->generator_is_running = 0;
     *image = img;
     return ISO_SUCCESS;
 }
@@ -86,7 +106,7 @@ void iso_image_ref(IsoImage *image)
 }
 
 /**
- * Decrements the reference couting of the given image.
+ * Decrements the reference counting of the given image.
  * If it reaches 0, the image is free, together with its tree nodes (whether
  * their refcount reach 0 too, of course).
  */
@@ -110,6 +130,7 @@ void iso_image_unref(IsoImage *image)
         iso_node_builder_unref(image->builder);
         iso_filesystem_unref(image->fs);
         el_torito_boot_catalog_free(image->bootcat);
+        iso_image_give_up_mips_boot(image, 0);
         free(image->volset_id);
         free(image->volume_id);
         free(image->publisher_id);
@@ -119,10 +140,26 @@ void iso_image_unref(IsoImage *image)
         free(image->copyright_file_id);
         free(image->abstract_file_id);
         free(image->biblio_file_id);
+        if (image->used_inodes != NULL)
+            free(image->used_inodes);
+        iso_image_free_checksums(image, 0);
         free(image);
     }
 }
 
+
+int iso_image_free_checksums(IsoImage *image, int flag)
+{
+    image->checksum_start_lba = 0;
+    image->checksum_end_lba = 0;
+    image->checksum_idx_count = 0;
+    if (image->checksum_array != NULL)
+        free(image->checksum_array);
+    image->checksum_array = NULL;
+    return 1;
+}
+
+
 /**
  * Attach user defined data to the image. Use this if your application needs
  * to store addition info together with the IsoImage. If the image already
@@ -138,13 +175,13 @@ void iso_image_unref(IsoImage *image)
  */
 int iso_image_attach_data(IsoImage *image, void *data, void (*give_up)(void*))
 {
-    if (image == NULL || (data != NULL && free == NULL)) {
+    if (image == NULL) {
         return ISO_NULL_POINTER;
     }
 
     if (image->user_data != NULL) {
         /* free previously attached data */
-        if (image->user_data_free) {
+        if (image->user_data_free != NULL) {
             image->user_data_free(image->user_data);
         }
         image->user_data = NULL;
@@ -179,6 +216,8 @@ void iso_image_set_volset_id(IsoImage *image, const char *volset_id)
 
 const char *iso_image_get_volset_id(const IsoImage *image)
 {
+    if (image->volset_id == NULL)
+        return "";
     return image->volset_id;
 }
 
@@ -190,6 +229,8 @@ void iso_image_set_volume_id(IsoImage *image, const char *volume_id)
 
 const char *iso_image_get_volume_id(const IsoImage *image)
 {
+    if (image->volume_id == NULL)
+        return "";
     return image->volume_id;
 }
 
@@ -201,6 +242,8 @@ void iso_image_set_publisher_id(IsoImage *image, const char *publisher_id)
 
 const char *iso_image_get_publisher_id(const IsoImage *image)
 {
+    if (image->publisher_id == NULL)
+        return "";
     return image->publisher_id;
 }
 
@@ -213,6 +256,8 @@ void iso_image_set_data_preparer_id(IsoImage *image,
 
 const char *iso_image_get_data_preparer_id(const IsoImage *image)
 {
+    if (image->data_preparer_id == NULL)
+        return "";
     return image->data_preparer_id;
 }
 
@@ -224,6 +269,8 @@ void iso_image_set_system_id(IsoImage *image, const char *system_id)
 
 const char *iso_image_get_system_id(const IsoImage *image)
 {
+    if (image->system_id == NULL)
+        return "";
     return image->system_id;
 }
 
@@ -235,6 +282,8 @@ void iso_image_set_application_id(IsoImage *image, const char *application_id)
 
 const char *iso_image_get_application_id(const IsoImage *image)
 {
+    if (image->application_id == NULL)
+        return "";
     return image->application_id;
 }
 
@@ -247,6 +296,8 @@ void iso_image_set_copyright_file_id(IsoImage *image,
 
 const char *iso_image_get_copyright_file_id(const IsoImage *image)
 {
+    if (image->copyright_file_id == NULL)
+        return "";
     return image->copyright_file_id;
 }
 
@@ -259,6 +310,8 @@ void iso_image_set_abstract_file_id(IsoImage *image,
 
 const char *iso_image_get_abstract_file_id(const IsoImage *image)
 {
+    if (image->abstract_file_id == NULL)
+        return "";
     return image->abstract_file_id;
 }
 
@@ -270,6 +323,8 @@ void iso_image_set_biblio_file_id(IsoImage *image, const char *biblio_file_id)
 
 const char *iso_image_get_biblio_file_id(const IsoImage *image)
 {
+    if (image->biblio_file_id == NULL)
+        return "";
     return image->biblio_file_id;
 }
 
@@ -278,6 +333,16 @@ int iso_image_get_msg_id(IsoImage *image)
     return image->id;
 }
 
+int iso_image_get_system_area(IsoImage *img, char system_area_data[32768],
+                              int *options, int flag)
+{
+    *options = img->system_area_options;
+    if (img->system_area_data == NULL)
+        return 0;
+    memcpy(system_area_data, img->system_area_data, 32768);
+    return 1;
+}
+
 static
 int dir_update_size(IsoImage *image, IsoDir *dir)
 {
@@ -318,3 +383,275 @@ void iso_image_set_ignore_aclea(IsoImage *image, int what)
     image->builder_ignore_ea = !!(what & 2);
 }
 
+
+static
+int img_register_ino(IsoImage *image, IsoNode *node, int flag)
+{
+    int ret;
+    ino_t ino;
+    unsigned int fs_id;
+    dev_t dev_id;
+
+    ret = iso_node_get_id(node, &fs_id, &dev_id, &ino, 1);
+    if (ret < 0)
+       return ret;
+    if (ret > 0 && ino >= image->used_inodes_start &&
+        ino <= image->used_inodes_start + (ISO_USED_INODE_RANGE - 1)) {
+                                   /* without -1 : rollover hazard on 32 bit */
+        image->used_inodes[(ino - image->used_inodes_start) / 8]
+                                                           |= (1 << (ino % 8));
+    }
+    return 1;
+}
+
+
+/* Collect the bitmap of used inode numbers in the range of
+   _ImageFsData.used_inodes_start + ISO_USED_INODE_RANGE
+   @param flag bit0= recursion is active
+*/
+int img_collect_inos(IsoImage *image, IsoDir *dir, int flag)
+{
+    int ret, register_dir = 1;
+    IsoDirIter *iter = NULL;
+    IsoNode *node;
+    IsoDir *subdir;
+
+    if (dir == NULL)
+        dir = image->root;
+    if (image->used_inodes == NULL) {
+        image->used_inodes = calloc(ISO_USED_INODE_RANGE / 8, 1);
+        if (image->used_inodes == NULL)
+            return ISO_OUT_OF_MEM;
+    } else if(!(flag & 1)) {
+        memset(image->used_inodes, 0, ISO_USED_INODE_RANGE / 8);
+    } else {
+        register_dir = 0;
+    }
+    if (register_dir) {
+        node = (IsoNode *) dir;
+        ret = img_register_ino(image, node, 0);
+        if (ret < 0)
+            return ret;
+    }
+
+    ret = iso_dir_get_children(dir, &iter);
+    if (ret < 0)
+        return ret;
+    while (iso_dir_iter_next(iter, &node) == 1 ) {
+        ret = img_register_ino(image, node, 0);
+        if (ret < 0)
+            goto ex;
+        if (iso_node_get_type(node) == LIBISO_DIR) {
+            subdir = (IsoDir *) node;
+            ret = img_collect_inos(image, subdir, flag | 1);
+            if (ret < 0)
+                goto ex;
+        }
+    }
+    ret = 1;
+ex:;
+    if (iter != NULL)
+        iso_dir_iter_free(iter);
+    return ret;
+}
+
+
+/**
+ * A global counter for inode numbers for the ISO image filesystem.
+ * On image import it gets maxed by the eventual inode numbers from PX
+ * entries. Up to the first 32 bit rollover it simply increments the counter.
+ * After the first rollover it uses a look ahead bitmap which gets filled
+ * by a full tree traversal. It covers the next inode numbers to come
+ * (somewhere between 1 and ISO_USED_INODE_RANGE which is quite many)
+ * and advances when being exhausted.
+ * @param image The image where the number shall be used
+ * @param flag  bit0= reset count (Caution: image must get new inos then)
+ * @return
+ *     Since ino_t 0 is used as default and considered self-unique,
+ *     the value 0 should only be returned in case of error.
+ */
+ino_t img_give_ino_number(IsoImage *image, int flag)
+{
+    int ret;
+    ino_t new_ino, ino_idx;
+    static uint64_t limit = 0xffffffff;
+
+    if (flag & 1) {
+        image->inode_counter = 0;
+        if (image->used_inodes != NULL)
+            free(image->used_inodes);
+        image->used_inodes = NULL;
+        image->used_inodes_start = 0;
+    }
+    new_ino = image->inode_counter + 1;
+    if (image->used_inodes == NULL) {
+        if (new_ino > 0 && new_ino <= limit) {
+            image->inode_counter = new_ino;
+            return image->inode_counter;
+        }
+    }
+    /* Look for free number in used territory */
+    while (1) {
+        if (new_ino <= 0 || new_ino > limit ||
+            new_ino >= image->used_inodes_start + ISO_USED_INODE_RANGE ) {
+
+            /* Collect a bitmap of used inode numbers ahead */
+
+            image->used_inodes_start += ISO_USED_INODE_RANGE;
+            if (image->used_inodes_start > 0xffffffff ||
+                image->used_inodes_start <= 0) 
+                image->used_inodes_start = 0;
+            ret = img_collect_inos(image, NULL, 0);
+            if (ret < 0)
+                goto return_result; /* >>> need error return value */
+
+            new_ino = image->used_inodes_start + !image->used_inodes_start;
+        }
+        ino_idx = (new_ino - image->used_inodes_start) / 8;
+        if (!(image->used_inodes[ino_idx] & (1 << (new_ino % 8)))) {
+            image->used_inodes[ino_idx] |= (1 << (new_ino % 8));
+    break;
+        }
+        new_ino++;
+    }
+return_result:;
+    image->inode_counter = new_ino;
+    return image->inode_counter;
+}
+
+
+/* @param flag bit0= overwrite any ino, else only ino == 0
+               bit1= install inode with non-data, non-directory files
+               bit2= install inode with directories
+*/
+static
+int img_update_ino(IsoImage *image, IsoNode *node, int flag)
+{
+    int ret;
+    ino_t ino;
+    unsigned int fs_id;
+    dev_t dev_id;
+
+    ret = iso_node_get_id(node, &fs_id, &dev_id, &ino, 1);
+    if (ret < 0)
+        return ret;
+    if (ret == 0)
+       ino = 0;
+    if (((flag & 1) || ino == 0) &&
+        (iso_node_get_type(node) == LIBISO_FILE || (flag & (2 | 4))) &&
+        ((flag & 4) || iso_node_get_type(node) != LIBISO_DIR)) {
+        ret = iso_node_set_unique_id(node, image, 0);
+        if (ret < 0)
+            return ret;
+    }
+    return 1;
+}
+
+
+/* @param flag bit0= overwrite any ino, else only ino == 0
+               bit1= install inode with non-data, non-directory files
+               bit2= install inode with directories
+               bit3= with bit2: install inode on parameter dir
+*/
+int img_make_inos(IsoImage *image, IsoDir *dir, int flag)
+{
+    int ret;
+    IsoDirIter *iter = NULL;
+    IsoNode *node;
+    IsoDir *subdir;
+
+    if (flag & 8) {
+        node = (IsoNode *) dir;
+        ret = img_update_ino(image, node, flag & 7);
+        if (ret < 0)
+            goto ex;
+    }
+    ret = iso_dir_get_children(dir, &iter);
+    if (ret < 0)
+        return ret;
+    while (iso_dir_iter_next(iter, &node) == 1) {
+        ret = img_update_ino(image, node, flag & 7);
+        if (ret < 0)
+            goto ex;
+        if (iso_node_get_type(node) == LIBISO_DIR) {
+            subdir = (IsoDir *) node;
+            ret = img_make_inos(image, subdir, flag & ~8);
+            if (ret < 0)
+                goto ex;
+        }
+    }
+    ret = 1;
+ex:;
+    if (iter != NULL)
+        iso_dir_iter_free(iter);
+    return ret;
+}
+
+
+/* API */
+int iso_image_get_session_md5(IsoImage *image, uint32_t *start_lba,
+                              uint32_t *end_lba, char md5[16], int flag)
+{
+    if (image->checksum_array == NULL || image->checksum_idx_count < 1)
+        return 0;
+    *start_lba = image->checksum_start_lba;
+    *end_lba = image->checksum_end_lba;
+    memcpy(md5, image->checksum_array, 16);
+    return ISO_SUCCESS;
+}
+
+int iso_image_set_checksums(IsoImage *image, char *checksum_array,
+                            uint32_t start_lba, uint32_t end_lba,
+                            uint32_t idx_count, int flag)
+{
+    iso_image_free_checksums(image, 0);
+    image->checksum_array = checksum_array;
+    image->checksum_start_lba = start_lba;
+    image->checksum_end_lba = end_lba;
+    image->checksum_idx_count = idx_count;
+    return 1;
+}
+
+int iso_image_generator_is_running(IsoImage *image)
+{
+    return image->generator_is_running;
+}
+
+
+/* API */
+int iso_image_add_mips_boot_file(IsoImage *image, char *path, int flag)
+{
+    if (image->num_mips_boot_files >= 15)
+        return ISO_BOOT_TOO_MANY_MIPS;
+    image->mips_boot_file_paths[image->num_mips_boot_files] = strdup(path);
+    if (image->mips_boot_file_paths[image->num_mips_boot_files] == NULL)
+        return ISO_OUT_OF_MEM;
+    image->num_mips_boot_files++;
+    return ISO_SUCCESS;
+}
+
+/* API */
+int iso_image_get_mips_boot_files(IsoImage *image, char *paths[15], int flag)
+{
+    int i;
+
+    for (i = 0; i < image->num_mips_boot_files; i++)
+         paths[i] = image->mips_boot_file_paths[i];
+    for (; i < 15; i++)
+         paths[i] = NULL;
+    return image->num_mips_boot_files;
+}
+
+/* API */
+int iso_image_give_up_mips_boot(IsoImage *image, int flag)
+{
+    int i;
+
+    for (i = 0; i < image->num_mips_boot_files; i++)
+        if (image->mips_boot_file_paths[i] != NULL) {
+            free(image->mips_boot_file_paths[i]);
+            image->mips_boot_file_paths[i] = NULL;
+        }
+    image->num_mips_boot_files = 0;
+    return ISO_SUCCESS;
+}

libisofs/image.h

@@ -1,9 +1,11 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_IMAGE_H_
 #define LIBISO_IMAGE_H_
@@ -13,6 +15,14 @@
 #include "fsource.h"
 #include "builder.h"
 
+/* Size of a inode recycling window. Each new window causes a tree traversal.
+   Window memory consumption is ISO_USED_INODE_RANGE / 8.
+   This must be a power of 2 smaller than 30 bit and larger than 8 bit.
+   Here: 32 kB memory for 256k inodes.
+*/
+#define ISO_USED_INODE_RANGE (1 << 18)
+
+
 /*
  * Image is a context for image manipulation.
  * Global objects such as the message_queues must belogn to that
@@ -43,6 +53,19 @@ struct Iso_Image
     /* el-torito boot catalog */
     struct el_torito_boot_catalog *bootcat;
 
+    /* Eventually loaded system area data, or NULL */
+    char *system_area_data;
+    /* Prescribed/detected options, see iso_write_opts_set_system_area() */
+    int system_area_options;
+
+   /*
+    * Up to 15 boot files can be referred by a MIPS Big Endian Volume Header.
+      The mips_boot_file_paths are ISO 9660 Rock Ridge paths.
+    */
+    int num_mips_boot_files;
+    char *mips_boot_file_paths[15]; /* ISO 9660 Rock Ridge Paths */
+
+
     /* image identifier, for message origin identifier */
     int id;
 
@@ -121,6 +144,87 @@ struct Iso_Image
      */
     void *user_data;
     void (*user_data_free)(void *ptr);
+
+    /**
+     * Inode number management. inode_counter is taken over from
+     * IsoImageFilesystem._ImageFsData after image import.
+     * It is to be used with img_give_ino_number()
+     */
+    ino_t inode_counter;
+    /*
+     * A bitmap of used inode numbers in an interval beginning at
+     * used_inodes_start and holding ISO_USED_INODE_RANGE bits.
+     * If a bit is set, then the corresponding inode number is occupied.
+     * This interval is kept around inode_counter and eventually gets
+     * advanced by ISO_USED_INODE_RANGE numbers in a tree traversal
+     * done by img_collect_inos().
+     */
+    uint8_t *used_inodes;
+    ino_t used_inodes_start;
+
+    /**
+     * Array of MD5 checksums as announced by xattr "isofs.ca" of the 
+     * root node. Array element 0 contains an overall image checksum for the
+     * block range checksum_start_lba,checksum_end_lba. Element size is
+     * 16 bytes. IsoFile objects in the image may have xattr "isofs.cx"
+     * which gives their index in checksum_array.
+     */
+    uint32_t checksum_start_lba;
+    uint32_t checksum_end_lba;
+    uint32_t checksum_idx_count;
+    char *checksum_array;
+
+    /**
+     * Whether a write run has been started by iso_image_create_burn_source()
+     * and has not yet been finished.
+     */
+    int generator_is_running;
+
 };
 
+
+    
+/* Collect the bitmap of used inode numbers in the range of
+   _ImageFsData.used_inodes_start + ISO_USED_INODE_RANGE
+   @param flag bit0= recursion is active
+*/
+int img_collect_inos(IsoImage *image, IsoDir *dir, int flag);
+
+/**
+ * A global counter for inode numbers for the ISO image filesystem.
+ * On image import it gets maxed by the eventual inode numbers from PX
+ * entries. Up to the first 32 bit rollover it simply increments the counter.
+ * After the first rollover it uses a look ahead bitmap which gets filled
+ * by a full tree traversal. It covers the next inode numbers to come
+ * (somewhere between 1 and ISO_USED_INODE_RANGE which is quite many)
+ * and advances when being exhausted.
+ * @param image The image where the number shall be used
+ * @param flag  bit0= reset count (Caution: image must get new inos then)
+ * @return
+ *     Since ino_t 0 is used as default and considered self-unique, 
+ *     the value 0 should only be returned in case of error.
+ */ 
+ino_t img_give_ino_number(IsoImage *image, int flag);
+
+/* @param flag bit0= overwrite any ino, else only ino == 0
+               bit1= install inode with non-data, non-directory files
+               bit2= install inode with directories
+               bit3= with bit2: install inode on parameter dir
+*/
+int img_make_inos(IsoImage *image, IsoDir *dir, int flag);
+
+
+/* Free the checksum array of an image and reset its layout parameters
+*/
+int iso_image_free_checksums(IsoImage *image, int flag);
+
+
+/* Equip an ISO image with a new checksum array buffer (after isofs.ca and
+   isofs.cx have already been adjusted).
+*/
+int iso_image_set_checksums(IsoImage *image, char *checksum_array, 
+                            uint32_t start_lba, uint32_t end_lba,
+                            uint32_t idx_count, int flag);
+
+
 #endif /*LIBISO_IMAGE_H_*/

libisofs/iso1999.c

@@ -2,10 +2,15 @@
  * Copyright (c) 2007 Vreixo Formoso
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "iso1999.h"
 #include "messages.h"
 #include "writer.h"
@@ -711,7 +716,7 @@ void write_one_dir_record(Ecma119Image *t, Iso1999Node *node, int file_id,
     iso_bb(rec->length, len, 4);
     iso_datetime_7(rec->recording_time, t->now, t->always_gmt);
     rec->flags[0] = ((node->type == ISO1999_DIR) ? 2 : 0) | (multi_extend ? 0x80 : 0);
-    iso_bb(rec->vol_seq_number, 1, 2);
+    iso_bb(rec->vol_seq_number, (uint32_t) 1, 2);
     rec->len_fi[0] = len_fi;
 }
 
@@ -762,9 +767,9 @@ int iso1999_writer_write_vol_desc(IsoImageWriter *writer)
     strncpy_pad((char*)vol.volume_id, vol_id, 32);
 
     iso_bb(vol.vol_space_size, t->vol_space_size, 4);
-    iso_bb(vol.vol_set_size, 1, 2);
+    iso_bb(vol.vol_set_size, (uint32_t) 1, 2);
-    iso_bb(vol.vol_seq_number, 1, 2);
+    iso_bb(vol.vol_seq_number, (uint32_t) 1, 2);
-    iso_bb(vol.block_size, BLOCK_SIZE, 2);
+    iso_bb(vol.block_size, (uint32_t) BLOCK_SIZE, 2);
     iso_bb(vol.path_table_size, t->iso1999_path_table_size, 4);
     iso_lsb(vol.l_path_table_pos, t->iso1999_l_path_table_pos, 4);
     iso_msb(vol.m_path_table_pos, t->iso1999_m_path_table_pos, 4);
@@ -1023,6 +1028,7 @@ int iso1999_writer_create(Ecma119Image *target)
                   "Creating low level ISO 9660:1999 tree...");
     ret = iso1999_tree_create(target);
     if (ret < 0) {
+        free((char *) writer);
         return ret;
     }
 

libisofs/iso1999.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /**

libisofs/joliet.c

@@ -3,10 +3,15 @@
  * Copyright (c) 2007 Mario Danic
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "joliet.h"
 #include "messages.h"
 #include "writer.h"
@@ -42,7 +47,7 @@ int get_joliet_name(Ecma119Image *t, IsoNode *iso, uint16_t **name)
     if (iso->type == LIBISO_DIR) {
         jname = iso_j_dir_id(ucs_name);
     } else {
-        jname = iso_j_file_id(ucs_name);
+        jname = iso_j_file_id(ucs_name, !!(t->no_force_dots & 2));
     }
     free(ucs_name);
     if (jname != NULL) {
@@ -243,7 +248,7 @@ int create_tree(Ecma119Image *t, IsoNode *iso, JolietNode **tree, int pathlen)
             char *ipath = iso_tree_get_node_path(iso);
             ret = iso_msg_submit(t->image->id, ISO_FILE_IGNORED, 0,
                  "Can't add %s to Joliet tree. %s can only be added to a "
-                 "Rock Ridget tree.", ipath, (iso->type == LIBISO_SYMLINK ?
+                 "Rock Ridge tree.", ipath, (iso->type == LIBISO_SYMLINK ?
                                              "Symlinks" : "Special files"));
             free(ipath);
         }
@@ -551,17 +556,19 @@ int joliet_tree_create(Ecma119Image *t)
     }
 
     /* the Joliet tree is stored in Ecma119Image target */
-    t->joliet_root = root;
+    if (t->eff_partition_offset > 0) {
+        t->j_part_root = root;
+    } else {
+        t->joliet_root = root;
+    }
 
     iso_msg_debug(t->image->id, "Sorting the Joliet tree...");
     sort_tree(root);
 
     iso_msg_debug(t->image->id, "Mangling Joliet names...");
-    ret = mangle_tree(t, t->joliet_root);
+    ret = mangle_tree(t, root);
-    if (ret < 0) {
+    if (ret < 0)
         return ret;
-    }
-
     return ISO_SUCCESS;
 }
 
@@ -573,7 +580,7 @@ size_t calc_dirent_len(Ecma119Image *t, JolietNode *n)
 {
     /* note than name len is always even, so we always need the pad byte */
     int ret = n->name ? ucslen(n->name) * 2 + 34 : 34;
-    if (n->type == JOLIET_FILE && !t->omit_version_numbers) {
+    if (n->type == JOLIET_FILE && !(t->omit_version_numbers & 3)) {
         /* take into account version numbers */
         ret += 4;
     }
@@ -668,6 +675,7 @@ int joliet_writer_compute_data_blocks(IsoImageWriter *writer)
 {
     Ecma119Image *t;
     uint32_t path_table_size;
+    size_t ndirs;
 
     if (writer == NULL) {
         return ISO_OUT_OF_MEM;
@@ -691,6 +699,24 @@ int joliet_writer_compute_data_blocks(IsoImageWriter *writer)
     t->curblock += DIV_UP(path_table_size, BLOCK_SIZE);
     t->joliet_path_table_size = path_table_size;
 
+    if (t->partition_offset > 0) {
+        /* Take into respect second directory tree */
+        ndirs = t->joliet_ndirs;
+        t->joliet_ndirs = 0;
+        calc_dir_pos(t, t->j_part_root);
+        if (t->joliet_ndirs != ndirs) {
+            iso_msg_submit(t->image->id, ISO_ASSERT_FAILURE, 0,
+                    "Number of directories differs in Joliet partiton_tree");
+            return ISO_ASSERT_FAILURE;
+        }
+        /* Take into respect second set of path tables */
+        path_table_size = calc_path_table_size(t->j_part_root);
+        t->j_part_l_path_table_pos = t->curblock;
+        t->curblock += DIV_UP(path_table_size, BLOCK_SIZE);
+        t->j_part_m_path_table_pos = t->curblock;
+        t->curblock += DIV_UP(path_table_size, BLOCK_SIZE);
+    }
+
     return ISO_SUCCESS;
 }
 
@@ -721,7 +747,7 @@ void write_one_dir_record(Ecma119Image *t, JolietNode *node, int file_id,
 
     memcpy(rec->file_id, name, len_fi);
 
-    if (node->type == JOLIET_FILE && !t->omit_version_numbers) {
+    if (node->type == JOLIET_FILE && !(t->omit_version_numbers & 3)) {
         len_dr += 4;
         rec->file_id[len_fi++] = 0;
         rec->file_id[len_fi++] = ';';
@@ -753,11 +779,11 @@ void write_one_dir_record(Ecma119Image *t, JolietNode *node, int file_id,
         node = node->parent;
 
     rec->len_dr[0] = len_dr;
-    iso_bb(rec->block, block, 4);
+    iso_bb(rec->block, block - t->eff_partition_offset, 4);
     iso_bb(rec->length, len, 4);
     iso_datetime_7(rec->recording_time, t->now, t->always_gmt);
     rec->flags[0] = ((node->type == JOLIET_DIR) ? 2 : 0) | (multi_extend ? 0x80 : 0);
-    iso_bb(rec->vol_seq_number, 1, 2);
+    iso_bb(rec->vol_seq_number, (uint32_t) 1, 2);
     rec->len_fi[0] = len_fi;
 }
 
@@ -789,7 +815,6 @@ void ucsncpy_pad(uint16_t *dest, const uint16_t *src, size_t max)
     }
 }
 
-static
 int joliet_writer_write_vol_desc(IsoImageWriter *writer)
 {
     IsoImage *image;
@@ -830,16 +855,25 @@ int joliet_writer_write_vol_desc(IsoImageWriter *writer)
 
     /* make use of UCS-2 Level 3 */
     memcpy(vol.esc_sequences, "%/E", 3);
-
+    iso_bb(vol.vol_space_size, t->vol_space_size  - t->eff_partition_offset,
-    iso_bb(vol.vol_space_size, t->vol_space_size, 4);
+           4);
-    iso_bb(vol.vol_set_size, 1, 2);
+    iso_bb(vol.vol_set_size, (uint32_t) 1, 2);
-    iso_bb(vol.vol_seq_number, 1, 2);
+    iso_bb(vol.vol_seq_number, (uint32_t) 1, 2);
-    iso_bb(vol.block_size, BLOCK_SIZE, 2);
+    iso_bb(vol.block_size, (uint32_t) BLOCK_SIZE, 2);
     iso_bb(vol.path_table_size, t->joliet_path_table_size, 4);
-    iso_lsb(vol.l_path_table_pos, t->joliet_l_path_table_pos, 4);
-    iso_msb(vol.m_path_table_pos, t->joliet_m_path_table_pos, 4);
 
-    write_one_dir_record(t, t->joliet_root, 0, vol.root_dir_record, 1, 0);
+    if (t->eff_partition_offset > 0) {
+        /* Point to second tables and second root */
+        iso_lsb(vol.l_path_table_pos,
+                t->j_part_l_path_table_pos - t->eff_partition_offset, 4);
+        iso_msb(vol.m_path_table_pos,
+                t->j_part_m_path_table_pos - t->eff_partition_offset, 4);
+        write_one_dir_record(t, t->j_part_root, 0, vol.root_dir_record, 1, 0);
+    } else {
+        iso_lsb(vol.l_path_table_pos, t->joliet_l_path_table_pos, 4);
+        iso_msb(vol.m_path_table_pos, t->joliet_m_path_table_pos, 4);
+        write_one_dir_record(t, t->joliet_root, 0, vol.root_dir_record, 1, 0);
+    }
 
     ucsncpy_pad((uint16_t*)vol.vol_set_id, volset_id, 128);
     ucsncpy_pad((uint16_t*)vol.publisher_id, pub_id, 128);
@@ -898,7 +932,7 @@ int write_one_dir(Ecma119Image *t, JolietNode *dir)
         /* compute len of directory entry */
         fi_len = ucslen(child->name) * 2;
         len = fi_len + 34;
-        if (child->type == JOLIET_FILE && !t->omit_version_numbers) {
+        if (child->type == JOLIET_FILE && !(t->omit_version_numbers & 3)) {
             len += 4;
         }
 
@@ -979,7 +1013,8 @@ int write_path_table(Ecma119Image *t, JolietNode **pathlist, int l_type)
         rec = (struct ecma119_path_table_record*) buf;
         rec->len_di[0] = dir->parent ? (uint8_t) ucslen(dir->name) * 2 : 1;
         rec->len_xa[0] = 0;
-        write_int(rec->block, dir->info.dir->block, 4);
+        write_int(rec->block, dir->info.dir->block - t->eff_partition_offset,
+                  4);
         write_int(rec->parent, parent + 1, 2);
         if (dir->parent) {
             memcpy(rec->dir_id, dir->name, rec->len_di[0]);
@@ -1018,7 +1053,12 @@ int write_path_tables(Ecma119Image *t)
     if (pathlist == NULL) {
         return ISO_OUT_OF_MEM;
     }
-    pathlist[0] = t->joliet_root;
+
+    if (t->eff_partition_offset > 0) {
+        pathlist[0] = t->j_part_root;
+    } else {
+        pathlist[0] = t->joliet_root;
+    }
     cur = 1;
 
     for (i = 0; i < t->joliet_ndirs; i++) {
@@ -1046,18 +1086,21 @@ int write_path_tables(Ecma119Image *t)
 }
 
 static
-int joliet_writer_write_data(IsoImageWriter *writer)
+int joliet_writer_write_dirs(IsoImageWriter *writer)
 {
     int ret;
     Ecma119Image *t;
+    JolietNode *root;
 
-    if (writer == NULL) {
-        return ISO_NULL_POINTER;
-    }
     t = writer->target;
 
     /* first of all, we write the directory structure */
-    ret = write_dirs(t, t->joliet_root);
+    if (t->eff_partition_offset > 0) {
+        root = t->j_part_root;
+    } else {
+        root = t->joliet_root;
+    }
+    ret = write_dirs(t, root);
     if (ret < 0) {
         return ret;
     }
@@ -1068,12 +1111,40 @@ int joliet_writer_write_data(IsoImageWriter *writer)
     return ret;
 }
 
+static
+int joliet_writer_write_data(IsoImageWriter *writer)
+{
+    int ret;
+    Ecma119Image *t;
+
+    if (writer == NULL) {
+        return ISO_NULL_POINTER;
+    }
+    t = writer->target;
+
+    ret = joliet_writer_write_dirs(writer);
+    if (ret < 0)
+        return ret;
+
+    if (t->partition_offset > 0) {
+        t->eff_partition_offset = t->partition_offset;
+        ret = joliet_writer_write_dirs(writer);
+        t->eff_partition_offset = 0;
+        if (ret < 0)
+            return ret;
+    }
+    return ISO_SUCCESS;
+}
+
 static
 int joliet_writer_free_data(IsoImageWriter *writer)
 {
     /* free the Joliet tree */
     Ecma119Image *t = writer->target;
     joliet_node_free(t->joliet_root);
+    if (t->j_part_root != NULL)
+        joliet_node_free(t->j_part_root);
+    t->j_part_root = NULL;
     return ISO_SUCCESS;
 }
 
@@ -1097,12 +1168,23 @@ int joliet_writer_create(Ecma119Image *target)
     iso_msg_debug(target->image->id, "Creating low level Joliet tree...");
     ret = joliet_tree_create(target);
     if (ret < 0) {
+        free((char *) writer);
         return ret;
     }
 
     /* add this writer to image */
     target->writers[target->nwriters++] = writer;
 
+    if(target->partition_offset > 0) {
+        /* Create second tree */
+        target->eff_partition_offset = target->partition_offset;
+        ret = joliet_tree_create(target);
+        if (ret < 0) {
+            return ret;
+        }
+        target->eff_partition_offset = 0;
+    }
+
     /* we need the volume descriptor */
     target->curblock++;
     return ISO_SUCCESS;

libisofs/joliet.h

@@ -3,8 +3,9 @@
  * Copyright (c) 2007 Mario Danic
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /**
@@ -53,4 +54,10 @@ struct joliet_node
  */
 int joliet_writer_create(Ecma119Image *target);
 
+
+/* Not to be called but only for comparison with target->writers[i]
+*/
+int joliet_writer_write_vol_desc(IsoImageWriter *writer);
+
+
 #endif /* LIBISO_JOLIET_H */

libisofs/libiso_msgs.c

@@ -2,9 +2,13 @@
 /* libiso_msgs   (generated from libdax_msgs : Fri Feb 22 19:42:52 CET 2008)
    Message handling facility of libisofs.
    Copyright (C) 2006 - 2008 Thomas Schmitt <scdbackup@gmx.net>,
-   provided under GPL version 2
+   provided under GPL version 2 or later
 */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <stdio.h>
 #include <sys/types.h>
 #include <unistd.h>

libisofs/libiso_msgs.h

@@ -2,7 +2,7 @@
 /* libiso_msgs   (generated from libdax_msgs : Fri Feb 22 19:42:52 CET 2008)
    Message handling facility of libisofs.
    Copyright (C) 2006-2008 Thomas Schmitt <scdbackup@gmx.net>,
-   provided under GPL version 2
+   provided under GPL version 2 or later
 */
 
 

libisofs/libisofs.ver

@@ -0,0 +1,313 @@
+LIBISOFS6 {
+global:
+aaip_xinfo_cloner;
+aaip_xinfo_func;
+el_torito_get_bootable;
+el_torito_get_boot_media_type;
+el_torito_get_boot_platform_id;
+el_torito_get_id_string;
+el_torito_get_isolinux_options;
+el_torito_get_load_seg;
+el_torito_get_load_size;
+el_torito_get_selection_crit;
+el_torito_patch_isolinux_image;
+el_torito_seems_boot_info_table;
+el_torito_set_boot_platform_id;
+el_torito_set_id_string;
+el_torito_set_isolinux_options;
+el_torito_set_load_seg;
+el_torito_set_load_size;
+el_torito_set_no_bootable;
+el_torito_set_selection_crit;
+iso_data_source_new_from_file;
+iso_data_source_ref;
+iso_data_source_unref;
+iso_dir_add_node;
+iso_dir_find_children;
+iso_dir_get_children;
+iso_dir_get_children_count;
+iso_dir_get_node;
+iso_dir_iter_free;
+iso_dir_iter_has_next;
+iso_dir_iter_next;
+iso_dir_iter_remove;
+iso_dir_iter_take;
+iso_error_get_code;
+iso_error_get_priority;
+iso_error_get_severity;
+iso_error_to_msg;
+iso_file_add_external_filter;
+iso_file_add_gzip_filter;
+iso_file_add_zisofs_filter;
+iso_file_get_md5;
+iso_file_get_old_image_lba;
+iso_file_get_old_image_sections;
+iso_file_get_size;
+iso_file_get_sort_weight;
+iso_file_get_stream;
+iso_file_make_md5;
+iso_file_remove_filter;
+iso_file_source_access;
+iso_file_source_close;
+iso_file_source_get_aa_string;
+iso_file_source_get_filesystem;
+iso_file_source_get_name;
+iso_file_source_get_path;
+iso_file_source_lseek;
+iso_file_source_lstat;
+iso_file_source_open;
+iso_file_source_read;
+iso_file_source_readdir;
+iso_file_source_readlink;
+iso_file_source_ref;
+iso_file_source_stat;
+iso_file_source_unref;
+iso_filesystem_ref;
+iso_filesystem_unref;
+iso_finish;
+iso_get_local_charset;
+iso_get_messenger;
+iso_gzip_get_refcounts;
+iso_image_add_boot_image;
+iso_image_add_mips_boot_file;
+iso_image_attach_data;
+iso_image_create_burn_source;
+iso_image_filesystem_new;
+iso_image_fs_get_abstract_file_id;
+iso_image_fs_get_application_id;
+iso_image_fs_get_biblio_file_id;
+iso_image_fs_get_copyright_file_id;
+iso_image_fs_get_data_preparer_id;
+iso_image_fs_get_publisher_id;
+iso_image_fs_get_system_id;
+iso_image_fs_get_volset_id;
+iso_image_fs_get_volume_id;
+iso_image_generator_is_running;
+iso_image_get_abstract_file_id;
+iso_image_get_all_boot_imgs;
+iso_image_get_application_id;
+iso_image_get_attached_data;
+iso_image_get_biblio_file_id;
+iso_image_get_boot_image;
+iso_image_get_copyright_file_id;
+iso_image_get_data_preparer_id;
+iso_image_get_mips_boot_files;
+iso_image_get_msg_id;
+iso_image_get_publisher_id;
+iso_image_get_root;
+iso_image_get_session_md5;
+iso_image_get_system_area;
+iso_image_get_system_id;
+iso_image_get_volset_id;
+iso_image_get_volume_id;
+iso_image_give_up_mips_boot;
+iso_image_import;
+iso_image_new;
+iso_image_ref;
+iso_image_remove_boot_image;
+iso_image_set_abstract_file_id;
+iso_image_set_application_id;
+iso_image_set_biblio_file_id;
+iso_image_set_boot_catalog_hidden;
+iso_image_set_boot_catalog_weight;
+iso_image_set_boot_image;
+iso_image_set_copyright_file_id;
+iso_image_set_data_preparer_id;
+iso_image_set_ignore_aclea;
+iso_image_set_publisher_id;
+iso_image_set_system_id;
+iso_image_set_volset_id;
+iso_image_set_volume_id;
+iso_image_unref;
+iso_image_update_sizes;
+iso_init;
+iso_init_with_flag;
+iso_lib_is_compatible;
+iso_lib_version;
+iso_local_get_acl_text;
+iso_local_get_attrs;
+iso_local_get_perms_wo_acl;
+iso_local_set_acl_text;
+iso_local_set_attrs;
+iso_md5_clone;
+iso_md5_compute;
+iso_md5_end;
+iso_md5_match;
+iso_md5_start;
+iso_memory_stream_new;
+iso_msgs_submit;
+iso_new_find_conditions_and;
+iso_new_find_conditions_atime;
+iso_new_find_conditions_ctime;
+iso_new_find_conditions_gid;
+iso_new_find_conditions_mode;
+iso_new_find_conditions_mtime;
+iso_new_find_conditions_name;
+iso_new_find_conditions_not;
+iso_new_find_conditions_or;
+iso_new_find_conditions_uid;
+iso_node_add_xinfo;
+iso_node_cmp_ino;
+iso_node_get_acl_text;
+iso_node_get_atime;
+iso_node_get_attrs;
+iso_node_get_ctime;
+iso_node_get_gid;
+iso_node_get_hidden;
+iso_node_get_mode;
+iso_node_get_mtime;
+iso_node_get_name;
+iso_node_get_next_xinfo;
+iso_node_get_old_image_lba;
+iso_node_get_parent;
+iso_node_get_permissions;
+iso_node_get_perms_wo_acl;
+iso_node_get_type;
+iso_node_get_uid;
+iso_node_get_xinfo;
+iso_node_lookup_attr;
+iso_node_ref;
+iso_node_remove;
+iso_node_remove_all_xinfo;
+iso_node_remove_tree;
+iso_node_remove_xinfo;
+iso_node_set_acl_text;
+iso_node_set_atime;
+iso_node_set_attrs;
+iso_node_set_ctime;
+iso_node_set_gid;
+iso_node_set_hidden;
+iso_node_set_mtime;
+iso_node_set_name;
+iso_node_set_permissions;
+iso_node_set_sort_weight;
+iso_node_set_uid;
+iso_node_take;
+iso_node_unref;
+iso_node_xinfo_get_cloner;
+iso_node_xinfo_make_clonable;
+iso_node_zf_by_magic;
+iso_obtain_msgs;
+iso_read_image_features_destroy;
+iso_read_image_features_get_size;
+iso_read_image_features_has_eltorito;
+iso_read_image_features_has_iso1999;
+iso_read_image_features_has_joliet;
+iso_read_image_features_has_rockridge;
+iso_read_opts_auto_input_charset;
+iso_read_opts_free;
+iso_read_opts_load_system_area;
+iso_read_opts_new;
+iso_read_opts_set_default_gid;
+iso_read_opts_set_default_permissions;
+iso_read_opts_set_default_uid;
+iso_read_opts_set_input_charset;
+iso_read_opts_set_new_inos;
+iso_read_opts_set_no_aaip;
+iso_read_opts_set_no_iso1999;
+iso_read_opts_set_no_joliet;
+iso_read_opts_set_no_md5;
+iso_read_opts_set_no_rockridge;
+iso_read_opts_set_preferjoliet;
+iso_read_opts_set_start_block;
+iso_ring_buffer_get_status;
+iso_set_abort_severity;
+iso_set_local_charset;
+iso_set_msgs_severities;
+iso_sev_to_text;
+iso_special_get_dev;
+iso_stream_clone;
+iso_stream_close;
+iso_stream_cmp_ino;
+iso_stream_get_external_filter;
+iso_stream_get_id;
+iso_stream_get_input_stream;
+iso_stream_get_size;
+iso_stream_get_source_path;
+iso_stream_is_repeatable;
+iso_stream_open;
+iso_stream_read;
+iso_stream_ref;
+iso_stream_unref;
+iso_stream_update_size;
+iso_symlink_get_dest;
+iso_symlink_set_dest;
+iso_text_to_sev;
+iso_tree_add_dir_rec;
+iso_tree_add_exclude;
+iso_tree_add_new_cut_out_node;
+iso_tree_add_new_dir;
+iso_tree_add_new_file;
+iso_tree_add_new_node;
+iso_tree_add_new_special;
+iso_tree_add_new_symlink;
+iso_tree_add_node;
+iso_tree_clone;
+iso_tree_get_follow_symlinks;
+iso_tree_get_ignore_hidden;
+iso_tree_get_ignore_special;
+iso_tree_get_node_path;
+iso_tree_get_replace_mode;
+iso_tree_path_to_node;
+iso_tree_remove_exclude;
+iso_tree_set_follow_symlinks;
+iso_tree_set_ignore_hidden;
+iso_tree_set_ignore_special;
+iso_tree_set_replace_mode;
+iso_tree_set_report_callback;
+iso_util_decode_md5_tag;
+iso_write_opts_attach_jte;
+iso_write_opts_detach_jte;
+iso_write_opts_free;
+iso_write_opts_get_data_start;
+iso_write_opts_new;
+iso_write_opts_set_aaip;
+iso_write_opts_set_aaip_susp_1_10;
+iso_write_opts_set_allow_deep_paths;
+iso_write_opts_set_allow_dir_id_ext;
+iso_write_opts_set_allow_full_ascii;
+iso_write_opts_set_allow_longer_paths;
+iso_write_opts_set_allow_lowercase;
+iso_write_opts_set_always_gmt;
+iso_write_opts_set_appendable;
+iso_write_opts_set_default_dir_mode;
+iso_write_opts_set_default_file_mode;
+iso_write_opts_set_default_gid;
+iso_write_opts_set_default_timestamp;
+iso_write_opts_set_default_uid;
+iso_write_opts_set_dir_rec_mtime;
+iso_write_opts_set_disc_label;
+iso_write_opts_set_fifo_size;
+iso_write_opts_set_hardlinks;
+iso_write_opts_set_iso1999;
+iso_write_opts_set_iso_level;
+iso_write_opts_set_joliet;
+iso_write_opts_set_joliet_longer_paths;
+iso_write_opts_set_max_37_char_filenames;
+iso_write_opts_set_ms_block;
+iso_write_opts_set_no_force_dots;
+iso_write_opts_set_old_empty;
+iso_write_opts_set_omit_version_numbers;
+iso_write_opts_set_output_charset;
+iso_write_opts_set_overwrite_buf;
+iso_write_opts_set_part_offset;
+iso_write_opts_set_partition_img;
+iso_write_opts_set_pvd_times;
+iso_write_opts_set_record_md5;
+iso_write_opts_set_relaxed_vol_atts;
+iso_write_opts_set_replace_mode;
+iso_write_opts_set_replace_timestamps;
+iso_write_opts_set_rockridge;
+iso_write_opts_set_rrip_1_10_px_ino;
+iso_write_opts_set_rrip_version_1_10;
+iso_write_opts_set_scdbackup_tag;
+iso_write_opts_set_sort_files;
+iso_write_opts_set_system_area;
+iso_write_opts_set_tail_blocks;
+iso_write_opts_set_untranslated_name_len;
+iso_write_opts_set_will_cancel;
+iso_zisofs_get_params;
+iso_zisofs_get_refcounts;
+iso_zisofs_set_params;
+local: *;
+};

libisofs/make_isohybrid_mbr.c

@@ -1,5 +1,18 @@
+
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <ctype.h>
+
+#ifdef HAVE_STDINT_H
 #include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
+
 #include <sys/types.h>
 #include <stdlib.h>
 #include <stdio.h>
@@ -10,6 +23,7 @@
 /* for gettimeofday() */
 #include <sys/time.h>
 
+
 /* This code stems from syslinux-3.72/utils/isohybrid, a perl script
 under GPL which is Copyright 2002-2008 H. Peter Anvin.
 
@@ -25,12 +39,13 @@ and is now under the licenses to which H.Peter Anvin agreed:
  or both, at your option.
  Sincerely, H. Peter Anvin
 
-In the context of libisofs this code derives its matching open source
+In the context of GNU xorriso, this code is under GPLv3+ derived from LGPL.
+In the context of libisofs this code derives its matching free software
 license from above stem licenses, typically from LGPL.
 In case its generosity is needed, here is the 2-clause BSD license:
 
 make_isohybrid_mbr.c is copyright 2002-2008 H. Peter Anvin
-                              and 2008 libburnia project.
+                              and 2008-2010 Thomas Schmitt
 
 1. Redistributions of source code must retain the above copyright notice,
    this list of conditions and the following disclaimer.
@@ -50,8 +65,9 @@ ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
 */
 
+
 /* A helper function. One could replace it by one or two macros. */
-static int lsb_to_buf(char **wpt, int value, int bits, int flag)
+static int lsb_to_buf(char **wpt, uint32_t value, int bits, int flag)
 {
     int b;
 
@@ -60,6 +76,11 @@ static int lsb_to_buf(char **wpt, int value, int bits, int flag)
     return (1);
 }
 
+
+/* ====================================================================== */
+/*                          Deprecated Function                           */
+/* ====================================================================== */
+
 /*
  * Create a MBR for an isohybrid enabled ISOLINUX boot image.
  *
@@ -227,3 +248,214 @@ int make_isohybrid_mbr(int bin_lba, int *img_blocks, char *mbr, int flag)
     return (1);
 }
 
+
+/* ====================================================================== */
+/*                          The New MBR Producer                          */
+/* ====================================================================== */
+
+/* The new MBR producer for isohybrid is a slightly generalized version of
+   the deprecated function make_isohybrid_mbr(). It complies to the urge
+   of H.Peter Anvin not to hardcode MBR templates but rather to read a
+   file from the Syslinux tree, and to patch it as was done with the old
+   MBR producer.
+
+   The old algorithm was clarified publicly by the following mail.
+   Changes towards the old algorithm:
+   - 512-byte LBA of boot image is extended to 64 bit (we stay with 32)
+   - check for a magic number is now gone
+
+   The new implementation tries to use similar terms as the mail in order
+   to facilitate its future discussion with Syslinux developers.
+
+From hpa@zytor.com Thu Apr  1 08:32:52 2010
+Date: Wed, 31 Mar 2010 14:53:51 -0700
+From: H. Peter Anvin <hpa@zytor.com>
+To: For discussion of Syslinux and tftp-hpa <syslinux@zytor.com>
+Cc: Thomas Schmitt <scdbackup@gmx.net>
+Subject: Re: [syslinux] port syslinux isohybrid perl script to C
+
+[...]
+
+[me:]
+> Currently i lack of blob and prescriptions.
+
+The blobs are available in the Syslinux build tree under the names:
+
+mbr/isohdp[fp]x*.bin
+
+The default probably should be mbr/isohdppx.bin, but it's ultimately up
+to the user.
+
+User definable parameters:
+
+-> MBR ID		(default random 32-bit number,
+			 or preserved from previous instance)
+-> Sector count		(default 32, range 1-63)
+-> Head count		(default 64, range 1-256)
+-> Partition offset	(default 0, range 0-64)
+-> Partition number	(default 1, range 1-4)
+-> Filesystem type	(default 0x17, range 1-255)
+
+Note: the filesystem type is largely arbitrary, in theory it can be any
+value other than 0x00, 0x05, 0x0f, 0x85, 0xee, or 0xef.  0x17 ("Windows
+IFS Hidden") seems safeish, some people believe 0x83 (Linux) is better.
+
+Here is the prescriptions for how to install it:
+
+All numbers are littleendian.  "word" means 16 bits, "dword" means 32
+bits, "qword" means 64 bits.
+
+Common subroutine LBA_to_CHS():
+	s = (lba % sector_count) + 1
+	t = (lba / sector_count)
+	h = (t % head_count)
+	c = (t / head_count)
+
+	if (c >= 1024):
+		c = 1023
+		h = head_count
+		s = sector_count
+
+	s = s | ((c & 0x300) >> 2)
+	c = c & 0xff
+
+	write byte h
+	write byte s
+	write byte c
+
+Main:
+	Pad image_size to a multiple of sector_count*head_count
+	Use the input file unmodified for bytes 0..431
+	write qword boot_lba		# Offset 432
+	write dword mbr_id		# Offset 440
+	write word 0			# Offset 444
+
+	# Offset 446
+	For each partition entry 1..4:
+		if this_partition != partition_number:
+			write 16 zero bytes
+		else:
+			write byte 0x80
+			write LBA_to_CHS(partition_offset)
+			write byte filesystem_type
+			write LBA_to_CHS(image_size-1)
+			write dword partition_offset
+			write dword image_size
+
+	# Offset 510
+	write word 0xaa55
+
+	Use the input file unmodified for bytes 512..32767
+	(pad with zero as necessary)
+
+[...]
+
+	-hpa
+*/
+
+
+static
+int lba512chs_to_buf(char **wpt, off_t lba, int head_count, int sector_count)
+{
+    int s, t, h, c;
+
+    s = (lba % sector_count) + 1;
+    t = (lba / sector_count);
+    h = (t % head_count);
+    c = (t / head_count);
+    if (c >= 1024) {
+        c = 1023;
+        h = head_count; /* >>> not -1 ? Limits head_count to 255 */
+        s = sector_count;
+    }
+    s = s | ((c & 0x300) >> 2);
+    c = c & 0xff;
+    (*((unsigned char **) wpt))[0] = h;
+    (*((unsigned char **) wpt))[1] = s;
+    (*((unsigned char **) wpt))[2] = c;
+    (*wpt)+= 3;
+    return(1); 
+}
+
+
+/*
+ * @param flag  bit0= make own random MBR Id from current time
+ */
+int make_isolinux_mbr(int32_t *img_blocks, uint32_t boot_lba,
+                      uint32_t mbr_id, int head_count, int sector_count,
+                      int part_offset, int part_number, int fs_type,
+                      uint8_t *buf, int flag)
+{
+    uint32_t spc, id, part, nominal_part_size;
+    off_t hd_img_blocks, hd_boot_lba;
+    char *wpt;
+    /* For generating a weak random number */
+    struct timeval tv;
+    struct timezone tz;
+
+    /* Pad image_size to a multiple of sector_count*head_count
+    */
+    spc = head_count * sector_count;
+    hd_img_blocks = ((off_t) *img_blocks) * (off_t) 4;
+    if (hd_img_blocks % spc) {
+        hd_img_blocks += spc - (hd_img_blocks % spc);
+        *img_blocks = hd_img_blocks / 4 + !!(hd_img_blocks % 4);
+    }
+
+    wpt = (char *) buf + 432;
+
+    /* write qword boot_lba            # Offset 432
+    */
+    hd_boot_lba = ((off_t) boot_lba) * (off_t) 4;
+    lsb_to_buf(&wpt, hd_boot_lba & 0xffffffff, 32, 0);
+    lsb_to_buf(&wpt, hd_boot_lba >> 32, 32, 0);
+
+    /* write dword mbr_id              # Offset 440 
+       (here some 32-bit random value with no crypto strength)
+    */
+    if (flag & 1) {
+        gettimeofday(&tv, &tz);
+        id = 0xffffffff & (tv.tv_sec ^ (tv.tv_usec * 2000));
+        lsb_to_buf(&wpt, id, 32, 0);
+    }
+
+    /* write word 0                    # Offset 444
+    */
+    lsb_to_buf(&wpt, 0, 16, 0);
+
+    /* # Offset 446
+    */
+    for (part = 1 ; part <= 4; part++) {
+        if (part != part_number) {
+            /* if this_partition != partition_number: write 16 zero bytes */
+            memset(wpt, 0, 16);
+            wpt+= 16;
+    continue;
+        }
+        /* write byte 0x80
+           write LBA_to_CHS(partition_offset)
+           write byte filesystem_type
+           write LBA_to_CHS(image_size-1)
+           write dword partition_offset
+           write dword image_size
+        */
+        lsb_to_buf(&wpt, 0x80, 8, 0);
+        lba512chs_to_buf(&wpt, part_offset, head_count, sector_count);
+        lsb_to_buf(&wpt, fs_type, 8, 0);
+        lba512chs_to_buf(&wpt, hd_img_blocks - 1, head_count, sector_count);
+        lsb_to_buf(&wpt, part_offset, 32, 0);
+        if (hd_img_blocks - (off_t) part_offset > (off_t) 0xffffffff)
+            nominal_part_size = 0xffffffff;
+        else
+            nominal_part_size = hd_img_blocks - (off_t) part_offset;
+        lsb_to_buf(&wpt, nominal_part_size, 32, 0);
+    }
+
+    /*  write word 0xaa55            # Offset 510
+    */
+    lsb_to_buf(&wpt, 0xaa55, 16, 0);
+
+    return(1);
+}
+
+

libisofs/md5.c

@@ -0,0 +1,876 @@
+/*
+ * Copyright (c) 2009 Thomas Schmitt
+ *
+ * This file is part of the libisofs project; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2 
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
+ */
+
+
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
+#ifdef HAVE_STDINT_H
+#include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
+
+#include <stdlib.h>
+#include <string.h>
+#include <stdio.h>
+
+#include "writer.h"
+#include "messages.h"
+#include "ecma119.h"
+#include "image.h"
+#include "util.h"
+
+#include "md5.h"
+
+
+/* This code is derived from RFC 1321 and implements computation of the
+    "RSA Data Security, Inc. MD5 Message-Digest Algorithm" */
+
+#define Libisofs_md5_S11 7
+#define Libisofs_md5_S12 12
+#define Libisofs_md5_S13 17
+#define Libisofs_md5_S14 22
+#define Libisofs_md5_S21 5
+#define Libisofs_md5_S22 9
+#define Libisofs_md5_S23 14
+#define Libisofs_md5_S24 20
+#define Libisofs_md5_S31 4
+#define Libisofs_md5_S32 11
+#define Libisofs_md5_S33 16
+#define Libisofs_md5_S34 23
+#define Libisofs_md5_S41 6
+#define Libisofs_md5_S42 10
+#define Libisofs_md5_S43 15
+#define Libisofs_md5_S44 21
+
+
+/* F, G, H and I are basic MD5 functions.
+ */
+#define Libisofs_md5_F(x, y, z) (((x) & (y)) | ((~x) & (z)))
+#define Libisofs_md5_G(x, y, z) (((x) & (z)) | ((y) & (~z)))
+#define Libisofs_md5_H(x, y, z) ((x) ^ (y) ^ (z))
+#define Libisofs_md5_I(x, y, z) ((y) ^ ((x) | (~z)))
+
+/* ROTATE_LEFT rotates x left n bits.
+ */
+#define Libisofs_md5_ROTATE_LEFT(x, n) (((x) << (n)) | ((x) >> (32-(n))))
+
+/* FF, GG, HH, and II transformations for rounds 1, 2, 3, and 4.
+Rotation is separate from addition to prevent recomputation.
+ */
+#define Libisofs_md5_FF(a, b, c, d, x, s, ac) { \
+ (a) += Libisofs_md5_F ((b), (c), (d)) + (x) + (uint32_t)(ac); \
+ (a) = Libisofs_md5_ROTATE_LEFT ((a), (s)); \
+ (a) += (b); \
+  }
+#define Libisofs_md5_GG(a, b, c, d, x, s, ac) { \
+ (a) += Libisofs_md5_G ((b), (c), (d)) + (x) + (uint32_t)(ac); \
+ (a) = Libisofs_md5_ROTATE_LEFT ((a), (s)); \
+ (a) += (b); \
+  }
+#define Libisofs_md5_HH(a, b, c, d, x, s, ac) { \
+ (a) += Libisofs_md5_H ((b), (c), (d)) + (x) + (uint32_t)(ac); \
+ (a) = Libisofs_md5_ROTATE_LEFT ((a), (s)); \
+ (a) += (b); \
+  }
+#define Libisofs_md5_II(a, b, c, d, x, s, ac) { \
+ (a) += Libisofs_md5_I ((b), (c), (d)) + (x) + (uint32_t)(ac); \
+ (a) = Libisofs_md5_ROTATE_LEFT ((a), (s)); \
+ (a) += (b); \
+  }
+
+
+/* MD5 context. */
+struct _libisofs_md5_ctx {
+  uint32_t state[4];                                   /* state (ABCD) */
+  uint32_t count[2];        /* number of bits, modulo 2^64 (lsb first) */
+  unsigned char buffer[64];                         /* input buffer */
+};
+
+typedef struct _libisofs_md5_ctx libisofs_md5_ctx;
+
+
+/* MD5 basic transformation. Transforms state based on block.
+ */
+static int md5__transform (uint32_t state[4], unsigned char block[64])
+{
+ uint32_t a = state[0], b = state[1], c = state[2], d = state[3], x[16];
+ unsigned int i, j;
+
+ for (i = 0, j = 0; j < 64; i++, j += 4)
+   x[i] = ((uint32_t)block[j]) | (((uint32_t)block[j+1]) << 8) |
+    (((uint32_t)block[j+2]) << 16) | (((uint32_t)block[j+3]) << 24);
+
+  /* Round 1 */
+  Libisofs_md5_FF (a, b, c, d, x[ 0], Libisofs_md5_S11, 0xd76aa478); /* 1 */
+  Libisofs_md5_FF (d, a, b, c, x[ 1], Libisofs_md5_S12, 0xe8c7b756); /* 2 */
+  Libisofs_md5_FF (c, d, a, b, x[ 2], Libisofs_md5_S13, 0x242070db); /* 3 */
+  Libisofs_md5_FF (b, c, d, a, x[ 3], Libisofs_md5_S14, 0xc1bdceee); /* 4 */
+  Libisofs_md5_FF (a, b, c, d, x[ 4], Libisofs_md5_S11, 0xf57c0faf); /* 5 */
+  Libisofs_md5_FF (d, a, b, c, x[ 5], Libisofs_md5_S12, 0x4787c62a); /* 6 */
+  Libisofs_md5_FF (c, d, a, b, x[ 6], Libisofs_md5_S13, 0xa8304613); /* 7 */
+  Libisofs_md5_FF (b, c, d, a, x[ 7], Libisofs_md5_S14, 0xfd469501); /* 8 */
+  Libisofs_md5_FF (a, b, c, d, x[ 8], Libisofs_md5_S11, 0x698098d8); /* 9 */
+  Libisofs_md5_FF (d, a, b, c, x[ 9], Libisofs_md5_S12, 0x8b44f7af); /* 10 */
+  Libisofs_md5_FF (c, d, a, b, x[10], Libisofs_md5_S13, 0xffff5bb1); /* 11 */
+  Libisofs_md5_FF (b, c, d, a, x[11], Libisofs_md5_S14, 0x895cd7be); /* 12 */
+  Libisofs_md5_FF (a, b, c, d, x[12], Libisofs_md5_S11, 0x6b901122); /* 13 */
+  Libisofs_md5_FF (d, a, b, c, x[13], Libisofs_md5_S12, 0xfd987193); /* 14 */
+  Libisofs_md5_FF (c, d, a, b, x[14], Libisofs_md5_S13, 0xa679438e); /* 15 */
+  Libisofs_md5_FF (b, c, d, a, x[15], Libisofs_md5_S14, 0x49b40821); /* 16 */
+
+  /* Round 2 */
+  Libisofs_md5_GG (a, b, c, d, x[ 1], Libisofs_md5_S21, 0xf61e2562); /* 17 */
+  Libisofs_md5_GG (d, a, b, c, x[ 6], Libisofs_md5_S22, 0xc040b340); /* 18 */
+  Libisofs_md5_GG (c, d, a, b, x[11], Libisofs_md5_S23, 0x265e5a51); /* 19 */
+  Libisofs_md5_GG (b, c, d, a, x[ 0], Libisofs_md5_S24, 0xe9b6c7aa); /* 20 */
+  Libisofs_md5_GG (a, b, c, d, x[ 5], Libisofs_md5_S21, 0xd62f105d); /* 21 */
+  Libisofs_md5_GG (d, a, b, c, x[10], Libisofs_md5_S22,  0x2441453); /* 22 */
+  Libisofs_md5_GG (c, d, a, b, x[15], Libisofs_md5_S23, 0xd8a1e681); /* 23 */
+  Libisofs_md5_GG (b, c, d, a, x[ 4], Libisofs_md5_S24, 0xe7d3fbc8); /* 24 */
+  Libisofs_md5_GG (a, b, c, d, x[ 9], Libisofs_md5_S21, 0x21e1cde6); /* 25 */
+  Libisofs_md5_GG (d, a, b, c, x[14], Libisofs_md5_S22, 0xc33707d6); /* 26 */
+  Libisofs_md5_GG (c, d, a, b, x[ 3], Libisofs_md5_S23, 0xf4d50d87); /* 27 */
+  Libisofs_md5_GG (b, c, d, a, x[ 8], Libisofs_md5_S24, 0x455a14ed); /* 28 */
+  Libisofs_md5_GG (a, b, c, d, x[13], Libisofs_md5_S21, 0xa9e3e905); /* 29 */
+  Libisofs_md5_GG (d, a, b, c, x[ 2], Libisofs_md5_S22, 0xfcefa3f8); /* 30 */
+  Libisofs_md5_GG (c, d, a, b, x[ 7], Libisofs_md5_S23, 0x676f02d9); /* 31 */
+  Libisofs_md5_GG (b, c, d, a, x[12], Libisofs_md5_S24, 0x8d2a4c8a); /* 32 */
+
+  /* Round 3 */
+  Libisofs_md5_HH (a, b, c, d, x[ 5], Libisofs_md5_S31, 0xfffa3942); /* 33 */
+  Libisofs_md5_HH (d, a, b, c, x[ 8], Libisofs_md5_S32, 0x8771f681); /* 34 */
+  Libisofs_md5_HH (c, d, a, b, x[11], Libisofs_md5_S33, 0x6d9d6122); /* 35 */
+  Libisofs_md5_HH (b, c, d, a, x[14], Libisofs_md5_S34, 0xfde5380c); /* 36 */
+  Libisofs_md5_HH (a, b, c, d, x[ 1], Libisofs_md5_S31, 0xa4beea44); /* 37 */
+  Libisofs_md5_HH (d, a, b, c, x[ 4], Libisofs_md5_S32, 0x4bdecfa9); /* 38 */
+  Libisofs_md5_HH (c, d, a, b, x[ 7], Libisofs_md5_S33, 0xf6bb4b60); /* 39 */
+  Libisofs_md5_HH (b, c, d, a, x[10], Libisofs_md5_S34, 0xbebfbc70); /* 40 */
+  Libisofs_md5_HH (a, b, c, d, x[13], Libisofs_md5_S31, 0x289b7ec6); /* 41 */
+  Libisofs_md5_HH (d, a, b, c, x[ 0], Libisofs_md5_S32, 0xeaa127fa); /* 42 */
+  Libisofs_md5_HH (c, d, a, b, x[ 3], Libisofs_md5_S33, 0xd4ef3085); /* 43 */
+  Libisofs_md5_HH (b, c, d, a, x[ 6], Libisofs_md5_S34,  0x4881d05); /* 44 */
+  Libisofs_md5_HH (a, b, c, d, x[ 9], Libisofs_md5_S31, 0xd9d4d039); /* 45 */
+  Libisofs_md5_HH (d, a, b, c, x[12], Libisofs_md5_S32, 0xe6db99e5); /* 46 */
+  Libisofs_md5_HH (c, d, a, b, x[15], Libisofs_md5_S33, 0x1fa27cf8); /* 47 */
+  Libisofs_md5_HH (b, c, d, a, x[ 2], Libisofs_md5_S34, 0xc4ac5665); /* 48 */
+
+  /* Round 4 */
+  Libisofs_md5_II (a, b, c, d, x[ 0], Libisofs_md5_S41, 0xf4292244); /* 49 */
+  Libisofs_md5_II (d, a, b, c, x[ 7], Libisofs_md5_S42, 0x432aff97); /* 50 */
+  Libisofs_md5_II (c, d, a, b, x[14], Libisofs_md5_S43, 0xab9423a7); /* 51 */
+  Libisofs_md5_II (b, c, d, a, x[ 5], Libisofs_md5_S44, 0xfc93a039); /* 52 */
+  Libisofs_md5_II (a, b, c, d, x[12], Libisofs_md5_S41, 0x655b59c3); /* 53 */
+  Libisofs_md5_II (d, a, b, c, x[ 3], Libisofs_md5_S42, 0x8f0ccc92); /* 54 */
+  Libisofs_md5_II (c, d, a, b, x[10], Libisofs_md5_S43, 0xffeff47d); /* 55 */
+  Libisofs_md5_II (b, c, d, a, x[ 1], Libisofs_md5_S44, 0x85845dd1); /* 56 */
+  Libisofs_md5_II (a, b, c, d, x[ 8], Libisofs_md5_S41, 0x6fa87e4f); /* 57 */
+  Libisofs_md5_II (d, a, b, c, x[15], Libisofs_md5_S42, 0xfe2ce6e0); /* 58 */
+  Libisofs_md5_II (c, d, a, b, x[ 6], Libisofs_md5_S43, 0xa3014314); /* 59 */
+  Libisofs_md5_II (b, c, d, a, x[13], Libisofs_md5_S44, 0x4e0811a1); /* 60 */
+  Libisofs_md5_II (a, b, c, d, x[ 4], Libisofs_md5_S41, 0xf7537e82); /* 61 */
+  Libisofs_md5_II (d, a, b, c, x[11], Libisofs_md5_S42, 0xbd3af235); /* 62 */
+  Libisofs_md5_II (c, d, a, b, x[ 2], Libisofs_md5_S43, 0x2ad7d2bb); /* 63 */
+  Libisofs_md5_II (b, c, d, a, x[ 9], Libisofs_md5_S44, 0xeb86d391); /* 64 */
+
+  state[0] += a;
+  state[1] += b;
+  state[2] += c;
+  state[3] += d;
+
+  /* Zeroize sensitive information. */
+  memset ((char *) x, 0, sizeof (x));
+  return(1);
+}
+
+
+static int md5__encode(unsigned char *output, uint32_t *input,
+                       unsigned int len)
+{
+  unsigned int i, j;
+
+  for (i = 0, j = 0; j < len; i++, j += 4) {
+    output[j] = (unsigned char)(input[i] & 0xff);
+    output[j+1] = (unsigned char)((input[i] >> 8) & 0xff);
+    output[j+2] = (unsigned char)((input[i] >> 16) & 0xff);
+    output[j+3] = (unsigned char)((input[i] >> 24) & 0xff);
+  }
+  return(1);
+}
+
+
+
+static int md5_init(libisofs_md5_ctx *ctx, int flag)
+{
+ ctx->count[0] = ctx->count[1] = 0;
+ /* Load magic initialization constants. */
+ ctx->state[0] = 0x67452301;
+ ctx->state[1] = 0xefcdab89;
+ ctx->state[2] = 0x98badcfe;
+ ctx->state[3] = 0x10325476;
+ return(1);
+}
+
+
+/* MD5 block update operation. Continues an MD5 message-digest
+  operation, processing another message block, and updating the
+  context.
+ */
+static int md5_update(libisofs_md5_ctx *ctx, unsigned char *data,
+                      int datalen, int flag)
+{
+ unsigned int i, index, partlen;
+
+ /* Compute number of bytes mod 64 */
+ index = (unsigned int)((ctx->count[0] >> 3) & 0x3F);
+ /* Update number of bits */
+ if ((ctx->count[0] += ((uint32_t) datalen << 3)) < 
+     ((uint32_t) datalen << 3))
+   ctx->count[1]++;
+ ctx->count[1] += ((uint32_t) datalen >> 29);
+ partlen = 64 - index;
+
+ /* Transform as many times as possible. */
+ if (datalen >= partlen) {
+   memcpy((char *) &ctx->buffer[index], (char *) data, partlen);
+   md5__transform(ctx->state, ctx->buffer);
+   for (i = partlen; i + 63 < datalen; i += 64)
+     md5__transform(ctx->state, &data[i]);
+   index = 0;
+ } else
+   i = 0;
+
+ /* Buffer remaining data */
+ memcpy((char *) &ctx->buffer[index], (char *) &data[i],datalen-i);
+
+ return(1);
+}
+
+
+static int md5_final(libisofs_md5_ctx *ctx, char result[16], int flag)
+{
+ unsigned char bits[8], *respt;
+ unsigned int index, padlen;
+ static unsigned char PADDING[64] = {
+   0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
+   0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+ };
+
+ /* Save number of bits */
+ md5__encode(bits, ctx->count, 8);
+ /* Pad out to 56 mod 64.  */
+ index = (unsigned int)((ctx->count[0] >> 3) & 0x3f);
+ padlen = (index < 56) ? (56 - index) : (120 - index);
+ md5_update(ctx, PADDING, padlen,0);
+ /* Append length (before padding) */
+ md5_update(ctx, bits, 8,0);
+ /* Store state in result */
+ respt= (unsigned char *) result;
+ md5__encode(respt, ctx->state, 16);
+ /* Zeroize sensitive information.  */
+ memset ((char *) ctx, 0, sizeof (*ctx));
+ return(1);
+}
+
+/** Compute a MD5 checksum from one or more calls of this function.
+    The first call has to be made with flag bit0 == 1. It may already submit
+    processing payload in data and datalen.
+    The last call has to be made with bit15 set. Normally bit1 will be set
+    too in order to receive the checksum before it gets disposed. 
+    bit1 may only be set in the last call or together with bit2.
+    The combination of bit1 and bit2 may be used to get an intermediate
+    result without hampering an ongoing checksum computation.
+
+    @param ctx      the checksum context which stores the state between calls.
+                    It gets created with flag bit0 and disposed with bit15.
+                    With flag bit0, *ctx has to be NULL or point to freeable
+                    memory.
+    @param data     the bytes to be checksummed
+    @param datalen  the number of bytes in data
+    @param result   returns the 16 bytes of checksum if called with bit1 set
+    @param flag     bit0= allocate and init *ctx
+                    bit1= transfer ctx to result
+                    bit2= with bit 0 : clone new *ctx from data
+                    bit15= free *ctx
+*/
+static
+int libisofs_md5(void **ctx_in, char *data, int datalen,
+                 char result[16], int flag)
+/* *ctx has to be NULL or point to freeable memory */
+/*
+ bit0= allocate and init *ctx
+ bit1= transfer ctx to result
+ bit2= with bit 0 : clone new *ctx from data
+ bit15= free *ctx
+*/
+{
+ unsigned char *datapt;
+ libisofs_md5_ctx **ctx;
+
+ ctx= (libisofs_md5_ctx **) ctx_in;
+ if(flag&1) {
+   if(*ctx!=NULL)
+     free((char *) *ctx);
+   *ctx= calloc(1, sizeof(libisofs_md5_ctx));
+   if(*ctx==NULL)
+     return(-1);
+   md5_init(*ctx,0);
+   if(flag&4)
+     memcpy((char *) *ctx,data,sizeof(libisofs_md5_ctx));
+ }
+ if(*ctx==NULL)
+   return(0);
+ if(datalen>0) {
+   datapt= (unsigned char *) data;
+   md5_update(*ctx, datapt, datalen, 0);
+ }
+ if(flag&2)
+   md5_final(*ctx, result, 0);
+ if(flag&(1<<15)) {
+   free((char *) *ctx);
+   *ctx= NULL;
+ }
+ return(1);
+}
+
+
+/* ----------------------------------------------------------------------- */
+
+/* Public MD5 computing facility */
+
+/* API */
+int iso_md5_start(void **md5_context)
+{
+    int ret;
+
+    ret = libisofs_md5(md5_context, NULL, 0, NULL, 1);
+    if (ret <= 0)
+        return ISO_OUT_OF_MEM;
+    return 1;
+}
+
+
+/* API */
+int iso_md5_compute(void *md5_context, char *data, int datalen)
+{
+    int ret;
+
+    ret = libisofs_md5(&md5_context, data, datalen, NULL, 0);
+    if (ret <= 0)
+        return ISO_NULL_POINTER;
+    return 1;
+}
+
+
+/* API */
+int iso_md5_clone(void *old_md5_context, void **new_md5_context)
+{
+    int ret;
+
+    ret = libisofs_md5(new_md5_context, old_md5_context, 0, NULL, 1 | 4);
+    if (ret < 0)
+        return ISO_OUT_OF_MEM;
+    if (ret == 0)
+        return ISO_NULL_POINTER;
+    return 1;
+}
+
+
+/* API */
+int iso_md5_end(void **md5_context, char result[16])
+{
+    int ret;
+
+    ret = libisofs_md5(md5_context, NULL, 0, result, 2 | (1 << 15));
+    if (ret <= 0)
+        return ISO_NULL_POINTER;
+    return 1;
+}
+
+
+/* API */
+int iso_md5_match(char first_md5[16], char second_md5[16])
+{
+    int i;
+
+    for (i= 0; i < 16; i++)
+        if (first_md5[i] != second_md5[i])
+            return 0;
+    return 1;
+}
+
+
+/* ----------------------------------------------------------------------- */
+
+
+/* Function to identify and manage md5sum indice of the old image.
+ * data is supposed to be a 4 byte integer, bit 31 shall be 0,
+ * value 0 of this integer means that it is not a valid index.
+ */
+int checksum_cx_xinfo_func(void *data, int flag)
+{
+    /* data is an int disguised as pointer. It does not point to memory. */
+    return 1;
+}
+
+/* The iso_node_xinfo_cloner function which gets associated to
+ * checksum_cx_xinfo_func by iso_init() resp. iso_init_with_flag() via
+ * iso_node_xinfo_make_clonable()
+ */
+int checksum_cx_xinfo_cloner(void *old_data, void **new_data, int flag)
+{
+    *new_data = NULL;
+    if (flag)
+        return ISO_XINFO_NO_CLONE;
+    if (old_data == NULL)
+        return 0;
+    /* data is an int disguised as pointer. It does not point to memory. */
+    *new_data = old_data;
+    return 0;
+}
+
+
+/* Function to identify and manage md5 sums of unspecified providence stored
+ * directly in this xinfo.
+ */
+int checksum_md5_xinfo_func(void *data, int flag)
+{
+    if (data == NULL)
+        return 1;
+    free(data);
+    return 1;
+}
+
+/* The iso_node_xinfo_cloner function which gets associated to
+ * checksum_md5_xinfo_func by iso_init() resp. iso_init_with_flag() via
+ * iso_node_xinfo_make_clonable()
+ */
+int checksum_md5_xinfo_cloner(void *old_data, void **new_data, int flag)
+{
+    *new_data = NULL;
+    if (flag)
+        return ISO_XINFO_NO_CLONE;
+    if (old_data == NULL)
+        return 0;
+    *new_data = calloc(1, 16);
+    if (*new_data == NULL)
+        return ISO_OUT_OF_MEM;
+    memcpy(*new_data, old_data, 16);
+    return 16;
+}
+
+
+/* ----------------------------------------------------------------------- */
+
+/* MD5 checksum image writer */
+
+/*
+  @flag bit0= recursion
+        bit1= session will be appended to an existing image
+*/
+static
+int checksum_copy_old_nodes(Ecma119Image *target, IsoNode *node, int flag)
+{
+    IsoNode *pos;
+    IsoFile *file;
+    IsoImage *img;
+    int ret, i;
+    size_t value_length;
+    unsigned int idx = 0, old_idx = 0;
+    char *value = NULL, *md5_pt = NULL;
+    void *xipt;
+
+    img = target->image;
+    if (target->checksum_buffer == NULL)
+        return 0;
+
+    if (node->type == LIBISO_FILE) {
+        file = (IsoFile *) node;
+        if (file->from_old_session && target->appendable) {
+            /* Look for checksums at various places */
+
+            /* Try checksum directly stored with node */
+            if (md5_pt == NULL) {
+                ret = iso_node_get_xinfo(node, checksum_md5_xinfo_func, &xipt);
+                if (ret < 0)
+                    return ret;
+                if (ret == 1)
+                    md5_pt = (char *) xipt;
+            }
+
+            /* Try checksum index to image checksum buffer */
+            if (md5_pt == NULL && img->checksum_array != NULL) {
+                ret = iso_node_get_xinfo(node, checksum_cx_xinfo_func, &xipt);
+                if (ret <= 0)
+                    return ret;
+                /* xipt is an int disguised as void pointer */
+                old_idx = 0;
+                for (i = 0; i < 4; i++)
+                    old_idx = (old_idx << 8) | ((unsigned char *) &xipt)[i];
+    
+                if (old_idx == 0 || old_idx > img->checksum_idx_count - 1)
+                    return 0;
+                md5_pt = img->checksum_array + 16 * old_idx;
+            }
+
+            if (md5_pt == NULL)
+                return 0;
+
+            ret = iso_node_lookup_attr(node, "isofs.cx", &value_length,
+                                       &value, 0);
+            if (ret == 1 && value_length == 4) {
+                for (i = 0; i < 4; i++)
+                    idx = (idx << 8) | ((unsigned char *) value)[i];
+                if (idx > 0 && idx <= target->checksum_idx_counter) {
+                    memcpy(target->checksum_buffer + 16 * idx, md5_pt, 16);
+                }
+            }
+            if (value != NULL)
+                free(value);
+            iso_node_remove_xinfo(node, checksum_md5_xinfo_func);
+            iso_node_remove_xinfo(node, checksum_cx_xinfo_func);
+        }
+    } else if (node->type == LIBISO_DIR) {
+        for (pos = ((IsoDir *) node)->children; pos != NULL; pos = pos->next) {
+            ret = checksum_copy_old_nodes(target, pos, 1);
+            if (ret < 0)
+                return ret;
+        }
+    }
+    return ISO_SUCCESS;
+}
+
+
+static
+int checksum_writer_compute_data_blocks(IsoImageWriter *writer)
+{
+    size_t size;
+    Ecma119Image *t;
+    int ret;
+
+    if (writer == NULL) {
+        return ISO_ASSERT_FAILURE;
+    }
+    t = writer->target;
+
+    t->checksum_array_pos = t->curblock;
+                               /* (t->curblock already contains t->ms_block) */ 
+    t->checksum_range_start = t->ms_block;
+    size = (t->checksum_idx_counter + 2) / 128;
+    if (size * 128 < t->checksum_idx_counter + 2)
+        size++;
+    t->curblock += size;
+    t->checksum_range_size = t->checksum_array_pos + size
+                             - t->checksum_range_start;
+
+    /* Extra block for stream detectable checksum tag */
+    t->checksum_tag_pos =  t->curblock;
+    t->curblock++;
+
+    /* Allocate array of MD5 sums */
+    t->checksum_buffer = calloc(size, 2048);
+    if (t->checksum_buffer == NULL)
+        return ISO_OUT_OF_MEM;
+
+    /* Copy MD5 from nodes of old image into writer->data */
+    ret = checksum_copy_old_nodes(t, (IsoNode *) t->image->root, 0);
+    if (ret < 0)
+        return ret;
+
+    /* Record lba,count,size,cecksum_type in "isofs.ca" of root node */
+    ret = iso_root_set_isofsca((IsoNode *) t->image->root,
+                               t->checksum_range_start, 
+                               t->checksum_array_pos,
+                               t->checksum_idx_counter + 2, 16, "MD5", 0);
+    if (ret < 0)
+        return ret;
+    return ISO_SUCCESS;
+}
+
+
+static
+int checksum_writer_write_vol_desc(IsoImageWriter *writer)
+{
+
+    /* The superblock checksum tag has to be written after
+       the Volume Descriptor Set Terminator and thus may not be
+       written by this function. (It would have been neat, though).
+    */
+
+    return ISO_SUCCESS;
+}
+
+
+static
+int checksum_writer_write_data(IsoImageWriter *writer)
+{
+    int wres, res;
+    size_t i, size;
+    Ecma119Image *t;
+    void *ctx = NULL;
+    char md5[16];
+
+#ifdef NIX
+    char tag_block[2048];
+    int l;
+#endif
+
+    if (writer == NULL) {
+        return ISO_ASSERT_FAILURE;
+    }
+
+    t = writer->target;
+    iso_msg_debug(t->image->id, "Writing Checksums...");
+
+    /* Write image checksum to index 0 */
+    if (t->checksum_ctx != NULL) {
+        res = iso_md5_clone(t->checksum_ctx, &ctx);
+        if (res > 0) {
+            res = iso_md5_end(&ctx, t->image_md5);
+            if (res > 0)
+                memcpy(t->checksum_buffer + 0 * 16, t->image_md5, 16);
+        }
+    }
+
+    size = (t->checksum_idx_counter + 2) / 128;
+    if (size * 128 < t->checksum_idx_counter + 2)
+        size++;
+ 
+    /* Write checksum of checksum array as index t->checksum_idx_counter + 1 */
+    res = iso_md5_start(&ctx);
+    if (res > 0) {
+        for (i = 0; i < t->checksum_idx_counter + 1; i++)
+            iso_md5_compute(ctx,
+                          t->checksum_buffer + ((size_t) i) * (size_t) 16, 16);
+        res = iso_md5_end(&ctx, md5);
+        if (res > 0)
+           memcpy(t->checksum_buffer + (t->checksum_idx_counter + 1) * 16,
+                  md5, 16);
+    }
+
+    for (i = 0; i < size; i++) {
+        wres = iso_write(t, t->checksum_buffer + ((size_t) 2048) * i, 2048);
+        if (wres < 0) {
+            res = wres;
+            goto ex;
+        }
+    }
+    if (t->checksum_ctx == NULL) {
+        res = ISO_SUCCESS;
+        goto ex;
+    }
+
+    /* Write stream detectable checksum tag to extra block */;
+    res = iso_md5_write_tag(t, 1);
+    if (res < 0)
+        goto ex;
+
+    res = ISO_SUCCESS;
+ex:;
+    if (ctx != NULL)
+        iso_md5_end(&ctx, md5);
+    return(res);
+}
+
+
+static
+int checksum_writer_free_data(IsoImageWriter *writer)
+{
+    /* nothing was allocated at writer->data */
+    return ISO_SUCCESS;
+}
+
+
+int checksum_writer_create(Ecma119Image *target)
+{
+    IsoImageWriter *writer;
+    
+    writer = malloc(sizeof(IsoImageWriter));
+    if (writer == NULL) {
+        return ISO_OUT_OF_MEM;
+    }
+    
+    writer->compute_data_blocks = checksum_writer_compute_data_blocks;
+    writer->write_vol_desc = checksum_writer_write_vol_desc;
+    writer->write_data = checksum_writer_write_data;
+    writer->free_data = checksum_writer_free_data;
+    writer->data = NULL;
+    writer->target = target;
+
+    /* add this writer to image */
+    target->writers[target->nwriters++] = writer;
+    /* Account for superblock checksum tag */
+    if (target->md5_session_checksum) {
+        target->checksum_sb_tag_pos = target->curblock;
+        target->curblock++;
+    }
+    return ISO_SUCCESS;
+}
+
+
+static
+int iso_md5_write_scdbackup_tag(Ecma119Image *t, char *tag_block, int flag)
+{
+    void *ctx = NULL;
+    off_t pos = 0, line_start;
+    int record_len, block_len, res, i;
+    char postext[40], md5[16], record[160];
+
+    line_start = strlen(tag_block);
+    iso_md5_compute(t->checksum_ctx, tag_block, line_start);
+    res = iso_md5_clone(t->checksum_ctx, &ctx);
+    if (res < 0)
+        goto ex;
+    res = iso_md5_end(&ctx, md5);
+
+    pos = (off_t) t->checksum_tag_pos * (off_t) 2048 + line_start;
+    if(pos >= 1000000000)
+        sprintf(postext, "%u%9.9u", (unsigned int) (pos / 1000000000),
+                                    (unsigned int) (pos % 1000000000));
+    else
+        sprintf(postext, "%u", (unsigned int) pos);
+    sprintf(record, "%s %s ", t->scdbackup_tag_parm, postext);
+    record_len = strlen(record);
+    for (i = 0; i < 16; i++)
+         sprintf(record + record_len + 2 * i,
+                 "%2.2x", ((unsigned char *) md5)[i]);
+    record_len += 32;
+
+    res = iso_md5_start(&ctx);
+    if (res < 0)
+        goto ex;
+    iso_md5_compute(ctx, record, record_len);
+    iso_md5_end(&ctx, md5);
+
+    sprintf(tag_block + line_start, "scdbackup_checksum_tag_v0.1 %s %d %s ",
+            postext, record_len, record);
+    block_len = strlen(tag_block);
+    for (i = 0; i < 16; i++)
+        sprintf(tag_block + block_len + 2 * i,
+                "%2.2x", ((unsigned char *) md5)[i]);
+    block_len+= 32;
+    tag_block[block_len++]= '\n';
+
+    if (t->scdbackup_tag_written != NULL)
+    	strncpy(t->scdbackup_tag_written, tag_block + line_start,
+                block_len - line_start);
+    res = ISO_SUCCESS;
+ex:;
+    if (ctx != NULL)
+        iso_md5_end(&ctx, md5);
+    return res;
+}
+
+
+/* Write stream detectable checksum tag to extra block.
+ * @flag bit0-7= tag type
+ *               1= session tag (End checksumming.)
+ *               2= superblock tag (System Area and Volume Descriptors)
+ *               3= tree tag (ECMA-119 and Rock Ridge tree)
+ *               4= relocated superblock tag (at LBA 0 of overwriteable media)
+ *                  Write to target->opts_overwrite rather than to iso_write().
+ */
+int iso_md5_write_tag(Ecma119Image *t, int flag)
+{
+    int res, mode, l, i, wres, tag_id_len;
+    void *ctx = NULL;
+    char md5[16], tag_block[2048], *tag_id;
+    uint32_t size = 0, pos = 0, start;
+
+    start = t->checksum_range_start;
+    memset(tag_block, 0, 2048);
+    mode = flag & 255;
+    if (mode < 1 || mode > 4)
+        return ISO_WRONG_ARG_VALUE;
+    res = iso_md5_clone(t->checksum_ctx, &ctx);
+    if (res < 0)
+        return res;
+    res = iso_md5_end(&ctx, md5);
+    if (mode == 1) {
+        size = t->checksum_range_size;
+        pos = t->checksum_tag_pos;
+    } else {
+        if (mode == 2) {
+            pos = t->checksum_sb_tag_pos;
+        } else if (mode == 3) {
+            pos = t->checksum_tree_tag_pos;
+        } else if (mode == 4) {
+            pos = t->checksum_rlsb_tag_pos;
+            start = pos - (pos % 32);
+        }
+        size = pos - start;
+    }
+    if (res < 0)
+        goto ex;
+
+    iso_util_tag_magic(mode, &tag_id, &tag_id_len, 0);
+    sprintf(tag_block, "%s pos=%u range_start=%u range_size=%u",
+            tag_id, pos, start, size);
+
+    l = strlen(tag_block);
+    if (mode == 2) {
+        sprintf(tag_block + l, " next=%u", t->checksum_tree_tag_pos);
+    } else if (mode == 3) {
+        sprintf(tag_block + l, " next=%u", t->checksum_tag_pos);
+    } else if (mode == 4) {
+        sprintf(tag_block + l, " session_start=%u", t->ms_block);
+    }
+    strcat(tag_block + l, " md5=");
+    l = strlen(tag_block);
+    for (i = 0; i < 16; i++)
+        sprintf(tag_block + l + 2 * i, "%2.2x",
+                ((unsigned char *) md5)[i]);
+    l+= 32;
+        
+    res = iso_md5_start(&ctx);
+    if (res > 0) {
+        iso_md5_compute(ctx, tag_block, l);
+        iso_md5_end(&ctx, md5);
+        strcpy(tag_block + l, " self=");
+        l += 6;
+        for (i = 0; i < 16; i++)
+            sprintf(tag_block + l + 2 * i, "%2.2x",
+                    ((unsigned char *) md5)[i]);
+    }
+    tag_block[l + 32] = '\n';
+
+    if (mode == 1 && t->scdbackup_tag_parm[0]) {
+        if (t->ms_block > 0) {
+            iso_msg_submit(t->image->id, ISO_SCDBACKUP_TAG_NOT_0, 0, NULL);
+        } else {
+            res = iso_md5_write_scdbackup_tag(t, tag_block, 0);
+            if (res < 0)
+                goto ex;
+        }
+    }
+
+    if (mode == 4) {
+        if (t->opts_overwrite != NULL)
+            memcpy(t->opts_overwrite + pos * 2048, tag_block, 2048);
+    } else {
+        wres = iso_write(t, tag_block, 2048);
+        if (wres < 0) {
+            res = wres;
+            goto ex;
+        }
+    }
+
+    res = ISO_SUCCESS;
+ex:;
+    if (ctx != NULL)
+        iso_md5_end(&ctx, md5);
+    return res;
+}
+
+

libisofs/md5.h

@@ -0,0 +1,41 @@
+/*
+ * Copyright (c) 2009 Thomas Schmitt
+ *
+ * This file is part of the libisofs project; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2 
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
+ */
+
+#ifndef LIBISO_MD5_H_
+#define LIBISO_MD5_H_
+
+
+/* The MD5 computation API is in libisofs.h : iso_md5_start() et.al. */
+
+
+/** Create a writer object for checksums and add it to the writer list of
+    the given Ecma119Image.
+*/
+int checksum_writer_create(Ecma119Image *target);
+
+
+/* Write stream detectable checksum tag to extra block.
+ * All tag ranges start at the beginning of the System Area (i.e. t->ms_block)
+ * and stem from the same MD5 computation context. Tag types 2 and 3 are
+ * intermediate checksums. Type 2 announces the existence of type 3.
+ * If both match, then at least the directory tree is trustworthy.
+ * Type 1 is written at the very end of the session. If it matches, then
+ * the whole image is trustworthy.
+ * @param t      The image being written
+ * @flag         bit0-7= tag type
+ *               1= session tag (End checksumming.)
+ *               2= superblock tag (System Area and Volume Descriptors)
+ *               3= tree tag (ECMA-119 and Rock Ridge tree)
+ */
+int iso_md5_write_tag(Ecma119Image *t, int flag);
+
+
+#endif /* ! LIBISO_MD5_H_ */
+
+

libisofs/messages.c

@@ -1,21 +1,43 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
+
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include <stdlib.h>
 #include <sys/types.h>
 #include <string.h>
 #include <stdio.h>
 #include <stdarg.h>
 
+#ifdef Xorriso_standalonE
+
+#ifdef Xorriso_with_libjtE
+#include "../libjte/libjte.h"
+#endif
+
+#else
+
+#ifdef Libisofs_with_libjtE
+#include <libjte/libjte.h>
+#endif
+
+#endif /* ! Xorriso_standalonE */
+
 #include "libiso_msgs.h"
 #include "libisofs.h"
 #include "messages.h"
 
 #include "util.h"
+#include "node.h"
 
 
 /*
@@ -59,28 +81,143 @@ int abort_threshold = LIBISO_MSGS_SEV_FAILURE;
 struct libiso_msgs *libiso_msgr = NULL;
 
 
+/* ------------- List of xinfo clone functions ----------- */
+
+struct iso_xinfo_cloner_assoc {
+    iso_node_xinfo_func proc;
+    iso_node_xinfo_cloner cloner;
+    struct iso_xinfo_cloner_assoc *next;
+};
+
+struct iso_xinfo_cloner_assoc *iso_xinfo_cloner_list = NULL;
+
+/* API */
+int iso_node_xinfo_make_clonable(iso_node_xinfo_func proc,
+                                 iso_node_xinfo_cloner cloner, int flag)
+{
+    struct iso_xinfo_cloner_assoc *assoc;
+
+    /* Look for existing assoc of proc */
+    for (assoc = iso_xinfo_cloner_list; assoc != NULL; assoc = assoc->next) 
+        if (assoc->proc == proc)
+    break;
+
+    if (assoc == NULL) {
+        assoc = calloc(1, sizeof(struct iso_xinfo_cloner_assoc));
+        if (assoc == NULL)
+            return ISO_OUT_OF_MEM;
+        assoc->proc = proc;
+        assoc->next = iso_xinfo_cloner_list;
+        iso_xinfo_cloner_list = assoc;
+    }
+    assoc->cloner = cloner;
+    return ISO_SUCCESS;
+}
+
+/* API */
+int iso_node_xinfo_get_cloner(iso_node_xinfo_func proc,
+                              iso_node_xinfo_cloner *cloner, int flag)
+{
+    struct iso_xinfo_cloner_assoc *assoc;
+
+    *cloner = NULL;
+    for (assoc = iso_xinfo_cloner_list; assoc != NULL; assoc = assoc->next) {
+        if (assoc->proc != proc)
+    continue;
+        *cloner = assoc->cloner;
+        return 1;
+    }
+    return 0;
+}
+
+static
+int iso_node_xinfo_dispose_cloners(int flag)
+{
+    struct iso_xinfo_cloner_assoc *assoc, *next;
+
+    for (assoc = iso_xinfo_cloner_list; assoc != NULL; assoc = next) {
+        next = assoc->next;
+        free((char *) assoc);
+    }
+    return(1);
+}
+
+/* ------------- End of xinfo clone functions list ----------- */
+
+
 /*
   @param flag  bit0= do not set up locale by LC_* environment variables
 */
 int iso_init_with_flag(int flag)
 {
+    int ret;
+
+#ifdef Libisofs_with_libjtE
+
+ /* Ugly compile time check for header version compatibility.
+    If everthing matches, then it produces no C code. In case of mismatch,
+    intentionally faulty C code will be inserted.
+ */ 
+ /* The indendation is an advise of man gcc to help old compilers ignoring */
+ #if iso_libjte_req_major > LIBJTE_VERSION_MAJOR
+ #define Libisofs_libjte_dot_h_too_olD 1
+ #endif
+ #if iso_libjte_req_major == LIBJTE_VERSION_MAJOR && iso_libjte_req_minor > LIBJTE_VERSION_MINOR
+ #define Libisofs_libjte_dot_h_too_olD 1
+ #endif
+ #if iso_libjte_req_minor == LIBJTE_VERSION_MINOR && iso_libjte_req_micro > LIBJTE_VERSION_MICRO
+ #define Libisofs_libjte_dot_h_too_olD 1
+ #endif
+
+#ifdef Libisofs_libjte_dot_h_too_olD
+LIBJTE_MISCONFIGURATION = 0;
+INTENTIONAL_ABORT_OF_COMPILATION__HEADERFILE_libjte_dot_h_TOO_OLD__SEE_libisofs_dot_h_AND_messages_c = 0;
+LIBJTE_MISCONFIGURATION_ = 0;
+#endif
+
+    if (! libjte__is_compatible(LIBJTE_VERSION_MAJOR, LIBJTE_VERSION_MINOR,
+                                LIBJTE_VERSION_MICRO, 0)) {
+        fprintf(stderr,
+              "\nlibisofs: libjte TOO OLD ! Need at least libjte-%d.%d.%d\n\n",
+              LIBJTE_VERSION_MAJOR, LIBJTE_VERSION_MINOR,
+              LIBJTE_VERSION_MICRO);
+        return ISO_FATAL_ERROR;
+    }
+
+#endif /* Libisofs_with_libjtE */
 
-#ifdef Libisofs_setlocale_in_iniT
     if (! (flag & 1)) {
         iso_init_locale(0);
     }
-#endif
-
     if (libiso_msgr == NULL) {
         if (libiso_msgs_new(&libiso_msgr, 0) <= 0)
             return ISO_FATAL_ERROR;
     }
     libiso_msgs_set_severities(libiso_msgr, LIBISO_MSGS_SEV_NEVER,
                    LIBISO_MSGS_SEV_FATAL, "libisofs: ", 0);
+
+    ret = iso_node_xinfo_make_clonable(aaip_xinfo_func, aaip_xinfo_cloner, 0);
+    if (ret < 0)
+        return ret;
+    ret = iso_node_xinfo_make_clonable(checksum_cx_xinfo_func,
+                                       checksum_cx_xinfo_cloner, 0);
+    if (ret < 0)
+        return ret;
+    ret = iso_node_xinfo_make_clonable(checksum_md5_xinfo_func,
+                                       checksum_md5_xinfo_cloner, 0);
+    if (ret < 0)
+        return ret;
+    ret = iso_node_xinfo_make_clonable(zisofs_zf_xinfo_func,
+                                       zisofs_zf_xinfo_cloner, 0);
+    if (ret < 0)
+        return ret;
+    ret = iso_node_xinfo_make_clonable(iso_px_ino_xinfo_func,
+                                       iso_px_ino_xinfo_cloner, 0);
+    if (ret < 0)
+        return ret;
     return 1;
 }
 
-
 int iso_init()
 {
     return iso_init_with_flag(0);
@@ -89,6 +226,7 @@ int iso_init()
 void iso_finish()
 {
     libiso_msgs_destroy(&libiso_msgr, 0);
+    iso_node_xinfo_dispose_cloners(0);
 }
 
 int iso_set_abort_severity(char *severity)
@@ -155,6 +293,8 @@ const char *iso_error_to_msg(int errcode)
         return "Try to set the boot image of an already bootable image";
     case ISO_BOOT_IMAGE_NOT_VALID:
         return "Trying to use an invalid file as boot image";
+    case ISO_BOOT_IMAGE_OVERFLOW:
+        return "Too many boot images added";
     case ISO_FILE_ERROR:
         return "Error on file operation";
     case ISO_FILE_ALREADY_OPENED:
@@ -263,11 +403,69 @@ const char *iso_error_to_msg(int errcode)
         return "Cannot set global zisofs parameters while filters exist";
     case ISO_ZLIB_EARLY_EOF:
         return "Premature EOF of zlib input stream";
+    case ISO_MD5_AREA_CORRUPTED:
+        return "Checksum area or checksum tag appear corrupted";
+    case ISO_MD5_TAG_MISMATCH:
+        return "Checksum mismatch between checksum tag and data blocks";
+    case ISO_SB_TREE_CORRUPTED:
+        return "Checksum mismatch in System Area, Volume Descriptors, or directory tree";
+    case ISO_MD5_TAG_UNEXPECTED:
+        return "Unexpected checksum tag type encountered";
+    case ISO_MD5_TAG_MISPLACED:
+        return "Misplaced checksum tag type encountered";
+    case ISO_MD5_TAG_OTHER_RANGE:
+        return "Checksum tag with unexpected address range encountered";
+    case ISO_MD5_STREAM_CHANGE:
+        return "Detected file content changes while it was written into the image";
+    case ISO_SCDBACKUP_TAG_NOT_0:
+        return "Session does not start at LBA 0. scdbackup checksum tag not written.";
+    case ISO_BOOT_NO_CATALOG:
+        return "No boot catalog created yet";
+    case ISO_OVWRT_MS_TOO_SMALL:
+        return "Multi-session offset too small for overwrite buffer";
+    case ISO_PART_OFFST_TOO_SMALL:
+        return "Partition offset too small for first tree root.";
+    case ISO_OVWRT_FIFO_TOO_SMALL:
+        return "The ring buffer is too small for overwrite buffer";
+    case ISO_LIBJTE_NOT_ENABLED:
+        return "Use of libjte was not enabled at compile time";
+    case ISO_LIBJTE_START_FAILED:
+        return "Failed to start up Jigdo Template Extraction";
+    case ISO_LIBJTE_END_FAILED:
+        return "Failed to finish Jigdo Template Extraction";
+    case ISO_LIBJTE_FILE_FAILED:
+        return "Failed to process file for Jigdo Template Extraction";
+    case ISO_BOOT_TOO_MANY_MIPS:
+        return "Too many MIPS Big Endian boot files given (max. 15)";
+    case ISO_BOOT_FILE_MISSING:
+        return "Boot file missing in image";
+    case ISO_BAD_PARTITION_NO:
+        return "Partition number out of range";
+    case ISO_BAD_PARTITION_FILE:
+        return "Cannot open data file for appended partition";
+    case ISO_NON_MBR_SYS_AREA:
+        return "May not combine appended partition with non-MBR system area";
+    case ISO_DISPLACE_ROLLOVER:
+        return "Displacement offset leads outside 32 bit range";
+    case ISO_NAME_NEEDS_TRANSL:
+        return "File name cannot be written into ECMA-119 untranslated";
+    case ISO_STREAM_NO_CLONE:
+        return "Data file input stream object offers no cloning method";
+    case ISO_XINFO_NO_CLONE:
+        return "Extended information class offers no cloning method";
     default:
         return "Unknown error";
     }
 }
 
+int iso_msg_is_abort(int errcode)
+{
+    if (ISO_ERR_SEV(errcode) >= abort_threshold)
+        return 1;
+    return 0;
+}
+       
+
 int iso_msg_submit(int imgid, int errcode, int causedby, const char *fmt, ...)
 {
     char msg[MAX_MSG_LEN];
@@ -298,7 +496,7 @@ int iso_msg_submit(int imgid, int errcode, int causedby, const char *fmt, ...)
         }
     }
 
-    if (ISO_ERR_SEV(errcode) >= abort_threshold) {
+    if (iso_msg_is_abort(errcode)) {
         return ISO_CANCELED;
     } else {
         return 0;
@@ -395,8 +593,6 @@ int iso_obtain_msgs(char *minimum_severity, int *error_code, int *imgid,
 }
 
 
-/* ts A80222 : derived from libburn/init.c:burn_msgs_submit()
-*/
 int iso_msgs_submit(int error_code, char msg_text[], int os_errno,
 			char severity[], int origin)
 {
@@ -424,8 +620,6 @@ int iso_msgs_submit(int error_code, char msg_text[], int os_errno,
 }
 
 
-/* ts A80222 : derived from libburn/init.c:burn_text_to_sev()
-*/
 int iso_text_to_sev(char *severity_name, int *sevno)
 {
     int ret;
@@ -437,8 +631,6 @@ int iso_text_to_sev(char *severity_name, int *sevno)
 }
 
 
-/* ts A80222 : derived from libburn/init.c:burn_sev_to_text()
-*/
 int iso_sev_to_text(int severity_number, char **severity_name)
 {
     int ret;
@@ -477,7 +669,6 @@ int iso_error_get_code(int e)
 }
 
 
-/* ts A80222 */
 int iso_report_errfile(char *path, int error_code, int os_errno, int flag)
 {
     libiso_msgs_submit(libiso_msgr, 0, error_code,
@@ -485,3 +676,33 @@ int iso_report_errfile(char *path, int error_code, int os_errno, int flag)
                        path, os_errno, 0);
     return(1);
 }
+
+
+int iso_libjte_forward_msgs(void *libjte_handle,
+                            int imgid, int errcode, int flag)
+{
+
+#ifdef Libisofs_with_libjtE
+
+    char *msg = NULL;
+    int res;
+    struct libjte_env *handle = (struct libjte_env *) libjte_handle;
+
+    res = ISO_SUCCESS;
+    while(1) {
+       msg= libjte_get_next_message(handle);
+       if(msg == NULL)
+    break;
+       res = iso_msg_submit(imgid, errcode, 0, msg);
+       free(msg);
+    }
+    return res;
+
+#else /* Libisofs_with_libjtE */
+
+    return ISO_SUCCESS;
+
+#endif /* ! Libisofs_with_libjtE */
+
+}
+

libisofs/messages.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /*
@@ -24,6 +25,13 @@ extern int iso_message_id;
  */
 void iso_msg_debug(int imgid, const char *fmt, ...);
 
+
+/**
+ * Inquire whether the given error code triggers the abort threshold
+ */
+int iso_msg_is_abort(int errcode);
+
+
 /**
  * 
  * @param errcode
@@ -33,12 +41,11 @@ void iso_msg_debug(int imgid, const char *fmt, ...);
  *      < 0 will be returned in any case. Use 0 if there is no previous 
  *      cause for the error.
  * @return
- *      1 on success, < 0 if function must abort asap.
+ *      0 on success, < 0 if function must abort asap.
  */
 int iso_msg_submit(int imgid, int errcode, int causedby, const char *fmt, ...);
 
 
-/* ts A80222 */
 /* To be called with events which report incidents with individual input 
    files from the local filesystem. Not with image nodes, files containing an
    image or similar file-like objects.
@@ -46,4 +53,10 @@ int iso_msg_submit(int imgid, int errcode, int causedby, const char *fmt, ...);
 int iso_report_errfile(char *path, int error_code, int os_errno, int flag);
 
 
+/* Drains the libjte message list and puts out the messages via
+   iso_msg_submit()
+*/
+int iso_libjte_forward_msgs(void *libjte_handle,
+                            int imgid, int errcode, int flag);
+
 #endif /*MESSAGES_H_*/

libisofs/node.c

@@ -1,17 +1,24 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
- * Copyright (c) 2009 Thomas Schmitt
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
+#include "image.h"
 #include "node.h"
 #include "stream.h"
 #include "aaip_0_2.h"
 #include "messages.h"
+#include "util.h"
 
 
 #include <stdlib.h>
@@ -20,6 +27,12 @@
 #include <limits.h>
 #include <stdio.h>
 
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
 struct dir_iter_data
 {
     /* points to the last visited child, to NULL before start */
@@ -199,6 +212,7 @@ int iso_node_get_xinfo(IsoNode *node, iso_node_xinfo_func proc, void **data)
         return ISO_NULL_POINTER;
     }
 
+    *data = NULL;
     pos = node->xinfo;
     while (pos != NULL) {
         if (pos->process == proc) {
@@ -212,6 +226,89 @@ int iso_node_get_xinfo(IsoNode *node, iso_node_xinfo_func proc, void **data)
     return 0;
 }
 
+/* API */
+int iso_node_get_next_xinfo(IsoNode *node, void **handle,
+                            iso_node_xinfo_func *proc, void **data)
+{
+    IsoExtendedInfo *xinfo;
+
+    if (node == NULL || handle == NULL || proc == NULL || data == NULL)
+        return ISO_NULL_POINTER;
+    *proc = NULL;
+    *data = NULL;
+    xinfo = (IsoExtendedInfo *) *handle;
+    if (xinfo == NULL)
+        xinfo = node->xinfo;
+    else
+        xinfo = xinfo->next;
+    *handle = xinfo;
+    if (xinfo == NULL)
+        return 0;
+    *proc = xinfo->process;
+    *data = xinfo->data;
+    return ISO_SUCCESS;
+}
+
+int iso_node_remove_all_xinfo(IsoNode *node, int flag)
+{
+    IsoExtendedInfo *pos, *next;
+
+    for (pos = node->xinfo; pos != NULL; pos = next) {
+        next = pos->next;
+        pos->process(pos->data, 1);
+        free((char *) pos);
+    }
+    node->xinfo = NULL;
+    return ISO_SUCCESS;
+}
+
+static
+int iso_node_revert_xinfo_list(IsoNode *node, int flag)
+{
+
+    IsoExtendedInfo *pos, *next, *prev = NULL;
+
+    for (pos = node->xinfo; pos != NULL; pos = next) {
+        next = pos->next;
+        pos->next = prev;
+        prev = pos;
+    }
+    node->xinfo = prev;
+    return ISO_SUCCESS;
+}
+
+int iso_node_clone_xinfo(IsoNode *from_node, IsoNode *to_node, int flag)
+{
+    void *handle = NULL, *data, *new_data;
+    iso_node_xinfo_func proc;
+    iso_node_xinfo_cloner cloner;
+    int ret;
+
+    iso_node_remove_all_xinfo(to_node, 0);
+    while (1) {
+        ret = iso_node_get_next_xinfo(from_node, &handle, &proc, &data);
+        if (ret <= 0)
+    break;
+        ret = iso_node_xinfo_get_cloner(proc, &cloner, 0);
+        if (ret == 0)
+            return ISO_XINFO_NO_CLONE;
+        if (ret < 0)
+            return ret;
+        ret = (*cloner)(data, &new_data, 0);
+        if (ret < 0)
+    break;
+        ret = iso_node_add_xinfo(to_node, proc, new_data);
+        if (ret < 0)
+    break;
+    }
+    if (ret < 0) {
+        iso_node_remove_all_xinfo(to_node, 0);
+    } else {
+        ret = iso_node_revert_xinfo_list(to_node, 0);
+    }
+    return ret;
+}
+
 /**
  * Get the type of an IsoNode.
  */
@@ -418,6 +515,12 @@ void iso_node_set_hidden(IsoNode *node, int hide_attrs)
     }
 }
 
+int iso_node_get_hidden(IsoNode *node)
+{
+    return node->hidden;
+}
+
+
 /**
  * Add a new node to a dir. Note that this function don't add a new ref to
  * the node, so you don't need to free it, it will be automatically freed
@@ -631,6 +734,9 @@ int iso_node_take(IsoNode *node)
     if (dir == NULL) {
         return ISO_NODE_NOT_ADDED_TO_DIR;
     }
+
+    /* >>> Do not take root directory ! (dir == node) ? */;
+
     pos = iso_dir_find_node(dir, node);
     if (pos == NULL) {
         /* should never occur */
@@ -666,6 +772,44 @@ int iso_node_remove(IsoNode *node)
     return ret;
 }
 
+/* API */
+int iso_node_remove_tree(IsoNode *node, IsoDirIter *boss_iter)
+{
+    IsoDirIter *iter = NULL;
+    IsoNode *sub_node;
+    int ret;
+
+    if (node->type != LIBISO_DIR) {
+
+        /* >>> Do not remove root directory ! (node->parent == node) ? */;
+
+        ret = iso_dir_get_children((IsoDir *) node, &iter);
+        if (ret < 0)
+            goto ex;
+        while(1) {
+            ret = iso_dir_iter_next(iter, &sub_node);
+            if (ret == 0)
+        break;
+            ret = iso_node_remove_tree(sub_node, iter);
+            if (ret < 0)
+                goto ex;
+        }
+        if (node->parent == NULL) {
+            /* node is not grafted into a boss directory */
+            iso_node_unref(node);
+            goto ex;
+        }
+    }
+    if (boss_iter != NULL)
+        ret = iso_dir_iter_remove(boss_iter);
+    else
+        ret = iso_node_remove(node);
+ex:;
+    if (iter != NULL)
+        iso_dir_iter_free(iter);
+    return ret;
+}
+
 /*
  * Get the parent of the given iso tree node. No extra ref is added to the
  * returned directory, you must take your ref. with iso_node_ref() if you
@@ -1274,6 +1418,7 @@ int iso_node_new_file(char *name, IsoStream *stream, IsoFile **file)
     new->node.type = LIBISO_FILE;
     new->node.name = name;
     new->node.mode = S_IFREG;
+    new->sort_weight = 0;
     new->stream = stream;
 
     *file = new;
@@ -1308,6 +1453,9 @@ int iso_node_new_symlink(char *name, char *dest, IsoSymlink **link)
     new->node.name = name;
     new->dest = dest;
     new->node.mode = S_IFLNK;
+    new->fs_id = 0;
+    new->st_dev = 0;
+    new->st_ino = 0;
     *link = new;
     return ISO_SUCCESS;
 }
@@ -1339,7 +1487,9 @@ int iso_node_new_special(char *name, mode_t mode, dev_t dev,
 
     new->node.mode = mode;
     new->dev = dev;
-
+    new->fs_id = 0;
+    new->st_dev = 0;
+    new->st_ino = 0;
     *special = new;
     return ISO_SUCCESS;
 }
@@ -2127,6 +2277,23 @@ int zisofs_zf_xinfo_func(void *data, int flag)
     return 1;
 }
 
+/* The iso_node_xinfo_cloner function which gets associated to
+ * zisofs_zf_xinfo_func by iso_init() resp. iso_init_with_flag() via
+ * iso_node_xinfo_make_clonable()
+ */
+int zisofs_zf_xinfo_cloner(void *old_data, void **new_data, int flag)
+{
+    *new_data = NULL;
+    if (flag)
+        return ISO_XINFO_NO_CLONE;
+    if (old_data == NULL)
+        return 0;
+    *new_data = calloc(1, sizeof(struct zisofs_zf_info));
+    if (*new_data == NULL)
+        return ISO_OUT_OF_MEM;
+    memcpy(*new_data, old_data, sizeof(struct zisofs_zf_info));
+    return (int) sizeof(struct zisofs_zf_info);
+}
 
 /* Checks whether a file effectively bears a zisofs file header and eventually
  * marks this by a struct zisofs_zf_info as xinfo of the file node.
@@ -2239,3 +2406,491 @@ int iso_node_zf_by_magic(IsoNode *node, int flag)
     return total_ret;
 }
 
+
+int iso_px_ino_xinfo_func(void *data, int flag)
+{
+    if (flag == 1) {
+        free(data);
+    }
+    return 1;
+}
+
+/* The iso_node_xinfo_cloner function which gets associated to
+ * iso_px_ino_xinfo_func by iso_init() resp. iso_init_with_flag() via
+ * iso_node_xinfo_make_clonable()
+ */
+int iso_px_ino_xinfo_cloner(void *old_data, void **new_data, int flag)
+{
+    *new_data = NULL;
+    if (flag)
+        return ISO_XINFO_NO_CLONE; 
+    *new_data = calloc(1, sizeof(ino_t));
+    if (*new_data == NULL)
+        return ISO_OUT_OF_MEM;
+    memcpy(*new_data, old_data, sizeof(ino_t));
+    return (int) sizeof(ino_t);
+}
+
+/*
+ * @param flag
+ *     bit0= do only retrieve id if node is in imported ISO image
+ *           or has an explicit xinfo inode number
+ * @return
+ *     1= reply is valid from stream, 2= reply is valid from xinfo
+ *     0= no id available,           <0= error
+ *     (fs_id, dev_id, ino_id) will be (0,0,0) in case of return <= 0
+ */
+int iso_node_get_id(IsoNode *node, unsigned int *fs_id, dev_t *dev_id,
+                    ino_t *ino_id, int flag)
+{
+    int ret;
+    IsoFile *file;
+    IsoSymlink *symlink;
+    IsoSpecial *special;
+    void *xipt;
+    
+    ret = iso_node_get_xinfo(node, iso_px_ino_xinfo_func, &xipt);
+    if (ret < 0)
+        goto no_id;
+    if (ret == 1) {
+        *fs_id = ISO_IMAGE_FS_ID;
+        *dev_id = 0;
+        *ino_id = *((ino_t *) xipt);
+        return 2;
+    }
+
+    if (node->type == LIBISO_FILE) {
+        file= (IsoFile *) node;
+        iso_stream_get_id(file->stream, fs_id, dev_id, ino_id);
+        if (*fs_id != ISO_IMAGE_FS_ID && (flag & 1)) {
+            ret = 0;
+            goto no_id;
+        }
+        return 1;
+
+    } else if (node->type == LIBISO_SYMLINK) {
+        symlink = (IsoSymlink *) node;
+        if (symlink->fs_id != ISO_IMAGE_FS_ID && (flag & 1)) {
+            ret = 0;
+            goto no_id;
+        }
+        *fs_id = symlink->fs_id;
+        *dev_id = symlink->st_dev;
+        *ino_id = symlink->st_ino;
+        return 1;
+
+    } else if (node->type == LIBISO_SPECIAL) {
+        special = (IsoSpecial *) node;
+        if (special->fs_id != ISO_IMAGE_FS_ID && (flag & 1)) {
+            ret = 0;
+            goto no_id;
+        }
+        *fs_id = special->fs_id;
+        *dev_id = special->st_dev;
+        *ino_id = special->st_ino;
+        return 1;
+
+    }
+
+    ret = 0;
+no_id:;
+    *fs_id = 0;
+    *dev_id = 0;
+    *ino_id = 0;
+    return ret;
+}
+
+
+static
+int iso_node_set_ino_xinfo(IsoNode *node, ino_t ino, int flag)
+{
+    int ret;
+    void *xipt;
+
+    if (flag & 1) {
+        ret = iso_node_remove_xinfo(node, iso_px_ino_xinfo_func);
+        if (ret < 0)
+            return ret;
+    }
+    xipt = calloc(1, sizeof(ino_t));
+    if (xipt == NULL)
+        return ISO_OUT_OF_MEM;
+    memcpy(xipt, &ino, sizeof(ino_t));
+    ret = iso_node_add_xinfo(node, iso_px_ino_xinfo_func, xipt);
+    return ret;
+}
+
+int iso_node_set_ino(IsoNode *node, ino_t ino, int flag)
+{
+    int ret;
+    IsoFile *file;
+    IsoSymlink *symlink;
+    IsoSpecial *special;
+    void *xipt;
+    
+    ret = iso_node_get_xinfo(node, iso_px_ino_xinfo_func, &xipt);
+    if (ret < 0)
+        return ret;
+    if (ret == 1) {
+        ret = iso_node_set_ino_xinfo(node, ino, 1);
+        if (ret < 0)
+            return ret;
+        return 2;
+    }
+    if (node->type == LIBISO_FILE) {
+        file= (IsoFile *) node;
+        ret = iso_stream_set_image_ino(file->stream, ino, 0);
+        if (ret < 0 || ret == 1)
+            return ret;
+
+    } else if (node->type == LIBISO_SYMLINK) {
+        symlink = (IsoSymlink *) node;
+        if (symlink->fs_id == ISO_IMAGE_FS_ID) {
+            symlink->st_ino = ino;
+            return 1;
+        }
+
+    } else if (node->type == LIBISO_SPECIAL) {
+        special = (IsoSpecial *) node;
+        if (special->fs_id == ISO_IMAGE_FS_ID) {
+            special->st_ino = ino;
+            return 1;
+        }
+
+    }
+    ret = iso_node_set_ino_xinfo(node, ino, 0);
+    if (ret < 0)
+        return ret;
+    return 2;
+}
+
+
+int iso_node_set_unique_id(IsoNode *node, IsoImage *image, int flag)
+{
+    int ret;
+    ino_t ino;
+
+    ino = img_give_ino_number(image, 0);
+    ret = iso_node_set_ino(node, ino, 0);
+    return ret;
+}
+
+/*
+ * Note to programmers: It is crucial not to break the following constraints.
+ * Anti-symmetry: cmp(X,Y) == - cmp(Y,X)
+ * Transitivity : if cmp(A,B) < 0 && cmp(B,C) < 0 then cmp(A,C) < 0
+ *                if cmp(A,B) == 0 && cmp(B,C) == 0 then cmp(A,C) == 0
+ * A big transitivity hazard are tests which do not apply to some nodes.
+ * In this case for any A that is applicable and any B that is not applicable
+ * the comparison must have the same non-zero result. I.e. a pair of applicable
+ * and non-applicable node must return that non-zero result before the test
+ * for a pair of applicable nodes would happen.
+ * 
+ * @param flag
+ *     bit0= compare stat properties and attributes 
+ *     bit1= treat all nodes with image ino == 0 as unique
+ */
+int iso_node_cmp_flag(IsoNode *n1, IsoNode *n2, int flag)
+{
+    int ret1, ret2;
+    unsigned int fs_id1, fs_id2;
+    dev_t dev_id1, dev_id2;
+    ino_t ino_id1, ino_id2;
+    IsoFile *f1 = NULL, *f2 = NULL;
+    IsoSymlink *l1 = NULL, *l2 = NULL;
+    IsoSpecial *s1 = NULL, *s2 = NULL;
+    void *x1, *x2;
+
+    if (n1 == n2)
+        return 0;
+    if (n1->type != n2->type)
+        return (n1->type < n2->type ? -1 : 1);
+
+    /* Imported or explicite ISO image node id has priority */
+    ret1 = (iso_node_get_id(n1, &fs_id1, &dev_id1, &ino_id1, 1) > 0);
+    ret2 = (iso_node_get_id(n2, &fs_id2, &dev_id2, &ino_id2, 1) > 0);
+    if (ret1 != ret2)
+        return (ret1 < ret2 ? -1 : 1);
+    if (ret1) {
+        /* fs_id and dev_id do not matter here.
+           Both nodes have explicit inode numbers of the emerging image.
+         */
+        if (ino_id1 != ino_id2)
+            return (ino_id1 < ino_id2 ? -1 : 1);
+        if (ino_id1 == 0) /* Image ino 0 is always unique */
+            return (n1 < n2 ? -1 : 1);
+        goto image_inode_match;
+    }
+
+    if (n1->type == LIBISO_FILE) {
+
+        f1 = (IsoFile *) n1;
+        f2 = (IsoFile *) n2;
+        ret1 = iso_stream_cmp_ino(f1->stream, f2->stream, 0);
+        if (ret1)
+            return ret1;
+        goto inode_match;
+
+    } else if (n1->type == LIBISO_SYMLINK) {
+
+        l1 = (IsoSymlink *) n1;
+        l2 = (IsoSymlink *) n2;
+        fs_id1 = l1->fs_id;
+        dev_id1 = l1->st_dev;
+        ino_id1 = l1->st_ino;
+        fs_id2 = l2->fs_id;
+        dev_id2 = l2->st_dev;
+        ino_id2 = l2->st_ino;
+
+    } else if (n1->type == LIBISO_SPECIAL) {
+
+        s1 = (IsoSpecial *) n1;
+        s2 = (IsoSpecial *) n2;
+        fs_id1 = s1->fs_id;
+        dev_id1 = s1->st_dev;
+        ino_id1 = s1->st_ino;
+        fs_id2 = s2->fs_id;
+        dev_id2 = s2->st_dev;
+        ino_id2 = s2->st_ino;
+
+    } else {
+        return (n1 < n2 ? -1 : 1); /* case n1 == n2 is handled above */
+    }
+    if (fs_id1 != fs_id2)
+        return (fs_id1 < fs_id2 ? -1 : 1);
+    if (dev_id1 != dev_id2)
+        return (dev_id1 < dev_id2 ? -1 : 1);
+    if (ino_id1 != ino_id2)
+        return (ino_id1 < ino_id2 ? -1 : 1);
+    if (fs_id1 == 0 && dev_id1 == 0 && ino_id1 == 0)
+        return (n1 < n2 ? -1 : 1);
+
+inode_match:;
+
+    if (flag & 2) {
+        /* What comes here has no predefined image ino resp. image_ino == 0 .
+           Regard this as not equal.
+        */
+        return (n1 < n2 ? -1 : 1);
+    }
+
+image_inode_match:;
+
+    if (!(flag & 1))
+        return 0;
+    if (n1->type == LIBISO_SYMLINK) {
+        l1 = (IsoSymlink *) n1;
+        l2 = (IsoSymlink *) n2;
+        ret1 = strcmp(l1->dest, l2->dest);
+        if (ret1)
+            return ret1;
+    } else if (n1->type == LIBISO_SPECIAL) {
+        s1 = (IsoSpecial *) n1;
+        s2 = (IsoSpecial *) n2;
+        if (s1->dev != s2->dev)
+            return (s1->dev < s2->dev ? -1 : 1);
+    }
+
+    if (n1->mode != n2->mode)
+        return (n1->mode < n2->mode ? -1 : 1);
+    if (n1->uid != n2->uid)
+        return (n1->uid < n2->uid ? -1 : 1);
+    if (n1->gid != n2->gid)
+        return (n1->gid < n2->gid ? -1 : 1);
+    if (n1->atime != n2->atime)
+        return (n1->atime < n2->atime ? -1 : 1);
+    if (n1->mtime != n2->mtime)
+        return (n1->mtime < n2->mtime ? -1 : 1);
+    if (n1->ctime != n2->ctime)
+        return (n1->ctime < n2->ctime ? -1 : 1);
+
+    /* Compare xinfo */
+    /* :( cannot compare general xinfo because data length is not known :( */
+
+    /* compare aa_string */
+    ret1 = iso_node_get_xinfo(n1, aaip_xinfo_func, &x1);
+    ret2 = iso_node_get_xinfo(n2, aaip_xinfo_func, &x2);
+    if (ret1 != ret2)
+        return (ret1 < ret2 ? -1 : 1);
+    if (ret1 == 1) {
+        ret1 = aaip_count_bytes((unsigned char *) x1, 0);
+        ret2 = aaip_count_bytes((unsigned char *) x2, 0);
+        if (ret1 != ret2)
+            return (ret1 < ret2 ? -1 : 1);
+        ret1 = memcmp(x1, x2, ret1);
+        if (ret1)
+            return ret1;
+    }
+
+    return 0;
+}
+
+/* API */
+int iso_node_cmp_ino(IsoNode *n1, IsoNode *n2, int flag)
+{
+    return iso_node_cmp_flag(n1, n2, 1);
+}
+
+
+int iso_file_set_isofscx(IsoFile *file, unsigned int checksum_index,
+                         int flag)
+{
+    static char *names = "isofs.cx";
+    static size_t value_lengths[1] = {4};
+    unsigned char value[4];
+    char *valuept;
+    int i, ret;
+
+    for(i = 0; i < 4; i++)
+        value[3 - i] = (checksum_index >> (8 * i)) & 0xff;
+    valuept= (char *) value;
+    ret = iso_node_set_attrs((IsoNode *) file, (size_t) 1,
+                             &names, value_lengths, &valuept, 2 | 8);
+    return ret;
+}
+
+
+int iso_root_set_isofsca(IsoNode *node, uint32_t start_lba, uint32_t end_lba,
+                         uint32_t count, uint32_t size, char *typetext,
+                         int flag)
+{
+    char buffer[5 + 5 + 5 + 2 + 81], *wpt = buffer, *valuept = buffer;
+    int result_len, ret;
+    static char *names = "isofs.ca";
+    static size_t value_lengths[1];
+
+    /* Set value of isofs.ca with
+       4 byte START, 4 byte END, 4 byte COUNT, SIZE = 16,  MD5 */
+    iso_util_encode_len_bytes(start_lba, wpt, 4, &result_len, 0);
+    wpt += result_len;
+    iso_util_encode_len_bytes(end_lba, wpt, 4, &result_len, 0);
+    wpt += result_len;
+    iso_util_encode_len_bytes(count, wpt, 4, &result_len, 0);
+    wpt += result_len;
+    iso_util_encode_len_bytes(size, wpt, 1, &result_len, 0);
+    wpt += result_len;
+    strncpy(wpt, typetext, 80);
+    if (strlen(typetext) > 80)
+        wpt += 80;
+    else
+        wpt += strlen(typetext);
+    value_lengths[0] = wpt - buffer;
+    ret = iso_node_set_attrs(node, (size_t) 1,
+                             &names, value_lengths, &valuept, 2 | 8);
+    return ret;
+}
+
+
+int iso_root_get_isofsca(IsoNode *node, uint32_t *start_lba, uint32_t *end_lba,
+                         uint32_t *count, uint32_t *size, char typetext[81],
+                         int flag)
+{
+    int ret, len;
+    size_t value_len;
+    char *value = NULL, *rpt;
+
+    ret = iso_node_lookup_attr(node, "isofs.ca", &value_len, &value, 0);
+    if (ret <= 0)
+        goto ex;
+
+    /* Parse value of isofs.ca with
+       4 byte START, 4 byte END, 4 byte COUNT, SIZE = 16,  MD5 */
+    rpt = value;
+    iso_util_decode_len_bytes(start_lba, rpt, &len,
+                              value_len - (rpt - value), 0);
+    rpt += len + 1;
+    iso_util_decode_len_bytes(end_lba, rpt, &len,
+                              value_len - (rpt - value), 0);
+    rpt += len + 1;
+    iso_util_decode_len_bytes(count, rpt, &len,
+                              value_len - (rpt - value), 0);
+    rpt += len + 1;
+    iso_util_decode_len_bytes(size, rpt, &len,
+                              value_len - (rpt - value), 0);
+    rpt += len + 1;
+    len = value_len - (rpt - value);
+    if (len > 80)
+        len = 80;
+    memcpy(typetext, rpt, len);
+    typetext[len] = 0;
+
+    ret= ISO_SUCCESS;
+ex:;
+    if (value != NULL)
+        free(value);
+    return ret;
+}
+
+
+/* API */
+int iso_file_get_md5(IsoImage *image, IsoFile *file, char md5[16], int flag)
+{
+    int ret, i;
+    size_t value_len;
+    char *value = NULL;
+    uint32_t idx = 0;
+    void *xipt;
+
+    /* xinfo MD5 overrides everything else */
+    ret = iso_node_get_xinfo((IsoNode *) file, checksum_md5_xinfo_func, &xipt);
+    if (ret == 1) {
+        memcpy(md5, (char *) xipt, 16);
+        return 1;
+    }
+
+    if (image->checksum_array == NULL)
+        return 0;
+    ret = iso_node_lookup_attr((IsoNode *) file, "isofs.cx",
+                               &value_len, &value, 0);
+    if (ret <= 0)
+        goto ex;
+    if (value_len > 4) {
+        ret = 0;
+        goto ex;
+    }
+    for (i = 0; i < value_len; i++)
+        idx = (idx << 8) | ((unsigned char *) value)[i];
+    if (idx == 0 || idx > image->checksum_idx_count - 1) {
+                                       /* (last index is not MD5 of a file) */
+        ret = 0;
+        goto ex;
+    }
+    if (!(flag & 1)) {
+        memcpy(md5, image->checksum_array + ((size_t) 16) * ((size_t) idx),
+               16);
+    }
+    ret = 1;
+ex:;
+    if (value != NULL)
+        free(value);
+    return ret;
+}
+
+
+/* API */
+int iso_file_make_md5(IsoFile *file, int flag)
+{
+    int ret, dig = 0;
+    char *md5 = NULL;
+
+    if (file->from_old_session)
+        dig = 1;
+    md5= calloc(16, 1);
+    ret = iso_stream_make_md5(file->stream, md5, dig);
+    if (ret < 0)
+        goto ex;
+    iso_node_remove_xinfo((IsoNode *) file, checksum_md5_xinfo_func);
+    ret = iso_node_add_xinfo((IsoNode *) file, checksum_md5_xinfo_func, md5);
+    if (ret == 0)
+        ret = ISO_ERROR; /* should not happen after iso_node_remove_xinfo() */
+    if (ret < 0) {
+        free(md5);
+        goto ex;
+    }
+    ret = 1;
+ex:;
+    return ret;
+}
+
+
+

libisofs/node.h

@@ -1,9 +1,11 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_NODE_H_
 #define LIBISO_NODE_H_
@@ -18,12 +20,20 @@
 #include <sys/types.h>
 #include <sys/stat.h>
 #include <unistd.h>
+
+#ifdef HAVE_STDINT_H
 #include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
+
 
 /**
  * The extended information is a way to attach additional information to each
  * IsoNode. External applications may want to use this extension system to
- * store application speficic information related to each node. On the other
+ * store application specific information related to each node. On the other
  * side, libisofs may make use of this struct to attach information to nodes in
  * some particular, uncommon, cases, without incrementing the size of the
  * IsoNode struct.
@@ -66,7 +76,7 @@ struct iso_extended_info {
 struct Iso_Node
 {
     /*
-     * Initilized to 1, originally owned by user, until added to another node.
+     * Initialized to 1, originally owned by user, until added to another node.
      * Then it is owned by the parent node, so the user must take his own ref
      * if needed. With the exception of the creation functions, none of the
      * other libisofs functions that return an IsoNode increment its
@@ -111,6 +121,7 @@ struct Iso_Dir
     IsoNode *children; /**< list of children. ptr to first child */
 };
 
+/* IMPORTANT: Any change must be reflected by iso_tree_clone_file. */
 struct Iso_File
 {
     IsoNode node;
@@ -122,7 +133,7 @@ struct Iso_File
      * Higher weighting files are written at the beginning of image
      */
     int sort_weight;
-    IsoStream *stream;
+    IsoStream *stream;                    /* Knows fs_id, st_dev, and st_ino */
 };
 
 struct Iso_Symlink
@@ -130,12 +141,28 @@ struct Iso_Symlink
     IsoNode node;
 
     char *dest;
+
+    /* If the IsoNode represents an object in an existing filesystem then
+       the following three numbers should unique identify it.
+       (0,0,0) will always be taken as unique.
+     */
+    unsigned int fs_id;
+    dev_t st_dev;
+    ino_t st_ino;
 };
 
 struct Iso_Special
 {
     IsoNode node;
     dev_t dev;
+
+    /* If the IsoNode represents an object in an existing filesystem then
+       the following three numbers should unique identify it.
+       (0,0,0) will always be taken as unique.
+     */
+    unsigned int fs_id;
+    dev_t st_dev;
+    ino_t st_ino;
 };
 
 struct iso_dir_iter_iface
@@ -398,5 +425,92 @@ struct zisofs_zf_info {
  */ 
 int iso_file_zf_by_magic(IsoFile *file, int flag);
 
+/*
+ * @param flag
+ *     bit0= do only retrieve id if node is in imported ISO image
+ *           or has an explicit xinfo inode number
+ * @return
+ *     1= reply is valid from stream, 2= reply is valid from xinfo
+ *     0= no id available,           <0= error
+ *     (fs_id, dev_id, ino_id) will be (0,0,0) in case of return <= 0
+ */
+int iso_node_get_id(IsoNode *node, unsigned int *fs_id, dev_t *dev_id,
+                    ino_t *ino_id, int flag);
+
+/* Set a new unique inode ISO image number to the given node.
+ * This number shall eventually persist during image generation.
+ */
+int iso_node_set_unique_id(IsoNode *node, IsoImage *image, int flag);
+
+/* Use this with extreme care. Duplicate inode numbers will indicate hardlink
+ * relationship between the nodes.
+ */
+int iso_node_set_ino(IsoNode *node, ino_t ino, int flag);
+
+/*  
+ * @param flag 
+ *     bit0= compare stat properties and attributes
+ *     bit1= treat all nodes with image ino == 0 as unique
+ *           (those with 0,0,0 are treated as unique anyway)
+ */
+int iso_node_cmp_flag(IsoNode *n1, IsoNode *n2, int flag);
+
+
+/**
+ * Set the checksum index (typically comming from IsoFileSrc.checksum_index)
+ * of a regular file node. The index is encoded as xattr "isofs.cx" with
+ * four bytes of value.
+ */
+int iso_file_set_isofscx(IsoFile *file, unsigned int checksum_index,
+                         int flag);
+
+
+/**
+ * Set the checksum area description. node should be the root node.
+ * It is encoded as xattr "isofs.ca".
+ */
+int iso_root_set_isofsca(IsoNode *node, uint32_t start_lba, uint32_t end_lba,
+                         uint32_t count, uint32_t size, char *typetext,
+                         int flag);
+
+/**
+ * Get the checksum area description. node should be the root node.
+ * It is encoded as xattr "isofs.ca".
+ */
+int iso_root_get_isofsca(IsoNode *node, uint32_t *start_lba, uint32_t *end_lba,
+                         uint32_t *count, uint32_t *size, char typetext[81],
+                         int flag);
+
+
+/**
+ * Copy the xinfo list from one node to the another.
+ */
+int iso_node_clone_xinfo(IsoNode *from_node, IsoNode *to_node, int flag);
+
+
+/**
+ * The iso_node_xinfo_func instance which governs the storing of the inode
+ * number from Rock Ridge field PX.
+ */
+int iso_px_ino_xinfo_func(void *data, int flag);
+
+/* The iso_node_xinfo_cloner function which gets associated to
+ * iso_px_ino_xinfo_func by iso_init() resp. iso_init_with_flag() via
+ * iso_node_xinfo_make_clonable()
+ */
+int iso_px_ino_xinfo_cloner(void *old_data, void **new_data, int flag);
+
+
+/* Function to identify and manage ZF parameters of zisofs compression.
+ * data is supposed to be a pointer to struct zisofs_zf_info
+ */
+int zisofs_zf_xinfo_func(void *data, int flag);
+
+/* The iso_node_xinfo_cloner function which gets associated to
+ * zisofs_zf_xinfo_func by iso_init() resp. iso_init_with_flag() via
+ * iso_node_xinfo_make_clonable()
+ */
+int zisofs_zf_xinfo_cloner(void *old_data, void **new_data, int flag);
+
 
 #endif /*LIBISO_NODE_H_*/

libisofs/rockridge.c

@@ -4,10 +4,18 @@
  * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
+#include <string.h>
+#include <stdio.h>
+
 #include "rockridge.h"
 #include "node.h"
 #include "ecma119_tree.h"
@@ -17,7 +25,12 @@
 #include "aaip_0_2.h"
 #include "libisofs.h"
 
-#include <string.h>
+
+#ifdef Libisofs_with_rrip_rR
+#define ISO_ROCKRIDGE_IN_DIR_REC 128
+#else
+#define ISO_ROCKRIDGE_IN_DIR_REC 124
+#endif
 
 
 static
@@ -102,18 +115,18 @@ int rrip_add_PX(Ecma119Image *t, Ecma119Node *n, struct susp_info *susp)
 
     PX[0] = 'P';
     PX[1] = 'X';
-    if (!t->rrip_version_1_10) {
+    if (t->rrip_1_10_px_ino || !t->rrip_version_1_10 ) {
         PX[2] = 44;
     } else {
         PX[2] = 36;
     }
     PX[3] = 1;
-    iso_bb(&PX[4], px_get_mode(t, n), 4);
+    iso_bb(&PX[4], (uint32_t) px_get_mode(t, n), 4);
-    iso_bb(&PX[12], n->nlink, 4);
+    iso_bb(&PX[12], (uint32_t) n->nlink, 4);
-    iso_bb(&PX[20], px_get_uid(t, n), 4);
+    iso_bb(&PX[20], (uint32_t) px_get_uid(t, n), 4);
-    iso_bb(&PX[28], px_get_gid(t, n), 4);
+    iso_bb(&PX[28], (uint32_t) px_get_gid(t, n), 4);
-    if (!t->rrip_version_1_10) {
+    if (t->rrip_1_10_px_ino || !t->rrip_version_1_10) {
-        iso_bb(&PX[36], n->ino, 4);
+        iso_bb(&PX[36], (uint32_t) n->ino, 4);
     }
 
     return susp_append(t, susp, PX);
@@ -179,7 +192,8 @@ int rrip_add_PL(Ecma119Image *t, Ecma119Node *n, struct susp_info *susp)
     PL[3] = 1;
 
     /* write the location of the real parent, already computed */
-    iso_bb(&PL[4], n->info.dir->real_parent->info.dir->block, 4);
+    iso_bb(&PL[4],
+       n->info.dir->real_parent->info.dir->block - t->eff_partition_offset, 4);
     return susp_append(t, susp, PL);
 }
 
@@ -242,10 +256,10 @@ int rrip_add_PN(Ecma119Image *t, Ecma119Node *n, struct susp_info *susp)
     */
     if (sizeof(node->dev) > 4) {
         high_shift = 32;
-        iso_bb(&PN[4], node->dev >> high_shift, 4);
+        iso_bb(&PN[4], (uint32_t) (node->dev >> high_shift), 4);
     } else
         iso_bb(&PN[4], 0, 4);
-    iso_bb(&PN[12], node->dev & 0xffffffff, 4);
+    iso_bb(&PN[12], (uint32_t) (node->dev & 0xffffffff), 4);
     return susp_append(t, susp, PN);
 }
 
@@ -272,7 +286,8 @@ int rrip_add_CL(Ecma119Image *t, Ecma119Node *n, struct susp_info *susp)
     CL[1] = 'L';
     CL[2] = 12;
     CL[3] = 1;
-    iso_bb(&CL[4], n->info.real_me->info.dir->block, 4);
+    iso_bb(&CL[4], n->info.real_me->info.dir->block - t->eff_partition_offset,
+           4);
     return susp_append(t, susp, CL);
 }
 
@@ -662,9 +677,6 @@ int aaip_add_ER(Ecma119Image *t, struct susp_info *susp, int flag)
     ER[5] = 81;
     ER[6] = 62;
     ER[7] = 1;
-
-#ifdef Libisofs_aaip_2_0
-
     memcpy(ER + 8, "AAIP_0200", 9);
     memcpy(ER + 17,
            "AL PROVIDES VIA AAIP 2.0 SUPPORT FOR ARBITRARY FILE ATTRIBUTES"
@@ -673,18 +685,6 @@ int aaip_add_ER(Ecma119Image *t, struct susp_info *susp, int flag)
            "PLEASE CONTACT THE LIBBURNIA PROJECT VIA LIBBURNIA-PROJECT.ORG",
            62);
 
-#else /* Libisofs_aaip_2_0 */
-
-    memcpy(ER + 8, "AAIP_0100", 9);
-    memcpy(ER + 17,
-           "AA PROVIDES VIA AAIP 1.0 SUPPORT FOR ARBITRARY FILE ATTRIBUTES"
-           " IN ISO 9660 IMAGES", 81);
-    memcpy(ER + 98,
-           "PLEASE CONTACT THE LIBBURNIA PROJECT VIA LIBBURNIA-PROJECT.ORG",
-           62);
-
-#endif /* ! Libisofs_aaip_2_0 */
-
     /** This always goes to continuation area */
     return susp_append_ce(t, susp, ER);
 }
@@ -707,9 +707,10 @@ int susp_add_CE(Ecma119Image *t, size_t ce_len, struct susp_info *susp)
     CE[1] = 'E';
     CE[2] = 28;
     CE[3] = 1;
-    iso_bb(&CE[4], susp->ce_block, 4);
+
+    iso_bb(&CE[4], susp->ce_block - t->eff_partition_offset, 4);
     iso_bb(&CE[12], susp->ce_len, 4);
-    iso_bb(&CE[20], ce_len, 4);
+    iso_bb(&CE[20], (uint32_t) ce_len, 4);
 
     return susp_append(t, susp, CE);
 }
@@ -904,7 +905,7 @@ int add_zf_field(Ecma119Image *t, Ecma119Node *n, struct susp_info *info,
     return 1;
 }
 
-
+/* API */
 int aaip_xinfo_func(void *data, int flag)
 {
     if (flag & 1) {
@@ -913,6 +914,23 @@ int aaip_xinfo_func(void *data, int flag)
     return 1;
 }
 
+/* API */
+int aaip_xinfo_cloner(void *old_data, void **new_data, int flag)
+{
+    size_t aa_size;
+
+    *new_data = NULL;
+    if (old_data == NULL)
+        return 0;
+    aa_size = aaip_count_bytes((unsigned char *) old_data, 0);
+    if (aa_size <= 0)
+        return ISO_AAIP_BAD_AASTRING;
+    *new_data = calloc(1, aa_size);
+    if (*new_data == NULL)
+        return ISO_OUT_OF_MEM;
+    memcpy(*new_data, old_data, aa_size);
+    return (int) aa_size;
+}
 
 /**
  * Compute SUA length and eventual Continuation Area length of field NM and
@@ -1104,6 +1122,44 @@ unannounced_ca:;
 }
 
 
+/* @param flag bit0= Do not add data but only count sua_free and ce_len
+                     param info may be NULL in this case
+*/
+static
+int add_aa_string(Ecma119Image *t, Ecma119Node *n, struct susp_info *info,
+                  size_t *sua_free, size_t *ce_len, int flag)
+{
+    int ret;
+    uint8_t *aapt;
+    void *xipt;
+    size_t num_aapt= 0;
+
+    if (!t->aaip)
+        return 1;
+
+    ret = iso_node_get_xinfo(n->node, aaip_xinfo_func, &xipt);
+    if (ret == 1) {
+        num_aapt = aaip_count_bytes((unsigned char *) xipt, 0);
+        if (num_aapt > 0) {
+            if (flag & 1) {
+                ret = aaip_add_AL(t, NULL,NULL, num_aapt, sua_free, ce_len, 1);
+            } else {
+                aapt = malloc(num_aapt);
+                if (aapt == NULL)
+                    return ISO_OUT_OF_MEM;
+                memcpy(aapt, xipt, num_aapt);
+                ret = aaip_add_AL(t, info, &aapt, num_aapt, sua_free, ce_len,
+                                  0);
+            }
+            if (ret < 0) 
+                return ret;
+            /* aapt is NULL now and the memory is owned by t */
+        }
+    }
+    return 1;
+}
+
+
 /**
  * Compute the length needed for write all RR and SUSP entries for a given
  * node.
@@ -1111,27 +1167,31 @@ unannounced_ca:;
  * @param type
  *      0 normal entry, 1 "." entry for that node (it is a dir), 2 ".."
  *      for that node (i.e., it will refer to the parent)
- * @param space
+ * @param used_up
- *      Available space in the System Use Area for the directory record.
+ *      Already occupied space in the directory record.
  * @param ce
  *      Will be filled with the space needed in a CE
  * @return
  *      The size needed for the RR entries in the System Use Area
  */
-size_t rrip_calc_len(Ecma119Image *t, Ecma119Node *n, int type, size_t space,
+size_t rrip_calc_len(Ecma119Image *t, Ecma119Node *n, int type, size_t used_up,
                      size_t *ce)
 {
-    size_t su_size;
+    size_t su_size, space;
     int ret;
+    size_t aaip_sua_free= 0, aaip_len= 0;
+
+    /* Directory record length must be even (ECMA-119, 9.1.13). Maximum is 254.
+    */
+    space = 254 - used_up - (used_up % 2);
+    if (type < 0 || type > 2 || space < ISO_ROCKRIDGE_IN_DIR_REC) {
+        iso_msg_submit(t->image->id, ISO_ASSERT_FAILURE, 0,
+          "Unknown node type %d or short RR space %d < %d in directory record",
+          type, (int) space, ISO_ROCKRIDGE_IN_DIR_REC);
+        return ISO_ASSERT_FAILURE;
+    }
 
-    /* space min is 255 - 33 - 37 = 185
-     * At the same time, it is always an odd number, but we need to pad it
-     * propertly to ensure the length of a directory record is a even number
-     * (ECMA-119, 9.1.13). Thus, in fact the real space is always space - 1
-     */
-    space--;
     *ce = 0;
-
     su_size = 0;
 
     /* If AAIP enabled and announced by ER : account for 5 bytes of ES */;
@@ -1144,7 +1204,7 @@ size_t rrip_calc_len(Ecma119Image *t, Ecma119Node *n, int type, size_t space,
 #endif
 
     /* PX and TF, we are sure they always fit in SUA */
-    if (!t->rrip_version_1_10) {
+    if (t->rrip_1_10_px_ino || !t->rrip_version_1_10) {
         su_size += 44 + 26;
     } else {
         su_size += 36 + 26;
@@ -1197,9 +1257,15 @@ size_t rrip_calc_len(Ecma119Image *t, Ecma119Node *n, int type, size_t space,
             } else {
                 *ce = 182;
             }
-            if (t->aaip) {
+            if (t->aaip && !t->aaip_susp_1_10) {
                 *ce += 160; /* ER of AAIP */
             }
+            /* Compute length of AAIP string of root node */
+            aaip_sua_free= 0;
+            ret = add_aa_string(t, n, NULL, &aaip_sua_free, &aaip_len, 1);
+            if (ret < 0)
+               return ret;
+            *ce += aaip_len;
         }
     }
 
@@ -1231,39 +1297,6 @@ void susp_info_free(struct susp_info* susp)
 }
 
 
-/* @param flag bit0= Do not add data but only count sua_free and ce_len
-*/
-static
-int add_aa_string(Ecma119Image *t, Ecma119Node *n, struct susp_info *info,
-                  size_t *sua_free, size_t *ce_len, int flag)
-{
-    int ret;
-    uint8_t *aapt;
-    void *xipt;
-    size_t num_aapt= 0;
-
-    if (!t->aaip)
-        return 1;
-
-    ret = iso_node_get_xinfo(n->node, aaip_xinfo_func, &xipt);
-    if (ret == 1) {
-        num_aapt = aaip_count_bytes((unsigned char *) xipt, 0);
-        if (num_aapt > 0) {
-            aapt = malloc(num_aapt);
-            if (aapt == NULL)
-                return ISO_OUT_OF_MEM;
-            memcpy(aapt, xipt, num_aapt);
-            ret = aaip_add_AL(t, info, &aapt, num_aapt, sua_free, ce_len,
-                              flag & 1);
-            if (ret < 0) 
-                return ret;
-            /* aapt is NULL now and the memory is owned by t */
-        }
-    }
-    return 1;
-}
-
-
 /**
  * Fill a struct susp_info with the RR/SUSP entries needed for a given
  * node.
@@ -1271,8 +1304,8 @@ int add_aa_string(Ecma119Image *t, Ecma119Node *n, struct susp_info *info,
  * @param type
  *      0 normal entry, 1 "." entry for that node (it is a dir), 2 ".."
  *      for that node (i.e., it will refer to the parent)
- * @param space
+ * @param used_up
- *      Available space in the System Use Area for the directory record.
+ *      Already occupied space in the directory record.
  * @param info
  *      Pointer to the struct susp_info where the entries will be stored.
  *      If some entries need to go to a Continuation Area, they will be added
@@ -1282,7 +1315,7 @@ int add_aa_string(Ecma119Image *t, Ecma119Node *n, struct susp_info *info,
  *      1 success, < 0 error
  */
 int rrip_get_susp_fields(Ecma119Image *t, Ecma119Node *n, int type,
-                         size_t space, struct susp_info *info)
+                         size_t used_up, struct susp_info *info)
 {
     int ret;
     size_t i;
@@ -1294,13 +1327,20 @@ int rrip_get_susp_fields(Ecma119Image *t, Ecma119Node *n, int type,
     size_t su_size_pd, ce_len_pd; /* predicted sizes of SUA and CA */
     int ce_is_predicted = 0;
     size_t aaip_sua_free= 0, aaip_len= 0;
+    size_t space;
 
     if (t == NULL || n == NULL || info == NULL) {
         return ISO_NULL_POINTER;
     }
-    if (type < 0 || type > 2 || space < 185) {
+
-        /* space min is 255 - 33 - 37 = 185 */
+    /* Directory record length must be even (ECMA-119, 9.1.13). Maximum is 254.
-        return ISO_WRONG_ARG_VALUE;
+    */
+    space = 254 - used_up - (used_up % 2);
+    if (type < 0 || type > 2 || space < ISO_ROCKRIDGE_IN_DIR_REC) {
+        iso_msg_submit(t->image->id, ISO_ASSERT_FAILURE, 0,
+          "Unknown node type %d or short RR space %d < %d in directory record",
+          type, (int) space, ISO_ROCKRIDGE_IN_DIR_REC);
+        return ISO_ASSERT_FAILURE;
     }
 
     if (type == 2 && n->parent != NULL) {
@@ -1309,13 +1349,6 @@ int rrip_get_susp_fields(Ecma119Image *t, Ecma119Node *n, int type,
         node = n;
     }
 
-    /* space min is 255 - 33 - 37 = 185
-     * At the same time, it is always an odd number, but we need to pad it
-     * propertly to ensure the length of a directory record is a even number
-     * (ECMA-119, 9.1.13). Thus, in fact the real space is always space - 1
-     */
-    space--;
-
     /* 
      * SP must be the first entry for the "." record of the root directory
      * (SUSP, 5.3)
@@ -1387,6 +1420,13 @@ int rrip_get_susp_fields(Ecma119Image *t, Ecma119Node *n, int type,
         }
     }
 
+    if (info->suf_len + 28 > space) {
+        iso_msg_submit(t->image->id, ISO_ASSERT_FAILURE, 0,
+         "Directory Record overflow. name='%s' , suf_len=%d > space=%d - 28\n", 
+         node->iso_name, (int) info->suf_len, (int) space);
+        return ISO_ASSERT_FAILURE;
+    }
+
     if (type == 0) {
         size_t sua_free; /* free space in the SUA */
         int nm_type = 0; /* 0 whole entry in SUA, 1 part in CE */
@@ -1673,7 +1713,7 @@ int rrip_get_susp_fields(Ecma119Image *t, Ecma119Node *n, int type,
 
             /* Compute length of AAIP string of root node */
             aaip_sua_free= 0;
-            ret = add_aa_string(t, n, info, &aaip_sua_free, &aaip_len, 1);
+            ret = add_aa_string(t, n, NULL, &aaip_sua_free, &aaip_len, 1);
             if (ret < 0)
                 goto add_susp_cleanup;
 

libisofs/rockridge.h

@@ -4,8 +4,9 @@
  * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /**
@@ -263,7 +264,9 @@ void susp_iter_free(SuspIterator *iter);
  * Fills a struct stat with the values of a Rock Ridge PX entry (RRIP, 4.1.1).
  * 
  * @return 
- *      1 on success, < 0 on error
+ *    < 0 on error
+ *      1 on success with no inode number,
+ *      2 on success with inode number,
  */
 int read_rr_PX(struct susp_sys_user_entry *px, struct stat *st);
 

libisofs/rockridge_read.c

@@ -3,8 +3,9 @@
  * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /*
@@ -12,6 +13,10 @@
  * Rock Ridge and AAIP extensions on an ECMA-119 image.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "ecma119.h"
 #include "util.h"
@@ -169,11 +174,14 @@ int read_rr_PX(struct susp_sys_user_entry *px, struct stat *st)
     st->st_nlink = iso_read_bb(px->data.PX.links, 4, NULL);
     st->st_uid = iso_read_bb(px->data.PX.uid, 4, NULL);
     st->st_gid = iso_read_bb(px->data.PX.gid, 4, NULL);
+    st->st_ino = 0;
     if (px->len_sue[0] == 44) {
         /* this corresponds to RRIP 1.12, so we have inode serial number */
         st->st_ino = iso_read_bb(px->data.PX.serial, 4, NULL);
+        /* Indicate that st_ino is valid */
+        return 2;
     }
-    return ISO_SUCCESS;
+    return 1;
 }
 
 /**
@@ -299,7 +307,7 @@ int read_rr_NM(struct susp_sys_user_entry *nm, char **name, int *cont)
         *name = realloc(*name, strlen(*name) + nm->len_sue[0] - 5 + 1);
         strncat(*name, (char*)nm->data.NM.name, nm->len_sue[0] - 5);
     } else {
-        *name = strcopy((char*)nm->data.NM.name, nm->len_sue[0] - 5);
+        *name = iso_util_strcopy((char*)nm->data.NM.name, nm->len_sue[0] - 5);
     }
     if (*name == NULL) {
         return ISO_OUT_OF_MEM;
@@ -377,7 +385,7 @@ int read_rr_SL(struct susp_sys_user_entry *sl, char **dest, int *cont)
             /* we don't have to add the '/' */
             strncat(*dest, comp, len);
         } else {
-            *dest = strcopy(comp, len);
+            *dest = iso_util_strcopy(comp, len);
         }
         if (*dest == NULL) {
             return ISO_OUT_OF_MEM;
@@ -435,7 +443,7 @@ int read_rr_PN(struct susp_sys_user_entry *pn, struct stat *st)
 }
 
 
-/* AA is the field signature of AAIP versions < 2.0
+/* AA is the obsolete field signature of AAIP versions < 2.0
 */
 int read_aaip_AA(struct susp_sys_user_entry *sue,
                  unsigned char **aa_string, size_t *aa_size, size_t *aa_len,
@@ -492,33 +500,21 @@ int read_aaip_AA(struct susp_sys_user_entry *sue,
      aapt = *aa_string + *aa_len;
 
      aapt[0] = 'A';
-
-#ifdef Libisofs_aaip_2_0
      aapt[1] = 'L';
-#else /* Libisofs_aaip_2_0 */
-     aapt[1] = 'A';
-#endif /* ! Libisofs_aaip_2_0 */
-
      aapt[2] = sue->len_sue[0];
      aapt[3] = 1;
      aapt[4] = 0;
-     /* Append sue payload */
 
-#ifdef Libisofs_aaip_2_0
+     /* Append sue payload */
      memcpy(aapt + 5, sue->data.AL.comps, sue->len_sue[0] - 5);
      *is_done = !(sue->data.AL.flags[0] & 1);
-#else /* Libisofs_aaip_2_0 */
-     memcpy(aapt + 5, sue->data.AA.comps, sue->len_sue[0] - 5);
-     *is_done = !(sue->data.AA.flags[0] & 1);
-#endif /* ! Libisofs_aaip_2_0 */
-
      *aa_len += sue->len_sue[0];
 
      return ISO_SUCCESS;
 }
 
 
-/* AL is the obsolete field signature of AAIP versions >= 2.0
+/* AL is the field signature of AAIP versions >= 2.0
 */
 int read_aaip_AL(struct susp_sys_user_entry *sue,
                  unsigned char **aa_string, size_t *aa_size, size_t *aa_len,
@@ -554,26 +550,14 @@ int read_aaip_AL(struct susp_sys_user_entry *sue,
      aapt = *aa_string + *aa_len;
 
      aapt[0] = 'A';
-
-#ifdef Libisofs_aaip_2_0
      aapt[1] = 'L';
-#else /* Libisofs_aaip_2_0 */
-     aapt[1] = 'A';
-#endif /* ! Libisofs_aaip_2_0 */
-
      aapt[2] = sue->len_sue[0];
      aapt[3] = 1;
      aapt[4] = 0;
-     /* Append sue payload */
 
-#ifdef Libisofs_aaip_2_0
+     /* Append sue payload */
      memcpy(aapt + 5, sue->data.AL.comps, sue->len_sue[0] - 5);
      *is_done = !(sue->data.AL.flags[0] & 1);
-#else /* Libisofs_aaip_2_0 */
-     memcpy(aapt + 5, sue->data.AA.comps, sue->len_sue[0] - 5);
-     *is_done = !(sue->data.AA.flags[0] & 1);
-#endif /* ! Libisofs_aaip_2_0 */
-
      *aa_len += sue->len_sue[0];
 
      return ISO_SUCCESS;

libisofs/stream.c

@@ -1,11 +1,17 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "stream.h"
 #include "fsource.h"
@@ -16,6 +22,12 @@
 #include <limits.h>
 #include <stdio.h>
 
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
 ino_t serial_id = (ino_t)1;
 ino_t mem_serial_id = (ino_t)1;
 ino_t cut_out_serial_id = (ino_t)1;
@@ -145,8 +157,64 @@ int fsrc_update_size(IsoStream *stream)
     return ISO_SUCCESS;
 }
 
+static 
+IsoStream *fsrc_get_input_stream(IsoStream *stream, int flag)
+{
+    return NULL;
+}
+
+static 
+int fsrc_cmp_ino(IsoStream *s1, IsoStream *s2)
+{
+    int ret;
+
+    ret = iso_stream_cmp_ino(s1, s2, 1);
+    return ret;
+}
+
+int fsrc_clone_stream(IsoStream *old_stream, IsoStream **new_stream,
+                      int flag)
+{
+    FSrcStreamData *data, *new_data;
+    IsoStream *stream;
+    int ret;
+
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    data = (FSrcStreamData*) old_stream->data;
+    if (data->src->class->version < 2)
+        return ISO_STREAM_NO_CLONE; /* No clone_src() method available */
+
+    *new_stream = NULL;
+    stream = calloc(1, sizeof(IsoStream));
+    if (stream == NULL)
+        return ISO_OUT_OF_MEM;
+    new_data = calloc(1, sizeof(FSrcStreamData));
+    if (new_data == NULL) {
+        free((char *) stream);
+        return ISO_OUT_OF_MEM;
+    }
+    *new_stream = stream;
+    stream->class = old_stream->class;
+    stream->refcount = 1;
+    stream->data = new_data;
+
+    ret = data->src->class->clone_src(data->src, &(new_data->src), 0);
+    if (ret < 0) {
+        free((char *) stream);
+        free((char *) new_data);
+        return ret;
+    }
+    new_data->dev_id = data->dev_id;
+    new_data->ino_id = data->ino_id;
+    new_data->size = data->size;
+
+    return ISO_SUCCESS;
+}
+
 IsoStreamIface fsrc_stream_class = {
-    1, /* update_size is defined for this stream */
+    4, /* version */
     "fsrc",
     fsrc_open,
     fsrc_close,
@@ -155,7 +223,10 @@ IsoStreamIface fsrc_stream_class = {
     fsrc_is_repeatable,
     fsrc_get_id,
     fsrc_free,
-    fsrc_update_size
+    fsrc_update_size,
+    fsrc_get_input_stream,
+    fsrc_cmp_ino,
+    fsrc_clone_stream
 };
 
 int iso_file_source_stream_new(IsoFileSource *src, IsoStream **stream)
@@ -363,11 +434,76 @@ void cut_out_free(IsoStream *stream)
     free(data);
 }
 
+static
+int cut_out_update_size(IsoStream *stream)
+{
+    return ISO_SUCCESS;
+}
+
+static 
+IsoStream* cut_out_get_input_stream(IsoStream *stream, int flag)
+{
+    return NULL;
+}
+
+static
+int cut_out_cmp_ino(IsoStream *s1, IsoStream *s2)
+{
+    int ret;
+
+    ret = iso_stream_cmp_ino(s1, s2, 1);
+    return ret;
+}
+
+static
+int cut_out_clone_stream(IsoStream *old_stream, IsoStream **new_stream,
+                      int flag)
+{
+    struct cut_out_stream *data, *new_data;
+    IsoStream *stream;
+    int ret;
+
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    data = (struct cut_out_stream *) old_stream->data;
+    if (data->src->class->version < 2)
+        return ISO_STREAM_NO_CLONE; /* No clone_src() method available */
+
+    *new_stream = NULL;
+    stream = calloc(1, sizeof(IsoStream));
+    if (stream == NULL)
+        return ISO_OUT_OF_MEM;
+    stream->refcount = 1;
+    stream->class = old_stream->class;
+    new_data = calloc(1, sizeof(struct cut_out_stream));
+    if (new_data == NULL) {
+        free((char *) stream);
+        return ISO_OUT_OF_MEM;
+    }
+    ret = data->src->class->clone_src(data->src, &(new_data->src), 0);
+    if (ret < 0) {
+        free((char *) stream);
+        free((char *) new_data);
+        return ret;
+    }
+
+    new_data->dev_id = (dev_t) 0;
+    new_data->ino_id = cut_out_serial_id++;
+    new_data->offset = data->offset;
+    new_data->size = data->size;
+    new_data->pos = 0;
+
+    stream->data = new_data;
+    *new_stream = stream;
+    return ISO_SUCCESS;
+}
+
 /*
  * TODO update cut out streams to deal with update_size(). Seems hard.
  */
 IsoStreamIface cut_out_stream_class = {
-    0,
+    4, /* version */
     "cout",
     cut_out_open,
     cut_out_close,
@@ -375,7 +511,12 @@ IsoStreamIface cut_out_stream_class = {
     cut_out_read,
     cut_out_is_repeatable,
     cut_out_get_id,
-    cut_out_free
+    cut_out_free,
+    cut_out_update_size,
+    cut_out_get_input_stream,
+    cut_out_cmp_ino,
+    cut_out_clone_stream
+    
 };
 
 int iso_cut_out_stream_new(IsoFileSource *src, off_t offset, off_t size,
@@ -537,12 +678,77 @@ void mem_free(IsoStream *stream)
 {
     MemStreamData *data;
     data = (MemStreamData*)stream->data;
-    free(data->buf);
+    if (data->buf != NULL)
+        free(data->buf);
     free(data);
 }
 
+static
+int mem_update_size(IsoStream *stream)
+{
+    return ISO_SUCCESS;
+}
+
+static 
+IsoStream* mem_get_input_stream(IsoStream *stream, int flag)
+{
+    return NULL;
+}
+
+static
+int mem_cmp_ino(IsoStream *s1, IsoStream *s2)
+{
+    int ret;
+
+    ret = iso_stream_cmp_ino(s1, s2, 1);
+    return ret;
+}
+
+static
+int mem_clone_stream(IsoStream *old_stream, IsoStream **new_stream,
+                      int flag)
+{
+    MemStreamData *data, *new_data;
+    IsoStream *stream;
+    uint8_t *new_buf = NULL;
+
+    if (flag)
+        return ISO_STREAM_NO_CLONE; /* unknown option required */
+
+    *new_stream = NULL;
+    stream = calloc(1, sizeof(IsoStream));
+    if (stream == NULL)
+        return ISO_OUT_OF_MEM;
+    stream->refcount = 1;
+    stream->class = old_stream->class;
+    new_data = calloc(1, sizeof(MemStreamData));
+    if (new_data == NULL) {
+        free((char *) stream);
+        return ISO_OUT_OF_MEM;
+    }
+    data = (MemStreamData *) old_stream->data;
+    if (data->size > 0) {
+        new_buf = calloc(1, data->size);
+        if (new_buf == NULL) {
+            free((char *) stream);
+            free((char *) new_data);
+            return ISO_OUT_OF_MEM;
+        }
+        memcpy(new_buf, data->buf, data->size);
+    }
+    new_data->buf = new_buf;
+    new_data->offset = -1;
+    new_data->ino_id = mem_serial_id++;
+    new_data->size = data->size;
+
+    stream->data = new_data;
+    *new_stream = stream;
+    return ISO_SUCCESS;
+}
+
+
 IsoStreamIface mem_stream_class = {
-    0,
+    4, /* version */
     "mem ",
     mem_open,
     mem_close,
@@ -550,7 +756,12 @@ IsoStreamIface mem_stream_class = {
     mem_read,
     mem_is_repeatable,
     mem_get_id,
-    mem_free
+    mem_free,
+    mem_update_size,
+    mem_get_input_stream,
+    mem_cmp_ino,
+    mem_clone_stream
+
 };
 
 /**
@@ -574,7 +785,7 @@ int iso_memory_stream_new(unsigned char *buf, size_t size, IsoStream **stream)
         return ISO_OUT_OF_MEM;
     }
     data = malloc(sizeof(MemStreamData));
-    if (str == NULL) {
+    if (data == NULL) {
         free(str);
         return ISO_OUT_OF_MEM;
     }
@@ -672,8 +883,12 @@ void iso_stream_get_file_name(IsoStream *stream, char *name)
 
 IsoStream *iso_stream_get_input_stream(IsoStream *stream, int flag)
 {
-    IsoStreamIface* class = stream->class;
+    IsoStreamIface* class;
 
+    if (stream == NULL) {
+        return NULL;
+    }
+    class = stream->class;
     if (class->version < 2)
         return NULL;
     return class->get_input_stream(stream, 0);
@@ -709,3 +924,241 @@ ex:;
     return path;
 }
 
+/* @return 1 = ok , 0 = not an ISO image stream , <0 = error */
+int iso_stream_set_image_ino(IsoStream *stream, ino_t ino, int flag)
+{
+    if (stream == NULL) {
+        return ISO_NULL_POINTER;
+    }
+    if (stream->class == &fsrc_stream_class) {
+        FSrcStreamData *fsrc_data = stream->data;
+        fsrc_data->ino_id = ino;
+        return 1;
+    }
+   return 0;
+}
+
+/* API */ 
+int iso_stream_cmp_ino(IsoStream *s1, IsoStream *s2, int flag)
+{
+    int ret;
+    unsigned int fs_id1, fs_id2;
+    dev_t dev_id1, dev_id2;
+    ino_t ino_id1, ino_id2;
+    off_t size1, size2;
+    FSrcStreamData *fssd1, *fssd2;
+
+
+/*
+   #define Libisofs_stream_cmp_ino_debuG 1
+*/
+#ifdef Libisofs_stream_cmp_ino_debuG
+    static int report_counter = 0;
+    static int debug = 1;
+#endif /* Libisofs_stream_cmp_ino_debuG */
+
+    if (s1 == s2)
+        return 0;
+    if (s1 == NULL)
+        return -1;
+    if (s2 == NULL)
+        return 1;
+
+    if (s1->class->version >= 3 && !(flag & 1)) {
+       /* Filters may have smarter methods to compare themselves with others */
+       ret = s1->class->cmp_ino(s1, s2);
+       return ret;
+    }
+
+    iso_stream_get_id(s1, &fs_id1, &dev_id1, &ino_id1);
+    iso_stream_get_id(s2, &fs_id2, &dev_id2, &ino_id2);
+    if (fs_id1 < fs_id2) {
+        return -1;
+    } else if (fs_id1 > fs_id2) {
+        return 1;
+    }
+    /* files belong to the same fs */
+    if (dev_id1 > dev_id2) {
+        return -1;
+    } else if (dev_id1 < dev_id2) {
+        return 1;
+    } else if (ino_id1 < ino_id2) {
+        return -1;
+    } else if (ino_id1 > ino_id2) {
+        return 1;
+    }
+    size1 = iso_stream_get_size(s1);
+    size2 = iso_stream_get_size(s2);
+    if (size1 < size2) {
+
+#ifdef Libisofs_stream_cmp_ino_debuG
+        if (debug) {
+            if (report_counter < 5)
+                fprintf(stderr,
+      "\n\nlibisofs_DEBUG : Program error: same ino but differing size\n\n\n");
+            else if (report_counter == 5)
+                fprintf(stderr,
+      "\n\nlibisofs_DEBUG : Inode error: more of same ino but differing size\n\n\n");
+            report_counter++;
+        }
+#endif /* Libisofs_stream_cmp_ino_debuG */
+
+        return -1;
+    } else if (size1 > size2) {
+
+#ifdef Libisofs_stream_cmp_ino_debuG
+        if (debug) {
+            if (report_counter < 5)
+                fprintf(stderr,
+      "\n\nlibisofs_DEBUG : Inode error: same ino but differing size\n\n\n");
+            else if (report_counter == 5)
+                fprintf(stderr,
+      "\n\nlibisofs_DEBUG : Program error: more of same ino but differing size\n\n\n");
+            report_counter++;
+        }
+#endif /* Libisofs_stream_cmp_ino_debuG */
+
+        return 1;
+    }
+
+    if (s1->class != s2->class)
+        return (s1->class < s2->class ? -1 : 1);
+    if (s1->class == &fsrc_stream_class) {
+        /* Compare eventual image data section LBA and sizes */
+        fssd1= (FSrcStreamData *) s1->data;
+        fssd2= (FSrcStreamData *) s2->data;
+        ret = iso_ifs_sections_cmp(fssd1->src, fssd2->src, 0);
+        if (ret != 0)
+            return ret;
+    }
+    if (fs_id1 == 0 && dev_id1 == 0 && ino_id1 == 0) {
+        return (s1 < s2 ? -1 : 1);
+    }
+    return 0;
+}
+
+
+/**
+ * @return
+ *     1 ok, 0 EOF, < 0 error
+ */    
+int iso_stream_read_buffer(IsoStream *stream, char *buf, size_t count,
+                           size_t *got)
+{
+    ssize_t result;
+
+    *got = 0;
+    do {
+        result = iso_stream_read(stream, buf + *got, count - *got);
+        if (result < 0) {
+            memset(buf + *got, 0, count - *got);
+            return result;
+        }
+        if (result == 0)
+            break;
+        *got += result;
+    } while (*got < count);
+
+    if (*got < count) {
+        /* eof */
+        memset(buf + *got, 0, count - *got);
+        return 0;
+    }
+    return 1;
+}
+
+/* @param flag bit0= dig out most original stream (e.g. because from old image)
+   @return 1=ok, md5 is valid,
+           0= not ok, 
+          <0 fatal error, abort 
+*/  
+int iso_stream_make_md5(IsoStream *stream, char md5[16], int flag)
+{
+    int res, is_open = 0;
+    char buffer[2048];
+    void *ctx= NULL;
+    off_t file_size;
+    uint32_t b, nblocks;
+    size_t got_bytes;
+    IsoStream *input_stream;
+
+    if (flag & 1) {
+        while(1) {
+           input_stream = iso_stream_get_input_stream(stream, 0);
+           if (input_stream == NULL)
+        break;
+           stream = input_stream;
+        }
+    }
+
+    if (! iso_stream_is_repeatable(stream))
+        return 0;
+    res = iso_md5_start(&ctx);
+    if (res < 0)
+        return res;
+    res = iso_stream_open(stream);
+    if (res < 0)
+        return 0;
+    is_open = 1;
+    file_size = iso_stream_get_size(stream);
+    nblocks = DIV_UP(file_size, 2048);
+    for (b = 0; b < nblocks; ++b) {
+        res = iso_stream_read_buffer(stream, buffer, 2048, &got_bytes);
+        if (res < 0) {
+            res = 0;
+            goto ex;
+        }
+        /* Do not use got_bytes to stay closer to IsoFileSrc processing */
+        if (file_size - b * 2048 > 2048)
+            res = 2048;
+        else
+            res = file_size - b * 2048;
+        iso_md5_compute(ctx, buffer, res);
+    }
+    res = 1;
+ex:;
+    if (is_open)
+        iso_stream_close(stream);
+    if (ctx != NULL)
+        iso_md5_end(&ctx, md5);
+    return res;
+}
+
+/* API */
+int iso_stream_clone(IsoStream *old_stream, IsoStream **new_stream, int flag)
+{
+    int ret;
+
+    if (old_stream->class->version < 4)
+        return ISO_STREAM_NO_CLONE;
+    ret = old_stream->class->clone_stream(old_stream, new_stream, 0);
+    return ret;
+}
+
+int iso_stream_clone_filter_common(IsoStream *old_stream,
+                                   IsoStream **new_stream,
+                                   IsoStream **new_input, int flag)
+{
+    IsoStream *stream, *input_stream;
+    int ret;
+
+    *new_stream = NULL;
+    *new_input = NULL;
+    input_stream = iso_stream_get_input_stream(old_stream, 0);
+    if (input_stream == NULL)
+        return ISO_STREAM_NO_CLONE;
+    stream = calloc(1, sizeof(IsoStream));
+    if (stream == NULL)
+        return ISO_OUT_OF_MEM;
+    ret = iso_stream_clone(input_stream, new_input, 0);
+    if (ret < 0) {
+        free((char *) stream);
+        return ret;
+    }
+    stream->class = old_stream->class;
+    stream->refcount = 1;
+    stream->data = NULL;
+    *new_stream = stream;
+    return ISO_SUCCESS;
+}
+

libisofs/stream.h

@@ -1,9 +1,11 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_STREAM_H_
 #define LIBISO_STREAM_H_
@@ -13,6 +15,7 @@
  */
 #include "fsource.h"
 
+/* IMPORTANT: Any change must be reflected by fsrc_clone_stream */
 typedef struct
 {
     IsoFileSource *src;
@@ -53,15 +56,6 @@ int iso_file_source_stream_new(IsoFileSource *src, IsoStream **stream);
 int iso_cut_out_stream_new(IsoFileSource *src, off_t offset, off_t size,
                            IsoStream **stream);
 
-/**
- * Create a stream for reading from a arbitrary memory buffer.
- * When the Stream refcount reach 0, the buffer is free(3).
- *
- * @return
- *      1 sucess, < 0 error
- */
-int iso_memory_stream_new(unsigned char *buf, size_t size, IsoStream **stream);
-
 /**
  * Obtain eventual zisofs ZF field entry parameters from a file source out
  * of a loaded ISO image.
@@ -74,4 +68,47 @@ int iso_stream_get_src_zf(IsoStream *stream, int *header_size_div4,
                           int *block_size_log2, uint32_t *uncompressed_size,
                           int flag);
 
+/**
+ * Set the inode number of a stream that is based on FSrcStreamData, i.e.
+ * stems from the imported ISO image.
+ * @return 1 = ok , 0 = not an ISO image stream , <0 = error
+ */
+int iso_stream_set_image_ino(IsoStream *stream, ino_t ino, int flag);
+
+
+/**
+ * Read the full required amount of data unless error or EOF occurs.
+ * Fill missing bytes by 0s.
+ * @param count   Required amount
+ * @param got     Returns number of actually read bytes
+ * @return
+ *     1 no problem encountered, 0 EOF encountered, < 0 error
+ */
+int iso_stream_read_buffer(IsoStream *stream, char *buf, size_t count,
+                           size_t *got);
+
+/**
+ * @return 1=ok, md5 is valid,
+ *        0= not ok
+ *       <0 fatal error, abort 
+ */
+int iso_stream_make_md5(IsoStream *stream, char md5[16], int flag);
+
+
+/**
+ * Create a clone of the input stream of old_stream and a roughly initialized
+ * clone of old_stream which has the same class and refcount 1. Its data
+ * pointer will be NULL and needs to be filled by an expert which knows how
+ * to clone the data of old_stream.
+ * @param old_stream  The existing stream which is in process of cloning
+ * @param new_stream  Will return the uninitialized memory object which shall
+ *                    later become the clone of old_stream.
+ * @param new_input   The clone of the input stream of old stream.
+ * @param flag        Submit 0 for now.
+ * @return            ISO_SUCCESS or an error code <0 
+ */
+int iso_stream_clone_filter_common(IsoStream *old_stream, 
+                                   IsoStream **new_stream, 
+                                   IsoStream **new_input, int flag);
+
 #endif /*STREAM_H_*/

libisofs/system_area.c

@@ -1,38 +1,695 @@
 /*
  * Copyright (c) 2008 Vreixo Formoso
+ * Copyright (c) 2010 Thomas Schmitt
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
+#include "libisofs.h"
 #include "system_area.h"
 #include "eltorito.h"
 #include "filesrc.h"
+#include "ecma119_tree.h"
+#include "image.h"
+#include "messages.h"
+#include "ecma119.h"
 
 #include <string.h>
-
+#include <stdio.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <unistd.h>
 
 
 /*
  * Create a MBR for an isohybrid enabled ISOLINUX boot image.
- *
+ * See libisofs/make_isohybrid_mbr.c
- * It is assumed that the caller has verified the readiness of the boot image
+ * Deprecated.
- * by checking for 0xfb 0xc0 0x78 0x70 at bytes 0x40 to 0x43 of isolinux.bin.
- *
- * @param bin_lba    The predicted LBA of isolinux.bin within the emerging ISO.
- * @param img_blocks The predicted number of 2048 byte blocks in the ISO.
- *                   It will get rounded up to full MBs and that many blocks
- *                   must really be written as ISO 9660 image.
- * @param mbr        A buffer of at least 512 bytes to take the result which is
- *                   to be written as the very beginning of the ISO.
- * @param flag    unused yet, submit 0
- * @return  <0 = fatal, 0 = failed , 1 = ok , 2 = ok with size warning
  */
 int make_isohybrid_mbr(int bin_lba, int *img_blocks, char *mbr, int flag);
 
+/*
+ * The New ISOLINUX MBR Producer.
+ * Be cautious with changing parameters. Only few combinations are tested.
+ *
+ */
+int make_isolinux_mbr(uint32_t *img_blocks, uint32_t boot_lba,
+                      uint32_t mbr_id, int head_count, int sector_count,
+                      int part_offset, int part_number, int fs_type,
+                      uint8_t *buf, int flag);
+
+
+/*
+ * @param flag bit0= img_blocks is start address rather than end address:
+                     do not subtract 1
+ */
+static
+void iso_compute_cyl_head_sec(uint32_t *img_blocks, int hpc, int sph,
+                              uint32_t *end_lba, uint32_t *end_sec,
+                              uint32_t *end_head, uint32_t *end_cyl, int flag)
+{
+    uint32_t secs;
+
+    /* Partition table unit is 512 bytes per sector, ECMA-119 unit is 2048 */
+    if (*img_blocks >= 0x40000000)
+      *img_blocks = 0x40000000 - 1;        /* truncate rather than roll over */
+    if (flag & 1)
+        secs = *end_lba = *img_blocks * 4;            /* first valid 512-lba */
+    else
+        secs = *end_lba = *img_blocks * 4 - 1;         /* last valid 512-lba */
+    *end_cyl = secs / (sph * hpc);
+    secs -= *end_cyl * sph * hpc;
+    *end_head = secs / sph;
+    *end_sec = secs - *end_head * sph + 1;   /* Sector count starts by 1 */
+    if (*end_cyl >= 1024) {
+        *end_cyl = 1023;
+        *end_head = hpc - 1;
+        *end_sec = sph;
+    }
+}
+
+
+/* Compute size and position of appended partitions.
+*/
+int iso_compute_append_partitions(Ecma119Image *t, int flag)
+{
+    int ret, i, sa_type;
+    uint32_t pos, size, add_pos = 0;
+    struct stat stbuf;
+
+    sa_type = (t->system_area_options >> 2) & 0x3f;
+    pos = (t->vol_space_size + t->ms_block);
+    for (i = 0; i < ISO_MAX_PARTITIONS; i++) {
+        if (t->appended_partitions[i] == NULL)
+    continue;
+        if (t->appended_partitions[i][0] == 0)
+    continue;
+        ret = stat(t->appended_partitions[i], &stbuf);
+        if (ret == -1)
+            return ISO_BAD_PARTITION_FILE;
+        if (! S_ISREG(stbuf.st_mode))
+            return ISO_BAD_PARTITION_FILE;
+        size = ((stbuf.st_size + 2047) / 2048);
+        if (sa_type == 3 && (pos % ISO_SUN_CYL_SIZE))
+            add_pos = ISO_SUN_CYL_SIZE - (pos % ISO_SUN_CYL_SIZE);
+        t->appended_part_prepad[i] = add_pos;
+        t->appended_part_start[i] = pos + add_pos;
+        t->appended_part_size[i] = size;
+        pos += add_pos + size;
+        t->total_size += (add_pos + size) * 2048;
+    }
+    return ISO_SUCCESS;
+}
+
+
+/* Note: partition_offset and partition_size are counted in 2048 blocks
+ */
+static int write_mbr_partition_entry(int partition_number, int partition_type,
+                  uint32_t partition_offset, uint32_t partition_size,
+                  int sph, int hpc, uint8_t *buf, int flag)
+{
+    uint8_t *wpt;
+    uint32_t end_lba, end_sec, end_head, end_cyl;
+    uint32_t start_lba, start_sec, start_head, start_cyl;
+    uint32_t after_end;
+    int i;
+
+    after_end = partition_offset + partition_size;
+    iso_compute_cyl_head_sec(&partition_offset, hpc, sph,
+                           &start_lba, &start_sec, &start_head, &start_cyl, 1);
+    iso_compute_cyl_head_sec(&after_end, hpc, sph,
+                             &end_lba, &end_sec, &end_head, &end_cyl, 0);
+    wpt = buf + 446 + (partition_number - 1) * 16;
+
+    /* Not bootable */
+    *(wpt++) = 0x00;
+
+    /* C/H/S of the start */
+    *(wpt++) = start_head;
+    *(wpt++) = start_sec | ((start_cyl & 0x300) >> 2);
+    *(wpt++) = start_cyl & 0xff;
+
+    /* (partition type) */
+    *(wpt++) = partition_type;
+
+    /* 3 bytes of C/H/S end */
+    *(wpt++) = end_head;
+    *(wpt++) = end_sec | ((end_cyl & 0x300) >> 2);
+    *(wpt++) = end_cyl & 0xff;
+    
+    /* LBA start in little endian */
+    for (i = 0; i < 4; i++)
+       *(wpt++) = (start_lba >> (8 * i)) & 0xff;
+
+    /* Number of sectors in partition, little endian */
+    end_lba = end_lba - start_lba + 1;
+    for (i = 0; i < 4; i++)
+       *(wpt++) = (end_lba >> (8 * i)) & 0xff;
+
+    /* Afaik, partition tables are recognize donly with MBR signature */
+    buf[510] = 0x55;
+    buf[511] = 0xAA;
+
+    return ISO_SUCCESS;
+}
+
+
+/* This is the gesture of grub-mkisofs --protective-msdos-label as explained by
+   Vladimir Serbinenko <phcoder@gmail.com>, 2 April 2010, on grub-devel@gnu.org
+   "Currently we use first and not last entry. You need to:
+    1) Zero-fill 446-510
+    2) Put 0x55, 0xAA into 510-512
+    3) Put 0x80 (for bootable partition), 0, 2, 0 (C/H/S of the start), 0xcd
+      (partition type), [3 bytes of C/H/S end], 0x01, 0x00, 0x00, 0x00 (LBA
+      start in little endian), [LBA end in little endian] at 446-462
+   "
+
+   "C/H/S end" means the CHS address of the last block in the partition.
+   It seems that not "[LBA end in little endian]" but "number of blocks"
+   should go into bytes 458-461. But with a start lba of 1, this is the
+   same number.
+   See also http://en.wikipedia.org/wiki/Master_boot_record
+
+   flag   bit0= do not write 0x55, 0xAA to 510,511
+          bit1= do not mark partition as bootable
+*/
+static
+int make_grub_msdos_label(uint32_t img_blocks, int sph, int hpc,
+                          uint8_t *buf, int flag)
+{
+    uint8_t *wpt;
+    uint32_t end_lba, end_sec, end_head, end_cyl;
+    int i;
+
+    iso_compute_cyl_head_sec(&img_blocks, hpc, sph,
+                             &end_lba, &end_sec, &end_head, &end_cyl, 0);
+
+    /* 1) Zero-fill 446-510 */
+    wpt = buf + 446;
+    memset(wpt, 0, 64);
+
+    if (!(flag & 1)) {
+        /* 2) Put 0x55, 0xAA into 510-512 (actually 510-511) */
+        buf[510] = 0x55;
+        buf[511] = 0xAA;
+    }
+    if (!(flag & 2)) {
+      /* 3) Put 0x80 (for bootable partition), */
+      *(wpt++) = 0x80;
+    } else {
+      *(wpt++) = 0;
+    }
+
+    /* 0, 2, 0 (C/H/S of the start), */
+    *(wpt++) = 0;
+    *(wpt++) = 2;
+    *(wpt++) = 0;
+
+    /* 0xcd (partition type) */
+    *(wpt++) = 0xcd;
+
+    /* [3 bytes of C/H/S end], */
+    *(wpt++) = end_head;
+    *(wpt++) = end_sec | ((end_cyl & 0x300) >> 2);
+    *(wpt++) = end_cyl & 0xff;
+    
+
+    /* 0x01, 0x00, 0x00, 0x00 (LBA start in little endian), */
+    *(wpt++) = 0x01;
+    *(wpt++) = 0x00;
+    *(wpt++) = 0x00;
+    *(wpt++) = 0x00;
+
+    /* [LBA end in little endian] */
+    for (i = 0; i < 4; i++)
+       *(wpt++) = (end_lba >> (8 * i)) & 0xff;
+
+    /* at 446-462 */
+    if (wpt - buf != 462) {
+        fprintf(stderr,
+        "libisofs: program error in make_grub_msdos_label: \"assert 462\"\n");
+        return ISO_ASSERT_FAILURE;
+    }
+    return ISO_SUCCESS;
+}
+
+
+/* @param flag bit0= zeroize partitions entries 2, 3, 4
+*/
+static
+int iso_offset_partition_start(uint32_t img_blocks, uint32_t partition_offset,
+                               int sph, int hpc, uint8_t *buf, int flag)
+{
+    uint8_t *wpt;
+    uint32_t end_lba, end_sec, end_head, end_cyl;
+    uint32_t start_lba, start_sec, start_head, start_cyl;
+    int i;
+
+    iso_compute_cyl_head_sec(&partition_offset, hpc, sph,
+                           &start_lba, &start_sec, &start_head, &start_cyl, 1);
+    iso_compute_cyl_head_sec(&img_blocks, hpc, sph,
+                             &end_lba, &end_sec, &end_head, &end_cyl, 0);
+    wpt = buf + 446;
+
+    /* Let pass only legal bootability values */
+    if (*wpt != 0 && *wpt != 0x80)
+        (*wpt) = 0;
+    wpt++;
+
+    /* C/H/S of the start */
+    *(wpt++) = start_head;
+    *(wpt++) = start_sec | ((start_cyl & 0x300) >> 2);
+    *(wpt++) = start_cyl & 0xff;
+
+    /* (partition type) */
+    wpt++;
+
+    /* 3 bytes of C/H/S end */
+    *(wpt++) = end_head;
+    *(wpt++) = end_sec | ((end_cyl & 0x300) >> 2);
+    *(wpt++) = end_cyl & 0xff;
+    
+    /* LBA start in little endian */
+    for (i = 0; i < 4; i++)
+       *(wpt++) = (start_lba >> (8 * i)) & 0xff;
+
+    /* Number of sectors in partition, little endian */
+    end_lba = end_lba - start_lba + 1;
+    for (i = 0; i < 4; i++)
+       *(wpt++) = (end_lba >> (8 * i)) & 0xff;
+
+    if (wpt - buf != 462) {
+        fprintf(stderr,
+    "libisofs: program error in iso_offset_partition_start: \"assert 462\"\n");
+        return ISO_ASSERT_FAILURE;
+    }
+
+    if (flag & 1) /* zeroize the other partition entries */
+        memset(wpt, 0, 3 * 16);
+
+    return ISO_SUCCESS;
+}
+
+
+static int boot_nodes_from_iso_path(Ecma119Image *t, char *path, 
+                                   IsoNode **iso_node, Ecma119Node **ecma_node,
+                                   char *purpose, int flag)
+{
+    int ret;
+
+    ret = iso_tree_path_to_node(t->image, path, iso_node);
+    if (ret < 0) {
+        iso_msg_submit(t->image->id, ISO_BOOT_FILE_MISSING, 0,
+                       "Cannot find %s '%s'", purpose, path);
+        return ISO_BOOT_FILE_MISSING;
+    }
+    if ((*iso_node)->type != LIBISO_FILE) {
+        iso_msg_submit(t->image->id, ISO_BOOT_IMAGE_NOT_VALID, 0,
+                       "Designated boot file is not a data file: '%s'", path);
+        return ISO_BOOT_IMAGE_NOT_VALID;
+    }
+
+    *ecma_node= ecma119_search_iso_node(t, *iso_node);
+    if (*ecma_node == NULL) {
+        iso_msg_submit(t->image->id, ISO_BOOT_IMAGE_NOT_VALID, 0,
+                      "Program error: IsoFile has no Ecma119Node: '%s'", path);
+        return ISO_ASSERT_FAILURE;
+    } else {
+        if ((*ecma_node)->type != ECMA119_FILE) {
+            iso_msg_submit(t->image->id, ISO_BOOT_IMAGE_NOT_VALID, 0,
+              "Program error: Ecma119Node of IsoFile is no ECMA119_FILE: '%s'",
+                          path);
+            return ISO_ASSERT_FAILURE;
+        }
+    }
+    return ISO_SUCCESS;
+}
+
+
+/* This function was implemented according to doc/boot_sectors.txt section
+   "MIPS Volume Header" which was derived by Thomas Schmitt from
+   cdrkit-1.1.10/genisoimage/boot-mips.c by Steve McIntyre which is based
+   on work of Florian Lohoff and Thiemo Seufer who possibly learned from
+   documents of MIPS Computer Systems, Inc. and Silicon Graphics Computer
+   Systems, Inc.
+   This function itself is entirely under copyright (C) 2010 Thomas Schmitt.
+*/
+static int make_mips_volume_header(Ecma119Image *t, uint8_t *buf, int flag)
+{
+    char *namept, *name_field;
+    uint32_t num_cyl, idx, blocks, num, checksum;
+    off_t image_size;
+    static uint32_t bps = 512, spt = 32;
+    Ecma119Node *ecma_node;
+    IsoNode *node;
+    IsoStream *stream;
+    off_t file_size;
+    uint32_t file_lba;
+    int ret;
+
+    /* Bytes 512 to 32767 may come from image or external file */
+    memset(buf, 0, 512);
+
+    image_size = t->curblock * 2048;
+
+    /* 0 -   3 | 0x0be5a941 | Magic number */
+    iso_msb(buf, 0x0be5a941, 4);
+
+    /* 28 -  29 |  num_cyl_l | Number of usable cylinder, lower two bytes */
+    num_cyl = (image_size + (bps * spt) - 1) / (bps * spt);
+    iso_msb(buf + 28, num_cyl & 0xffff, 2);
+
+    /* 32 -  33 |          1 | Number of tracks per cylinder */
+    iso_msb(buf + 32, 1, 2);
+
+    /* 35 -  35 |  num_cyl_h | Number of usable cylinders, high byte */
+    buf[35] = (num_cyl >> 16) & 0xff;
+    
+    /* 38 -  39 |         32 | Sectors per track */
+    iso_msb(buf + 38, spt, 2);
+
+    /* 40 -  41 |        512 | Bytes per sector */
+    iso_msb(buf + 40, bps, 2);
+
+    /* 44 -  47 | 0x00000034 | Controller characteristics */
+    iso_msb(buf + 44, 0x00000034, 4);
+
+    /*  72 -  87 | ========== | Volume Directory Entry 1 */
+    /*  72 -  79 |  boot_name | Boot file basename */
+    /*  80 -  83 | boot_block | ISO 9660 LBA of boot file * 4 */
+    /*  84 -  87 | boot_bytes | File length in bytes */
+    /*  88 - 311 |          0 | Volume Directory Entries 2 to 15 */
+
+    for (idx = 0; idx < t->image->num_mips_boot_files; idx++) {
+        ret = boot_nodes_from_iso_path(t, t->image->mips_boot_file_paths[idx], 
+                                       &node, &ecma_node, "MIPS boot file", 0);
+        if (ret < 0)
+            return ret;
+
+        namept = (char *) iso_node_get_name(node);
+        name_field = (char *) (buf + (72 + 16 * idx));
+        strncpy(name_field, namept, 8);
+
+        file_lba = ecma_node->info.file->sections[0].block;
+
+        iso_msb(buf + (72 + 16 * idx) + 8, file_lba * 4, 4);
+
+        stream = iso_file_get_stream((IsoFile *) node);
+        file_size = iso_stream_get_size(stream);
+
+        /* Shall i really round up to 2048 ? Steve says yes.*/
+        iso_msb(buf + (72 + 16 * idx) + 12,
+                ((file_size + 2047) / 2048 ) * 2048, 4);
+        
+    }
+
+    /* 408 - 411 |  part_blks | Number of 512 byte blocks in partition */
+    blocks = (image_size + bps - 1) / bps;
+    iso_msb(buf + 408, blocks, 4);
+    /* 416 - 419 |          0 | Partition is volume header */
+    iso_msb(buf + 416, 0, 4);
+
+    /* 432 - 435 |  part_blks | Number of 512 byte blocks in partition */
+    iso_msb(buf + 432, blocks, 4);
+    iso_msb(buf + 444, 6, 4);
+
+    /* 504 - 507 |   head_chk | Volume header checksum  
+                                The two's complement of bytes 0 to 503 read
+                                as big endian unsigned 32 bit:
+                                  sum(32-bit-words) + head_chk == 0
+    */
+    checksum = 0;
+    for (idx = 0; idx < 504; idx += 4) {
+        num = iso_read_msb(buf + idx, 4);
+        /* Addition modulo a natural number is commutative and associative.
+           Thus the inverse of a sum is the sum of the inverses of the addends.
+        */
+        checksum -= num;
+    }
+    iso_msb(buf + 504, checksum, 4);
+
+    return ISO_SUCCESS;
+}
+
+
+/* The following two functions were implemented according to
+   doc/boot_sectors.txt section "MIPS Little Endian" which was derived
+   by Thomas Schmitt from
+   cdrkit-1.1.10/genisoimage/boot-mipsel.c by Steve McIntyre which is based
+   on work of Florian Lohoff and Thiemo Seufer,
+   and from <elf.h> by Free Software Foundation, Inc.
+
+   Both functions are entirely under copyright (C) 2010 Thomas Schmitt.
+*/
+
+/**
+ * Read the necessary ELF information from the first MIPS boot file.
+ * This is done before image writing starts.
+ */
+int iso_read_mipsel_elf(Ecma119Image *t, int flag)
+{
+    uint32_t phdr_adr, todo, count;
+    int ret;
+    uint8_t elf_buf[2048];
+    IsoNode *iso_node;
+    Ecma119Node *ecma_node;
+    IsoStream *stream;
+    
+    if (t->image->num_mips_boot_files <= 0)
+        return ISO_SUCCESS;
+
+    ret = boot_nodes_from_iso_path(t, t->image->mips_boot_file_paths[0],
+                                   &iso_node, &ecma_node, "MIPS boot file", 0);
+    if (ret < 0)
+        return ret;
+    stream = iso_file_get_stream((IsoFile *) iso_node);
+
+    ret = iso_stream_open(stream);
+    if (ret < 0) {
+        iso_msg_submit(t->image->id, ret, 0,
+                       "Cannot open designated MIPS boot file '%s'",
+                       t->image->mips_boot_file_paths[0]);
+        return ret;
+    }
+    ret = iso_stream_read(stream, elf_buf, 32);
+    if (ret != 32) {
+cannot_read:;
+        iso_stream_close(stream);
+        iso_msg_submit(t->image->id, ret, 0,
+                       "Cannot read from designated MIPS boot file '%s'",
+                       t->image->mips_boot_file_paths[0]);
+        return ret;
+    }
+
+
+    /*  24 -  27 |    e_entry | Entry point virtual address */
+    t->mipsel_e_entry = iso_read_lsb(elf_buf + 24, 4);
+
+    /* 28 -  31 |    e_phoff | Program header table file offset */
+    phdr_adr = iso_read_lsb(elf_buf + 28, 4);
+
+    /* Skip stream up to byte address phdr_adr */
+    todo = phdr_adr - 32;
+    while (todo > 0) {
+        if (todo > 2048)
+            count = 2048;
+        else
+            count = todo;
+        todo -= count;
+        ret = iso_stream_read(stream, elf_buf, count);
+        if (ret != count)
+            goto cannot_read;
+    }
+    ret = iso_stream_read(stream, elf_buf, 20);
+    if (ret != 20)
+        goto cannot_read;
+
+    /*  4 -   7 |   p_offset | Segment file offset */
+    t->mipsel_p_offset = iso_read_lsb(elf_buf + 4, 4);
+
+    /*  8 -  11 |    p_vaddr | Segment virtual address */
+    t->mipsel_p_vaddr = iso_read_lsb(elf_buf + 8, 4);
+
+    /* 16 -  19 |   p_filesz | Segment size in file */
+    t->mipsel_p_filesz = iso_read_lsb(elf_buf + 16, 4);
+
+    iso_stream_close(stream);
+    return ISO_SUCCESS;
+}
+
+
+/**
+ * Write DEC Bootblock from previously read ELF parameters.
+ * This is done when image writing has already begun.
+ */
+static int make_mipsel_boot_block(Ecma119Image *t, uint8_t *buf, int flag)
+{
+    int ret;
+    uint32_t seg_size, seg_start;
+    IsoNode *iso_node;
+    Ecma119Node *ecma_node;
+    
+    /* Bytes 512 to 32767 may come from image or external file */
+    memset(buf, 0, 512);
+
+    if (t->image->num_mips_boot_files <= 0)
+        return ISO_SUCCESS;
+
+    ret = boot_nodes_from_iso_path(t, t->image->mips_boot_file_paths[0],
+                                   &iso_node, &ecma_node, "MIPS boot file", 0);
+    if (ret < 0)
+        return ret;
+
+    /*  8 -  11 | 0x0002757a | Magic number */
+    iso_lsb(buf + 8, 0x0002757a, 4);
+
+    /* 12 -  15 |          1 | Mode  1: Multi extent boot */
+    iso_lsb(buf + 12, 1, 4);
+
+    /* 16 -  19 |   load_adr | Load address */
+    iso_lsb(buf + 16, t->mipsel_p_vaddr, 4);
+
+    /* 20 -  23 |   exec_adr | Execution address */
+    iso_lsb(buf + 20, t->mipsel_e_entry, 4);
+
+    /* 24 -  27 |   seg_size | Segment size in file. */
+    seg_size = (t->mipsel_p_filesz + 511) / 512;
+    iso_lsb(buf + 24, seg_size, 4);
+    
+    /* 28 -  31 |  seg_start | Segment file offset */
+    seg_start = ecma_node->info.file->sections[0].block * 4
+                + (t->mipsel_p_offset + 511) / 512;
+    iso_lsb(buf + 28, seg_start, 4);
+    
+    return ISO_SUCCESS;
+}
+
+
+/* The following two functions were implemented according to
+   doc/boot_sectors.txt section "SUN Disk Label and boot images" which
+   was derived by Thomas Schmitt from
+     cdrtools-2.01.01a77/mkisofs/sunlabel.h
+     cdrtools-2.01.01a77/mkisofs/mkisofs.8
+     by Joerg Schilling
+
+   Both functions are entirely under copyright (C) 2010 Thomas Schmitt.
+*/
+
+/* @parm flag bit0= copy from next lower valid partition table entry
+ */
+static int write_sun_partition_entry(int partition_number,
+                  char *appended_partitions[],
+                  uint32_t partition_offset[], uint32_t partition_size[],
+                  uint32_t cyl_size, uint8_t *buf, int flag)
+{
+    uint8_t *wpt;
+    int read_idx, i;
+
+    if (partition_number < 1 || partition_number > 8)
+        return ISO_ASSERT_FAILURE;
+    
+    /* 142 - 173 | ========== | 8 partition entries of 4 bytes */
+    wpt = buf + 142 + (partition_number - 1) * 4;
+    if (partition_number == 1)
+        iso_msb(wpt, 4, 2);                            /* 4 = User partition */
+    else
+        iso_msb(wpt, 2, 2);            /* 2 = Root partition with boot image */
+    iso_msb(wpt + 2, 0x10, 2);              /* Permissions: 0x10 = read-only */
+
+    /* 444 - 507 | ========== | Partition table */
+    wpt = buf + 444 + (partition_number - 1) * 8;
+    read_idx = partition_number - 1;
+    if (flag & 1) {
+        /* Search next lower valid partition table entry. #1 is default */
+        for (read_idx = partition_number - 2; read_idx > 0; read_idx--)
+            if (appended_partitions[read_idx] != NULL)
+                if (appended_partitions[read_idx][0] != 0)
+        break;
+    }
+    iso_msb(wpt, partition_offset[read_idx] / (uint32_t) ISO_SUN_CYL_SIZE, 4);
+    iso_msb(wpt + 4, partition_size[read_idx] * 4, 4);
+
+    /* 510 - 511 |   checksum | The result of exoring 2-byte words 0 to 254 */
+    buf[510] = buf[511] = 0;
+    for (i = 0; i < 510; i += 2) {
+        buf[510] ^= buf[i];
+        buf[511] ^= buf[i + 1];
+    }
+
+    return ISO_SUCCESS;
+}
+
+/**
+ * Write SUN Disk Label with ISO in partition 1 and unused 2 to 8
+ */
+static int make_sun_disk_label(Ecma119Image *t, uint8_t *buf, int flag)
+{
+    int ret;
+
+    /* Bytes 512 to 32767 may come from image or external file */
+    memset(buf, 0, 512);
+
+    /* 0 - 127 |      label | ASCII Label */
+    if (t->ascii_disc_label[0])
+        strncpy((char *) buf, t->ascii_disc_label, 128);
+    else
+        strcpy((char *) buf,
+               "CD-ROM Disc with Sun sparc boot created by libisofs");
+    
+    /* 128 - 131 |          1 | Layout version */
+    iso_msb(buf + 128, 1, 4);
+
+    /* 140 - 141 |          8 | Number of partitions */
+    iso_msb(buf + 140, 8, 2);
+
+    /* 188 - 191 | 0x600ddeee | vtoc sanity */
+    iso_msb(buf + 188, 0x600ddeee, 4);
+
+    /* 420 - 421 |        350 | Rotations per minute */
+    iso_msb(buf + 420, 350, 2);
+
+    /* 422 - 423 |       2048 | Number of physical cylinders (fixely 640 MB) */
+    iso_msb(buf + 422, 2048, 2);
+
+    /* 430 - 431 |          1 | interleave factor */
+    iso_msb(buf + 430, 1, 2);
+
+    /* 432 - 433 |       2048 | Number of data cylinders (fixely 640 MB) */
+    iso_msb(buf + 432, 2048, 2);
+
+    /* 436 - 437 |          1 | Number of heads per cylinder (1 cyl = 320 kB)*/
+    iso_msb(buf + 436, 1, 2);
+
+    /* 438 - 439 |        640 | Number of sectors per head (1 head = 320 kB) */
+    iso_msb(buf + 438, 640, 2);
+
+    /* 508 - 509 |     0xdabe | Magic Number */
+    iso_msb(buf + 508, 0xdabe, 2);
+
+    /* Set partition 1 to describe ISO image and compute checksum */
+    t->appended_part_start[0] = 0;
+    t->appended_part_size[0] = t->curblock;
+    ret = write_sun_partition_entry(1, t->appended_partitions,
+                  t->appended_part_start, t->appended_part_size,
+                  ISO_SUN_CYL_SIZE, buf, 0);
+ 
+    return ISO_SUCCESS;
+}
+
+
 int iso_write_system_area(Ecma119Image *t, uint8_t *buf)
 {
+    int ret, int_img_blocks, sa_type, i, will_append = 0;
+    int first_partition = 1, last_partition = 4;
+    uint32_t img_blocks;
+
     if ((t == NULL) || (buf == NULL)) {
         return ISO_NULL_POINTER;
     }
@@ -40,26 +697,115 @@ int iso_write_system_area(Ecma119Image *t, uint8_t *buf)
     /* set buf to 0s */
     memset(buf, 0, 16 * BLOCK_SIZE);
 
-    if (t->catalog != NULL && t->catalog->image->isolinux_options & 0x02) {
+    sa_type = (t->system_area_options >> 2) & 0x3f;
-        /* We need to write a MBR for an hybrid image */
+    if (sa_type == 3) {
-        int ret;
+        first_partition = 2;
-        int img_blocks;
+        last_partition = 8;
+    }
+    for (i = first_partition - 1; i <= last_partition - 1; i++)
+        if (t->appended_partitions[i] != NULL) {
+            will_append = 1;
+    break;
+        }
 
-        img_blocks = t->curblock;
+    img_blocks = t->curblock;
-        ret = make_isohybrid_mbr(t->bootimg->sections[0].block, &img_blocks, (char*)buf, 0);
+    if (t->system_area_data != NULL) {
-
+        /* Write more or less opaque boot image */
-/* 
+        memcpy(buf, t->system_area_data, 16 * BLOCK_SIZE);
-        API description of el_torito_set_isolinux_options() prescribes
-        to pad to full MB.
-        So this is not urgent any more :
-
-        // FIXME the new img_blocks size should be taken into account
-*/
 
+    } else if (sa_type == 0 && t->catalog != NULL &&
+               (t->catalog->bootimages[0]->isolinux_options & 0x0a) == 0x02) {
+        /* Check for isolinux image with magic number of 3.72 and produce
+           an MBR from our built-in template. (Deprecated since 31 Mar 2010)
+        */
+        if (img_blocks < 0x80000000) {
+            int_img_blocks= img_blocks;
+        } else {
+            int_img_blocks= 0x7ffffff0;
+        }
+        ret = make_isohybrid_mbr(t->bootsrc[0]->sections[0].block,
+                                 &int_img_blocks, (char*)buf, 0);
         if (ret != 1) {
             /* error, it should never happen */
             return ISO_ASSERT_FAILURE;
         }
+        return ISO_SUCCESS;
     }
+    if (sa_type == 0 && (t->system_area_options & 1)) {
+        /* Write GRUB protective msdos label, i.e. a simple partition table */
+        ret = make_grub_msdos_label(img_blocks, t->partition_secs_per_head,
+                                    t->partition_heads_per_cyl, buf, 0);
+        if (ret != ISO_SUCCESS) /* error should never happen */
+            return ISO_ASSERT_FAILURE;
+    } else if(sa_type == 0 && (t->system_area_options & 2)) {
+        /* Patch externally provided system area as isohybrid MBR */
+        if (t->catalog == NULL || t->system_area_data == NULL) {
+            /* isohybrid makes only sense together with ISOLINUX boot image
+               and externally provided System Area.
+            */
+            return ISO_ISOLINUX_CANT_PATCH;
+        }
+        ret = make_isolinux_mbr(&img_blocks, t->bootsrc[0]->sections[0].block,
+                               (uint32_t) 0, t->partition_heads_per_cyl,
+                               t->partition_secs_per_head, 0, 1, 0x17, buf, 1);
+        if (ret != 1)
+            return ret;
+    } else if (sa_type == 1) {
+        ret = make_mips_volume_header(t, buf, 0);
+        if (ret != ISO_SUCCESS)
+            return ret;
+    } else if (sa_type == 2) {
+        ret = make_mipsel_boot_block(t, buf, 0);
+        if (ret != ISO_SUCCESS)
+            return ret;
+    } else if ((t->partition_offset > 0 || will_append) && sa_type == 0) {
+        /* Write a simple partition table. */
+        ret = make_grub_msdos_label(img_blocks, t->partition_secs_per_head,
+                                    t->partition_heads_per_cyl, buf, 2);
+        if (ret != ISO_SUCCESS) /* error should never happen */
+            return ISO_ASSERT_FAILURE;
+        if (t->partition_offset == 0) {
+            /* Re-write partion entry 1 : start at 0, type Linux */
+            ret = write_mbr_partition_entry(1, 0x83, 0, img_blocks,
+                        t->partition_secs_per_head, t->partition_heads_per_cyl,
+                        buf, 0);
+            if (ret < 0)
+                return ret;
+        }
+    } else if (sa_type == 3) {
+        ret = make_sun_disk_label(t, buf, 0);
+        if (ret != ISO_SUCCESS)
+            return ret;
+    }
+
+    if (t->partition_offset > 0 && sa_type == 0) {
+        /* Adjust partition table to partition offset */
+        img_blocks = t->curblock;                  /* value might be altered */
+        ret = iso_offset_partition_start(img_blocks, t->partition_offset,
+                                         t->partition_secs_per_head,
+                                         t->partition_heads_per_cyl, buf, 1);
+        if (ret != ISO_SUCCESS) /* error should never happen */
+            return ISO_ASSERT_FAILURE;
+    }
+
+    /* This eventually overwrites the partition table entries made so far */
+    for (i = first_partition - 1; i <= last_partition - 1; i++) {
+        if (t->appended_partitions[i] == NULL)
+    continue;
+        if (sa_type == 3) {
+          ret = write_sun_partition_entry(i + 1, t->appended_partitions,
+                        t->appended_part_start, t->appended_part_size,
+                        ISO_SUN_CYL_SIZE,
+                        buf, t->appended_partitions[i][0] == 0);
+        } else {
+          ret = write_mbr_partition_entry(i + 1, t->appended_part_types[i],
+                        t->appended_part_start[i], t->appended_part_size[i],
+                        t->partition_secs_per_head, t->partition_heads_per_cyl,
+                        buf, 0);
+        }
+        if (ret < 0)
+            return ret;
+    }
+
     return ISO_SUCCESS;
 }

libisofs/system_area.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2008 Vreixo Formoso
  *
  * This file is part of the libisofs project; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License version 2 as
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /*
@@ -45,4 +46,16 @@ int make_isohybrid_mbr(int bin_lba, int *img_blocks, char *mbr, int flag);
  */
 int iso_write_system_area(Ecma119Image *t, uint8_t *buf);
 
+
+/** 
+ * Read the necessary ELF information from the first MIPS boot file.
+ * See doc/boot_sectors.txt "DEC Boot Block" for "MIPS Little Endian".
+ */
+int iso_read_mipsel_elf(Ecma119Image *t, int flag);
+
+
+/* Compute size and position of appended partitions.
+*/
+int iso_compute_append_partitions(Ecma119Image *t, int flag);
+
 #endif /* SYSTEM_AREA_H_ */

libisofs/tree.c

@@ -1,15 +1,21 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2011 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 /*
  * Functions that act on the iso tree.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "libisofs.h"
 #include "node.h"
 #include "image.h"
@@ -25,6 +31,12 @@
 #include <stdio.h>
 #include <fnmatch.h>
 
+
+#ifndef PATH_MAX
+#define PATH_MAX Libisofs_default_path_maX
+#endif
+
+
 /**
  * Add a new directory to the iso tree.
  * 
@@ -914,7 +926,7 @@ int iso_tree_path_to_node(IsoImage *image, const char *path, IsoNode **node)
     int result;
     IsoNode *n;
     IsoDir *dir;
-    char *ptr, *brk_info, *component;
+    char *ptr, *brk_info = NULL, *component;
 
     if (image == NULL || path == NULL) {
         return ISO_NULL_POINTER;
@@ -982,3 +994,194 @@ char *iso_tree_get_node_path(IsoNode *node)
         return strdup(path);
     }
 }
+
+/* ------------------------- tree cloning ------------------------------ */
+
+static
+int iso_tree_copy_node_attr(IsoNode *old_node, IsoNode *new_node, int flag)
+{
+    int ret;
+
+    new_node->mode = old_node->mode;
+    new_node->uid = old_node->uid;
+    new_node->gid = old_node->gid;
+    new_node->atime = old_node->atime;
+    new_node->mtime = old_node->mtime;
+    new_node->ctime = old_node->ctime;
+    new_node->hidden = old_node->hidden;
+    ret = iso_node_clone_xinfo(old_node, new_node, 0);
+    if (ret < 0)
+        return ret;
+    return ISO_SUCCESS;
+}
+
+/*
+  @param flag bit0= merge directory with *new_node
+*/
+static
+int iso_tree_clone_dir(IsoDir *old_dir,
+                       IsoDir *new_parent, char *new_name, IsoNode **new_node,
+                       int flag)
+{
+    IsoDir *new_dir = NULL;
+    IsoNode *sub_node = NULL, *new_sub_node = NULL;
+    IsoDirIter *iter = NULL;
+    int ret;
+
+    if (flag & 1) {
+        new_dir = (IsoDir *) *new_node;
+    } else {
+        *new_node = NULL;
+        ret = iso_tree_add_new_dir(new_parent, new_name, &new_dir);
+        if (ret < 0)
+           return ret;
+    }
+    /* Avoid traversal of target directory to allow cloning of old_dir to a
+       subordinate of old_dir.
+    */
+    iso_node_take((IsoNode *) new_dir);
+
+    ret = iso_dir_get_children(old_dir, &iter);
+    if (ret < 0)
+        goto ex;
+    while(1) {
+        ret = iso_dir_iter_next(iter, &sub_node);
+        if (ret == 0)
+    break;
+        ret = iso_tree_clone(sub_node, new_dir, sub_node->name, &new_sub_node,
+                             flag & 1);
+        if (ret < 0)
+            goto ex;
+    }
+
+    /* Now graft in the new tree resp. graft back the merged tree */
+    ret = iso_dir_add_node(new_parent, (IsoNode *) new_dir, 0);
+    if (ret < 0)
+        goto ex;
+
+    if (!(flag & 1))
+        *new_node = (IsoNode *) new_dir;
+    ret = ISO_SUCCESS;
+ex:;
+    if (iter != NULL)
+        iso_dir_iter_free(iter);
+    if (ret < 0 && new_dir != NULL) {
+	if (flag & 1) {
+            /* graft back the merged tree (eventually with half copy) */
+            iso_dir_add_node(new_parent, (IsoNode *) new_dir, 0);
+        } else {
+            iso_node_remove_tree((IsoNode *) new_dir, NULL);
+            *new_node = NULL;
+        }
+    }
+    return ret;
+}
+
+static
+int iso_tree_clone_file(IsoFile *old_file,
+                        IsoDir *new_parent, char *new_name, IsoNode **new_node,
+                        int flag)
+{
+    IsoStream *new_stream = NULL;
+    IsoFile *new_file = NULL;
+    int ret;
+
+    *new_node = NULL;
+
+    ret = iso_stream_clone(old_file->stream, &new_stream, 0);
+    if (ret < 0)
+        return ret;
+
+    ret = iso_tree_add_new_file(new_parent, new_name, new_stream, &new_file);
+    if (ret < 0)
+        goto ex;
+    new_stream = NULL; /* now owned by new_file */
+    new_file->sort_weight = old_file->sort_weight;
+    *new_node = (IsoNode *) new_file;
+    ret = ISO_SUCCESS;
+ex:;
+    if (new_stream != NULL)
+        iso_stream_unref(new_stream);
+    return ret;
+}
+
+static
+int iso_tree_clone_symlink(IsoSymlink *node,
+                        IsoDir *new_parent, char *new_name, IsoNode **new_node,
+                        int flag)
+{
+    IsoSymlink *new_sym;
+    int ret;
+
+    *new_node = NULL;
+
+    ret = iso_tree_add_new_symlink(new_parent, new_name, node->dest, &new_sym);
+    if (ret < 0)
+        return ret;
+    new_sym->fs_id = node->fs_id;
+    new_sym->st_dev = node->st_dev;
+    new_sym->st_ino = node->st_ino;
+    *new_node = (IsoNode *) new_sym;
+    return ISO_SUCCESS;
+}
+
+static
+int iso_tree_clone_special(IsoSpecial *node,
+                        IsoDir *new_parent, char *new_name, IsoNode **new_node,
+                        int flag)
+{
+    IsoSpecial *new_spec;
+    IsoNode *iso_node;
+    int ret;
+
+    iso_node = (IsoNode *) node;
+    ret = iso_tree_add_new_special(new_parent, new_name, iso_node->mode,
+                                   node->dev, &new_spec);
+    if (ret < 0)
+        return ret;
+    new_spec->fs_id = node->fs_id;
+    new_spec->st_dev = node->st_dev;
+    new_spec->st_ino = node->st_ino;
+    *new_node = (IsoNode *) new_spec;
+    return ISO_SUCCESS;
+}
+
+/* API */
+int iso_tree_clone(IsoNode *node,
+                   IsoDir *new_parent, char *new_name, IsoNode **new_node,
+                   int flag)
+{
+    int ret = ISO_SUCCESS;
+
+    if (iso_dir_get_node(new_parent, new_name, new_node) == 1) {
+        if (! (node->type == LIBISO_DIR && (*new_node)->type == LIBISO_DIR &&
+               (flag & 1))) {
+            *new_node = NULL;
+            return ISO_NODE_NAME_NOT_UNIQUE;
+        }
+    } else
+        flag &= ~1;
+
+    if (node->type == LIBISO_DIR) {
+        ret = iso_tree_clone_dir((IsoDir *) node, new_parent, new_name,
+                                 new_node, flag & 1);
+    } else if (node->type == LIBISO_FILE) {
+        ret = iso_tree_clone_file((IsoFile *) node, new_parent, new_name, 
+                                  new_node, 0);
+    } else if (node->type == LIBISO_SYMLINK) {
+        ret = iso_tree_clone_symlink((IsoSymlink *) node, new_parent, new_name,
+                                  new_node, 0);
+    } else if (node->type == LIBISO_SPECIAL) {
+        ret = iso_tree_clone_special((IsoSpecial *) node, new_parent, new_name,
+                                    new_node, 0);
+    } else if (node->type == LIBISO_BOOT) {
+        ret = ISO_SUCCESS; /* API says they are silently ignored */
+    }
+    if (ret < 0)
+        return ret;
+    if (flag & 1)
+        return 2; /* merged two directories, *new_node is not new */
+    ret = iso_tree_copy_node_attr(node, *new_node, 0);
+    return ret;
+}
+

libisofs/tree.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_IMAGE_TREE_H_
 #define LIBISO_IMAGE_TREE_H_

libisofs/util.c

@@ -1,14 +1,21 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
  * Copyright (c) 2007 Mario Danic
+ * Copyright (c) 2009 - 2011 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "util.h"
 #include "libisofs.h"
+#include "messages.h"
 #include "../version.h"
 
 #include <stdlib.h>
@@ -32,9 +39,6 @@
 #endif
 
 
-#ifdef Libisofs_with_iso_iconV
-
-
 /* Produce possibly inflationary error messages directly to stderr */
 static int iso_iconv_debug = 0;
 
@@ -78,6 +82,15 @@ size_t iso_iconv(struct iso_iconv_handle *handle,
                  char **outbuf, size_t *outbytesleft, int flag)
 {
     size_t ret;
+/* The build system might indicate iconv(,const char **inbuf,) by
+   defining ICONV_CONST const
+*/
+#ifndef ICONV_CONST
+#define ICONV_CONST
+#endif
+    ICONV_CONST char **local_inbuf;
+
+    local_inbuf = (ICONV_CONST char **) inbuf;
 
     if (!(handle->status & 1)) {
         if (iso_iconv_debug)
@@ -104,7 +117,7 @@ null_buf:;
             return (size_t) -1;
         return (size_t) 0;
     }
-    ret = iconv(handle->descr, inbuf, inbytesleft, outbuf, outbytesleft);
+    ret = iconv(handle->descr, local_inbuf, inbytesleft, outbuf, outbytesleft);
     if (ret == (size_t) -1) {
         if (iso_iconv_debug)
             fprintf(stderr, "libisofs_DEBUG: iconv() failed: errno= %d %s\n",
@@ -141,8 +154,6 @@ int iso_iconv_close(struct iso_iconv_handle *handle, int flag)
     return ret;
 }
 
-#endif /* Libisofs_with_iso_iconV */
-
 
 int int_pow(int base, int power)
 {
@@ -183,13 +194,8 @@ int strconv(const char *str, const char *icharset, const char *ocharset,
     size_t inbytes;
     size_t outbytes;
     size_t n;
-
-#ifdef Libisofs_with_iso_iconV
     struct iso_iconv_handle conv;
     int conv_ret;
-#else
-    iconv_t conv;
-#endif
 
     char *out = NULL;
     char *src;
@@ -204,43 +210,22 @@ int strconv(const char *str, const char *icharset, const char *ocharset,
         goto ex;
     }
 
-#ifdef Libisofs_with_iso_iconV
     conv_ret = iso_iconv_open(&conv, (char *) ocharset, (char *) icharset, 0);
     if (conv_ret <= 0) {
-#else
-    conv = iconv_open(ocharset, icharset);
-    if (conv == (iconv_t)(-1)) {
-#endif
-
         retval = ISO_CHARSET_CONV_ERROR;
         goto ex;
     }
-
     src = (char *)str;
     ret = (char *)out;
-
-#ifdef Libisofs_with_iso_iconV
     n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
     if (n == -1) {
         /* error */
         iso_iconv_close(&conv, 0);
-#else
-    n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-    if (n == -1) {
-        /* error */
-        iconv_close(conv);
-#endif
-
         retval = ISO_CHARSET_CONV_ERROR;
         goto ex;
     }
     *ret = '\0';
-
-#ifdef Libisofs_with_iso_iconV
     iso_iconv_close(&conv, 0);
-#else
-    iconv_close(conv);
-#endif
 
     *output = malloc(ret - out + 1);
     if (*output == NULL) {
@@ -261,14 +246,8 @@ int strnconv(const char *str, const char *icharset, const char *ocharset,
     size_t inbytes;
     size_t outbytes;
     size_t n;
-
-#ifdef Libisofs_with_iso_iconV
     struct iso_iconv_handle conv;
     int conv_ret;
-#else
-    iconv_t conv;
-#endif
-
     char *out = NULL;
     char *src;
     char *ret;
@@ -281,43 +260,22 @@ int strnconv(const char *str, const char *icharset, const char *ocharset,
         retval = ISO_OUT_OF_MEM;
         goto ex;
     }
-
-#ifdef Libisofs_with_iso_iconV
     conv_ret = iso_iconv_open(&conv, (char *) ocharset, (char *) icharset, 0);
     if (conv_ret <= 0) {
-#else
-    conv = iconv_open(ocharset, icharset);
-    if (conv == (iconv_t)(-1)) {
-#endif
-
         retval = ISO_CHARSET_CONV_ERROR;
         goto ex;
     }
     src = (char *)str;
     ret = (char *)out;
-
-#ifdef Libisofs_with_iso_iconV
     n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
     if (n == -1) {
         /* error */
         iso_iconv_close(&conv, 0);
-#else
-    n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-    if (n == -1) {
-        /* error */
-        iconv_close(conv);
-#endif
-
         retval = ISO_CHARSET_CONV_ERROR;
         goto ex;
     }
     *ret = '\0';
-
-#ifdef Libisofs_with_iso_iconV
     iso_iconv_close(&conv, 0);
-#else
-    iconv_close(conv);
-#endif
 
     *output = malloc(ret - out + 1);
     if (*output == NULL) {
@@ -342,13 +300,8 @@ ex:;
 static
 int str2wchar(const char *icharset, const char *input, wchar_t **output)
 {
-    
-#ifdef Libisofs_with_iso_iconV
     struct iso_iconv_handle conv;
     int conv_ret;
-#else
-    iconv_t conv;
-#endif
 
     /* That while loop smells like a potential show stopper */
     size_t loop_counter = 0, loop_limit = 3;
@@ -364,15 +317,8 @@ int str2wchar(const char *icharset, const char *input, wchar_t **output)
         return ISO_NULL_POINTER;
     }
 
-
-#ifdef Libisofs_with_iso_iconV
     conv_ret = iso_iconv_open(&conv, "WCHAR_T", (char *) icharset, 0);
     if (conv_ret <= 0) {
-#else
-    conv = iconv_open("WCHAR_T", icharset);
-    if (conv == (iconv_t)-1) {
-#endif
-
         return ISO_CHARSET_CONV_ERROR;
     }
 
@@ -388,13 +334,7 @@ int str2wchar(const char *icharset, const char *input, wchar_t **output)
     ret = (char *)wstr;
     src = (char *)input;
 
-
-#ifdef Libisofs_with_iso_iconV
     n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
-#else
-    n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-#endif
-
     while (n == -1) {
 
         if (errno == E2BIG) {
@@ -424,34 +364,16 @@ int str2wchar(const char *icharset, const char *input, wchar_t **output)
             loop_counter++;
             if (loop_counter > loop_limit)
                 goto conv_error;
-
-#ifdef Libisofs_with_iso_iconV
             n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
-#else
-            n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-#endif
-
         }
     }
-
-#ifdef Libisofs_with_iso_iconV
     iso_iconv_close(&conv, 0);
-#else
-    iconv_close(conv);
-#endif
-
     *( (wchar_t *)ret )='\0';
     *output = wstr;
     return ISO_SUCCESS;
 
 conv_error:;
-
-#ifdef Libisofs_with_iso_iconV
     iso_iconv_close(&conv, 0);
-#else
-    iconv_close(conv);
-#endif
-
     free(wstr);
     return ISO_CHARSET_CONV_ERROR;
 }
@@ -463,13 +385,8 @@ int str2ascii(const char *icharset, const char *input, char **output)
     char *ret;
     char *ret_;
     char *src;
-    
-#ifdef Libisofs_with_iso_iconV
     struct iso_iconv_handle conv;
     int conv_ret;
-#else
-    iconv_t conv;
-#endif
 
     /* That while loop smells like a potential show stopper */
     size_t loop_counter = 0, loop_limit = 3;
@@ -509,27 +426,14 @@ int str2ascii(const char *icharset, const char *input, char **output)
     ret = ret_;
 
     /* initialize iconv */
-
-#ifdef Libisofs_with_iso_iconV
     conv_ret = iso_iconv_open(&conv, "ASCII", "WCHAR_T", 0);
     if (conv_ret <= 0) {
-#else
-    conv = iconv_open("ASCII", "WCHAR_T");
-    if (conv == (iconv_t)-1) {
-#endif
-
         free(wsrc_);
         free(ret_);
         goto fallback;
     }
 
-
-#ifdef Libisofs_with_iso_iconV
     n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
-#else
-    n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-#endif
-
     while (n == -1) {
         /* The destination buffer is too small. Stops here. */
         if (errno == E2BIG)
@@ -563,21 +467,9 @@ int str2ascii(const char *icharset, const char *input, char **output)
         loop_counter++;
         if (loop_counter > loop_limit)
             break;
-
-#ifdef Libisofs_with_iso_iconV
         n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
-#else
-        n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-#endif
-
     }
-
-#ifdef Libisofs_with_iso_iconV
     iso_iconv_close(&conv, 0);
-#else
-    iconv_close(conv);
-#endif
-
     *ret='\0';
     free(wsrc_);
 
@@ -628,13 +520,8 @@ int str2ucs(const char *icharset, const char *input, uint16_t **output)
     char *src;
     char *ret;
     char *ret_;
-
-#ifdef Libisofs_with_iso_iconV
     struct iso_iconv_handle conv;
     int conv_ret;
-#else
-    iconv_t conv;
-#endif
     
     /* That while loop smells like a potential show stopper */
     size_t loop_counter = 0, loop_limit = 3;
@@ -670,27 +557,14 @@ int str2ucs(const char *icharset, const char *input, uint16_t **output)
     ret = ret_;
 
     /* initialize iconv */
-
-#ifdef Libisofs_with_iso_iconV
     conv_ret = iso_iconv_open(&conv, "UCS-2BE", "WCHAR_T", 0);
     if (conv_ret <= 0) {
-#else
-    conv = iconv_open("UCS-2BE", "WCHAR_T");
-    if (conv == (iconv_t)-1) {
-#endif
-
         free(wsrc_);
         free(ret_);
         return ISO_CHARSET_CONV_ERROR;
     }
 
-
-#ifdef Libisofs_with_iso_iconV
     n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
-#else
-    n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-#endif
-
     while (n == -1) {
         /* The destination buffer is too small. Stops here. */
         if (errno == E2BIG)
@@ -724,20 +598,9 @@ int str2ucs(const char *icharset, const char *input, uint16_t **output)
         loop_counter++;
         if (loop_counter > loop_limit)
             break;
-
-#ifdef Libisofs_with_iso_iconV
         n = iso_iconv(&conv, &src, &inbytes, &ret, &outbytes, 0);
-#else
-        n = iconv(conv, &src, &inbytes, &ret, &outbytes);
-#endif
-
     }
-
-#ifdef Libisofs_with_iso_iconV
     iso_iconv_close(&conv, 0);
-#else
-    iconv_close(conv);
-#endif
 
     /* close the ucs string */
     set_ucsbe((uint16_t*) ret, '\0');
@@ -910,6 +773,8 @@ char *iso_r_dirid(const char *src, int size, int relaxed)
         len = size;
     }
     dest = malloc(len + 1);
+    if (dest == NULL)
+        return NULL;
     for (i = 0; i < len; i++) {
         char c= src[i];
         if (relaxed == 2) {
@@ -1049,7 +914,10 @@ ex:;
     return retval;
 }
 
-uint16_t *iso_j_file_id(const uint16_t *src)
+/*
+   bit0= no_force_dots
+*/
+uint16_t *iso_j_file_id(const uint16_t *src, int flag)
 {
     uint16_t *dot;
     size_t lname, lext, lnname, lnext, pos, i;
@@ -1095,6 +963,10 @@ uint16_t *iso_j_file_id(const uint16_t *src)
             pos++;
         }
     }
+
+    if ((flag & 1) && lnext <= 0)
+        goto is_done;
+
     set_ucsbe(dest + pos, '.');
     pos++;
 
@@ -1108,6 +980,8 @@ uint16_t *iso_j_file_id(const uint16_t *src)
             pos++;
         }
     }
+
+is_done:;
     set_ucsbe(dest + pos, '\0');
     return ucsdup(dest);
 }
@@ -1164,6 +1038,8 @@ uint16_t *ucsdup(const uint16_t *str)
     size_t len = ucslen(str);
     
     ret = malloc(2 * (len + 1));
+    if (ret == NULL)
+        return NULL;
     if (ret != NULL) {
         memcpy(ret, str, 2 * (len + 1));
     }
@@ -1347,7 +1223,7 @@ void iso_datetime_7(unsigned char *buf, time_t t, int always_gmt)
     }
 
     memset(&tm, 0, sizeof(tm));
-    tm.tm_isdst = -1;  /* some Linuxes change tm_isdst only if it is -1 */
+    tm.tm_isdst = -1;  /* some OSes change tm_isdst only if it is -1 */
     localtime_r(&t, &tm);
 
 #ifdef HAVE_TM_GMTOFF
@@ -1391,7 +1267,7 @@ void iso_datetime_17(unsigned char *buf, time_t t, int always_gmt)
     }
 
     memset(&tm, 0, sizeof(tm));
-    tm.tm_isdst = -1;  /* some Linuxes change tm_isdst only if it is -1 */
+    tm.tm_isdst = -1;  /* some OSes change tm_isdst only if it is -1 */
     localtime_r(&t, &tm);
 
     localtime_r(&t, &tm);
@@ -1422,12 +1298,41 @@ void iso_datetime_17(unsigned char *buf, time_t t, int always_gmt)
 }
 
 #ifndef HAVE_TIMEGM
+
+/* putenv is SVr4, POSIX.1-2001, 4.3BSD , setenv is 4.3BSD, POSIX.1-2001.
+   So putenv is more widely available.
+   Also, setenv spoils eventual putenv expectation of applications because
+   putenv installs the original string which then may be altered from
+   its owner. setenv installs a copy that may not be altered.
+   Both are slow.
+   Thus first try with a naive implementation that assumes no leap seconds.
+   If it fails a test with gmtime() then use the slow function with mktime().
+*/
+#define Libisofs_use_putenV yes
+
 static
-time_t timegm(struct tm *tm)
+time_t env_timegm(struct tm *tm)
 {
     time_t ret;
     char *tz;
 
+#ifdef Libisofs_use_putenV
+
+    static char unset_name[] = {"TZ"};
+
+    tz = getenv("TZ");
+    putenv("TZ=");
+    tzset();
+    ret = mktime(tm);
+    if (tz != NULL) {
+        /* tz is a pointer to the value part in a string of form "TZ="value */
+        putenv(tz - 3);
+    } else
+        putenv(unset_name); /* not daring to submit constant */
+    tzset();
+
+#else /* Libisofs_use_putenV */
+
     tz = getenv("TZ");
     setenv("TZ", "", 1);
     tzset();
@@ -1437,9 +1342,92 @@ time_t timegm(struct tm *tm)
     else
         unsetenv("TZ");
     tzset();
+
+#endif /* ! Libisofs_use_putenV */
+
     return ret;
 }
-#endif
+
+static
+int ts_is_leapyear(int tm_year) /* years since 1900 */
+{
+  return ((tm_year % 4) == 0 && ((tm_year % 100) != 0 ||
+                                 (tm_year % 400) == 100));
+}
+
+/* Fast implementation without leap seconds.
+   Inspired by but not copied from code by Kungliga Tekniska Hgskolan
+   (Royal Institute of Technology, Stockholm, Sweden),
+   which was modified by Andrew Tridgell for Samba4.
+   I claim own copyright 2011 Thomas Schmitt <scdbackup@gmx.net>.
+*/ 
+static
+time_t ts_timegm(struct tm *tm)
+{
+    time_t ret;
+    static int month_length_normal[12] =
+                {31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31};
+    static int month_length_leap[12] =
+                {31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31};
+    int *month_length_pt;
+    int years, i;
+
+    ret = 0;
+
+    years = tm->tm_year - 70; /* Years since 1970 */
+    if (years < 0)
+        return ret;
+    for (i = 0; i < years; i++) {
+        ret += 365 * 86400;
+        if (ts_is_leapyear(70 + i))
+            ret += 86400;
+    }
+    if (ts_is_leapyear(tm->tm_year))
+        month_length_pt = month_length_leap;
+    else
+        month_length_pt = month_length_normal;
+    for (i = 0; i < tm->tm_mon; i++)
+        ret += month_length_pt[i] * 86400;
+    ret += (tm->tm_mday - 1) * 86400;
+    ret += tm->tm_hour * 3600;
+    ret += tm->tm_min * 60;
+    ret += tm->tm_sec;
+    return ret;
+}
+
+static
+time_t timegm(struct tm *tm)
+{
+    time_t raw_t, ret;
+    struct tm *test_tm, input_tm_copy;
+
+    /* Beware of ill effects if tm is result of gmtime() or alike */
+    memcpy(&input_tm_copy, tm, sizeof(struct tm));
+
+    /* Try without leapseconds (which are rarely implemented, as it seems) */
+    raw_t = ts_timegm(tm);
+    if (raw_t == 0)
+        return raw_t;
+
+    /* Check whether this translates back to the input values */
+    test_tm = gmtime(&raw_t);
+    if (input_tm_copy.tm_sec == test_tm->tm_sec &&
+        input_tm_copy.tm_min == test_tm->tm_min &&
+        input_tm_copy.tm_hour == test_tm->tm_hour &&
+        input_tm_copy.tm_mday == test_tm->tm_mday &&
+        input_tm_copy.tm_mon == test_tm->tm_mon &&
+        input_tm_copy.tm_year == test_tm->tm_year) {
+        ret = raw_t;
+    } else {
+        /* Mismatch. Use slow method around mktime() */
+        ret = env_timegm(&input_tm_copy);
+    }
+    return ret;
+}
+
+
+#endif /* ! HAVE_TIMEGM */
+
 
 time_t iso_datetime_read_7(const uint8_t *buf)
 {
@@ -1451,6 +1439,7 @@ time_t iso_datetime_read_7(const uint8_t *buf)
     tm.tm_hour = buf[3];
     tm.tm_min = buf[4];
     tm.tm_sec = buf[5];
+
     return timegm(&tm) - ((int8_t)buf[6]) * 60 * 15;
 }
 
@@ -1524,21 +1513,33 @@ int iso_eaccess(const char *path)
     return ISO_SUCCESS;
 }
 
-char *strcopy(const char *buf, size_t len)
+char *iso_util_strcopy(const char *buf, size_t len)
 {
     char *str;
     
-    str = malloc((len + 1) * sizeof(char));
+    str = calloc(len + 1, 1);
     if (str == NULL) {
         return NULL;
     }
     strncpy(str, buf, len);
     str[len] = '\0';
+    return str;
+}
+
+char *iso_util_strcopy_untail(const char *buf, size_t len)
+{
+    char *str;
     
+    str = iso_util_strcopy(buf, len);
+    if (str == NULL) {
+        return NULL;
+    }
     /* remove trailing spaces */
-    for (len = len-1; str[len] == ' ' && len > 0; --len)
+    for (len = len-1; len >= 0; --len) {
-        str[len] = '\0'; 
+        if (str[len] != ' ')
-    
+    break;
+        str[len] = 0; 
+    }
     return str;
 }
 
@@ -1566,14 +1567,8 @@ char *ucs2str(const char *buf, size_t len)
 {
     size_t outbytes, inbytes;
     char *str, *src, *out = NULL, *retval = NULL;
-
-#ifdef Libisofs_with_iso_iconV
     struct iso_iconv_handle conv;
     int conv_ret;
-#else
-    iconv_t conv;
-#endif
-
     size_t n;
     
     inbytes = len;
@@ -1582,38 +1577,19 @@ char *ucs2str(const char *buf, size_t len)
     
     /* ensure enought space */
     out = calloc(outbytes, 1);
+    if (out == NULL)
+        return NULL;
 
     /* convert to local charset */
-
-#ifndef Libisofs_setlocale_in_iniT
-    /* ??? ts Nov 25 2008 :
-       Shouldn't this go to library initialization or even to app ?
-    */
-    setlocale(LC_CTYPE, "");
-#endif
-
-#ifdef Libisofs_with_iso_iconV
     conv_ret = iso_iconv_open(&conv, iso_get_local_charset(0), "UCS-2BE", 0);
     if (conv_ret <= 0) {
-#else
-    conv = iconv_open(iso_get_local_charset(0), "UCS-2BE");
-    if (conv == (iconv_t)(-1)) {
-#endif
-
         goto ex;
     }
     src = (char *)buf;
     str = (char *)out;
 
-    
-#ifdef Libisofs_with_iso_iconV
     n = iso_iconv(&conv, &src, &inbytes, &str, &outbytes, 0);
     iso_iconv_close(&conv, 0);
-#else
-    n = iconv(conv, &src, &inbytes, &str, &outbytes);
-    iconv_close(conv);
-#endif
-
     if (n == -1) {
         /* error */
         goto ex;
@@ -1660,3 +1636,229 @@ int iso_init_locale(int flag)
 }
 
 
+int iso_util_encode_len_bytes(uint32_t data, char *buffer, int data_len,
+                              int *result_len, int flag)
+{
+    uint32_t x;
+    int i, l;
+    char *wpt = buffer;
+
+    if (data_len <= 0) {
+        x = data;
+        for (i = 0; i < 4 && x != 0; i++)
+            x = x >> 8;
+        l = i;
+        if (l == 0)
+            l = 1;
+    } else
+        l = data_len;
+    *((unsigned char *) (wpt++)) = l;
+    for (i = 0; i < l; i++)
+        *((unsigned char *) (wpt++)) = data >> (8 * (l - i - 1));
+    *result_len = l + 1;
+    return ISO_SUCCESS;
+}
+
+
+int iso_util_decode_len_bytes(uint32_t *data, char *buffer, int *data_len,
+                              int buffer_len, int flag)
+{
+    int i;
+
+    *data = 0;
+    *data_len = ((unsigned char *) buffer)[0];
+    if (*data_len > buffer_len - 1)
+        *data_len = buffer_len - 1;
+    for (i = 1; i <= *data_len; i++)
+        *data = (*data << 8) | ((unsigned char *) buffer)[i];
+    return ISO_SUCCESS;
+}
+
+
+int iso_util_dec_to_uint32(char *dec, uint32_t *value, int flag)
+{
+    double num;
+
+    sscanf(dec, "%lf", &num);
+    if (num < 0 || num > 4294967295.0)
+        return 0;
+    *value = num;
+    return 1;
+}
+
+
+int iso_util_hex_to_bin(char *hex, char *bin, int bin_size, int *bin_count,
+                        int flag)
+{
+    static char *allowed = {"0123456789ABCDEFabcdef"};
+    char b[3];
+    int i;
+    unsigned int u;
+
+    b[2] = 0;
+    *bin_count = 0;
+    for (i = 0; i < bin_size; i++) {
+        b[0] = hex[2 * i];
+        b[1] = hex[2 * i + 1];
+        if (strchr(allowed, b[0]) == NULL || strchr(allowed, b[1]) == NULL)
+    break;
+        sscanf(b, "%x", &u);
+        ((unsigned char *) bin)[i] = u;
+        (*bin_count)++;
+    }
+    return (*bin_count > 0);
+}
+
+
+int iso_util_tag_magic(int tag_type, char **tag_magic, int *len, int flag)
+{
+    static char *magic[] = {"",
+        "libisofs_checksum_tag_v1",
+        "libisofs_sb_checksum_tag_v1",
+        "libisofs_tree_checksum_tag_v1",
+        "libisofs_rlsb32_checksum_tag_v1"};
+    static int magic_len[]= {0, 24, 27, 29, 31};
+    static int magic_max = 4;
+
+    *tag_magic = NULL;
+    *len = 0;
+    if (tag_type < 0 || tag_type > magic_max)
+        return ISO_WRONG_ARG_VALUE;
+    *tag_magic = magic[tag_type];
+    *len = magic_len[tag_type];
+    return magic_max;
+}
+
+
+int iso_util_decode_md5_tag(char data[2048], int *tag_type, uint32_t *pos,
+                            uint32_t *range_start, uint32_t *range_size,
+                            uint32_t *next_tag, char md5[16], int flag)
+{
+    int ret, bin_count, i, mode, magic_first = 1, magic_last = 4;
+    int magic_len = 0;
+    char *cpt, self_md5[16], tag_md5[16], *tag_magic;
+    void *ctx = NULL;
+
+    *next_tag = 0;
+    mode = flag & 255;
+    if (mode > magic_last)
+        return ISO_WRONG_ARG_VALUE;
+    if (mode > 0)
+        magic_first = magic_last = mode;
+    for (i = magic_first; i <= magic_last; i++) {
+        iso_util_tag_magic(i, &tag_magic, &magic_len, 0);
+        if (strncmp(data, tag_magic, magic_len) == 0)
+    break;
+    }
+    if (i > magic_last )
+        return 0;
+    *tag_type = i;
+    cpt = data + magic_len + 1;
+    if (strncmp(cpt, "pos=", 4) != 0)
+        return 0;
+    cpt+= 4;
+    ret = iso_util_dec_to_uint32(cpt, pos, 0);
+    if (ret <= 0)
+        return 0;
+    cpt = strstr(cpt, "range_start=");
+    if (cpt == NULL)
+        return(0);
+    ret = iso_util_dec_to_uint32(cpt + 12, range_start, 0);
+    if (ret <= 0)
+        return 0;
+    cpt = strstr(cpt, "range_size=");
+    if (cpt == NULL)
+        return(0);
+    ret = iso_util_dec_to_uint32(cpt + 11, range_size, 0);
+    if (ret <= 0)
+        return 0;
+    if (*tag_type == 2 || *tag_type == 3) {
+        cpt = strstr(cpt, "next=");
+        if (cpt == NULL)
+            return(0);
+        ret = iso_util_dec_to_uint32(cpt + 5, next_tag, 0);
+        if (ret <= 0)
+            return 0;
+    } else if (*tag_type == 4) {
+        cpt = strstr(cpt, "session_start=");
+        if (cpt == NULL)
+            return(0);
+        ret = iso_util_dec_to_uint32(cpt + 14, next_tag, 0);
+        if (ret <= 0)
+            return 0;
+    }
+    cpt = strstr(cpt, "md5=");
+    if (cpt == NULL)
+        return(0);
+    ret = iso_util_hex_to_bin(cpt + 4, md5, 16, &bin_count, 0);
+    if (ret <= 0 || bin_count != 16)
+        return 0;
+
+    cpt += 4 + 32;
+    ret = iso_md5_start(&ctx);
+    if (ret < 0)
+        return ret;
+    iso_md5_compute(ctx, data , cpt - data);
+    iso_md5_end(&ctx, tag_md5);
+    cpt = strstr(cpt, "self=");
+    if (cpt == NULL)
+        return(0);
+    ret = iso_util_hex_to_bin(cpt + 5, self_md5, 16, &bin_count, 0);
+    if (ret <= 0 || bin_count != 16)
+        return 0;
+    for(i= 0; i < 16; i++)
+      if(self_md5[i] != tag_md5[i])
+        return ISO_MD5_AREA_CORRUPTED;
+    if (*(cpt + 5 + 32) != '\n')
+        return 0;
+    return(1);
+}
+
+
+int iso_util_eval_md5_tag(char *block, int desired, uint32_t lba,
+                          void *ctx, uint32_t ctx_start_lba, 
+                          int *tag_type, uint32_t *next_tag, int flag)
+{
+    int decode_ret, ret;
+    char md5[16], cloned_md5[16];
+    uint32_t pos, range_start, range_size;
+    void *cloned_ctx = NULL;
+
+    *tag_type = 0;
+    decode_ret = iso_util_decode_md5_tag(block, tag_type, &pos,
+                                  &range_start, &range_size, next_tag, md5, 0);
+    if (decode_ret != 1 && decode_ret != ISO_MD5_AREA_CORRUPTED)
+        return 0;
+    if (*tag_type > 30)
+        goto unexpected_type;
+
+    if (decode_ret == ISO_MD5_AREA_CORRUPTED) {
+        ret = decode_ret; 
+        goto ex;
+    } else if (!((1 << *tag_type) & desired)) {
+unexpected_type:;
+        iso_msg_submit(-1, ISO_MD5_TAG_UNEXPECTED, 0, NULL);
+        ret = 0;
+        goto ex;
+    } else if(pos != lba) {
+        ret = ISO_MD5_TAG_MISPLACED;
+        goto ex;
+    } else if(range_start != ctx_start_lba) {
+        ret = ISO_MD5_TAG_MISPLACED;
+    }
+    ret = iso_md5_clone(ctx, &cloned_ctx);
+    if (ret < 0)
+        goto ex;
+    iso_md5_end(&cloned_ctx, cloned_md5);
+    if (! iso_md5_match(cloned_md5, md5)) {
+        ret = ISO_MD5_TAG_MISMATCH;
+        goto ex;
+    }
+    ret = 1;
+ex:;
+    if (ret < 0)
+        iso_msg_submit(-1, ret, 0, NULL);
+    return ret;
+}
+
+ 

libisofs/util.h

@@ -1,15 +1,24 @@
 /*
  * Copyright (c) 2007 Vreixo Formoso
+ * Copyright (c) 2009 Thomas Schmitt
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
 #ifndef LIBISO_UTIL_H_
 #define LIBISO_UTIL_H_
 
+#ifdef HAVE_STDINT_H
 #include <stdint.h>
+#else
+#ifdef HAVE_INTTYPES_H
+#include <inttypes.h>
+#endif
+#endif
+
 #include <time.h>
 
 #ifndef MAX
@@ -149,11 +158,12 @@ char *iso_r_fileid(const char *src, size_t len, int relaxed, int forcedot);
  * 2 bytes and the resulting string is NULL-terminated by a 2-byte NULL. 
  * 
  * Note that version number and (;1) is not appended.
- *
+ * @param flag
+ *        bit0= no_force_dots
  * @return 
  *        NULL if the original name and extension both are of length 0.
  */
-uint16_t *iso_j_file_id(const uint16_t *src);
+uint16_t *iso_j_file_id(const uint16_t *src, int flag);
 
 /**
  * Create a Joliet directory identifier that consists of name and optionally
@@ -250,7 +260,14 @@ int iso_eaccess(const char *path);
  * Copy up to \p len chars from \p buf and return this newly allocated
  * string. The new string is null-terminated.
  */
-char *strcopy(const char *buf, size_t len);
+char *iso_util_strcopy(const char *buf, size_t len);
+
+/**
+ * Copy up to \p len chars from \p buf and return this newly allocated
+ * string. The new string is null-terminated.
+ * Any trailing blanks will be removed.
+ */
+char *iso_util_strcopy_untail(const char *buf, size_t len);
 
 /**
  * Copy up to \p max characters from \p src to \p dest. If \p src has less than
@@ -440,4 +457,82 @@ void iso_htable_destroy(IsoHTable *table, hfree_data_t free_data);
  */
 unsigned int iso_str_hash(const void *key);
 
+/**
+ * Encode an integer as LEN,BYTES for being a component in certain AAIP
+ * attribute values.
+ */
+int iso_util_encode_len_bytes(uint32_t data, char *buffer, int data_len,
+                              int *result_len, int flag);
+
+/**
+ * Decode an integer as LEN,BYTES for being a component in certain AAIP
+ * attribute values.
+ * @param data        returns the decoded value
+ * @param buffer      contains the encoded value
+ * @param data_len    returns the number of value bytes (without len byte)
+ * @param buffer_len  tells the number of valid buffer bytes
+ */
+int iso_util_decode_len_bytes(uint32_t *data, char *buffer, int *data_len,
+                              int buffer_len, int flag);
+
+    
+/* Evaluate a data block whether it is a libisofs session checksum tag of
+   desired type and eventually use it to verify the MD5 checksum computed
+   so far.
+   @param block      The data block to be evaluated
+   @param desired    Bit map which tells what tag types are expected
+                     (0 to 30)
+   @param lba        The address from where block was read
+   @param ctx        The checksum context computed so far
+   @param ctx_start_lba  The block address where checksum computing started
+   @param tag_type   Returns the tag type (0 means invalid tag type)
+   @param flag       Bitfield for control purposes, unused yet, submit 0
+   @return           1= tag is desired and matches
+                     0= not a recognizable tag or a undesired tag
+                    <0 is error or mismatch
+*/
+int iso_util_eval_md5_tag(char *block, int desired, uint32_t lba,
+                          void *ctx, uint32_t ctx_start_lba,
+                          int *tag_type, uint32_t *next_tag, int flag);
+
+
+int iso_util_tag_magic(int tag_type, char **tag_magic, int *len, int flag);
+
+
+/* ------------------------------------------------------------------------- */
+
+/* In md5.h these function prototypes would be neighbors of (Ecma119Image *)
+   which needs inclusion of ecma119.h and more. So, being generic, they ended
+   up here.
+*/
+
+/* Function to identify and manage md5sum indice of the old image.
+ * data is supposed to be a 4 byte integer, bit 31 shall be 0,
+ * value 0 of this integer means that it is not a valid index.
+ */
+int checksum_cx_xinfo_func(void *data, int flag);
+
+/* The iso_node_xinfo_cloner function which gets associated to 
+ * checksum_cx_xinfo_func by iso_init() resp. iso_init_with_flag() via 
+ * iso_node_xinfo_make_clonable()
+ */
+int checksum_cx_xinfo_cloner(void *old_data, void **new_data, int flag);
+
+
+/* Function to identify and manage md5 sums of unspecified providence stored
+ * directly in this xinfo. This is supposed to override any other recorded
+ * MD5 of the node unless data get copied and checksummed during that copying.
+ */
+int checksum_md5_xinfo_func(void *data, int flag);
+
+/* The iso_node_xinfo_cloner function which gets associated to 
+ * checksum_md5_xinfo_func by iso_init() resp. iso_init_with_flag() via 
+ * iso_node_xinfo_make_clonable()
+ */
+int checksum_md5_xinfo_cloner(void *old_data, void **new_data, int flag);
+
+
+/* ------------------------------------------------------------------------- */
+
+
 #endif /*LIBISO_UTIL_H_*/

libisofs/util_htable.c

@@ -2,10 +2,15 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "util.h"
 #include "libisofs.h"
 

libisofs/util_rbtree.c

@@ -2,10 +2,15 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 
+#ifdef HAVE_CONFIG_H
+#include "../config.h"
+#endif
+
 #include "util.h"
 #include "libisofs.h"
 
@@ -198,7 +203,11 @@ int iso_rbtree_insert(IsoRBTree *tree, void *data, void **item)
                 }
             }
 
-            comp = tree->compare(q->data, data);
+            if (q->data == data) {
+                comp = 0;
+            } else {
+                comp = tree->compare(q->data, data);
+            }
 
             /* Stop if found */
             if (comp == 0) {
@@ -276,7 +285,7 @@ void ** iso_rbtree_to_array(IsoRBTree *tree, int (*include_item)(void *),
                             size_t *size)
 {
     size_t pos;
-    void **array;
+    void **array, **new_array;
 
     array = malloc((tree->size + 1) * sizeof(void*));
     if (array == NULL) {
@@ -287,7 +296,12 @@ void ** iso_rbtree_to_array(IsoRBTree *tree, int (*include_item)(void *),
     pos = rbtree_to_array_aux(tree->root, array, 0, include_item);
     array[pos] = NULL;
 
-    array = realloc(array, (pos + 1) * sizeof(void*));
+    new_array = realloc(array, (pos + 1) * sizeof(void*));
+    if (new_array == NULL) {
+        free((char *) array);
+        return NULL;
+    }
+    array= new_array;
     if (size) {
         *size = pos;
     }

libisofs/writer.h

@@ -2,8 +2,9 @@
  * Copyright (c) 2007 Vreixo Formoso
  * 
  * This file is part of the libisofs project; you can redistribute it and/or 
- * modify it under the terms of the GNU General Public License version 2 as 
+ * modify it under the terms of the GNU General Public License version 2 
- * published by the Free Software Foundation. See COPYING file for details.
+ * or later as published by the Free Software Foundation. 
+ * See COPYING file for details.
  */
 #ifndef LIBISO_IMAGE_WRITER_H_
 #define LIBISO_IMAGE_WRITER_H_