7 * lzmadec and liblzma's lzma_alone_decoder(): Support decompressing
8 .lzma files that have less common settings in the headers
9 (dictionary size other than 2^n or 2^n + 2^(n-1), or uncompressed
10 size greater than 256 GiB). The limitations existed to avoid false
11 positives when detecting .lzma files. The lc + lp <= 4 limitation
12 still remains since liblzma's LZMA decoder has that limitation.
14 NOTE: xz's .lzma support or liblzma's lzma_auto_decoder() are NOT
15 affected by this change. They still consider uncommon .lzma headers
16 as not being in the .lzma format. Changing this would give way too
21 - Interaction of preset and custom filter chain options was
22 made less illogical. This affects only certain less typical
23 uses cases so few people are expected to notice this change.
25 Now when a custom filter chain option (e.g. --lzma2) is
26 specified, all preset options (-0 ... -9, -e) earlier are on
27 the command line are completely forgotten. Similarly, when
28 a preset option is specified, all custom filter chain options
29 earlier on the command line are completely forgotten.
31 Example 1: "xz -9 --lzma2=preset=5 -e" is equivalent to "xz -e"
32 which is equivalent to "xz -6e". Earlier -e didn't put xz back
33 into preset mode and thus the example command was equivalent
34 to "xz --lzma2=preset=5".
36 Example 2: "xz -9e --lzma2=preset=5 -7" is equivalent to
37 "xz -7". Earlier a custom filter chain option didn't make
38 xz forget the -e option so the example was equivalent to
41 - Fixes and improvements to error handling.
43 - Various fixes to the man page.
45 * xzless: Fixed to work with "less" versions 448 and later.
47 * xzgrep: Made -h an alias for --no-filename.
49 * Include the previously missing debug/translation.bash which can
50 be useful for translators.
52 * Include a build script for Mac OS X. This has been in the Git
53 repository since 2010 but due to a mistake in Makefile.am the
54 script hasn't been included in a release tarball before.
61 - Fix lzma_index_init(). It could crash if memory allocation
64 - Fix the possibility of an incorrect LZMA_BUF_ERROR when a BCJ
65 filter is used and the application only provides exactly as
66 much output space as is the uncompressed size of the file.
68 - Fix a bug in doc/examples_old/xz_pipe_decompress.c. It didn't
69 check if the last call to lzma_code() really returned
70 LZMA_STREAM_END, which made the program think that truncated
73 - New example programs in doc/examples (old programs are now in
74 doc/examples_old). These have more comments and more detailed
77 * Fix "xz -lvv foo.xz". It could crash on some corrupted files.
79 * Fix output of "xz --robot -lv" and "xz --robot -lvv" which
80 incorrectly printed the filename also in the "foo (x/x)" format.
82 * Fix exit status of "xzdiff foo.xz bar.xz".
84 * Fix exit status of "xzgrep foo binary_file".
86 * Fix portability to EBCDIC systems.
88 * Fix a configure issue on AIX with the XL C compiler. See INSTALL
91 * Update French, German, Italian, and Polish translations.
98 - A memory leak was fixed.
100 - lzma_stream_buffer_encode() no longer creates an empty .xz
101 Block if encoding an empty buffer. Such an empty Block with
102 LZMA2 data would trigger a bug in 5.0.1 and older (see the
103 first bullet point in 5.0.2 notes). When releasing 5.0.2,
104 I thought that no encoder creates this kind of files but
107 - Validate function arguments better in a few functions. Most
108 importantly, specifying an unsupported integrity check to
109 lzma_stream_buffer_encode() no longer creates a corrupt .xz
110 file. Probably no application tries to do that, so this
111 shouldn't be a big problem in practice.
113 - Document that lzma_block_buffer_encode(),
114 lzma_easy_buffer_encode(), lzma_stream_encoder(), and
115 lzma_stream_buffer_encode() may return LZMA_UNSUPPORTED_CHECK.
117 - The return values of the _memusage() functions are now
120 * Fix command name detection in xzgrep. xzegrep and xzfgrep now
121 correctly use egrep and fgrep instead of grep.
123 * French translation was added.
128 * LZMA2 decompressor now correctly accepts LZMA2 streams with no
129 uncompressed data. Previously it considered them corrupt. The
130 bug can affect applications that use raw LZMA2 streams. It is
131 very unlikely to affect .xz files because no compressor creates
132 .xz files with empty LZMA2 streams. (Empty .xz files are a
133 different thing than empty LZMA2 streams.)
135 * "xz --suffix=.foo filename.foo" now refuses to compress the
136 file due to it already having the suffix .foo. It was already
137 documented on the man page, but the code lacked the test.
139 * "xzgrep -l foo bar.xz" works now.
141 * Polish translation was added.
146 * xz --force now (de)compresses files that have setuid, setgid,
147 or sticky bit set and files that have multiple hard links.
148 The man page had it documented this way already, but the code
151 * gzip and bzip2 support in xzdiff was fixed.
155 * Minor fix to Czech translation
160 Only the most important changes compared to 4.999.9beta are listed
161 here. One change is especially important:
163 * The memory usage limit is now disabled by default. Some scripts
164 written before this change may have used --memory=max on xz command
165 line or in XZ_OPT. THESE USES OF --memory=max SHOULD BE REMOVED
166 NOW, because they interfere with user's ability to set the memory
167 usage limit himself. If user-specified limit causes problems to
168 your script, blame the user.
170 Other significant changes:
172 * Added support for XZ_DEFAULTS environment variable. This variable
173 allows users to set default options for xz, e.g. default memory
174 usage limit or default compression level. Scripts that use xz
175 must never set or unset XZ_DEFAULTS. Scripts should use XZ_OPT
176 instead if they need a way to pass options to xz via an
177 environment variable.
179 * The compression settings associated with the preset levels
180 -0 ... -9 have been changed. --extreme was changed a little too.
181 It is now less likely to make compression worse, but with some
182 files the new --extreme may compress slightly worse than the old
185 * If a preset level (-0 ... -9) is specified after a custom filter
186 chain options have been used (e.g. --lzma2), the custom filter
187 chain will be forgotten. Earlier the preset options were
188 completely ignored after custom filter chain options had been
191 * xz will create sparse files when decompressing if the uncompressed
192 data contains long sequences of binary zeros. This is done even
193 when writing to standard output that is connected to a regular
194 file and certain additional conditions are met to make it safe.
196 * Support for "xz --list" was added. Combine with --verbose or
197 --verbose --verbose (-vv) for detailed output.
199 * I had hoped that liblzma API would have been stable after
200 4.999.9beta, but there have been a couple of changes in the
201 advanced features, which don't affect most applications:
203 - Index handling code was revised. If you were using the old
204 API, you will get a compiler error (so it's easy to notice).
206 - A subtle but important change was made to the Block handling
207 API. lzma_block.version has to be initialized even for
208 lzma_block_header_decode(). Code that doesn't do it will work
209 for now, but might break in the future, which makes this API
212 * The major soname has been bumped to 5.0.0. liblzma API and ABI
213 are now stable, so the need to recompile programs linking against
214 liblzma shouldn't arise soon.