ViewVC logotype

Diff of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 360 by ph10, Wed Jul 9 20:00:28 2008 UTC revision 1350 by ph10, Fri Jul 26 10:03:38 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4  Version 8.0 02 Jul-08  Version 8.34 xx-xxxx-201x
5    --------------------------
7    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
8        executable memory. Patch inspired by Carsten Klein.
10    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
11        this macro is never tested and has no effect, because the work to support
12        coverage involves only compiling and linking options and special targets in
13        the Makefile. The comment in config.h implied that defining the macro would
14        enable coverage support, which is totally false. There was also support for
15        setting this macro in the CMake files (my fault, I just copied it from
16        configure). SUPPORT_GCOV has now been removed.
18    3.  Make a small performance improvement in strlen16() and strlen32() in
19        pcretest.
21    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
22        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
23        these cases). There was also one in pcretest.c.
25    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
27    6.  In UTF mode, the code for checking whether a group could match an empty
28        string (which is used for indefinitely repeated groups to allow for
29        breaking an infinite loop) was broken when the group contained a repeated
30        negated single-character class with a character that occupied more than one
31        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
32        UTF-8 mode). The effect was undefined: the group might or might not be
33        deemed as matching an empty string, or the program might have crashed.
35    7.  The code for checking whether a group could match an empty string was not
36        recognizing that \h, \H, \v, \V, and \R must match a character.
38    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
39        an empty string. If it can, pcretest shows this in its information output.
41    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
42        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
43        matched by pcre_exec() without using JIT:
45        (a) If the rest of the pattern did not match after a maximal run of
46            grapheme clusters, the code for backing up to try with fewer of them
47            did not always back up over a full grapheme when characters that do not
48            have the modifier quality were involved, e.g. Hangul syllables.
50        (b) If the match point in a subject started with modifier character, and
51            there was no match, the code could incorrectly back up beyond the match
52            point, and potentially beyond the first character in the subject,
53            leading to a segfault or an incorrect match result.
56    Version 8.33 28-May-2013
57    --------------------------
59    1.  Added 'U' to some constants that are compared to unsigned integers, to
60        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
61        variables that are added to signed variables, to ensure the result is
62        signed and can be negated.
64    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
65        CMake config files.
67    3.  Revise the creation of config.h.generic so that all boolean macros are
68        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
69        overriding via -D on the command line possible.
71    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
72        to unsigned int is reported to make a quite noticeable speed difference in
73        a specific Windows environment. Testing on Linux did also appear to show
74        some benefit (and it is clearly not harmful). Also fixed the definition of
75        Xop which should be unsigned.
77    5.  Related to (4), changing the definition of the intermediate variable cc
78        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
79        improvements.
81    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
82        unnecessary spaces.
84    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
85        and later.
87    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
89    9.  Optimizing fast_forward_start_bits in JIT.
91    10. Adding support for callouts in JIT, and fixing some issues revealed
92        during this work. Namely:
94        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
96        (b) Minimum length was not checked before the matching is started.
98    11. The value of capture_last that is passed to callouts was incorrect in some
99        cases when there was a capture on one path that was subsequently abandoned
100        after a backtrack. Also, the capture_last value is now reset after a
101        recursion, since all captures are also reset in this case.
103    12. The interpreter no longer returns the "too many substrings" error in the
104        case when an overflowing capture is in a branch that is subsequently
105        abandoned after a backtrack.
107    13. In the pathological case when an offset vector of size 2 is used, pcretest
108        now prints out the matched string after a yield of 0 or 1.
110    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
111        Only supported by the JIT compiler at the moment.
113    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
115    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
116        offset of the starting point of the matching process, provided the offsets
117        vector is large enough.
119    17. The \A escape now records a lookbehind value of 1, though its execution
120        does not actually inspect the previous character. This is to ensure that,
121        in partial multi-segment matching, at least one character from the old
122        segment is retained when a new segment is processed. Otherwise, if there
123        are no lookbehinds in the pattern, \A might match incorrectly at the start
124        of a new segment.
126    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
128    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
129        gives some modest performance improvement in 8-bit mode.
131    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
132        be expressed in certain programming languages using Universal Character
133        Names.
135    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
136        which points out that "non characters" are not "characters that may not
137        appear in Unicode strings" but rather "characters that are reserved for
138        internal use and have only local meaning".
140    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
141        there was a conditional group that depended on an assertion, if the
142        assertion was false, the callout that immediately followed the alternation
143        in the condition was skipped when pcre_exec() was used for matching.
145    23. Allow an explicit callout to be inserted before an assertion that is the
146        condition for a conditional group, for compatibility with automatic
147        callouts, which always insert a callout at this point.
149    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
150        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
152    25. (*PRUNE) is now supported by the JIT compiler.
154    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
156    27. Fix the case where there are two or more SKIPs with arguments that may be
157        ignored.
159    28. (*SKIP) is now supported by the JIT compiler.
161    29. (*THEN) is now supported by the JIT compiler.
163    30. Update RunTest with additional test selector options.
165    31. The way PCRE handles backtracking verbs has been changed in two ways.
167        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
168        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
169        backtracking. In some cases this makes it more Perl-compatible, but Perl's
170        rather obscure rules do not always do the same thing.
172        (2) Previously, backtracking verbs were confined within assertions. This is
173        no longer the case for positive assertions, except for (*ACCEPT). Again,
174        this sometimes improves Perl compatibility, and sometimes does not.
176    32. A number of tests that were in test 2 because Perl did things differently
177        have been moved to test 1, because either Perl or PCRE has changed, and
178        these tests are now compatible.
180    32. Backtracking control verbs are now handled in the same way in JIT and
181        interpreter.
183    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
184        contained a forward subroutine reference caused a compile error.
186    34. Auto-detect and optimize limited repetitions in JIT.
188    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
189        blocking (*UTF) etc.
191    36. In the interpreter, maximizing pattern repetitions for characters and
192        character types now use tail recursion, which reduces stack usage.
194    37. The value of the max lookbehind was not correctly preserved if a compiled
195        and saved regex was reloaded on a host of different endianness.
197    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
198        of the compiled pattern block, expand the flags field from 16 to 32 bits
199        because it was almost full.
201    39. Try madvise first before posix_madvise.
203    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
204        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
205        original change locked out all matches of empty strings. This has been
206        changed so that one match of an empty string per line is recognized.
207        Subsequent searches on the same line (for colouring or for --only-matching,
208        for example) do not recognize empty strings.
210    41. Applied a user patch to fix a number of spelling mistakes in comments.
212    42. Data lines longer than 65536 caused pcretest to crash.
214    43. Clarified the data type for length and startoffset arguments for pcre_exec
215        and pcre_dfa_exec in the function-specific man pages, where they were
216        explicitly stated to be in bytes, never having been updated. I also added
217        some clarification to the pcreapi man page.
219    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
220        a segmentation fault.
223    Version 8.32 30-November-2012
224    -----------------------------
226    1.  Improved JIT compiler optimizations for first character search and single
227        character iterators.
229    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
230        Patch by Daniel Richard G.
232    3.  Single character iterator optimizations in the JIT compiler.
234    4.  Improved JIT compiler optimizations for character ranges.
236    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
237        Reported by Giuseppe D'Angelo.
239    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
240        of a line, was being set incorrectly in cases where .* appeared inside
241        atomic brackets at the start of a pattern, or where there was a subsequent
242        *PRUNE or *SKIP.
244    7.  Improved instruction cache flush for POWER/PowerPC.
245        Patch by Daniel Richard G.
247    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
248        grep:
250        (a) There is now no limit to the number of patterns to be matched.
252        (b) An error is given if a pattern is too long.
254        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
255            are now supported.
257        (d) --exclude-from and --include-from (multiple use) have been added.
259        (e) Exclusions and inclusions now apply to all files and directories, not
260            just to those obtained from scanning a directory recursively.
262        (f) Multiple uses of -f and --file-list are now supported.
264        (g) In a Windows environment, the default for -d has been changed from
265            "read" (the GNU grep default) to "skip", because otherwise the presence
266            of a directory in the file list provokes an error.
268        (h) The documentation has been revised and clarified in places.
270    9.  Improve the matching speed of capturing brackets.
272    10. Changed the meaning of \X so that it now matches a Unicode extended
273        grapheme cluster.
275    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
276        out POSIX threads when JIT support is configured.
278    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
280    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
281        structure could go wrong in environments where size_t is not the same size
282        as int.
284    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
285        set.
287    15. The EBCDIC support had decayed; later updates to the code had included
288        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
289        general tidy up of EBCDIC-related issues, and the documentation was also
290        not quite right. There is now a test that can be run on ASCII systems to
291        check some of the EBCDIC-related things (but is it not a full test).
293    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
294        in a small tidy to the code.
296    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
298    18. If the --only-matching (-o) option in pcregrep is specified multiple
299        times, each one causes appropriate output. For example, -o1 -o2 outputs the
300        substrings matched by the 1st and 2nd capturing parentheses. A separating
301        string can be specified by --om-separator (default empty).
303    19. Improving the first n character searches.
305    20. Turn case lists for horizontal and vertical white space into macros so that
306        they are defined only once.
308    21. This set of changes together give more compatible Unicode case-folding
309        behaviour for characters that have more than one other case when UCP
310        support is available.
312        (a) The Unicode property table now has offsets into a new table of sets of
313            three or more characters that are case-equivalent. The MultiStage2.py
314            script that generates these tables (the pcre_ucd.c file) now scans
315            CaseFolding.txt instead of UnicodeData.txt for character case
316            information.
318        (b) The code for adding characters or ranges of characters to a character
319            class has been abstracted into a generalized function that also handles
320            case-independence. In UTF-mode with UCP support, this uses the new data
321            to handle characters with more than one other case.
323        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
324            whose other case is greater than 256 are now correctly matched
325            caselessly. Previously, the high codepoint matched the low one, but not
326            vice versa.
328        (d) The processing of \h, \H, \v, and \ in character classes now makes use
329            of the new class addition function, using character lists defined as
330            macros alongside the case definitions of 20 above.
332        (e) Caseless back references now work with characters that have more than
333            one other case.
335        (f) General caseless matching of characters with more than one other case
336            is supported.
338    22. Unicode character properties were updated from Unicode 6.2.0
340    23. Improved CMake support under Windows. Patch by Daniel Richard G.
342    24. Add support for 32-bit character strings, and UTF-32
344    25. Major JIT compiler update (code refactoring and bugfixing).
345        Experimental Sparc 32 support is added.
347    26. Applied a modified version of Daniel Richard G's patch to create
348        pcre.h.generic and config.h.generic by "make" instead of in the
349        PrepareRelease script.
351    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
352        pcre_compile.c when checking for a zero character.
354    28. Introducing a native interface for JIT. Through this interface, the compiled
355        machine code can be directly executed. The purpose of this interface is to
356        provide fast pattern matching, so several sanity checks are not performed.
357        However, feature tests are still performed. The new interface provides
358        1.4x speedup compared to the old one.
360    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
361        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
362        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
364    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
365        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
366        the "old" RFC 2279). Instead, it ended up passing a negative length to
367        pcre_exec().
369    31. Add support for GCC's visibility feature to hide internal functions.
371    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
372        "unknown -C option" after outputting 0 or 1.
374    33. There is now support for generating a code coverage report for the test
375        suite in environments where gcc is the compiler and lcov is installed. This
376        is mainly for the benefit of the developers.
378    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
379        unaddressable using valgrind annotations, allowing valgrind to detect
380        invalid memory accesses. This is mainly for the benefit of the developers.
382    25. (*UTF) can now be used to start a pattern in any of the three libraries.
384    26. Give configure error if --enable-cpp but no C++ compiler found.
387    Version 8.31 06-July-2012
388    -------------------------
390    1.  Fixing a wrong JIT test case and some compiler warnings.
392    2.  Removed a bashism from the RunTest script.
394    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
395        to unsigned type, result still unsigned" that was given by an MS compiler
396        on encountering the code "-sizeof(xxx)".
398    4.  Partial matching support is added to the JIT compiler.
400    5.  Fixed several bugs concerned with partial matching of items that consist
401        of more than one character:
403        (a) /^(..)\1/ did not partially match "aba" because checking references was
404            done on an "all or nothing" basis. This also applied to repeated
405            references.
407        (b) \R did not give a hard partial match if \r was found at the end of the
408            subject.
410        (c) \X did not give a hard partial match after matching one or more
411            characters at the end of the subject.
413        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
414            a partial match for the string "\r".
416        (e) When newline was set to CRLF, the metacharacter "." did not recognize
417            a partial match for a CR character at the end of the subject string.
419    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
420        running pcretest, the text "(JIT)" added to the output whenever JIT is
421        actually used to run the match.
423    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
424        or /S+[+] with a digit between 1 and 7.
426    8.  OP_NOT now supports any UTF character not just single-byte ones.
428    9.  (*MARK) control verb is now supported by the JIT compiler.
430    10. The command "./RunTest list" lists the available tests without actually
431        running any of them. (Because I keep forgetting what they all are.)
435    12. Applied a (slightly modified) user-supplied patch that improves performance
436        when the heap is used for recursion (compiled with --disable-stack-for-
437        recursion). Instead of malloc and free for each heap frame each time a
438        logical recursion happens, frames are retained on a chain and re-used where
439        possible. This sometimes gives as much as 30% improvement.
441    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
442        call.
444    14. As documented, (*COMMIT) is now confined to within a positive assertion.
446    15. It is now possible to link pcretest with libedit as an alternative to
447        libreadline.
449    16. (*COMMIT) control verb is now supported by the JIT compiler.
451    17. The Unicode data tables have been updated to Unicode 6.1.0.
453    18. Added --file-list option to pcregrep.
455    19. Added binary file support to pcregrep, including the -a, --binary-files,
456        -I, and --text options.
458    20. The madvise function is renamed for posix_madvise for QNX compatibility
459        reasons. Fixed by Giuseppe D'Angelo.
461    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
463    22. Changed the option for creating symbolic links for 16-bit man pages from
464        -s to -sf so that re-installing does not cause issues.
466    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
468    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
469        matches in certain environments (the workspace was not being correctly
470        retained). Also added to pcre_dfa_exec() a simple plausibility check on
471        some of the workspace data at the beginning of a restart.
473    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
474        was not doing so when it should - probably a typo introduced by SVN 528
475        (change 8.10/14).
477    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
478        \w+ when the character tables indicated that \x{c4} was a word character.
479        There were several related cases, all because the tests for doing a table
480        lookup were testing for characters less than 127 instead of 255.
482    27. If a pattern contains capturing parentheses that are not used in a match,
483        their slots in the ovector are set to -1. For those that are higher than
484        any matched groups, this happens at the end of processing. In the case when
485        there were back references that the ovector was too small to contain
486        (causing temporary malloc'd memory to be used during matching), and the
487        highest capturing number was not used, memory off the end of the ovector
488        was incorrectly being set to -1. (It was using the size of the temporary
489        memory instead of the true size.)
491    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
492        ovector size, it uses memory at the end of the block that it has got.
494    29. Check for an overlong MARK name and give an error at compile time. The
495        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
497    30. JIT compiler update.
499    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
500        Rill for the patch.
502    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
504    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
506    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
507        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
508        was enabled.
510    35. Improve JIT code generation for greedy plus quantifier.
512    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
513        1 to "aa" instead of to an empty string. The bug affected repeated groups
514        that could potentially match an empty string.
516    37. Optimizing single character iterators in JIT.
518    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
519        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
520        codepoints that are too big for the mode are faulted, and in a UTF mode,
521        disallowed codepoints are also faulted.
523    39. If PCRE was compiled with UTF support, in three places in the DFA
524        matcher there was code that should only have been obeyed in UTF mode, but
525        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
526        processing when bytes with values greater than 127 were present. In 16-bit
527        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
528        both cases the values are those that cannot be the first data item in a UTF
529        character. The three items that might have provoked this were recursions,
530        possessively repeated groups, and atomic groups.
532    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
533        and pcregrep, because some OS require shared objects to be explicitly
534        passed to ld, causing the link step to fail if they are not.
536    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
537        mode, patterns that started with \h* or \R* might be incorrectly matched.
540    Version 8.30 04-February-2012
541    -----------------------------
543    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
544        name is defined in ctype.h.
546    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
547        only in quite long subpatterns.
549    3.  Removed the function pcre_info(), which has been obsolete and deprecated
550        since it was replaced by pcre_fullinfo() in February 2000.
552    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
553        match a (*MARK), and the match failed at the start of the subject, a
554        reference to memory before the start of the subject could occur. This bug
555        was introduced by fix 17 of release 8.21.
557    5.  A reference to an unset group with zero minimum repetition was giving
558        totally wrong answers (in non-JavaScript-compatibility mode). For example,
559        /(another)?(\1?)test/ matched against "hello world test". This bug was
560        introduced in release 8.13.
562    6.  Add support for 16-bit character strings (a large amount of work involving
563        many changes and refactorings).
565    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
566        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
567        from a file.
569    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
570        rounding is not applied in this particular case).
572    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
573        if they appear, or are escaped, in patterns.
575    10. Get rid of a number of -Wunused-but-set-variable warnings.
577    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
578        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
579        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
580        also returns the mark "x". This bug applied to capturing parentheses,
581        non-capturing parentheses, and atomic parentheses. It also applied to some
582        assertions.
584    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
585        information out of configure.ac instead of relying on pcre.h.generic, which
586        is not stored in the repository.
588    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
589        -lreadline.
591    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
593    15. Applied Graycode's patch to put the top-level frame on the stack rather
594        than the heap when not using the stack for recursion. This gives a
595        performance improvement in many cases when recursion is not deep.
597    16. Experimental code added to "pcretest -C" to output the stack frame size.
600    Version 8.21 12-Dec-2011
601    ------------------------
603    1.  Updating the JIT compiler.
605    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
606        are added as well.
608    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
609        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
610        calling _pcre_jit_exec. Some extra comments are added.
612    4.  (*MARK) settings inside atomic groups that do not contain any capturing
613        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
614        was introduced by change 18 for 8.20.
616    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
617        ECMA-262 standard.
619    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
620        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
621        This bug was probably introduced by change 9 of 8.13.
623    7.  While fixing 6 above, I noticed that a number of other items were being
624        incorrectly rejected as "not fixed length". This arose partly because newer
625        opcodes had not been added to the fixed-length checking code. I have (a)
626        corrected the bug and added tests for these items, and (b) arranged for an
627        error to occur if an unknown opcode is encountered while checking for fixed
628        length instead of just assuming "not fixed length". The items that were
629        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
630        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
631        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
633    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
634        being incorrectly compiled and would have given unpredicatble results.
636    9.  A possessively repeated subpattern with minimum repeat count greater than
637        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
638        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
639        the first (A) could occur when it should not.
641    10. Add a cast and remove a redundant test from the code.
643    11. JIT should use pcre_malloc/pcre_free for allocation.
645    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
646        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
647        is anything other than /usr, -L is still shown).
649    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
651    14. Perl does not support \N without a following name in a [] class; PCRE now
652        also gives an error.
654    15. If a forward reference was repeated with an upper limit of around 2000,
655        it caused the error "internal error: overran compiling workspace". The
656        maximum number of forward references (including repeats) was limited by the
657        internal workspace, and dependent on the LINK_SIZE. The code has been
658        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
659        the default depends on LINK_SIZE. There is a new upper limit (for safety)
660        of around 200,000 forward references. While doing this, I also speeded up
661        the filling in of repeated forward references.
663    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
664        incorrectly expecting the subject to contain another "a" after the start.
666    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
667        in the match, the SKIP should be ignored. This was not happening; instead
668        the SKIP was being treated as NOMATCH. For patterns such as
669        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
670        tested.
672    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
673        now much more compatible with Perl, in particular in cases where the result
674        is a non-match for a non-anchored pattern. For example, if
675        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
676        "m", where previously it did not return a name. A side effect of this
677        change is that for partial matches, the last encountered mark name is
678        returned, as for non matches. A number of tests that were previously not
679        Perl-compatible have been moved into the Perl-compatible test files. The
680        refactoring has had the pleasing side effect of removing one argument from
681        the match() function, thus reducing its stack requirements.
683    19. If the /S+ option was used in pcretest to study a pattern using JIT,
684        subsequent uses of /S (without +) incorrectly behaved like /S+.
686    21. Retrieve executable code size support for the JIT compiler and fixing
687        some warnings.
689    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
690        not work when the shorter character appeared right at the end of the
691        subject string.
693    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
694        systems.
696    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
697        output it when the /M option is used in pcretest.
699    25. The CheckMan script was not being included in the distribution. Also, added
700        an explicit "perl" to run Perl scripts from the PrepareRelease script
701        because this is reportedly needed in Windows.
703    26. If study data was being save in a file and studying had not found a set of
704        "starts with" bytes for the pattern, the data written to the file (though
705        never used) was taken from uninitialized memory and so caused valgrind to
706        complain.
708    27. Updated RunTest.bat as provided by Sheri Pierce.
710    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
712    29. Computation of memory usage for the table of capturing group names was
713        giving an unnecessarily large value.
716    Version 8.20 21-Oct-2011
717    ------------------------
719    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
720        a POSIX class. After further experiments with Perl, which convinced me that
721        Perl has bugs and confusions, a closing square bracket is no longer allowed
722        in a POSIX name. This bug also affected patterns with classes that started
723        with full stops.
725    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
726        captured substring, but while checking the failing first alternative,
727        substring 1 is temporarily captured. If the output vector supplied to
728        pcre_exec() was not big enough for this capture, the yield of the function
729        was still zero ("insufficient space for captured substrings"). This cannot
730        be totally fixed without adding another stack variable, which seems a lot
731        of expense for a edge case. However, I have improved the situation in cases
732        such as /(a)(b)x|abc/ matched against "abc", where the return code
733        indicates that fewer than the maximum number of slots in the ovector have
734        been set.
736    3.  Related to (2) above: when there are more back references in a pattern than
737        slots in the output vector, pcre_exec() uses temporary memory during
738        matching, and copies in the captures as far as possible afterwards. It was
739        using the entire output vector, but this conflicts with the specification
740        that only 2/3 is used for passing back captured substrings. Now it uses
741        only the first 2/3, for compatibility. This is, of course, another edge
742        case.
744    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
745        main code base, and can be used by building with --enable-jit. When this is
746        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
747        runtime --no-jit option is given.
749    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
750        ovector, the return from the function was zero, implying that there were
751        other matches that did not fit. The correct "exactly full" value is now
752        returned.
754    6.  If a subpattern that was called recursively or as a subroutine contained
755        (*PRUNE) or any other control that caused it to give a non-standard return,
756        invalid errors such as "Error -26 (nested recursion at the same subject
757        position)" or even infinite loops could occur.
759    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
760        computing the minimum length on reaching *ACCEPT, and so ended up with the
761        wrong value of 1 rather than 0. Further investigation indicates that
762        computing a minimum subject length in the presence of *ACCEPT is difficult
763        (think back references, subroutine calls), and so I have changed the code
764        so that no minimum is registered for a pattern that contains *ACCEPT.
766    8.  If (*THEN) was present in the first (true) branch of a conditional group,
767        it was not handled as intended. [But see 16 below.]
769    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
770        Sheri Pierce.
772    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
773        the first byte in a match must be "a".
775    11. Change 17 for 8.13 increased the recursion depth for patterns like
776        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
777        pattern contains any instances of (*THEN). If it does not, the old
778        optimizations are restored. It would be nice to do this on a per-group
779        basis, but at the moment that is not feasible.
781    12. In some environments, the output of pcretest -C is CRLF terminated. This
782        broke RunTest's code that checks for the link size. A single white space
783        character after the value is now allowed for.
785    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
786        For "fr", it uses the Windows-specific input and output files.
788    14. If (*THEN) appeared in a group that was called recursively or as a
789        subroutine, it did not work as intended. [But see next item.]
791    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
792        pattern fragments (but not containing any | characters). If A and B are
793        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
794        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
795        D. In other words, Perl considers parentheses that do not contain any |
796        characters to be part of a surrounding alternative, whereas PCRE was
797        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
798        differently. PCRE now behaves in the same way as Perl, except in the case
799        of subroutine/recursion calls such as (?1) which have in any case always
800        been different (but PCRE had them first :-).
802    16. Related to 15 above: Perl does not treat the | in a conditional group as
803        creating alternatives. Such a group is treated in the same way as an
804        ordinary group without any | characters when processing (*THEN). PCRE has
805        been changed to match Perl's behaviour.
807    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
808        RunGrepTest script failed.
810    18. Change 22 for version 13 caused atomic groups to use more stack. This is
811        inevitable for groups that contain captures, but it can lead to a lot of
812        stack use in large patterns. The old behaviour has been restored for atomic
813        groups that do not contain any capturing parentheses.
815    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
816        suppress the check for a minimum subject length at run time. (If it was
817        given to pcre_exec() or pcre_dfa_exec() it did work.)
819    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
820        fail to work when decoding hex characters in data strings in EBCDIC
821        environments.
823    21. It appears that in at least one Mac OS environment, the isxdigit() function
824        is implemented as a macro that evaluates to its argument more than once,
825        contravening the C 90 Standard (I haven't checked a later standard). There
826        was an instance in pcretest which caused it to go wrong when processing
827        \x{...} escapes in subject strings. The has been rewritten to avoid using
828        things like p++ in the argument of isxdigit().
831    Version 8.13 16-Aug-2011
832    ------------------------
834    1.  The Unicode data tables have been updated to Unicode 6.0.0.
836    2.  Two minor typos in pcre_internal.h have been fixed.
838    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
839        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
840        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
842    4.  There were a number of related bugs in the code for matching backrefences
843        caselessly in UTF-8 mode when codes for the characters concerned were
844        different numbers of bytes. For example, U+023A and U+2C65 are an upper
845        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
846        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
847        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
848        2-byte code at the end of the subject (it thought there wasn't enough data
849        left).
851    5.  Comprehensive information about what went wrong is now returned by
852        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
853        as the output vector has at least 2 elements. The offset of the start of
854        the failing character and a reason code are placed in the vector.
856    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
857        now returned is for the first byte of the failing character, instead of the
858        last byte inspected. This is an incompatible change, but I hope it is small
859        enough not to be a problem. It makes the returned offset consistent with
860        pcre_exec() and pcre_dfa_exec().
862    7.  pcretest now gives a text phrase as well as the error number when
863        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
864        failure, the offset and reason code are output.
866    8.  When \R was used with a maximizing quantifier it failed to skip backwards
867        over a \r\n pair if the subsequent match failed. Instead, it just skipped
868        back over a single character (\n). This seems wrong (because it treated the
869        two characters as a single entity when going forwards), conflicts with the
870        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
871        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
872        has been changed.
874    9.  Some internal refactoring has changed the processing so that the handling
875        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
876        time (the PCRE_DOTALL option was changed this way some time ago: version
877        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
878        which was always a bit of a fudge. It also means that there is one less
879        argument for the match() function, which reduces its stack requirements
880        slightly. This change also fixes an incompatibility with Perl: the pattern
881        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
883    10. More internal refactoring has drastically reduced the number of recursive
884        calls to match() for possessively repeated groups such as (abc)++ when
885        using pcre_exec().
887    11. While implementing 10, a number of bugs in the handling of groups were
888        discovered and fixed:
890        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
891        (a|)*(?1) gave a compile-time internal error.
892        ((a|)+)+  did not notice that the outer group could match an empty string.
893        (^a|^)+   was not marked as anchored.
894        (.*a|.*)+ was not marked as matching at start or after a newline.
896    12. Yet more internal refactoring has removed another argument from the match()
897        function. Special calls to this function are now indicated by setting a
898        value in a variable in the "match data" data block.
900    13. Be more explicit in pcre_study() instead of relying on "default" for
901        opcodes that mean there is no starting character; this means that when new
902        ones are added and accidentally left out of pcre_study(), testing should
903        pick them up.
905    14. The -s option of pcretest has been documented for ages as being an old
906        synonym of -m (show memory usage). I have changed it to mean "force study
907        for every regex", that is, assume /S for every regex. This is similar to -i
908        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
909        using it. It makes it easier to run collections of tests with and without
910        study enabled, and thereby test pcre_study() more easily. All the standard
911        tests are now run with and without -s (but some patterns can be marked as
912        "never study" - see 20 below).
914    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
915        restoration of the capturing data to the outer values was not happening
916        correctly.
918    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
919        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
920        pattern had matched an empty string, and so incorrectly returned a no
921        match.
923    17. There was optimizing code for the last branch of non-capturing parentheses,
924        and also for the obeyed branch of a conditional subexpression, which used
925        tail recursion to cut down on stack usage. Unfortunately, now that there is
926        the possibility of (*THEN) occurring in these branches, tail recursion is
927        no longer possible because the return has to be checked for (*THEN). These
928        two optimizations have therefore been removed. [But see 8.20/11 above.]
930    18. If a pattern containing \R was studied, it was assumed that \R always
931        matched two bytes, thus causing the minimum subject length to be
932        incorrectly computed because \R can also match just one byte.
934    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
935        was incorrectly computed.
937    20. If /S is present twice on a test pattern in pcretest input, it now
938        *disables* studying, thereby overriding the use of -s on the command line
939        (see 14 above). This is necessary for one or two tests to keep the output
940        identical in both cases.
942    21. When (*ACCEPT) was used in an assertion that matched an empty string and
943        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
945    22. When an atomic group that contained a capturing parenthesis was
946        successfully matched, but the branch in which it appeared failed, the
947        capturing was not being forgotten if a higher numbered group was later
948        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
949        group 1 to "a", when in fact it should be unset. This applied to multi-
950        branched capturing and non-capturing groups, repeated or not, and also to
951        positive assertions (capturing in negative assertions does not happen
952        in PCRE) and also to nested atomic groups.
954    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
955        subject after a captured substring, to make it easier to tell which of a
956        number of identical substrings has been captured.
958    24. The way atomic groups are processed by pcre_exec() has been changed so that
959        if they are repeated, backtracking one repetition now resets captured
960        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
961        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
962        "aaa". Previously, it would have been "a". As part of this code
963        refactoring, the way recursive calls are handled has also been changed.
965    25. If an assertion condition captured any substrings, they were not passed
966        back unless some other capturing happened later. For example, if
967        (?(?=(a))a) was matched against "a", no capturing was returned.
969    26. When studying a pattern that contained subroutine calls or assertions,
970        the code for finding the minimum length of a possible match was handling
971        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
972        group 1 called group 2 while simultaneously a separate group 2 called group
973        1). A stack overflow occurred in this case. I have fixed this by limiting
974        the recursion depth to 10.
976    27. Updated RunTest.bat in the distribution to the version supplied by Tom
977        Fortmann. This supports explicit test numbers on the command line, and has
978        argument validation and error reporting.
980    28. An instance of \X with an unlimited repeat could fail if at any point the
981        first character it looked at was a mark character.
983    29. Some minor code refactoring concerning Unicode properties and scripts
984        should reduce the stack requirement of match() slightly.
986    30. Added the '=' option to pcretest to check the setting of unused capturing
987        slots at the end of the pattern, which are documented as being -1, but are
988        not included in the return count.
990    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
991        compiled something random. Now it gives a compile-time error (as does
992        Perl).
994    32. A *MARK encountered during the processing of a positive assertion is now
995        recorded and passed back (compatible with Perl).
997    33. If --only-matching or --colour was set on a pcregrep call whose pattern
998        had alternative anchored branches, the search for a second match in a line
999        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1000        matched the line "0102" twice. The same bug affected patterns that started
1001        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1002        twice.
1004    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1005        does, and because of capturing effects, quantifying parenthesized
1006        assertions may at times be useful. Quantifiers are now allowed for
1007        parenthesized assertions.
1009    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1011    36. \g was being checked for fancy things in a character class, when it should
1012        just be a literal "g".
1014    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1015        appearance of a nested POSIX class supersedes an apparent external class.
1016        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1017        unescaped square brackets may also appear as part of class names. For
1018        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1019        more like Perl. (But see 8.20/1 above.)
1021    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1022        was because it thought it was \N{name}, which is not supported).
1024    39. Add minix to OS list not supporting the -S option in pcretest.
1026    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1027        cannot analyze patterns in sufficient detail to catch mutual recursions
1028        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1029        subgroup is called recursively as a subpattern for a second time at the
1030        same position in the subject string. In previous releases this might have
1031        been caught by the recursion limit, or it might have run out of stack.
1033    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1034        happen only once. PCRE was, however incorrectly giving a compile time error
1035        "recursive call could loop indefinitely" because it cannot analyze the
1036        pattern in sufficient detail. The compile time test no longer happens when
1037        PCRE is compiling a conditional subpattern, but actual runaway loops are
1038        now caught at runtime (see 40 above).
1040    42. It seems that Perl allows any characters other than a closing parenthesis
1041        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1042        has been changed to be the same.
1044    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1045        as not to get warnings when autogen.sh is called. Also changed
1046        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1048    44. To help people who use pcregrep to scan files containing exceedingly long
1049        lines, the following changes have been made:
1051        (a) The default value of the buffer size parameter has been increased from
1052            8K to 20K. (The actual buffer used is three times this size.)
1054        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1055            PCRE is built.
1057        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1058            to be set at run time.
1060        (d) Numerical values in pcregrep options can be followed by K or M, for
1061            example --buffer-size=50K.
1063        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1064            given and the return code is set to 2.
1066    45. Add a pointer to the latest mark to the callout data block.
1068    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1069        partial match of an empty string instead of no match. This was specific to
1070        the use of ".".
1072    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1073        complete match instead of a partial match. This bug was dependent on both
1074        the PCRE_UTF8 and PCRE_DOTALL options being set.
1076    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1077        starting byte set, because \b was not being ignored.
1080    Version 8.12 15-Jan-2011
1081    ------------------------
1083    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1084        checks for such things as part of the documentation building process.
1086    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1087        --match-limit and --recursion-limit options (added for 8.11). In
1088        particular, this made one of the standard tests fail. (The integer value
1089        went into the wrong half of a long int.)
1091    3.  If the --colour option was given to pcregrep with -v (invert match), it
1092        did strange things, either producing crazy output, or crashing. It should,
1093        of course, ignore a request for colour when reporting lines that do not
1094        match.
1096    4.  Another pcregrep bug caused similar problems if --colour was specified with
1097        -M (multiline) and the pattern match finished with a line ending.
1099    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1100        matched in multiline mode, the following line was shown as part of the
1101        match. This seems wrong, so I have changed it.
1103    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1104        the check for further matches in the same line (so they could be coloured)
1105        to overrun the end of the current line. If another match was found, it was
1106        incorrectly shown (and then shown again when found in the next line).
1108    7.  If pcregrep was compiled under Windows, there was a reference to the
1109        function pcregrep_exit() before it was defined. I am assuming this was
1110        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1111        reported by a user. I've moved the definition above the reference.
1114    Version 8.11 10-Dec-2010
1115    ------------------------
1117    1.  (*THEN) was not working properly if there were untried alternatives prior
1118        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1119        backtracked to try for "b" instead of moving to the next alternative branch
1120        at the same level (in this case, to look for "c"). The Perl documentation
1121        is clear that when (*THEN) is backtracked onto, it goes to the "next
1122        alternative in the innermost enclosing group".
1124    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1125        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1126        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1127        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1128        (*THEN).
1130    3.  If \s appeared in a character class, it removed the VT character from
1131        the class, even if it had been included by some previous item, for example
1132        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1133        of \s, but is part of the POSIX "space" class.)
1135    4.  A partial match never returns an empty string (because you can always
1136        match an empty string at the end of the subject); however the checking for
1137        an empty string was starting at the "start of match" point. This has been
1138        changed to the "earliest inspected character" point, because the returned
1139        data for a partial match starts at this character. This means that, for
1140        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1141        (previously it gave "no match").
1143    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1144        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1145        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1146        has an implication that the given string is incomplete (because a partial
1147        match is preferred over a full match). For this reason, these items now
1148        give a partial match in this situation. [Aside: previously, the one case
1149        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1150        match rather than a full match, which was wrong by the old rules, but is
1151        now correct.]
1153    6.  There was a bug in the handling of #-introduced comments, recognized when
1154        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1155        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1156        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1157        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1158        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1159        places in pcre_compile().
1161    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1162        comments when looking ahead for named forward references to subpatterns,
1163        the only newline sequence it recognized was NL. It now handles newlines
1164        according to the set newline convention.
1166    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1167        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1168        cater for a lack of strerror(). These oversights have been fixed.
1170    9.  Added --match-limit and --recursion-limit to pcregrep.
1172    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1174    11. When the -o option was used, pcregrep was setting a return code of 1, even
1175        when matches were found, and --line-buffered was not being honoured.
1177    12. Added an optional parentheses number to the -o and --only-matching options
1178        of pcregrep.
1180    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1181        can match an empty string. The code to do it in pcretest and pcredemo
1182        needed fixing:
1184        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1185            only one byte after an empty string match just before CRLF (this case
1186            just got forgotten; "any" and "anycrlf" were OK).
1188        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1189            mode when an empty string match preceded an ASCII character followed by
1190            a non-ASCII character. (The code for advancing by one character rather
1191            than one byte was nonsense.)
1193        (c) The pcredemo.c sample program did not have any code at all to handle
1194            the cases when CRLF is a valid newline sequence.
1196    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1197        as a starting offset was within the subject string. There is now a new
1198        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1199        negative or greater than the length of the string. In order to test this,
1200        pcretest is extended to allow the setting of negative starting offsets.
1202    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1203        starting offset points to the beginning of a UTF-8 character was
1204        unnecessarily clumsy. I tidied it up.
1206    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1207        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1209    17. Nobody had reported that the --include_dir option, which was added in
1210        release 7.7 should have been called --include-dir (hyphen, not underscore)
1211        for compatibility with GNU grep. I have changed it to --include-dir, but
1212        left --include_dir as an undocumented synonym, and the same for
1213        --exclude-dir, though that is not available in GNU grep, at least as of
1214        release 2.5.4.
1216    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1217        characters from a string of bytes) have been redefined so as not to use
1218        loops, in order to improve performance in some environments. At the same
1219        time, I abstracted some of the common code into auxiliary macros to save
1220        repetition (this should not affect the compiled code).
1222    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1223        compile-time error is now given if \c is not followed by an ASCII
1224        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1225        different, and any byte value is allowed.)
1227    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1228        START_OPTIMIZE option, which is now allowed at compile time - but just
1229        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1230        to pcregrep and other applications that have no direct access to PCRE
1231        options. The new /Y option in pcretest sets this option when calling
1232        pcre_compile().
1234    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1235        back references. Groups containing recursive back references were forced to
1236        be atomic by that change, but in the case of named groups, the amount of
1237        memory required was incorrectly computed, leading to "Failed: internal
1238        error: code overflow". This has been fixed.
1240    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1241        pcretest.c, to avoid build problems in some Borland environments.
1244    Version 8.10 25-Jun-2010
1245    ------------------------
1247    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1248        THEN.
1250    2.  (*ACCEPT) was not working when inside an atomic group.
1252    3.  Inside a character class, \B is treated as a literal by default, but
1253        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1254        causes the error). The code is unchanged, but I tidied the documentation.
1256    4.  Inside a character class, PCRE always treated \R and \X as literals,
1257        whereas Perl faults them if its -w option is set. I have changed PCRE so
1258        that it faults them when PCRE_EXTRA is set.
1260    5.  Added support for \N, which always matches any character other than
1261        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1263    6.  When compiling pcregrep with newer versions of gcc which may have
1264        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1265        declared with attribute warn_unused_result" were given. Just casting the
1266        result to (void) does not stop the warnings; a more elaborate fudge is
1267        needed. I've used a macro to implement this.
1269    7.  Minor change to pcretest.c to avoid a compiler warning.
1271    8.  Added four artifical Unicode properties to help with an option to make
1272        \s etc use properties (see next item). The new properties are: Xan
1273        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1275    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1276        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1277        this option. Modified pcretest to add /W to test this facility. Added
1278        REG_UCP to make it available via the POSIX interface.
1280    10. Added --line-buffered to pcregrep.
1282    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1283        studied, and the match started with a letter with a code point greater than
1284        127 whose first byte was different to the first byte of the other case of
1285        the letter, the other case of this starting letter was not recognized
1286        (#976).
1288    12. If a pattern that was studied started with a repeated Unicode property
1289        test, for example, \p{Nd}+, there was the theoretical possibility of
1290        setting up an incorrect bitmap of starting bytes, but fortunately it could
1291        not have actually happened in practice until change 8 above was made (it
1292        added property types that matched character-matching opcodes).
1294    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1295        possible starting bytes for non-anchored patterns.
1297    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1298        \R, and also a number of cases that involve Unicode properties, both
1299        explicit and implicit when PCRE_UCP is set.
1301    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1302        input, it could crash or give wrong results if characters with values
1303        greater than 0xc0 were present in the subject string. (Detail: it assumed
1304        UTF-8 input when processing these items.)
1306    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1307        size_t is 64-bit (#991).
1309    17. Added a check for running out of memory when PCRE is compiled with
1310        --disable-stack-for-recursion (#990).
1312    18. If the last data line in a file for pcretest does not have a newline on
1313        the end, a newline was missing in the output.
1315    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1316        less than 128) in its various bitmaps. However, there is a facility for
1317        generating tables according to the current locale when PCRE is compiled. It
1318        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1319        characters, are recognized by isspace() and therefore were getting set in
1320        these tables, and indeed these tables seem to approximate to ISO 8859. This
1321        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1322        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1323        which of course cannot start UTF-8 characters. I have changed the code so
1324        that only real ASCII characters (less than 128) and the correct starting
1325        bytes for UTF-8 encodings are set for characters greater than 127 when in
1326        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1327        altogether.)
1329    20. Added the /T option to pcretest so as to be able to run tests with non-
1330        standard character tables, thus making it possible to include the tests
1331        used for 19 above in the standard set of tests.
1333    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1334        reference to a subpattern the other side of a comment that contains an
1335        opening parenthesis caused either an internal compiling error, or a
1336        reference to the wrong subpattern.
1339    Version 8.02 19-Mar-2010
1340    ------------------------
1342    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1344    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1345        configured.
1347    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1348        original author of that file, following a query about its status.
1350    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1351        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1353    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1354        quantifier applied to a forward-referencing subroutine call, could compile
1355        incorrect code or give the error "internal error: previously-checked
1356        referenced subpattern not found".
1358    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1359        variables to point to external functions. For these systems, therefore,
1360        pcre_malloc etc. are now initialized to local functions that call the
1361        relevant global functions.
1363    7.  There were two entries missing in the vectors called coptable and poptable
1364        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1365        I've fixed the data, and added a kludgy way of testing at compile time that
1366        the lengths are correct (equal to the number of opcodes).
1368    8.  Following on from 7, I added a similar kludge to check the length of the
1369        eint vector in pcreposix.c.
1371    9.  Error texts for pcre_compile() are held as one long string to avoid too
1372        much relocation at load time. To find a text, the string is searched,
1373        counting zeros. There was no check for running off the end of the string,
1374        which could happen if a new error number was added without updating the
1375        string.
1377    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1379    11. \K was not working if it appeared in an atomic group or in a group that
1380        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1381        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1382        the assertion is positive, but not if it is negative.
1384    12. Change 11 fortuitously reduced the size of the stack frame used in the
1385        "match()" function of pcre_exec.c by one pointer. Forthcoming
1386        implementation of support for (*MARK) will need an extra pointer on the
1387        stack; I have reserved it now, so that the stack frame size does not
1388        decrease.
1390    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1391        item in branch that calls a recursion is a subroutine call - as in the
1392        second branch in the above example - was incorrectly given the compile-
1393        time error "recursive call could loop indefinitely" because pcre_compile()
1394        was not correctly checking the subroutine for matching a non-empty string.
1396    14. The checks for overrunning compiling workspace could trigger after an
1397        overrun had occurred. This is a "should never occur" error, but it can be
1398        triggered by pathological patterns such as hundreds of nested parentheses.
1399        The checks now trigger 100 bytes before the end of the workspace.
1401    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1404    Version 8.01 19-Jan-2010
1405    ------------------------
1407    1.  If a pattern contained a conditional subpattern with only one branch (in
1408        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1409        computed the wrong minimum data length (which is of course zero for such
1410        subpatterns). This could cause incorrect "no match" results.
1412    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1413        the pattern is reset in the first branch, pcre_compile() failed with
1414        "internal error: code overflow at offset...". This happened only when
1415        the reset was to the original external option setting. (An optimization
1416        abstracts leading options settings into an external setting, which was the
1417        cause of this.)
1419    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1420        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1421        assertion pattern did not match (meaning that the assertion was true), it
1422        was incorrectly treated as false if the SKIP had been reached during the
1423        matching. This also applied to assertions used as conditions.
1425    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1426        assertion subpattern, including such a pattern used as a condition,
1427        unpredictable results occurred, instead of the error return
1430    5.  The C++ GlobalReplace function was not working like Perl for the special
1431        situation when an empty string is matched. It now does the fancy magic
1432        stuff that is necessary.
1434    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1435        removed. (These were left over from very, very early versions of PCRE.)
1437    7.  Some cosmetic changes to the code to make life easier when compiling it
1438        as part of something else:
1440        (a) Change DEBUG to PCRE_DEBUG.
1442        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1443            called "current" as "current_branch", to prevent a collision with the
1444            Linux macro when compiled as a kernel module.
1446        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1447            prevent a collision with the Linux macro when compiled as a kernel
1448            module.
1450    8.  In pcre_compile() there are some checks for integer overflows that used to
1451        cast potentially large values to (double). This has been changed to that
1452        when building, a check for int64_t is made, and if it is found, it is used
1453        instead, thus avoiding the use of floating point arithmetic. (There is no
1454        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1455        double.
1457    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1458        2005 (difference between two addresses compared to an unsigned value).
1460    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1461        custom one, because of the following reported problem in Windows:
1463          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1464              under Win32.
1465          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1466              therefore missing the function definition.
1467          - The compiler thus generates a "C" signature for the test function.
1468          - The linker fails to find the "C" function.
1469          - PCRE fails to configure if asked to do so against libbz2.
1471    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1472        messages were output:
1474          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1475          rerunning libtoolize, to keep the correct libtool macros in-tree.
1476          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1478        I have done both of these things.
1480    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1481        most of the time, it *can* run out if it is given a pattern that contains a
1482        runaway infinite recursion. I updated the discussion in the pcrestack man
1483        page.
1485    13. Now that we have gone to the x.xx style of version numbers, the minor
1486        version may start with zero. Using 08 or 09 is a bad idea because users
1487        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1488        interpreted as invalid octal numbers. I've updated the previous comment in
1489        configure.ac, and also added a check that gives an error if 08 or 09 are
1490        used.
1492    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1493        causing partial matching to fail when the end of the subject matched \W
1494        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1496    15. There were some discrepancies between the declarations in pcre_internal.h
1497        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1498        their definitions. The declarations used "const uschar *" and the
1499        definitions used USPTR. Even though USPTR is normally defined as "const
1500        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1501        reported that: "This difference in casting confuses some C++ compilers, for
1502        example, SunCC recognizes above declarations as different functions and
1503        generates broken code for hbpcre." I have changed the declarations to use
1504        USPTR.
1506    16. GNU libtool is named differently on some systems. The autogen.sh script now
1507        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1508        (FreeBSD).
1510    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1511        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1512        comment: "Figure out how to create a longlong from a string: strtoll and
1513        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1514        instance, but it only takes 2 args instead of 3!"
1516    18. A subtle bug concerned with back references has been fixed by a change of
1517        specification, with a corresponding code fix. A pattern such as
1518        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1519        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1520        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1521        same bug. Such groups have to be quantified to be useful, or contained
1522        inside another quantified group. (If there's no repetition, the reference
1523        can never match.) The problem arises because, having left the group and
1524        moved on to the rest of the pattern, a later failure that backtracks into
1525        the group uses the captured value from the final iteration of the group
1526        rather than the correct earlier one. I have fixed this in PCRE by forcing
1527        any group that contains a reference to itself to be an atomic group; that
1528        is, there cannot be any backtracking into it once it has completed. This is
1529        similar to recursive and subroutine calls.
1532    Version 8.00 19-Oct-09
1533    ----------------------
1535    1.  The table for translating pcre_compile() error codes into POSIX error codes
1536        was out-of-date, and there was no check on the pcre_compile() error code
1537        being within the table. This could lead to an OK return being given in
1538        error.
1540    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1541        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1542        in a Windows environment.
1544    3.  The pcregrep --count option prints the count for each file even when it is
1545        zero, as does GNU grep. However, pcregrep was also printing all files when
1546        --files-with-matches was added. Now, when both options are given, it prints
1547        counts only for those files that have at least one match. (GNU grep just
1548        prints the file name in this circumstance, but including the count seems
1549        more useful - otherwise, why use --count?) Also ensured that the
1550        combination -clh just lists non-zero counts, with no names.
1552    4.  The long form of the pcregrep -F option was incorrectly implemented as
1553        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1554        but it seems right to fix it, and I didn't think it was worth preserving
1555        the old behaviour.
1557    5.  The command line items --regex=pattern and --regexp=pattern were not
1558        recognized by pcregrep, which required --regex pattern or --regexp pattern
1559        (with a space rather than an '='). The man page documented the '=' forms,
1560        which are compatible with GNU grep; these now work.
1562    6.  No libpcreposix.pc file was created for pkg-config; there was just
1563        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1565    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1566        when UCP support is not needed, by modifying the Python script that
1567        generates it from Unicode data files. This should not matter if the module
1568        is correctly used as a library, but I received one complaint about 50K of
1569        unwanted data. My guess is that the person linked everything into his
1570        program rather than using a library. Anyway, it does no harm.
1572    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1573        was a minimum greater than 1 for a wide character in a possessive
1574        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1575        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1576        character. Chaos in the form of incorrect output or a compiling loop could
1577        result.
1579    9.  The restrictions on what a pattern can contain when partial matching is
1580        requested for pcre_exec() have been removed. All patterns can now be
1581        partially matched by this function. In addition, if there are at least two
1582        slots in the offset vector, the offset of the earliest inspected character
1583        for the match and the offset of the end of the subject are set in them when
1584        PCRE_ERROR_PARTIAL is returned.
1586    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1587        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1588        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1589        and may be more useful for multi-segment matching.
1591    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1592        used to be given if ever the end of the subject was reached; now it is
1593        given only if matching could not proceed because another character was
1594        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1595        string "Z", which now yields "no match" instead of "partial match". In the
1596        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1597        final character ended with (*FAIL).
1599    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1600        if the pattern had a "must contain" character that was already found in the
1601        earlier partial match, unless partial matching was again requested. For
1602        example, with the pattern /dog.(body)?/, the "must contain" character is
1603        "g". If the first part-match was for the string "dog", restarting with
1604        "sbody" failed. This bug has been fixed.
1606    13. The string returned by pcre_dfa_exec() after a partial match has been
1607        changed so that it starts at the first inspected character rather than the
1608        first character of the match. This makes a difference only if the pattern
1609        starts with a lookbehind assertion or \b or \B (\K is not supported by
1610        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1611        matching functions compatible, and I think it's the right thing to do.
1613    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1614        so that the demonstration program is easily available in environments where
1615        PCRE has not been installed from source.
1617    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1618        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1619        library.
1621    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1622        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1623        is not the first non-POSIX option to be added. Clearly some people find
1624        these options useful.
1626    17. If a caller to the POSIX matching function regexec() passes a non-zero
1627        value for nmatch with a NULL value for pmatch, the value of
1628        nmatch is forced to zero.
1630    18. RunGrepTest did not have a test for the availability of the -u option of
1631        the diff command, as RunTest does. It now checks in the same way as
1632        RunTest, and also checks for the -b option.
1634    19. If an odd number of negated classes containing just a single character
1635        interposed, within parentheses, between a forward reference to a named
1636        subpattern and the definition of the subpattern, compilation crashed with
1637        an internal error, complaining that it could not find the referenced
1638        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1639        [The bug was that it was starting one character too far in when skipping
1640        over the character class, thus treating the ] as data rather than
1641        terminating the class. This meant it could skip too much.]
1643    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1644        /g option in pcretest when the pattern contains \K, which makes it possible
1645        to have an empty string match not at the start, even when the pattern is
1646        anchored. Updated pcretest and pcredemo to use this option.
1648    21. If the maximum number of capturing subpatterns in a recursion was greater
1649        than the maximum at the outer level, the higher number was returned, but
1650        with unset values at the outer level. The correct (outer level) value is
1651        now given.
1653    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1654        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1655        make it do so. The string so far is captured, making this feature
1656        compatible with Perl.
1658    23. The tests have been re-organized, adding tests 11 and 12, to make it
1659        possible to check the Perl 5.10 features against Perl 5.10.
1661    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1662        pattern matches a fixed length string. PCRE did not allow this; now it
1663        does. Neither allows recursion.
1665    25. I finally figured out how to implement a request to provide the minimum
1666        length of subject string that was needed in order to match a given pattern.
1667        (It was back references and recursion that I had previously got hung up
1668        on.) This code has now been added to pcre_study(); it finds a lower bound
1669        to the length of subject needed. It is not necessarily the greatest lower
1670        bound, but using it to avoid searching strings that are too short does give
1671        some useful speed-ups. The value is available to calling programs via
1672        pcre_fullinfo().
1674    26. While implementing 25, I discovered to my embarrassment that pcretest had
1675        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1676        study optimizations had never been tested with that matching function.
1677        Oops. What is worse, even when it was passed study data, there was a bug in
1678        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1679        were also very few tests of studied patterns with pcre_dfa_exec().
1681    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1682        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1683        on the other side of the coin, they are no longer allowed to have different
1684        names, because these cannot be distinguished in PCRE, and this has caused
1685        confusion. (This is a difference from Perl.)
1687    28. When duplicate subpattern names are present (necessarily with different
1688        numbers, as required by 27 above), and a test is made by name in a
1689        conditional pattern, either for a subpattern having been matched, or for
1690        recursion in such a pattern, all the associated numbered subpatterns are
1691        tested, and the overall condition is true if the condition is true for any
1692        one of them. This is the way Perl works, and is also more like the way
1693        testing by number works.
1696    Version 7.9 11-Apr-09
1697    ---------------------
1699    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1700        (pcretest), all targets were linked against these libraries. This included
1701        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1702        libraries. This caused unwanted dependencies to be created. This problem
1703        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1704        pcretest is linked with readline.
1706    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1707        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1708        moved outside it again, because FALSE and TRUE are already defined in AIX,
1709        but BOOL is not.
1711    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1712        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1714    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1715        hyphens) following file names and line numbers when outputting matching
1716        lines. This is not true; no spaces are inserted. I have also clarified the
1717        wording for the --colour (or --color) option.
1719    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1720        was not coloured; this is different to GNU grep, so I have changed it to be
1721        the same.
1723    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1724        each matching line was coloured. Now it goes on to look for further matches
1725        of any of the test patterns, which is the same behaviour as GNU grep.
1727    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1728        doesn't make sense to accept an empty string match in pcregrep, so I have
1729        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1730        seems to be how GNU grep behaves. [But see later change 40 for release
1731        8.33.]
1733    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1734        start or after a newline", because the conditional assertion was not being
1735        correctly handled. The rule now is that both the assertion and what follows
1736        in the first alternative must satisfy the test.
1738    9.  If auto-callout was enabled in a pattern with a conditional group whose
1739        condition was an assertion, PCRE could crash during matching, both with
1740        pcre_exec() and pcre_dfa_exec().
1742    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1743        used for matching.
1745    11. Unicode property support in character classes was not working for
1746        characters (bytes) greater than 127 when not in UTF-8 mode.
1748    12. Added the -M command line option to pcretest.
1750    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1752    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1754    16. Added comments and documentation about mis-use of no_arg in the C++
1755        wrapper.
1757    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1758        from Martin Jerabek that uses macro names for all relevant character and
1759        string constants.
1761    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1762        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1763        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1764        these, but not everybody uses configure.
1766    19. A conditional group that had only one branch was not being correctly
1767        recognized as an item that could match an empty string. This meant that an
1768        enclosing group might also not be so recognized, causing infinite looping
1769        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1770        with the subject "ab", where knowledge that the repeated group can match
1771        nothing is needed in order to break the loop.
1773    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1774        exec(), but without supplying a callout function, matching went wrong.
1776    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1777        leak if the size of the offset vector was greater than 30. When the vector
1778        is smaller, the saved offsets during recursion go onto a local stack
1779        vector, but for larger vectors malloc() is used. It was failing to free
1780        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1781        error, in fact).
1783    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1784        heapframe that is used only when UTF-8 support is enabled. This caused no
1785        problem, but was untidy.
1787    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1788        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1789        included within another project.
1791    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1792        slightly modified by me:
1794          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1795              not building pcregrep.
1797          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1798              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1800    25. Forward references, both numeric and by name, in patterns that made use of
1801        duplicate group numbers, could behave incorrectly or give incorrect errors,
1802        because when scanning forward to find the reference group, PCRE was not
1803        taking into account the duplicate group numbers. A pattern such as
1804        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1806    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1807        the feature of a custom pointer more persuasive (as requested by a user).
1809    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1810        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1811        pre-defined.
1813    28. Added support for (*UTF8) at the start of a pattern.
1815    29. Arrange for flags added by the "release type" setting in CMake to be shown
1816        in the configuration summary.
1819    Version 7.8 05-Sep-08
1820  ---------------------  ---------------------
1822  1.  Replaced UCP searching code with optimized version as implemented for Ad  1.  Replaced UCP searching code with optimized version as implemented for Ad
# Line 31  Version 8.0 02 Jul-08 Line 1846  Version 8.0 02 Jul-08
1847  7.  Added two (int) casts to pcregrep when printing the difference of two  7.  Added two (int) casts to pcregrep when printing the difference of two
1848      pointers, in case they are 64-bit values.      pointers, in case they are 64-bit values.
1850  8.  Added comments about Mac OS X stack usage to the pcrestack man page and to  8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1851      test 2 if it fails.      test 2 if it fails.
1853  9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,  9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1854      and a #define of that name to empty if it is not externally set. This is to      and a #define of that name to empty if it is not externally set. This is to
1855      allow users of MSVC to set it if necessary.      allow users of MSVC to set it if necessary.
1857  10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from  10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1858      the convenience functions in the pcre_get.c source file.      the convenience functions in the pcre_get.c source file.
1860  11. An option change at the start of a pattern that had top-level alternatives  11. An option change at the start of a pattern that had top-level alternatives
1861      could cause overwriting and/or a crash. This command provoked a crash in      could cause overwriting and/or a crash. This command provoked a crash in
1862      some environments:      some environments:
1864        printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest        printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1866      This potential security problem was recorded as CVE-2008-2371.      This potential security problem was recorded as CVE-2008-2371.
1868    12. For a pattern where the match had to start at the beginning or immediately
1869        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1870        pcre_dfa_exec() could read past the end of the passed subject if there was
1871        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1872        pcretest so that it places the subject at the end of its malloc-ed buffer.
1874    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1875        exec() might read past the end of the data buffer in UTF-8 mode.
1877    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1878        the data contained the byte 0x85 as part of a UTF-8 character within its
1879        first line. This applied both to normal and DFA matching.
1881    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1882        /^[^d]*?$/8 failed to match "abc".
1884    16. Added a missing copyright notice to pcrecpp_internal.h.
1886    17. Make it more clear in the documentation that values returned from
1887        pcre_exec() in ovector are byte offsets, not character counts.
1889    18. Tidied a few places to stop certain compilers from issuing warnings.
1891    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1892        supplied by Stefan Weber. I made a further small update for 7.8 because
1893        there is a change of source arrangements: the pcre_searchfuncs.c module is
1894        replaced by pcre_ucd.c.
1897  Version 7.7 07-May-08  Version 7.7 07-May-08
1898  ---------------------  ---------------------
# Line 128  Version 7.7 07-May-08 Line 1971  Version 7.7 07-May-08
1971      containing () gave an internal compiling error instead of "reference to      containing () gave an internal compiling error instead of "reference to
1972      non-existent subpattern". Fortunately, when the pattern did exist, the      non-existent subpattern". Fortunately, when the pattern did exist, the
1973      compiled code was correct. (When scanning forwards to check for the      compiled code was correct. (When scanning forwards to check for the
1974      existencd of the subpattern, it was treating the data ']' as terminating      existence of the subpattern, it was treating the data ']' as terminating
1975      the class, so got the count wrong. When actually compiling, the reference      the class, so got the count wrong. When actually compiling, the reference
1976      was subsequently set up correctly.)      was subsequently set up correctly.)

Removed from v.360  
changed lines
  Added in v.1350

  ViewVC Help
Powered by ViewVC 1.1.5