/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 193 by ph10, Mon Jul 30 11:34:18 2007 UTC revision 1335 by ph10, Tue May 28 09:13:59 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.3 05-Jul-07  Version 8.33 28-May-2013
5    --------------------------
6    
7    1.  Added 'U' to some constants that are compared to unsigned integers, to
8        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9        variables that are added to signed variables, to ensure the result is
10        signed and can be negated.
11    
12    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
13        CMake config files.
14    
15    3.  Revise the creation of config.h.generic so that all boolean macros are
16        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17        overriding via -D on the command line possible.
18    
19    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20        to unsigned int is reported to make a quite noticeable speed difference in
21        a specific Windows environment. Testing on Linux did also appear to show
22        some benefit (and it is clearly not harmful). Also fixed the definition of
23        Xop which should be unsigned.
24    
25    5.  Related to (4), changing the definition of the intermediate variable cc
26        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27        improvements.
28    
29    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
30        unnecessary spaces.
31    
32    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33        and later.
34    
35    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
36    
37    9.  Optimizing fast_forward_start_bits in JIT.
38    
39    10. Adding support for callouts in JIT, and fixing some issues revealed
40        during this work. Namely:
41    
42        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
43    
44        (b) Minimum length was not checked before the matching is started.
45    
46    11. The value of capture_last that is passed to callouts was incorrect in some
47        cases when there was a capture on one path that was subsequently abandoned
48        after a backtrack. Also, the capture_last value is now reset after a
49        recursion, since all captures are also reset in this case.
50    
51    12. The interpreter no longer returns the "too many substrings" error in the
52        case when an overflowing capture is in a branch that is subsequently
53        abandoned after a backtrack.
54    
55    13. In the pathological case when an offset vector of size 2 is used, pcretest
56        now prints out the matched string after a yield of 0 or 1.
57    
58    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59        Only supported by the JIT compiler at the moment.
60    
61    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
62    
63    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64        offset of the starting point of the matching process, provided the offsets
65        vector is large enough.
66    
67    17. The \A escape now records a lookbehind value of 1, though its execution
68        does not actually inspect the previous character. This is to ensure that,
69        in partial multi-segment matching, at least one character from the old
70        segment is retained when a new segment is processed. Otherwise, if there
71        are no lookbehinds in the pattern, \A might match incorrectly at the start
72        of a new segment.
73    
74    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
75    
76    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77        gives some modest performance improvement in 8-bit mode.
78    
79    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80        be expressed in certain programming languages using Universal Character
81        Names.
82    
83    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84        which points out that "non characters" are not "characters that may not
85        appear in Unicode strings" but rather "characters that are reserved for
86        internal use and have only local meaning".
87    
88    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89        there was a conditional group that depended on an assertion, if the
90        assertion was false, the callout that immediately followed the alternation
91        in the condition was skipped when pcre_exec() was used for matching.
92    
93    23. Allow an explicit callout to be inserted before an assertion that is the
94        condition for a conditional group, for compatibility with automatic
95        callouts, which always insert a callout at this point.
96    
97    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
98        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
99    
100    25. (*PRUNE) is now supported by the JIT compiler.
101    
102    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
103    
104    27. Fix the case where there are two or more SKIPs with arguments that may be
105        ignored.
106    
107    28. (*SKIP) is now supported by the JIT compiler.
108    
109    29. (*THEN) is now supported by the JIT compiler.
110    
111    30. Update RunTest with additional test selector options.
112    
113    31. The way PCRE handles backtracking verbs has been changed in two ways.
114    
115        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
116        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
117        backtracking. In some cases this makes it more Perl-compatible, but Perl's
118        rather obscure rules do not always do the same thing.
119    
120        (2) Previously, backtracking verbs were confined within assertions. This is
121        no longer the case for positive assertions, except for (*ACCEPT). Again,
122        this sometimes improves Perl compatibility, and sometimes does not.
123    
124    32. A number of tests that were in test 2 because Perl did things differently
125        have been moved to test 1, because either Perl or PCRE has changed, and
126        these tests are now compatible.
127    
128    32. Backtracking control verbs are now handled in the same way in JIT and
129        interpreter.
130    
131    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
132        contained a forward subroutine reference caused a compile error.
133    
134    34. Auto-detect and optimize limited repetitions in JIT.
135    
136    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
137        blocking (*UTF) etc.
138    
139    36. In the interpreter, maximizing pattern repetitions for characters and
140        character types now use tail recursion, which reduces stack usage.
141    
142    37. The value of the max lookbehind was not correctly preserved if a compiled
143        and saved regex was reloaded on a host of different endianness.
144    
145    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
146        of the compiled pattern block, expand the flags field from 16 to 32 bits
147        because it was almost full.
148    
149    39. Try madvise first before posix_madvise.
150    
151    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
152        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
153        original change locked out all matches of empty strings. This has been
154        changed so that one match of an empty string per line is recognized.
155        Subsequent searches on the same line (for colouring or for --only-matching,
156        for example) do not recognize empty strings.
157    
158    41. Applied a user patch to fix a number of spelling mistakes in comments.
159    
160    42. Data lines longer than 65536 caused pcretest to crash.
161    
162    43. Clarified the data type for length and startoffset arguments for pcre_exec
163        and pcre_dfa_exec in the function-specific man pages, where they were
164        explicitly stated to be in bytes, never having been updated. I also added
165        some clarification to the pcreapi man page.
166    
167    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
168        a segmentation fault.
169    
170    
171    Version 8.32 30-November-2012
172    -----------------------------
173    
174    1.  Improved JIT compiler optimizations for first character search and single
175        character iterators.
176    
177    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
178        Patch by Daniel Richard G.
179    
180    3.  Single character iterator optimizations in the JIT compiler.
181    
182    4.  Improved JIT compiler optimizations for character ranges.
183    
184    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
185        Reported by Giuseppe D'Angelo.
186    
187    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
188        of a line, was being set incorrectly in cases where .* appeared inside
189        atomic brackets at the start of a pattern, or where there was a subsequent
190        *PRUNE or *SKIP.
191    
192    7.  Improved instruction cache flush for POWER/PowerPC.
193        Patch by Daniel Richard G.
194    
195    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
196        grep:
197    
198        (a) There is now no limit to the number of patterns to be matched.
199    
200        (b) An error is given if a pattern is too long.
201    
202        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
203            are now supported.
204    
205        (d) --exclude-from and --include-from (multiple use) have been added.
206    
207        (e) Exclusions and inclusions now apply to all files and directories, not
208            just to those obtained from scanning a directory recursively.
209    
210        (f) Multiple uses of -f and --file-list are now supported.
211    
212        (g) In a Windows environment, the default for -d has been changed from
213            "read" (the GNU grep default) to "skip", because otherwise the presence
214            of a directory in the file list provokes an error.
215    
216        (h) The documentation has been revised and clarified in places.
217    
218    9.  Improve the matching speed of capturing brackets.
219    
220    10. Changed the meaning of \X so that it now matches a Unicode extended
221        grapheme cluster.
222    
223    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
224        out POSIX threads when JIT support is configured.
225    
226    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
227    
228    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
229        structure could go wrong in environments where size_t is not the same size
230        as int.
231    
232    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
233        set.
234    
235    15. The EBCDIC support had decayed; later updates to the code had included
236        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
237        general tidy up of EBCDIC-related issues, and the documentation was also
238        not quite right. There is now a test that can be run on ASCII systems to
239        check some of the EBCDIC-related things (but is it not a full test).
240    
241    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
242        in a small tidy to the code.
243    
244    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
245    
246    18. If the --only-matching (-o) option in pcregrep is specified multiple
247        times, each one causes appropriate output. For example, -o1 -o2 outputs the
248        substrings matched by the 1st and 2nd capturing parentheses. A separating
249        string can be specified by --om-separator (default empty).
250    
251    19. Improving the first n character searches.
252    
253    20. Turn case lists for horizontal and vertical white space into macros so that
254        they are defined only once.
255    
256    21. This set of changes together give more compatible Unicode case-folding
257        behaviour for characters that have more than one other case when UCP
258        support is available.
259    
260        (a) The Unicode property table now has offsets into a new table of sets of
261            three or more characters that are case-equivalent. The MultiStage2.py
262            script that generates these tables (the pcre_ucd.c file) now scans
263            CaseFolding.txt instead of UnicodeData.txt for character case
264            information.
265    
266        (b) The code for adding characters or ranges of characters to a character
267            class has been abstracted into a generalized function that also handles
268            case-independence. In UTF-mode with UCP support, this uses the new data
269            to handle characters with more than one other case.
270    
271        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
272            whose other case is greater than 256 are now correctly matched
273            caselessly. Previously, the high codepoint matched the low one, but not
274            vice versa.
275    
276        (d) The processing of \h, \H, \v, and \ in character classes now makes use
277            of the new class addition function, using character lists defined as
278            macros alongside the case definitions of 20 above.
279    
280        (e) Caseless back references now work with characters that have more than
281            one other case.
282    
283        (f) General caseless matching of characters with more than one other case
284            is supported.
285    
286    22. Unicode character properties were updated from Unicode 6.2.0
287    
288    23. Improved CMake support under Windows. Patch by Daniel Richard G.
289    
290    24. Add support for 32-bit character strings, and UTF-32
291    
292    25. Major JIT compiler update (code refactoring and bugfixing).
293        Experimental Sparc 32 support is added.
294    
295    26. Applied a modified version of Daniel Richard G's patch to create
296        pcre.h.generic and config.h.generic by "make" instead of in the
297        PrepareRelease script.
298    
299    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
300        pcre_compile.c when checking for a zero character.
301    
302    28. Introducing a native interface for JIT. Through this interface, the compiled
303        machine code can be directly executed. The purpose of this interface is to
304        provide fast pattern matching, so several sanity checks are not performed.
305        However, feature tests are still performed. The new interface provides
306        1.4x speedup compared to the old one.
307    
308    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
309        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
310        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
311    
312    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
313        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
314        the "old" RFC 2279). Instead, it ended up passing a negative length to
315        pcre_exec().
316    
317    31. Add support for GCC's visibility feature to hide internal functions.
318    
319    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
320        "unknown -C option" after outputting 0 or 1.
321    
322    33. There is now support for generating a code coverage report for the test
323        suite in environments where gcc is the compiler and lcov is installed. This
324        is mainly for the benefit of the developers.
325    
326    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
327        unaddressable using valgrind annotations, allowing valgrind to detect
328        invalid memory accesses. This is mainly for the benefit of the developers.
329    
330    25. (*UTF) can now be used to start a pattern in any of the three libraries.
331    
332    26. Give configure error if --enable-cpp but no C++ compiler found.
333    
334    
335    Version 8.31 06-July-2012
336    -------------------------
337    
338    1.  Fixing a wrong JIT test case and some compiler warnings.
339    
340    2.  Removed a bashism from the RunTest script.
341    
342    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
343        to unsigned type, result still unsigned" that was given by an MS compiler
344        on encountering the code "-sizeof(xxx)".
345    
346    4.  Partial matching support is added to the JIT compiler.
347    
348    5.  Fixed several bugs concerned with partial matching of items that consist
349        of more than one character:
350    
351        (a) /^(..)\1/ did not partially match "aba" because checking references was
352            done on an "all or nothing" basis. This also applied to repeated
353            references.
354    
355        (b) \R did not give a hard partial match if \r was found at the end of the
356            subject.
357    
358        (c) \X did not give a hard partial match after matching one or more
359            characters at the end of the subject.
360    
361        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
362            a partial match for the string "\r".
363    
364        (e) When newline was set to CRLF, the metacharacter "." did not recognize
365            a partial match for a CR character at the end of the subject string.
366    
367    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
368        running pcretest, the text "(JIT)" added to the output whenever JIT is
369        actually used to run the match.
370    
371    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
372        or /S+[+] with a digit between 1 and 7.
373    
374    8.  OP_NOT now supports any UTF character not just single-byte ones.
375    
376    9.  (*MARK) control verb is now supported by the JIT compiler.
377    
378    10. The command "./RunTest list" lists the available tests without actually
379        running any of them. (Because I keep forgetting what they all are.)
380    
381    11. Add PCRE_INFO_MAXLOOKBEHIND.
382    
383    12. Applied a (slightly modified) user-supplied patch that improves performance
384        when the heap is used for recursion (compiled with --disable-stack-for-
385        recursion). Instead of malloc and free for each heap frame each time a
386        logical recursion happens, frames are retained on a chain and re-used where
387        possible. This sometimes gives as much as 30% improvement.
388    
389    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
390        call.
391    
392    14. As documented, (*COMMIT) is now confined to within a positive assertion.
393    
394    15. It is now possible to link pcretest with libedit as an alternative to
395        libreadline.
396    
397    16. (*COMMIT) control verb is now supported by the JIT compiler.
398    
399    17. The Unicode data tables have been updated to Unicode 6.1.0.
400    
401    18. Added --file-list option to pcregrep.
402    
403    19. Added binary file support to pcregrep, including the -a, --binary-files,
404        -I, and --text options.
405    
406    20. The madvise function is renamed for posix_madvise for QNX compatibility
407        reasons. Fixed by Giuseppe D'Angelo.
408    
409    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
410    
411    22. Changed the option for creating symbolic links for 16-bit man pages from
412        -s to -sf so that re-installing does not cause issues.
413    
414    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
415    
416    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
417        matches in certain environments (the workspace was not being correctly
418        retained). Also added to pcre_dfa_exec() a simple plausibility check on
419        some of the workspace data at the beginning of a restart.
420    
421    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
422        was not doing so when it should - probably a typo introduced by SVN 528
423        (change 8.10/14).
424    
425    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
426        \w+ when the character tables indicated that \x{c4} was a word character.
427        There were several related cases, all because the tests for doing a table
428        lookup were testing for characters less than 127 instead of 255.
429    
430    27. If a pattern contains capturing parentheses that are not used in a match,
431        their slots in the ovector are set to -1. For those that are higher than
432        any matched groups, this happens at the end of processing. In the case when
433        there were back references that the ovector was too small to contain
434        (causing temporary malloc'd memory to be used during matching), and the
435        highest capturing number was not used, memory off the end of the ovector
436        was incorrectly being set to -1. (It was using the size of the temporary
437        memory instead of the true size.)
438    
439    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
440        ovector size, it uses memory at the end of the block that it has got.
441    
442    29. Check for an overlong MARK name and give an error at compile time. The
443        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
444    
445    30. JIT compiler update.
446    
447    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
448        Rill for the patch.
449    
450    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
451    
452    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
453    
454    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
455        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
456        was enabled.
457    
458    35. Improve JIT code generation for greedy plus quantifier.
459    
460    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
461        1 to "aa" instead of to an empty string. The bug affected repeated groups
462        that could potentially match an empty string.
463    
464    37. Optimizing single character iterators in JIT.
465    
466    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
467        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
468        codepoints that are too big for the mode are faulted, and in a UTF mode,
469        disallowed codepoints are also faulted.
470    
471    39. If PCRE was compiled with UTF support, in three places in the DFA
472        matcher there was code that should only have been obeyed in UTF mode, but
473        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
474        processing when bytes with values greater than 127 were present. In 16-bit
475        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
476        both cases the values are those that cannot be the first data item in a UTF
477        character. The three items that might have provoked this were recursions,
478        possessively repeated groups, and atomic groups.
479    
480    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
481        and pcregrep, because some OS require shared objects to be explicitly
482        passed to ld, causing the link step to fail if they are not.
483    
484    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
485        mode, patterns that started with \h* or \R* might be incorrectly matched.
486    
487    
488    Version 8.30 04-February-2012
489    -----------------------------
490    
491    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
492        name is defined in ctype.h.
493    
494    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
495        only in quite long subpatterns.
496    
497    3.  Removed the function pcre_info(), which has been obsolete and deprecated
498        since it was replaced by pcre_fullinfo() in February 2000.
499    
500    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
501        match a (*MARK), and the match failed at the start of the subject, a
502        reference to memory before the start of the subject could occur. This bug
503        was introduced by fix 17 of release 8.21.
504    
505    5.  A reference to an unset group with zero minimum repetition was giving
506        totally wrong answers (in non-JavaScript-compatibility mode). For example,
507        /(another)?(\1?)test/ matched against "hello world test". This bug was
508        introduced in release 8.13.
509    
510    6.  Add support for 16-bit character strings (a large amount of work involving
511        many changes and refactorings).
512    
513    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
514        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
515        from a file.
516    
517    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
518        rounding is not applied in this particular case).
519    
520    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
521        if they appear, or are escaped, in patterns.
522    
523    10. Get rid of a number of -Wunused-but-set-variable warnings.
524    
525    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
526        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
527        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
528        also returns the mark "x". This bug applied to capturing parentheses,
529        non-capturing parentheses, and atomic parentheses. It also applied to some
530        assertions.
531    
532    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
533        information out of configure.ac instead of relying on pcre.h.generic, which
534        is not stored in the repository.
535    
536    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
537        -lreadline.
538    
539    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
540    
541    15. Applied Graycode's patch to put the top-level frame on the stack rather
542        than the heap when not using the stack for recursion. This gives a
543        performance improvement in many cases when recursion is not deep.
544    
545    16. Experimental code added to "pcretest -C" to output the stack frame size.
546    
547    
548    Version 8.21 12-Dec-2011
549    ------------------------
550    
551    1.  Updating the JIT compiler.
552    
553    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
554        are added as well.
555    
556    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
557        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
558        calling _pcre_jit_exec. Some extra comments are added.
559    
560    4.  (*MARK) settings inside atomic groups that do not contain any capturing
561        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
562        was introduced by change 18 for 8.20.
563    
564    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
565        ECMA-262 standard.
566    
567    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
568        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
569        This bug was probably introduced by change 9 of 8.13.
570    
571    7.  While fixing 6 above, I noticed that a number of other items were being
572        incorrectly rejected as "not fixed length". This arose partly because newer
573        opcodes had not been added to the fixed-length checking code. I have (a)
574        corrected the bug and added tests for these items, and (b) arranged for an
575        error to occur if an unknown opcode is encountered while checking for fixed
576        length instead of just assuming "not fixed length". The items that were
577        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
578        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
579        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
580    
581    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
582        being incorrectly compiled and would have given unpredicatble results.
583    
584    9.  A possessively repeated subpattern with minimum repeat count greater than
585        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
586        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
587        the first (A) could occur when it should not.
588    
589    10. Add a cast and remove a redundant test from the code.
590    
591    11. JIT should use pcre_malloc/pcre_free for allocation.
592    
593    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
594        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
595        is anything other than /usr, -L is still shown).
596    
597    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
598    
599    14. Perl does not support \N without a following name in a [] class; PCRE now
600        also gives an error.
601    
602    15. If a forward reference was repeated with an upper limit of around 2000,
603        it caused the error "internal error: overran compiling workspace". The
604        maximum number of forward references (including repeats) was limited by the
605        internal workspace, and dependent on the LINK_SIZE. The code has been
606        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
607        the default depends on LINK_SIZE. There is a new upper limit (for safety)
608        of around 200,000 forward references. While doing this, I also speeded up
609        the filling in of repeated forward references.
610    
611    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
612        incorrectly expecting the subject to contain another "a" after the start.
613    
614    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
615        in the match, the SKIP should be ignored. This was not happening; instead
616        the SKIP was being treated as NOMATCH. For patterns such as
617        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
618        tested.
619    
620    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
621        now much more compatible with Perl, in particular in cases where the result
622        is a non-match for a non-anchored pattern. For example, if
623        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
624        "m", where previously it did not return a name. A side effect of this
625        change is that for partial matches, the last encountered mark name is
626        returned, as for non matches. A number of tests that were previously not
627        Perl-compatible have been moved into the Perl-compatible test files. The
628        refactoring has had the pleasing side effect of removing one argument from
629        the match() function, thus reducing its stack requirements.
630    
631    19. If the /S+ option was used in pcretest to study a pattern using JIT,
632        subsequent uses of /S (without +) incorrectly behaved like /S+.
633    
634    21. Retrieve executable code size support for the JIT compiler and fixing
635        some warnings.
636    
637    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
638        not work when the shorter character appeared right at the end of the
639        subject string.
640    
641    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
642        systems.
643    
644    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
645        output it when the /M option is used in pcretest.
646    
647    25. The CheckMan script was not being included in the distribution. Also, added
648        an explicit "perl" to run Perl scripts from the PrepareRelease script
649        because this is reportedly needed in Windows.
650    
651    26. If study data was being save in a file and studying had not found a set of
652        "starts with" bytes for the pattern, the data written to the file (though
653        never used) was taken from uninitialized memory and so caused valgrind to
654        complain.
655    
656    27. Updated RunTest.bat as provided by Sheri Pierce.
657    
658    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
659    
660    29. Computation of memory usage for the table of capturing group names was
661        giving an unnecessarily large value.
662    
663    
664    Version 8.20 21-Oct-2011
665    ------------------------
666    
667    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
668        a POSIX class. After further experiments with Perl, which convinced me that
669        Perl has bugs and confusions, a closing square bracket is no longer allowed
670        in a POSIX name. This bug also affected patterns with classes that started
671        with full stops.
672    
673    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
674        captured substring, but while checking the failing first alternative,
675        substring 1 is temporarily captured. If the output vector supplied to
676        pcre_exec() was not big enough for this capture, the yield of the function
677        was still zero ("insufficient space for captured substrings"). This cannot
678        be totally fixed without adding another stack variable, which seems a lot
679        of expense for a edge case. However, I have improved the situation in cases
680        such as /(a)(b)x|abc/ matched against "abc", where the return code
681        indicates that fewer than the maximum number of slots in the ovector have
682        been set.
683    
684    3.  Related to (2) above: when there are more back references in a pattern than
685        slots in the output vector, pcre_exec() uses temporary memory during
686        matching, and copies in the captures as far as possible afterwards. It was
687        using the entire output vector, but this conflicts with the specification
688        that only 2/3 is used for passing back captured substrings. Now it uses
689        only the first 2/3, for compatibility. This is, of course, another edge
690        case.
691    
692    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
693        main code base, and can be used by building with --enable-jit. When this is
694        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
695        runtime --no-jit option is given.
696    
697    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
698        ovector, the return from the function was zero, implying that there were
699        other matches that did not fit. The correct "exactly full" value is now
700        returned.
701    
702    6.  If a subpattern that was called recursively or as a subroutine contained
703        (*PRUNE) or any other control that caused it to give a non-standard return,
704        invalid errors such as "Error -26 (nested recursion at the same subject
705        position)" or even infinite loops could occur.
706    
707    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
708        computing the minimum length on reaching *ACCEPT, and so ended up with the
709        wrong value of 1 rather than 0. Further investigation indicates that
710        computing a minimum subject length in the presence of *ACCEPT is difficult
711        (think back references, subroutine calls), and so I have changed the code
712        so that no minimum is registered for a pattern that contains *ACCEPT.
713    
714    8.  If (*THEN) was present in the first (true) branch of a conditional group,
715        it was not handled as intended. [But see 16 below.]
716    
717    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
718        Sheri Pierce.
719    
720    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
721        the first byte in a match must be "a".
722    
723    11. Change 17 for 8.13 increased the recursion depth for patterns like
724        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
725        pattern contains any instances of (*THEN). If it does not, the old
726        optimizations are restored. It would be nice to do this on a per-group
727        basis, but at the moment that is not feasible.
728    
729    12. In some environments, the output of pcretest -C is CRLF terminated. This
730        broke RunTest's code that checks for the link size. A single white space
731        character after the value is now allowed for.
732    
733    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
734        For "fr", it uses the Windows-specific input and output files.
735    
736    14. If (*THEN) appeared in a group that was called recursively or as a
737        subroutine, it did not work as intended. [But see next item.]
738    
739    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
740        pattern fragments (but not containing any | characters). If A and B are
741        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
742        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
743        D. In other words, Perl considers parentheses that do not contain any |
744        characters to be part of a surrounding alternative, whereas PCRE was
745        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
746        differently. PCRE now behaves in the same way as Perl, except in the case
747        of subroutine/recursion calls such as (?1) which have in any case always
748        been different (but PCRE had them first :-).
749    
750    16. Related to 15 above: Perl does not treat the | in a conditional group as
751        creating alternatives. Such a group is treated in the same way as an
752        ordinary group without any | characters when processing (*THEN). PCRE has
753        been changed to match Perl's behaviour.
754    
755    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
756        RunGrepTest script failed.
757    
758    18. Change 22 for version 13 caused atomic groups to use more stack. This is
759        inevitable for groups that contain captures, but it can lead to a lot of
760        stack use in large patterns. The old behaviour has been restored for atomic
761        groups that do not contain any capturing parentheses.
762    
763    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
764        suppress the check for a minimum subject length at run time. (If it was
765        given to pcre_exec() or pcre_dfa_exec() it did work.)
766    
767    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
768        fail to work when decoding hex characters in data strings in EBCDIC
769        environments.
770    
771    21. It appears that in at least one Mac OS environment, the isxdigit() function
772        is implemented as a macro that evaluates to its argument more than once,
773        contravening the C 90 Standard (I haven't checked a later standard). There
774        was an instance in pcretest which caused it to go wrong when processing
775        \x{...} escapes in subject strings. The has been rewritten to avoid using
776        things like p++ in the argument of isxdigit().
777    
778    
779    Version 8.13 16-Aug-2011
780    ------------------------
781    
782    1.  The Unicode data tables have been updated to Unicode 6.0.0.
783    
784    2.  Two minor typos in pcre_internal.h have been fixed.
785    
786    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
787        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
788        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
789    
790    4.  There were a number of related bugs in the code for matching backrefences
791        caselessly in UTF-8 mode when codes for the characters concerned were
792        different numbers of bytes. For example, U+023A and U+2C65 are an upper
793        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
794        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
795        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
796        2-byte code at the end of the subject (it thought there wasn't enough data
797        left).
798    
799    5.  Comprehensive information about what went wrong is now returned by
800        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
801        as the output vector has at least 2 elements. The offset of the start of
802        the failing character and a reason code are placed in the vector.
803    
804    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
805        now returned is for the first byte of the failing character, instead of the
806        last byte inspected. This is an incompatible change, but I hope it is small
807        enough not to be a problem. It makes the returned offset consistent with
808        pcre_exec() and pcre_dfa_exec().
809    
810    7.  pcretest now gives a text phrase as well as the error number when
811        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
812        failure, the offset and reason code are output.
813    
814    8.  When \R was used with a maximizing quantifier it failed to skip backwards
815        over a \r\n pair if the subsequent match failed. Instead, it just skipped
816        back over a single character (\n). This seems wrong (because it treated the
817        two characters as a single entity when going forwards), conflicts with the
818        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
819        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
820        has been changed.
821    
822    9.  Some internal refactoring has changed the processing so that the handling
823        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
824        time (the PCRE_DOTALL option was changed this way some time ago: version
825        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
826        which was always a bit of a fudge. It also means that there is one less
827        argument for the match() function, which reduces its stack requirements
828        slightly. This change also fixes an incompatibility with Perl: the pattern
829        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
830    
831    10. More internal refactoring has drastically reduced the number of recursive
832        calls to match() for possessively repeated groups such as (abc)++ when
833        using pcre_exec().
834    
835    11. While implementing 10, a number of bugs in the handling of groups were
836        discovered and fixed:
837    
838        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
839        (a|)*(?1) gave a compile-time internal error.
840        ((a|)+)+  did not notice that the outer group could match an empty string.
841        (^a|^)+   was not marked as anchored.
842        (.*a|.*)+ was not marked as matching at start or after a newline.
843    
844    12. Yet more internal refactoring has removed another argument from the match()
845        function. Special calls to this function are now indicated by setting a
846        value in a variable in the "match data" data block.
847    
848    13. Be more explicit in pcre_study() instead of relying on "default" for
849        opcodes that mean there is no starting character; this means that when new
850        ones are added and accidentally left out of pcre_study(), testing should
851        pick them up.
852    
853    14. The -s option of pcretest has been documented for ages as being an old
854        synonym of -m (show memory usage). I have changed it to mean "force study
855        for every regex", that is, assume /S for every regex. This is similar to -i
856        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
857        using it. It makes it easier to run collections of tests with and without
858        study enabled, and thereby test pcre_study() more easily. All the standard
859        tests are now run with and without -s (but some patterns can be marked as
860        "never study" - see 20 below).
861    
862    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
863        restoration of the capturing data to the outer values was not happening
864        correctly.
865    
866    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
867        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
868        pattern had matched an empty string, and so incorrectly returned a no
869        match.
870    
871    17. There was optimizing code for the last branch of non-capturing parentheses,
872        and also for the obeyed branch of a conditional subexpression, which used
873        tail recursion to cut down on stack usage. Unfortunately, now that there is
874        the possibility of (*THEN) occurring in these branches, tail recursion is
875        no longer possible because the return has to be checked for (*THEN). These
876        two optimizations have therefore been removed. [But see 8.20/11 above.]
877    
878    18. If a pattern containing \R was studied, it was assumed that \R always
879        matched two bytes, thus causing the minimum subject length to be
880        incorrectly computed because \R can also match just one byte.
881    
882    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
883        was incorrectly computed.
884    
885    20. If /S is present twice on a test pattern in pcretest input, it now
886        *disables* studying, thereby overriding the use of -s on the command line
887        (see 14 above). This is necessary for one or two tests to keep the output
888        identical in both cases.
889    
890    21. When (*ACCEPT) was used in an assertion that matched an empty string and
891        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
892    
893    22. When an atomic group that contained a capturing parenthesis was
894        successfully matched, but the branch in which it appeared failed, the
895        capturing was not being forgotten if a higher numbered group was later
896        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
897        group 1 to "a", when in fact it should be unset. This applied to multi-
898        branched capturing and non-capturing groups, repeated or not, and also to
899        positive assertions (capturing in negative assertions does not happen
900        in PCRE) and also to nested atomic groups.
901    
902    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
903        subject after a captured substring, to make it easier to tell which of a
904        number of identical substrings has been captured.
905    
906    24. The way atomic groups are processed by pcre_exec() has been changed so that
907        if they are repeated, backtracking one repetition now resets captured
908        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
909        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
910        "aaa". Previously, it would have been "a". As part of this code
911        refactoring, the way recursive calls are handled has also been changed.
912    
913    25. If an assertion condition captured any substrings, they were not passed
914        back unless some other capturing happened later. For example, if
915        (?(?=(a))a) was matched against "a", no capturing was returned.
916    
917    26. When studying a pattern that contained subroutine calls or assertions,
918        the code for finding the minimum length of a possible match was handling
919        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
920        group 1 called group 2 while simultaneously a separate group 2 called group
921        1). A stack overflow occurred in this case. I have fixed this by limiting
922        the recursion depth to 10.
923    
924    27. Updated RunTest.bat in the distribution to the version supplied by Tom
925        Fortmann. This supports explicit test numbers on the command line, and has
926        argument validation and error reporting.
927    
928    28. An instance of \X with an unlimited repeat could fail if at any point the
929        first character it looked at was a mark character.
930    
931    29. Some minor code refactoring concerning Unicode properties and scripts
932        should reduce the stack requirement of match() slightly.
933    
934    30. Added the '=' option to pcretest to check the setting of unused capturing
935        slots at the end of the pattern, which are documented as being -1, but are
936        not included in the return count.
937    
938    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
939        compiled something random. Now it gives a compile-time error (as does
940        Perl).
941    
942    32. A *MARK encountered during the processing of a positive assertion is now
943        recorded and passed back (compatible with Perl).
944    
945    33. If --only-matching or --colour was set on a pcregrep call whose pattern
946        had alternative anchored branches, the search for a second match in a line
947        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
948        matched the line "0102" twice. The same bug affected patterns that started
949        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
950        twice.
951    
952    34. Previously, PCRE did not allow quantification of assertions. However, Perl
953        does, and because of capturing effects, quantifying parenthesized
954        assertions may at times be useful. Quantifiers are now allowed for
955        parenthesized assertions.
956    
957    35. A minor code tidy in pcre_compile() when checking options for \R usage.
958    
959    36. \g was being checked for fancy things in a character class, when it should
960        just be a literal "g".
961    
962    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
963        appearance of a nested POSIX class supersedes an apparent external class.
964        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
965        unescaped square brackets may also appear as part of class names. For
966        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
967        more like Perl. (But see 8.20/1 above.)
968    
969    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
970        was because it thought it was \N{name}, which is not supported).
971    
972    39. Add minix to OS list not supporting the -S option in pcretest.
973    
974    40. PCRE tries to detect cases of infinite recursion at compile time, but it
975        cannot analyze patterns in sufficient detail to catch mutual recursions
976        such as ((?1))((?2)). There is now a runtime test that gives an error if a
977        subgroup is called recursively as a subpattern for a second time at the
978        same position in the subject string. In previous releases this might have
979        been caught by the recursion limit, or it might have run out of stack.
980    
981    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
982        happen only once. PCRE was, however incorrectly giving a compile time error
983        "recursive call could loop indefinitely" because it cannot analyze the
984        pattern in sufficient detail. The compile time test no longer happens when
985        PCRE is compiling a conditional subpattern, but actual runaway loops are
986        now caught at runtime (see 40 above).
987    
988    42. It seems that Perl allows any characters other than a closing parenthesis
989        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
990        has been changed to be the same.
991    
992    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
993        as not to get warnings when autogen.sh is called. Also changed
994        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
995    
996    44. To help people who use pcregrep to scan files containing exceedingly long
997        lines, the following changes have been made:
998    
999        (a) The default value of the buffer size parameter has been increased from
1000            8K to 20K. (The actual buffer used is three times this size.)
1001    
1002        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1003            PCRE is built.
1004    
1005        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1006            to be set at run time.
1007    
1008        (d) Numerical values in pcregrep options can be followed by K or M, for
1009            example --buffer-size=50K.
1010    
1011        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1012            given and the return code is set to 2.
1013    
1014    45. Add a pointer to the latest mark to the callout data block.
1015    
1016    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1017        partial match of an empty string instead of no match. This was specific to
1018        the use of ".".
1019    
1020    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1021        complete match instead of a partial match. This bug was dependent on both
1022        the PCRE_UTF8 and PCRE_DOTALL options being set.
1023    
1024    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1025        starting byte set, because \b was not being ignored.
1026    
1027    
1028    Version 8.12 15-Jan-2011
1029    ------------------------
1030    
1031    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1032        checks for such things as part of the documentation building process.
1033    
1034    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1035        --match-limit and --recursion-limit options (added for 8.11). In
1036        particular, this made one of the standard tests fail. (The integer value
1037        went into the wrong half of a long int.)
1038    
1039    3.  If the --colour option was given to pcregrep with -v (invert match), it
1040        did strange things, either producing crazy output, or crashing. It should,
1041        of course, ignore a request for colour when reporting lines that do not
1042        match.
1043    
1044    4.  Another pcregrep bug caused similar problems if --colour was specified with
1045        -M (multiline) and the pattern match finished with a line ending.
1046    
1047    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1048        matched in multiline mode, the following line was shown as part of the
1049        match. This seems wrong, so I have changed it.
1050    
1051    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1052        the check for further matches in the same line (so they could be coloured)
1053        to overrun the end of the current line. If another match was found, it was
1054        incorrectly shown (and then shown again when found in the next line).
1055    
1056    7.  If pcregrep was compiled under Windows, there was a reference to the
1057        function pcregrep_exit() before it was defined. I am assuming this was
1058        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1059        reported by a user. I've moved the definition above the reference.
1060    
1061    
1062    Version 8.11 10-Dec-2010
1063    ------------------------
1064    
1065    1.  (*THEN) was not working properly if there were untried alternatives prior
1066        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1067        backtracked to try for "b" instead of moving to the next alternative branch
1068        at the same level (in this case, to look for "c"). The Perl documentation
1069        is clear that when (*THEN) is backtracked onto, it goes to the "next
1070        alternative in the innermost enclosing group".
1071    
1072    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1073        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1074        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1075        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1076        (*THEN).
1077    
1078    3.  If \s appeared in a character class, it removed the VT character from
1079        the class, even if it had been included by some previous item, for example
1080        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1081        of \s, but is part of the POSIX "space" class.)
1082    
1083    4.  A partial match never returns an empty string (because you can always
1084        match an empty string at the end of the subject); however the checking for
1085        an empty string was starting at the "start of match" point. This has been
1086        changed to the "earliest inspected character" point, because the returned
1087        data for a partial match starts at this character. This means that, for
1088        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1089        (previously it gave "no match").
1090    
1091    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1092        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1093        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1094        has an implication that the given string is incomplete (because a partial
1095        match is preferred over a full match). For this reason, these items now
1096        give a partial match in this situation. [Aside: previously, the one case
1097        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1098        match rather than a full match, which was wrong by the old rules, but is
1099        now correct.]
1100    
1101    6.  There was a bug in the handling of #-introduced comments, recognized when
1102        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1103        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1104        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1105        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1106        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1107        places in pcre_compile().
1108    
1109    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1110        comments when looking ahead for named forward references to subpatterns,
1111        the only newline sequence it recognized was NL. It now handles newlines
1112        according to the set newline convention.
1113    
1114    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1115        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1116        cater for a lack of strerror(). These oversights have been fixed.
1117    
1118    9.  Added --match-limit and --recursion-limit to pcregrep.
1119    
1120    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1121    
1122    11. When the -o option was used, pcregrep was setting a return code of 1, even
1123        when matches were found, and --line-buffered was not being honoured.
1124    
1125    12. Added an optional parentheses number to the -o and --only-matching options
1126        of pcregrep.
1127    
1128    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1129        can match an empty string. The code to do it in pcretest and pcredemo
1130        needed fixing:
1131    
1132        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1133            only one byte after an empty string match just before CRLF (this case
1134            just got forgotten; "any" and "anycrlf" were OK).
1135    
1136        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1137            mode when an empty string match preceded an ASCII character followed by
1138            a non-ASCII character. (The code for advancing by one character rather
1139            than one byte was nonsense.)
1140    
1141        (c) The pcredemo.c sample program did not have any code at all to handle
1142            the cases when CRLF is a valid newline sequence.
1143    
1144    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1145        as a starting offset was within the subject string. There is now a new
1146        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1147        negative or greater than the length of the string. In order to test this,
1148        pcretest is extended to allow the setting of negative starting offsets.
1149    
1150    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1151        starting offset points to the beginning of a UTF-8 character was
1152        unnecessarily clumsy. I tidied it up.
1153    
1154    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1155        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1156    
1157    17. Nobody had reported that the --include_dir option, which was added in
1158        release 7.7 should have been called --include-dir (hyphen, not underscore)
1159        for compatibility with GNU grep. I have changed it to --include-dir, but
1160        left --include_dir as an undocumented synonym, and the same for
1161        --exclude-dir, though that is not available in GNU grep, at least as of
1162        release 2.5.4.
1163    
1164    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1165        characters from a string of bytes) have been redefined so as not to use
1166        loops, in order to improve performance in some environments. At the same
1167        time, I abstracted some of the common code into auxiliary macros to save
1168        repetition (this should not affect the compiled code).
1169    
1170    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1171        compile-time error is now given if \c is not followed by an ASCII
1172        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1173        different, and any byte value is allowed.)
1174    
1175    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1176        START_OPTIMIZE option, which is now allowed at compile time - but just
1177        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1178        to pcregrep and other applications that have no direct access to PCRE
1179        options. The new /Y option in pcretest sets this option when calling
1180        pcre_compile().
1181    
1182    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1183        back references. Groups containing recursive back references were forced to
1184        be atomic by that change, but in the case of named groups, the amount of
1185        memory required was incorrectly computed, leading to "Failed: internal
1186        error: code overflow". This has been fixed.
1187    
1188    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1189        pcretest.c, to avoid build problems in some Borland environments.
1190    
1191    
1192    Version 8.10 25-Jun-2010
1193    ------------------------
1194    
1195    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1196        THEN.
1197    
1198    2.  (*ACCEPT) was not working when inside an atomic group.
1199    
1200    3.  Inside a character class, \B is treated as a literal by default, but
1201        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1202        causes the error). The code is unchanged, but I tidied the documentation.
1203    
1204    4.  Inside a character class, PCRE always treated \R and \X as literals,
1205        whereas Perl faults them if its -w option is set. I have changed PCRE so
1206        that it faults them when PCRE_EXTRA is set.
1207    
1208    5.  Added support for \N, which always matches any character other than
1209        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1210    
1211    6.  When compiling pcregrep with newer versions of gcc which may have
1212        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1213        declared with attribute warn_unused_result" were given. Just casting the
1214        result to (void) does not stop the warnings; a more elaborate fudge is
1215        needed. I've used a macro to implement this.
1216    
1217    7.  Minor change to pcretest.c to avoid a compiler warning.
1218    
1219    8.  Added four artifical Unicode properties to help with an option to make
1220        \s etc use properties (see next item). The new properties are: Xan
1221        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1222    
1223    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1224        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1225        this option. Modified pcretest to add /W to test this facility. Added
1226        REG_UCP to make it available via the POSIX interface.
1227    
1228    10. Added --line-buffered to pcregrep.
1229    
1230    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1231        studied, and the match started with a letter with a code point greater than
1232        127 whose first byte was different to the first byte of the other case of
1233        the letter, the other case of this starting letter was not recognized
1234        (#976).
1235    
1236    12. If a pattern that was studied started with a repeated Unicode property
1237        test, for example, \p{Nd}+, there was the theoretical possibility of
1238        setting up an incorrect bitmap of starting bytes, but fortunately it could
1239        not have actually happened in practice until change 8 above was made (it
1240        added property types that matched character-matching opcodes).
1241    
1242    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1243        possible starting bytes for non-anchored patterns.
1244    
1245    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1246        \R, and also a number of cases that involve Unicode properties, both
1247        explicit and implicit when PCRE_UCP is set.
1248    
1249    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1250        input, it could crash or give wrong results if characters with values
1251        greater than 0xc0 were present in the subject string. (Detail: it assumed
1252        UTF-8 input when processing these items.)
1253    
1254    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1255        size_t is 64-bit (#991).
1256    
1257    17. Added a check for running out of memory when PCRE is compiled with
1258        --disable-stack-for-recursion (#990).
1259    
1260    18. If the last data line in a file for pcretest does not have a newline on
1261        the end, a newline was missing in the output.
1262    
1263    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1264        less than 128) in its various bitmaps. However, there is a facility for
1265        generating tables according to the current locale when PCRE is compiled. It
1266        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1267        characters, are recognized by isspace() and therefore were getting set in
1268        these tables, and indeed these tables seem to approximate to ISO 8859. This
1269        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1270        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1271        which of course cannot start UTF-8 characters. I have changed the code so
1272        that only real ASCII characters (less than 128) and the correct starting
1273        bytes for UTF-8 encodings are set for characters greater than 127 when in
1274        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1275        altogether.)
1276    
1277    20. Added the /T option to pcretest so as to be able to run tests with non-
1278        standard character tables, thus making it possible to include the tests
1279        used for 19 above in the standard set of tests.
1280    
1281    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1282        reference to a subpattern the other side of a comment that contains an
1283        opening parenthesis caused either an internal compiling error, or a
1284        reference to the wrong subpattern.
1285    
1286    
1287    Version 8.02 19-Mar-2010
1288    ------------------------
1289    
1290    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1291    
1292    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1293        configured.
1294    
1295    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1296        original author of that file, following a query about its status.
1297    
1298    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1299        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1300    
1301    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1302        quantifier applied to a forward-referencing subroutine call, could compile
1303        incorrect code or give the error "internal error: previously-checked
1304        referenced subpattern not found".
1305    
1306    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1307        variables to point to external functions. For these systems, therefore,
1308        pcre_malloc etc. are now initialized to local functions that call the
1309        relevant global functions.
1310    
1311    7.  There were two entries missing in the vectors called coptable and poptable
1312        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1313        I've fixed the data, and added a kludgy way of testing at compile time that
1314        the lengths are correct (equal to the number of opcodes).
1315    
1316    8.  Following on from 7, I added a similar kludge to check the length of the
1317        eint vector in pcreposix.c.
1318    
1319    9.  Error texts for pcre_compile() are held as one long string to avoid too
1320        much relocation at load time. To find a text, the string is searched,
1321        counting zeros. There was no check for running off the end of the string,
1322        which could happen if a new error number was added without updating the
1323        string.
1324    
1325    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1326    
1327    11. \K was not working if it appeared in an atomic group or in a group that
1328        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1329        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1330        the assertion is positive, but not if it is negative.
1331    
1332    12. Change 11 fortuitously reduced the size of the stack frame used in the
1333        "match()" function of pcre_exec.c by one pointer. Forthcoming
1334        implementation of support for (*MARK) will need an extra pointer on the
1335        stack; I have reserved it now, so that the stack frame size does not
1336        decrease.
1337    
1338    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1339        item in branch that calls a recursion is a subroutine call - as in the
1340        second branch in the above example - was incorrectly given the compile-
1341        time error "recursive call could loop indefinitely" because pcre_compile()
1342        was not correctly checking the subroutine for matching a non-empty string.
1343    
1344    14. The checks for overrunning compiling workspace could trigger after an
1345        overrun had occurred. This is a "should never occur" error, but it can be
1346        triggered by pathological patterns such as hundreds of nested parentheses.
1347        The checks now trigger 100 bytes before the end of the workspace.
1348    
1349    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1350    
1351    
1352    Version 8.01 19-Jan-2010
1353    ------------------------
1354    
1355    1.  If a pattern contained a conditional subpattern with only one branch (in
1356        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1357        computed the wrong minimum data length (which is of course zero for such
1358        subpatterns). This could cause incorrect "no match" results.
1359    
1360    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1361        the pattern is reset in the first branch, pcre_compile() failed with
1362        "internal error: code overflow at offset...". This happened only when
1363        the reset was to the original external option setting. (An optimization
1364        abstracts leading options settings into an external setting, which was the
1365        cause of this.)
1366    
1367    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1368        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1369        assertion pattern did not match (meaning that the assertion was true), it
1370        was incorrectly treated as false if the SKIP had been reached during the
1371        matching. This also applied to assertions used as conditions.
1372    
1373    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1374        assertion subpattern, including such a pattern used as a condition,
1375        unpredictable results occurred, instead of the error return
1376        PCRE_ERROR_DFA_UITEM.
1377    
1378    5.  The C++ GlobalReplace function was not working like Perl for the special
1379        situation when an empty string is matched. It now does the fancy magic
1380        stuff that is necessary.
1381    
1382    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1383        removed. (These were left over from very, very early versions of PCRE.)
1384    
1385    7.  Some cosmetic changes to the code to make life easier when compiling it
1386        as part of something else:
1387    
1388        (a) Change DEBUG to PCRE_DEBUG.
1389    
1390        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1391            called "current" as "current_branch", to prevent a collision with the
1392            Linux macro when compiled as a kernel module.
1393    
1394        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1395            prevent a collision with the Linux macro when compiled as a kernel
1396            module.
1397    
1398    8.  In pcre_compile() there are some checks for integer overflows that used to
1399        cast potentially large values to (double). This has been changed to that
1400        when building, a check for int64_t is made, and if it is found, it is used
1401        instead, thus avoiding the use of floating point arithmetic. (There is no
1402        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1403        double.
1404    
1405    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1406        2005 (difference between two addresses compared to an unsigned value).
1407    
1408    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1409        custom one, because of the following reported problem in Windows:
1410    
1411          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1412              under Win32.
1413          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1414              therefore missing the function definition.
1415          - The compiler thus generates a "C" signature for the test function.
1416          - The linker fails to find the "C" function.
1417          - PCRE fails to configure if asked to do so against libbz2.
1418    
1419    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1420        messages were output:
1421    
1422          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1423          rerunning libtoolize, to keep the correct libtool macros in-tree.
1424          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1425    
1426        I have done both of these things.
1427    
1428    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1429        most of the time, it *can* run out if it is given a pattern that contains a
1430        runaway infinite recursion. I updated the discussion in the pcrestack man
1431        page.
1432    
1433    13. Now that we have gone to the x.xx style of version numbers, the minor
1434        version may start with zero. Using 08 or 09 is a bad idea because users
1435        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1436        interpreted as invalid octal numbers. I've updated the previous comment in
1437        configure.ac, and also added a check that gives an error if 08 or 09 are
1438        used.
1439    
1440    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1441        causing partial matching to fail when the end of the subject matched \W
1442        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1443    
1444    15. There were some discrepancies between the declarations in pcre_internal.h
1445        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1446        their definitions. The declarations used "const uschar *" and the
1447        definitions used USPTR. Even though USPTR is normally defined as "const
1448        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1449        reported that: "This difference in casting confuses some C++ compilers, for
1450        example, SunCC recognizes above declarations as different functions and
1451        generates broken code for hbpcre." I have changed the declarations to use
1452        USPTR.
1453    
1454    16. GNU libtool is named differently on some systems. The autogen.sh script now
1455        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1456        (FreeBSD).
1457    
1458    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1459        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1460        comment: "Figure out how to create a longlong from a string: strtoll and
1461        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1462        instance, but it only takes 2 args instead of 3!"
1463    
1464    18. A subtle bug concerned with back references has been fixed by a change of
1465        specification, with a corresponding code fix. A pattern such as
1466        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1467        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1468        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1469        same bug. Such groups have to be quantified to be useful, or contained
1470        inside another quantified group. (If there's no repetition, the reference
1471        can never match.) The problem arises because, having left the group and
1472        moved on to the rest of the pattern, a later failure that backtracks into
1473        the group uses the captured value from the final iteration of the group
1474        rather than the correct earlier one. I have fixed this in PCRE by forcing
1475        any group that contains a reference to itself to be an atomic group; that
1476        is, there cannot be any backtracking into it once it has completed. This is
1477        similar to recursive and subroutine calls.
1478    
1479    
1480    Version 8.00 19-Oct-09
1481    ----------------------
1482    
1483    1.  The table for translating pcre_compile() error codes into POSIX error codes
1484        was out-of-date, and there was no check on the pcre_compile() error code
1485        being within the table. This could lead to an OK return being given in
1486        error.
1487    
1488    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1489        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1490        in a Windows environment.
1491    
1492    3.  The pcregrep --count option prints the count for each file even when it is
1493        zero, as does GNU grep. However, pcregrep was also printing all files when
1494        --files-with-matches was added. Now, when both options are given, it prints
1495        counts only for those files that have at least one match. (GNU grep just
1496        prints the file name in this circumstance, but including the count seems
1497        more useful - otherwise, why use --count?) Also ensured that the
1498        combination -clh just lists non-zero counts, with no names.
1499    
1500    4.  The long form of the pcregrep -F option was incorrectly implemented as
1501        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1502        but it seems right to fix it, and I didn't think it was worth preserving
1503        the old behaviour.
1504    
1505    5.  The command line items --regex=pattern and --regexp=pattern were not
1506        recognized by pcregrep, which required --regex pattern or --regexp pattern
1507        (with a space rather than an '='). The man page documented the '=' forms,
1508        which are compatible with GNU grep; these now work.
1509    
1510    6.  No libpcreposix.pc file was created for pkg-config; there was just
1511        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1512    
1513    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1514        when UCP support is not needed, by modifying the Python script that
1515        generates it from Unicode data files. This should not matter if the module
1516        is correctly used as a library, but I received one complaint about 50K of
1517        unwanted data. My guess is that the person linked everything into his
1518        program rather than using a library. Anyway, it does no harm.
1519    
1520    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1521        was a minimum greater than 1 for a wide character in a possessive
1522        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1523        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1524        character. Chaos in the form of incorrect output or a compiling loop could
1525        result.
1526    
1527    9.  The restrictions on what a pattern can contain when partial matching is
1528        requested for pcre_exec() have been removed. All patterns can now be
1529        partially matched by this function. In addition, if there are at least two
1530        slots in the offset vector, the offset of the earliest inspected character
1531        for the match and the offset of the end of the subject are set in them when
1532        PCRE_ERROR_PARTIAL is returned.
1533    
1534    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1535        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1536        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1537        and may be more useful for multi-segment matching.
1538    
1539    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1540        used to be given if ever the end of the subject was reached; now it is
1541        given only if matching could not proceed because another character was
1542        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1543        string "Z", which now yields "no match" instead of "partial match". In the
1544        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1545        final character ended with (*FAIL).
1546    
1547    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1548        if the pattern had a "must contain" character that was already found in the
1549        earlier partial match, unless partial matching was again requested. For
1550        example, with the pattern /dog.(body)?/, the "must contain" character is
1551        "g". If the first part-match was for the string "dog", restarting with
1552        "sbody" failed. This bug has been fixed.
1553    
1554    13. The string returned by pcre_dfa_exec() after a partial match has been
1555        changed so that it starts at the first inspected character rather than the
1556        first character of the match. This makes a difference only if the pattern
1557        starts with a lookbehind assertion or \b or \B (\K is not supported by
1558        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1559        matching functions compatible, and I think it's the right thing to do.
1560    
1561    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1562        so that the demonstration program is easily available in environments where
1563        PCRE has not been installed from source.
1564    
1565    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1566        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1567        library.
1568    
1569    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1570        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1571        is not the first non-POSIX option to be added. Clearly some people find
1572        these options useful.
1573    
1574    17. If a caller to the POSIX matching function regexec() passes a non-zero
1575        value for nmatch with a NULL value for pmatch, the value of
1576        nmatch is forced to zero.
1577    
1578    18. RunGrepTest did not have a test for the availability of the -u option of
1579        the diff command, as RunTest does. It now checks in the same way as
1580        RunTest, and also checks for the -b option.
1581    
1582    19. If an odd number of negated classes containing just a single character
1583        interposed, within parentheses, between a forward reference to a named
1584        subpattern and the definition of the subpattern, compilation crashed with
1585        an internal error, complaining that it could not find the referenced
1586        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1587        [The bug was that it was starting one character too far in when skipping
1588        over the character class, thus treating the ] as data rather than
1589        terminating the class. This meant it could skip too much.]
1590    
1591    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1592        /g option in pcretest when the pattern contains \K, which makes it possible
1593        to have an empty string match not at the start, even when the pattern is
1594        anchored. Updated pcretest and pcredemo to use this option.
1595    
1596    21. If the maximum number of capturing subpatterns in a recursion was greater
1597        than the maximum at the outer level, the higher number was returned, but
1598        with unset values at the outer level. The correct (outer level) value is
1599        now given.
1600    
1601    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1602        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1603        make it do so. The string so far is captured, making this feature
1604        compatible with Perl.
1605    
1606    23. The tests have been re-organized, adding tests 11 and 12, to make it
1607        possible to check the Perl 5.10 features against Perl 5.10.
1608    
1609    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1610        pattern matches a fixed length string. PCRE did not allow this; now it
1611        does. Neither allows recursion.
1612    
1613    25. I finally figured out how to implement a request to provide the minimum
1614        length of subject string that was needed in order to match a given pattern.
1615        (It was back references and recursion that I had previously got hung up
1616        on.) This code has now been added to pcre_study(); it finds a lower bound
1617        to the length of subject needed. It is not necessarily the greatest lower
1618        bound, but using it to avoid searching strings that are too short does give
1619        some useful speed-ups. The value is available to calling programs via
1620        pcre_fullinfo().
1621    
1622    26. While implementing 25, I discovered to my embarrassment that pcretest had
1623        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1624        study optimizations had never been tested with that matching function.
1625        Oops. What is worse, even when it was passed study data, there was a bug in
1626        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1627        were also very few tests of studied patterns with pcre_dfa_exec().
1628    
1629    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1630        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1631        on the other side of the coin, they are no longer allowed to have different
1632        names, because these cannot be distinguished in PCRE, and this has caused
1633        confusion. (This is a difference from Perl.)
1634    
1635    28. When duplicate subpattern names are present (necessarily with different
1636        numbers, as required by 27 above), and a test is made by name in a
1637        conditional pattern, either for a subpattern having been matched, or for
1638        recursion in such a pattern, all the associated numbered subpatterns are
1639        tested, and the overall condition is true if the condition is true for any
1640        one of them. This is the way Perl works, and is also more like the way
1641        testing by number works.
1642    
1643    
1644    Version 7.9 11-Apr-09
1645    ---------------------
1646    
1647    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1648        (pcretest), all targets were linked against these libraries. This included
1649        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1650        libraries. This caused unwanted dependencies to be created. This problem
1651        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1652        pcretest is linked with readline.
1653    
1654    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1655        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1656        moved outside it again, because FALSE and TRUE are already defined in AIX,
1657        but BOOL is not.
1658    
1659    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1660        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1661    
1662    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1663        hyphens) following file names and line numbers when outputting matching
1664        lines. This is not true; no spaces are inserted. I have also clarified the
1665        wording for the --colour (or --color) option.
1666    
1667    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1668        was not coloured; this is different to GNU grep, so I have changed it to be
1669        the same.
1670    
1671    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1672        each matching line was coloured. Now it goes on to look for further matches
1673        of any of the test patterns, which is the same behaviour as GNU grep.
1674    
1675    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1676        doesn't make sense to accept an empty string match in pcregrep, so I have
1677        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1678        seems to be how GNU grep behaves. [But see later change 40 for release
1679        8.33.]
1680    
1681    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1682        start or after a newline", because the conditional assertion was not being
1683        correctly handled. The rule now is that both the assertion and what follows
1684        in the first alternative must satisfy the test.
1685    
1686    9.  If auto-callout was enabled in a pattern with a conditional group whose
1687        condition was an assertion, PCRE could crash during matching, both with
1688        pcre_exec() and pcre_dfa_exec().
1689    
1690    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1691        used for matching.
1692    
1693    11. Unicode property support in character classes was not working for
1694        characters (bytes) greater than 127 when not in UTF-8 mode.
1695    
1696    12. Added the -M command line option to pcretest.
1697    
1698    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1699    
1700    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1701    
1702    16. Added comments and documentation about mis-use of no_arg in the C++
1703        wrapper.
1704    
1705    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1706        from Martin Jerabek that uses macro names for all relevant character and
1707        string constants.
1708    
1709    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1710        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1711        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1712        these, but not everybody uses configure.
1713    
1714    19. A conditional group that had only one branch was not being correctly
1715        recognized as an item that could match an empty string. This meant that an
1716        enclosing group might also not be so recognized, causing infinite looping
1717        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1718        with the subject "ab", where knowledge that the repeated group can match
1719        nothing is needed in order to break the loop.
1720    
1721    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1722        exec(), but without supplying a callout function, matching went wrong.
1723    
1724    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1725        leak if the size of the offset vector was greater than 30. When the vector
1726        is smaller, the saved offsets during recursion go onto a local stack
1727        vector, but for larger vectors malloc() is used. It was failing to free
1728        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1729        error, in fact).
1730    
1731    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1732        heapframe that is used only when UTF-8 support is enabled. This caused no
1733        problem, but was untidy.
1734    
1735    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1736        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1737        included within another project.
1738    
1739    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1740        slightly modified by me:
1741    
1742          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1743              not building pcregrep.
1744    
1745          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1746              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1747    
1748    25. Forward references, both numeric and by name, in patterns that made use of
1749        duplicate group numbers, could behave incorrectly or give incorrect errors,
1750        because when scanning forward to find the reference group, PCRE was not
1751        taking into account the duplicate group numbers. A pattern such as
1752        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1753    
1754    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1755        the feature of a custom pointer more persuasive (as requested by a user).
1756    
1757    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1758        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1759        pre-defined.
1760    
1761    28. Added support for (*UTF8) at the start of a pattern.
1762    
1763    29. Arrange for flags added by the "release type" setting in CMake to be shown
1764        in the configuration summary.
1765    
1766    
1767    Version 7.8 05-Sep-08
1768    ---------------------
1769    
1770    1.  Replaced UCP searching code with optimized version as implemented for Ad
1771        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1772        stage table and inline lookup instead of a function, giving speed ups of 2
1773        to 5 times on some simple patterns that I tested. Permission was given to
1774        distribute the MultiStage2.py script that generates the tables (it's not in
1775        the tarball, but is in the Subversion repository).
1776    
1777    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1778        scripts.
1779    
1780    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1781        a group with a zero qualifier. The result of the study could be incorrect,
1782        or the function might crash, depending on the pattern.
1783    
1784    4.  Caseless matching was not working for non-ASCII characters in back
1785        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1786        It now works when Unicode Property Support is available.
1787    
1788    5.  In pcretest, an escape such as \x{de} in the data was always generating
1789        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1790        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1791        truncation.
1792    
1793    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1794    
1795    7.  Added two (int) casts to pcregrep when printing the difference of two
1796        pointers, in case they are 64-bit values.
1797    
1798    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1799        test 2 if it fails.
1800    
1801    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1802        and a #define of that name to empty if it is not externally set. This is to
1803        allow users of MSVC to set it if necessary.
1804    
1805    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1806        the convenience functions in the pcre_get.c source file.
1807    
1808    11. An option change at the start of a pattern that had top-level alternatives
1809        could cause overwriting and/or a crash. This command provoked a crash in
1810        some environments:
1811    
1812          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1813    
1814        This potential security problem was recorded as CVE-2008-2371.
1815    
1816    12. For a pattern where the match had to start at the beginning or immediately
1817        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1818        pcre_dfa_exec() could read past the end of the passed subject if there was
1819        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1820        pcretest so that it places the subject at the end of its malloc-ed buffer.
1821    
1822    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1823        exec() might read past the end of the data buffer in UTF-8 mode.
1824    
1825    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1826        the data contained the byte 0x85 as part of a UTF-8 character within its
1827        first line. This applied both to normal and DFA matching.
1828    
1829    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1830        /^[^d]*?$/8 failed to match "abc".
1831    
1832    16. Added a missing copyright notice to pcrecpp_internal.h.
1833    
1834    17. Make it more clear in the documentation that values returned from
1835        pcre_exec() in ovector are byte offsets, not character counts.
1836    
1837    18. Tidied a few places to stop certain compilers from issuing warnings.
1838    
1839    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1840        supplied by Stefan Weber. I made a further small update for 7.8 because
1841        there is a change of source arrangements: the pcre_searchfuncs.c module is
1842        replaced by pcre_ucd.c.
1843    
1844    
1845    Version 7.7 07-May-08
1846    ---------------------
1847    
1848    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1849        a string to a long long, pretend we don't even have a long long." This is
1850        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1851    
1852    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1853        pre-7.6 versions, which defined a global no_arg variable instead of putting
1854        it in the RE class. (See also #8 below.)
1855    
1856    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1857        Lopes.
1858    
1859    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1860    
1861        (1) The include/exclude patterns were being applied to the whole pathnames
1862            of files, instead of just to the final components.
1863    
1864        (2) If there was more than one level of directory, the subdirectories were
1865            skipped unless they satisfied the include/exclude conditions. This is
1866            inconsistent with GNU grep (and could even be seen as contrary to the
1867            pcregrep specification - which I improved to make it absolutely clear).
1868            The action now is always to scan all levels of directory, and just
1869            apply the include/exclude patterns to regular files.
1870    
1871    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1872        --exclude_dir in the tests to avoid scanning .svn directories.
1873    
1874    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1875        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1876        doesn't support NULs in patterns.
1877    
1878    7.  Added some missing "const"s to declarations of static tables in
1879        pcre_compile.c and pcre_dfa_exec.c.
1880    
1881    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1882        caused by fix #2  above. (Subsequently also a second patch to fix the
1883        first patch. And a third patch - this was a messy problem.)
1884    
1885    9.  Applied Craig's patch to remove the use of push_back().
1886    
1887    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1888        matching function regexec().
1889    
1890    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1891        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1892        references. PCRE supports relative numbers with this syntax (I don't think
1893        Oniguruma does).
1894    
1895    12. Previously, a group with a zero repeat such as (...){0} was completely
1896        omitted from the compiled regex. However, this means that if the group
1897        was called as a subroutine from elsewhere in the pattern, things went wrong
1898        (an internal error was given). Such groups are now left in the compiled
1899        pattern, with a new opcode that causes them to be skipped at execution
1900        time.
1901    
1902    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1903        to the way PCRE behaves:
1904    
1905        (a) A lone ] character is dis-allowed (Perl treats it as data).
1906    
1907        (b) A back reference to an unmatched subpattern matches an empty string
1908            (Perl fails the current match path).
1909    
1910        (c) A data ] in a character class must be notated as \] because if the
1911            first data character in a class is ], it defines an empty class. (In
1912            Perl it is not possible to have an empty class.) The empty class []
1913            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1914            The negative empty class [^] matches any one character, independently
1915            of the DOTALL setting.
1916    
1917    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1918        non-existent subpattern following a character class starting with ']' and
1919        containing () gave an internal compiling error instead of "reference to
1920        non-existent subpattern". Fortunately, when the pattern did exist, the
1921        compiled code was correct. (When scanning forwards to check for the
1922        existence of the subpattern, it was treating the data ']' as terminating
1923        the class, so got the count wrong. When actually compiling, the reference
1924        was subsequently set up correctly.)
1925    
1926    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1927        it was being rejected as not supported by pcre_dfa_exec(), even though
1928        other assertions are supported. I have made pcre_dfa_exec() support
1929        (*FAIL).
1930    
1931    16. The implementation of 13c above involved the invention of a new opcode,
1932        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1933        cannot be changed at match time, I realized I could make a small
1934        improvement to matching performance by compiling OP_ALLANY instead of
1935        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1936        on the OP_ANY path.
1937    
1938    17. Compiling pcretest on Windows with readline support failed without the
1939        following two fixes: (1) Make the unistd.h include conditional on
1940        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1941    
1942    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1943        ncurses library to be included for pcretest when ReadLine support is
1944        requested, but also to allow for it to be overridden. This patch came from
1945        Daniel Bergström.
1946    
1947    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1948        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1949        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1950        spotting this.
1951    
1952    
1953    Version 7.6 28-Jan-08
1954    ---------------------
1955    
1956    1.  A character class containing a very large number of characters with
1957        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1958        overflow.
1959    
1960    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1961        HAVE_LONG_LONG is not defined.
1962    
1963    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1964        bring them up to date and include new features. This patch includes:
1965    
1966        - Fixed PH's badly added libz and libbz2 support.
1967        - Fixed a problem with static linking.
1968        - Added pcredemo. [But later removed - see 7 below.]
1969        - Fixed dftables problem and added an option.
1970        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1971            HAVE_LONG_LONG.
1972        - Added readline support for pcretest.
1973        - Added an listing of the option settings after cmake has run.
1974    
1975    4.  A user submitted a patch to Makefile that makes it easy to create
1976        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1977        Makefile.am that cause it to include this special target, without
1978        affecting anything else. Note that the same mingw target plus all
1979        the other distribution libraries and programs are now supported
1980        when configuring with CMake (see 6 below) instead of with
1981        Configure/Make.
1982    
1983    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1984        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1985        exported in the Windows port". It has not yet been confirmed that the patch
1986        solves the problem, but it does no harm.
1987    
1988    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1989        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1990        with CMake, and also correct the comment about stack recursion.
1991    
1992    7.  Remove the automatic building of pcredemo from the ./configure system and
1993        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1994        of a program that users should build themselves after PCRE is installed, so
1995        building it automatically is not really right. What is more, it gave
1996        trouble in some build environments.
1997    
1998    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1999    
2000    
2001    Version 7.5 10-Jan-08
2002    ---------------------
2003    
2004    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2005        values in parens when parsing an RE using the C++ wrapper."
2006    
2007    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2008        Characters greater than 255 were excluded from the class instead of being
2009        included.
2010    
2011    3.  The same bug as (2) above applied to negated POSIX classes such as
2012        [:^space:].
2013    
2014    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2015        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2016        I have changed it.
2017    
2018    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2019        first named subpattern) and a construct such as (?&a) would reference the
2020        first named subpattern whose name started with "a" (in other words, the
2021        length check was missing). Both these problems are fixed. "Subpattern name
2022        expected" is now given for (?&) (a zero-length name), and this patch also
2023        makes it give the same error for \k'' (previously it complained that that
2024        was a reference to a non-existent subpattern).
2025    
2026    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2027        this is right because (?- can be followed by option settings as well as by
2028        digits. I have, however, made the messages clearer.
2029    
2030    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2031        than the number used in the conditional) now cause a compile-time error.
2032        This is actually not compatible with Perl, which accepts such patterns, but
2033        treats the conditional as always being FALSE (as PCRE used to), but it
2034        seems to me that giving a diagnostic is better.
2035    
2036    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2037        and messages.
2038    
2039    9.  Fix two occurrences of "backslash" in comments that should have been
2040        "backspace".
2041    
2042    10. Remove two redundant lines of code that can never be obeyed (their function
2043        was moved elsewhere).
2044    
2045    11. The program that makes PCRE's Unicode character property table had a bug
2046        which caused it to generate incorrect table entries for sequences of
2047        characters that have the same character type, but are in different scripts.
2048        It amalgamated them into a single range, with the script of the first of
2049        them. In other words, some characters were in the wrong script. There were
2050        thirteen such cases, affecting characters in the following ranges:
2051    
2052          U+002b0 - U+002c1
2053          U+0060c - U+0060d
2054          U+0061e - U+00612
2055          U+0064b - U+0065e
2056          U+0074d - U+0076d
2057          U+01800 - U+01805
2058          U+01d00 - U+01d77
2059          U+01d9b - U+01dbf
2060          U+0200b - U+0200f
2061          U+030fc - U+030fe
2062          U+03260 - U+0327f
2063          U+0fb46 - U+0fbb1
2064          U+10450 - U+1049d
2065    
2066    12. The -o option (show only the matching part of a line) for pcregrep was not
2067        compatible with GNU grep in that, if there was more than one match in a
2068        line, it showed only the first of them. It now behaves in the same way as
2069        GNU grep.
2070    
2071    13. If the -o and -v options were combined for pcregrep, it printed a blank
2072        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2073        does the same. The return code can be used to tell if there were any
2074        non-matching lines.
2075    
2076    14. Added --file-offsets and --line-offsets to pcregrep.
2077    
2078    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2079        infinitely looping recursion. The bug was that positive lookaheads were not
2080        being skipped when checking for a possible empty match (negative lookaheads
2081        and both kinds of lookbehind were skipped).
2082    
2083    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2084        inclusion of <windows.h> to before rather than after the definition of
2085        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2086    
2087    17. Specifying a possessive quantifier with a specific limit for a Unicode
2088        character property caused pcre_compile() to compile bad code, which led at
2089        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2090        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2091        caused the error; without that there was no problem.
2092    
2093    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2094    
2095    19. Added --enable-pcretest-libreadline.
2096    
2097    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2098        RE::GlobalReplace(). As a result, the number of replacements returned was
2099        double what it should be. I removed one of the increments, but Craig sent a
2100        later patch that removed the other one (the right fix) and added unit tests
2101        that check the return values (which was not done before).
2102    
2103    21. Several CMake things:
2104    
2105        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2106            the names libpcre and libpcreposix, not just pcre and pcreposix.
2107    
2108        (2) The above change means that pcretest and pcregrep are now correctly
2109            linked with the newly-built libraries, not previously installed ones.
2110    
2111        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2112    
2113    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2114        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2115        UTF-8 newline character). The key issue is that the pattern starts .*;
2116        this means that the match must be either at the beginning, or after a
2117        newline. The bug was in the code for advancing after a failed match and
2118        checking that the new position followed a newline. It was not taking
2119        account of UTF-8 characters correctly.
2120    
2121    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2122        character classes. PCRE was not treating the sequence [:...:] as a
2123        character class unless the ... were all letters. Perl, however, seems to
2124        allow any characters between [: and :], though of course it rejects as
2125        unknown any "names" that contain non-letters, because all the known class
2126        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2127        for example, whereas PCRE did not - it did not recognize a POSIX character
2128        class. This seemed a bit dangerous, so the code has been changed to be
2129        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2130        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2131        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2132        Perl does, and where it didn't before.
2133    
2134    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2135        Windows environments %n is disabled by default.
2136    
2137    
2138    Version 7.4 21-Sep-07
2139    ---------------------
2140    
2141    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2142        means that a class such as [\s] counted as "explicit reference to CR or
2143        LF". That isn't really right - the whole point of the change was to try to
2144        help when there was an actual mention of one of the two characters. So now
2145        the change happens only if \r or \n (or a literal CR or LF) character is
2146        encountered.
2147    
2148    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2149        of both had grown to the point where there were only 3 bits left.
2150        Fortunately, there was spare space in the data structure, and so I have
2151        moved the internal flags into a new 16-bit field to free up more option
2152        bits.
2153    
2154    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2155        but did not set the internal JCHANGED flag - either of these is enough to
2156        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2157        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2158        start sets both bits.
2159    
2160    4.  Added options (at build time, compile time, exec time) to change \R from
2161        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2162    
2163    5.  doc/pcresyntax.html was missing from the distribution.
2164    
2165    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2166        compatibility, even though it is no longer used.
2167    
2168    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2169        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2170        windows.h file is present (where different names are used). [This was
2171        reversed later after testing - see 16 below.]
2172    
2173    8.  Changed all #include <config.h> to #include "config.h". There were also
2174        some further <pcre.h> cases that I changed to "pcre.h".
2175    
2176    9.  When pcregrep was used with the --colour option, it missed the line ending
2177        sequence off the lines that it output.
2178    
2179    10. It was pointed out to me that arrays of string pointers cause lots of
2180        relocations when a shared library is dynamically loaded. A technique of
2181        using a single long string with a table of offsets can drastically reduce
2182        these. I have refactored PCRE in four places to do this. The result is
2183        dramatic:
2184    
2185          Originally:                          290
2186          After changing UCP table:            187
2187          After changing error message table:   43
2188          After changing table of "verbs"       36
2189          After changing table of Posix names   22
2190    
2191        Thanks to the folks working on Gregex for glib for this insight.
2192    
2193    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2194        unicode-properties was also set.
2195    
2196    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2197    
2198    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2199        checked only for CRLF.
2200    
2201    14. Added casts to pcretest.c to avoid compiler warnings.
2202    
2203    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2204    
2205    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2206        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2207        entirely. This removes changes made in 7 above.
2208    
2209    17. The CMake files have been updated, and there is now more information about
2210        building with CMake in the NON-UNIX-USE document.
2211    
2212    
2213    Version 7.3 28-Aug-07
2214  ---------------------  ---------------------
2215    
2216   1. In the rejigging of the build system that eventually resulted in 7.1, the   1. In the rejigging of the build system that eventually resulted in 7.1, the
# Line 9  Version 7.3 05-Jul-07 Line 2218  Version 7.3 05-Jul-07
2218      brackets there is not right, since it causes compilers to look for an      brackets there is not right, since it causes compilers to look for an
2219      installed pcre.h, not the version that is in the source that is being      installed pcre.h, not the version that is in the source that is being
2220      compiled (which of course may be different). I have changed it back to:      compiled (which of course may be different). I have changed it back to:
2221    
2222        #include "pcre.h"        #include "pcre.h"
2223    
2224      I have a vague recollection that the change was concerned with compiling in      I have a vague recollection that the change was concerned with compiling in
2225      different directories, but in the new build system, that is taken care of      different directories, but in the new build system, that is taken care of
2226      by the VPATH setting the Makefile.      by the VPATH setting the Makefile.
2227    
2228   2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed   2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2229      when the subject happened to end in the byte 0x85 (e.g. if the last      when the subject happened to end in the byte 0x85 (e.g. if the last
2230      character was \x{1ec5}). *Character* 0x85 is one of the "any" newline      character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
# Line 23  Version 7.3 05-Jul-07 Line 2232  Version 7.3 05-Jul-07
2232      of another character. The bug was that, for an unlimited repeat of . in      of another character. The bug was that, for an unlimited repeat of . in
2233      not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by      not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2234      characters when looking for a newline.      characters when looking for a newline.
   
  3. A small performance improvement in the DOTALL UTF-8 mode .* case.  
   
  4. Debugging: adjusted the names of opcodes for different kinds of parentheses  
     in debug output.  
   
  5. Arrange to use "%I64d" instead of "%lld" for long printing in the pcrecpp  
     unittest when running under MinGW.  
2235    
2236     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2237    
2238     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2239        in debug output.
2240    
2241     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2242        long printing in the pcrecpp unittest when running under MinGW.
2243    
2244     6. ESC_K was left out of the EBCDIC table.
2245    
2246     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2247        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2248        limit also applies to "virtual nesting" when a pattern is recursive, and in
2249        this case 1000 isn't so big. I have been able to remove this limit at the
2250        expense of backing off one optimization in certain circumstances. Normally,
2251        when pcre_exec() would call its internal match() function recursively and
2252        immediately return the result unconditionally, it uses a "tail recursion"
2253        feature to save stack. However, when a subpattern that can match an empty
2254        string has an unlimited repetition quantifier, it no longer makes this
2255        optimization. That gives it a stack frame in which to save the data for
2256        checking that an empty string has been matched. Previously this was taken
2257        from the 1000-entry workspace that had been reserved. So now there is no
2258        explicit limit, but more stack is used.
2259    
2260     8. Applied Daniel's patches to solve problems with the import/export magic
2261        syntax that is required for Windows, and which was going wrong for the
2262        pcreposix and pcrecpp parts of the library. These were overlooked when this
2263        problem was solved for the main library.
2264    
2265     9. There were some crude static tests to avoid integer overflow when computing
2266        the size of patterns that contain repeated groups with explicit upper
2267        limits. As the maximum quantifier is 65535, the maximum group length was
2268        set at 30,000 so that the product of these two numbers did not overflow a
2269        32-bit integer. However, it turns out that people want to use groups that
2270        are longer than 30,000 bytes (though not repeat them that many times).
2271        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2272        made it possible to implement the integer overflow checks in a much more
2273        dynamic way, which I have now done. The artificial limitation on group
2274        length has been removed - we now have only the limit on the total length of
2275        the compiled pattern, which depends on the LINK_SIZE setting.
2276    
2277    10. Fixed a bug in the documentation for get/copy named substring when
2278        duplicate names are permitted. If none of the named substrings are set, the
2279        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2280        empty string.
2281    
2282    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2283        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2284        because the ] is interpreted as the first data character and the
2285        terminating ] is not found. PCRE has been made compatible with Perl in this
2286        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2287        cause memory overwriting.
2288    
2289    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2290        string has been matched (to stop an infinite loop). It was not recognizing
2291        a conditional subpattern that could match an empty string if that
2292        subpattern was within another subpattern. For example, it looped when
2293        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2294        condition was not nested. This bug has been fixed.
2295    
2296    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2297        past the start of the subject in the presence of bytes with the top bit
2298        set, for example "\x8aBCD".
2299    
2300    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2301        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2302    
2303    14. Optimized (?!) to (*FAIL).
2304    
2305    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2306        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2307        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2308        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2309        does: it's just the validity check that is more restrictive.
2310    
2311    16. Inserted checks for integer overflows during escape sequence (backslash)
2312        processing, and also fixed erroneous offset values for syntax errors during
2313        backslash processing.
2314    
2315    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2316        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2317    
2318    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2319        caused an overrun.
2320    
2321    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2322        something other than just ASCII characters) inside a group that had an
2323        unlimited repeat caused a loop at compile time (while checking to see
2324        whether the group could match an empty string).
2325    
2326    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2327        [\P{Any}] did so. (Error in the code for printing property names.)
2328    
2329    21. An orphan \E inside a character class could cause a crash.
2330    
2331    22. A repeated capturing bracket such as (A)? could cause a wild memory
2332        reference during compilation.
2333    
2334    23. There are several functions in pcre_compile() that scan along a compiled
2335        expression for various reasons (e.g. to see if it's fixed length for look
2336        behind). There were bugs in these functions when a repeated \p or \P was
2337        present in the pattern. These operators have additional parameters compared
2338        with \d, etc, and these were not being taken into account when moving along
2339        the compiled data. Specifically:
2340    
2341        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2342            length.
2343    
2344        (b) An item such as \pL+ within a repeated group could cause crashes or
2345            loops.
2346    
2347        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2348            "reference to non-existent subpattern" error.
2349    
2350        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2351    
2352    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2353        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2354    
2355    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2356    
2357    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2358        character were causing crashes (broken optimization).
2359    
2360    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2361        \p or \P) caused a compile-time loop.
2362    
2363    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2364        break. For example, the unstudied pattern [\r\n]A does not match the string
2365        "\r\nA" because change 7.0/46 below moves the current point on by two
2366        characters after failing to match at the start. However, the pattern \nA
2367        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2368        the same is true. There doesn't seem any very clean way out of this, but
2369        what I have chosen to do makes the common cases work: PCRE now takes note
2370        of whether there can be an explicit match for \r or \n anywhere in the
2371        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2372        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2373        pattern has explicit CR or LF references.
2374    
2375    29. Added (*CR) etc for changing newline setting at start of pattern.
2376    
2377    
2378  Version 7.2 19-Jun-07  Version 7.2 19-Jun-07

Legend:
Removed from v.193  
changed lines
  Added in v.1335

  ViewVC Help
Powered by ViewVC 1.1.5