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

Diff of /code/trunk/ChangeLog

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

revision 98 by ph10, Mon Mar 5 15:55:02 2007 UTC revision 1361 by ph10, Fri Sep 6 17:47:32 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 05-Mar-07  Version 8.34 xx-xxxx-201x
5    --------------------------
6    
7    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
8        executable memory. Patch inspired by Carsten Klein.
9    
10    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
11        this macro is never tested and has no effect, because the work to support
12        coverage involves only compiling and linking options and special targets in
13        the Makefile. The comment in config.h implied that defining the macro would
14        enable coverage support, which is totally false. There was also support for
15        setting this macro in the CMake files (my fault, I just copied it from
16        configure). SUPPORT_GCOV has now been removed.
17    
18    3.  Make a small performance improvement in strlen16() and strlen32() in
19        pcretest.
20    
21    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
22        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
23        these cases). There was also one in pcretest.c.
24    
25    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
26    
27    6.  In UTF mode, the code for checking whether a group could match an empty
28        string (which is used for indefinitely repeated groups to allow for
29        breaking an infinite loop) was broken when the group contained a repeated
30        negated single-character class with a character that occupied more than one
31        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
32        UTF-8 mode). The effect was undefined: the group might or might not be
33        deemed as matching an empty string, or the program might have crashed.
34    
35    7.  The code for checking whether a group could match an empty string was not
36        recognizing that \h, \H, \v, \V, and \R must match a character.
37    
38    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
39        an empty string. If it can, pcretest shows this in its information output.
40    
41    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
42        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
43        matched by pcre_exec() without using JIT:
44    
45        (a) If the rest of the pattern did not match after a maximal run of
46            grapheme clusters, the code for backing up to try with fewer of them
47            did not always back up over a full grapheme when characters that do not
48            have the modifier quality were involved, e.g. Hangul syllables.
49    
50        (b) If the match point in a subject started with modifier character, and
51            there was no match, the code could incorrectly back up beyond the match
52            point, and potentially beyond the first character in the subject,
53            leading to a segfault or an incorrect match result.
54    
55    10. A conditional group with an assertion condition could lead to PCRE
56        recording an incorrect first data item for a match if no other first data
57        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
58        first data item, and therefore matched "ca" after "c" instead of at the
59        start.
60    
61    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
62        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
63    
64    12. The source of pcregrep now includes z/OS-specific code so that it can be
65        compiled for z/OS as part of the special z/OS distribution.
66    
67    13. Added the -T and -TM options to pcretest.
68    
69    14. The code in pcre_compile.c for creating the table of named capturing groups
70        has been refactored. Instead of creating the table dynamically during the
71        actual compiling pass, the information is remembered during the pre-compile
72        pass (on the stack unless there are more than 20 named groups, in which
73        case malloc() is used) and the whole table is created before the actual
74        compile happens. This has simplified the code (it is now nearly 150 lines
75        shorter) and prepared the way for better handling of references to groups
76        with duplicate names.
77    
78    15. A back reference to a named subpattern when there is more than one of the
79        same name now checks them in the order in which they appear in the pattern.
80        The first one that is set is used for the reference. Previously only the
81        first one was inspected. This change makes PCRE more compatible with Perl.
82    
83    
84    Version 8.33 28-May-2013
85    ------------------------
86    
87    1.  Added 'U' to some constants that are compared to unsigned integers, to
88        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
89        variables that are added to signed variables, to ensure the result is
90        signed and can be negated.
91    
92    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
93        CMake config files.
94    
95    3.  Revise the creation of config.h.generic so that all boolean macros are
96        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
97        overriding via -D on the command line possible.
98    
99    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
100        to unsigned int is reported to make a quite noticeable speed difference in
101        a specific Windows environment. Testing on Linux did also appear to show
102        some benefit (and it is clearly not harmful). Also fixed the definition of
103        Xop which should be unsigned.
104    
105    5.  Related to (4), changing the definition of the intermediate variable cc
106        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
107        improvements.
108    
109    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
110        unnecessary spaces.
111    
112    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
113        and later.
114    
115    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
116    
117    9.  Optimizing fast_forward_start_bits in JIT.
118    
119    10. Adding support for callouts in JIT, and fixing some issues revealed
120        during this work. Namely:
121    
122        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
123    
124        (b) Minimum length was not checked before the matching is started.
125    
126    11. The value of capture_last that is passed to callouts was incorrect in some
127        cases when there was a capture on one path that was subsequently abandoned
128        after a backtrack. Also, the capture_last value is now reset after a
129        recursion, since all captures are also reset in this case.
130    
131    12. The interpreter no longer returns the "too many substrings" error in the
132        case when an overflowing capture is in a branch that is subsequently
133        abandoned after a backtrack.
134    
135    13. In the pathological case when an offset vector of size 2 is used, pcretest
136        now prints out the matched string after a yield of 0 or 1.
137    
138    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
139        Only supported by the JIT compiler at the moment.
140    
141    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
142    
143    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
144        offset of the starting point of the matching process, provided the offsets
145        vector is large enough.
146    
147    17. The \A escape now records a lookbehind value of 1, though its execution
148        does not actually inspect the previous character. This is to ensure that,
149        in partial multi-segment matching, at least one character from the old
150        segment is retained when a new segment is processed. Otherwise, if there
151        are no lookbehinds in the pattern, \A might match incorrectly at the start
152        of a new segment.
153    
154    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
155    
156    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
157        gives some modest performance improvement in 8-bit mode.
158    
159    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
160        be expressed in certain programming languages using Universal Character
161        Names.
162    
163    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
164        which points out that "non characters" are not "characters that may not
165        appear in Unicode strings" but rather "characters that are reserved for
166        internal use and have only local meaning".
167    
168    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
169        there was a conditional group that depended on an assertion, if the
170        assertion was false, the callout that immediately followed the alternation
171        in the condition was skipped when pcre_exec() was used for matching.
172    
173    23. Allow an explicit callout to be inserted before an assertion that is the
174        condition for a conditional group, for compatibility with automatic
175        callouts, which always insert a callout at this point.
176    
177    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
178        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
179    
180    25. (*PRUNE) is now supported by the JIT compiler.
181    
182    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
183    
184    27. Fix the case where there are two or more SKIPs with arguments that may be
185        ignored.
186    
187    28. (*SKIP) is now supported by the JIT compiler.
188    
189    29. (*THEN) is now supported by the JIT compiler.
190    
191    30. Update RunTest with additional test selector options.
192    
193    31. The way PCRE handles backtracking verbs has been changed in two ways.
194    
195        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
196        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
197        backtracking. In some cases this makes it more Perl-compatible, but Perl's
198        rather obscure rules do not always do the same thing.
199    
200        (2) Previously, backtracking verbs were confined within assertions. This is
201        no longer the case for positive assertions, except for (*ACCEPT). Again,
202        this sometimes improves Perl compatibility, and sometimes does not.
203    
204    32. A number of tests that were in test 2 because Perl did things differently
205        have been moved to test 1, because either Perl or PCRE has changed, and
206        these tests are now compatible.
207    
208    32. Backtracking control verbs are now handled in the same way in JIT and
209        interpreter.
210    
211    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
212        contained a forward subroutine reference caused a compile error.
213    
214    34. Auto-detect and optimize limited repetitions in JIT.
215    
216    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
217        blocking (*UTF) etc.
218    
219    36. In the interpreter, maximizing pattern repetitions for characters and
220        character types now use tail recursion, which reduces stack usage.
221    
222    37. The value of the max lookbehind was not correctly preserved if a compiled
223        and saved regex was reloaded on a host of different endianness.
224    
225    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
226        of the compiled pattern block, expand the flags field from 16 to 32 bits
227        because it was almost full.
228    
229    39. Try madvise first before posix_madvise.
230    
231    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
232        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
233        original change locked out all matches of empty strings. This has been
234        changed so that one match of an empty string per line is recognized.
235        Subsequent searches on the same line (for colouring or for --only-matching,
236        for example) do not recognize empty strings.
237    
238    41. Applied a user patch to fix a number of spelling mistakes in comments.
239    
240    42. Data lines longer than 65536 caused pcretest to crash.
241    
242    43. Clarified the data type for length and startoffset arguments for pcre_exec
243        and pcre_dfa_exec in the function-specific man pages, where they were
244        explicitly stated to be in bytes, never having been updated. I also added
245        some clarification to the pcreapi man page.
246    
247    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
248        a segmentation fault.
249    
250    
251    Version 8.32 30-November-2012
252    -----------------------------
253    
254    1.  Improved JIT compiler optimizations for first character search and single
255        character iterators.
256    
257    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
258        Patch by Daniel Richard G.
259    
260    3.  Single character iterator optimizations in the JIT compiler.
261    
262    4.  Improved JIT compiler optimizations for character ranges.
263    
264    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
265        Reported by Giuseppe D'Angelo.
266    
267    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
268        of a line, was being set incorrectly in cases where .* appeared inside
269        atomic brackets at the start of a pattern, or where there was a subsequent
270        *PRUNE or *SKIP.
271    
272    7.  Improved instruction cache flush for POWER/PowerPC.
273        Patch by Daniel Richard G.
274    
275    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
276        grep:
277    
278        (a) There is now no limit to the number of patterns to be matched.
279    
280        (b) An error is given if a pattern is too long.
281    
282        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
283            are now supported.
284    
285        (d) --exclude-from and --include-from (multiple use) have been added.
286    
287        (e) Exclusions and inclusions now apply to all files and directories, not
288            just to those obtained from scanning a directory recursively.
289    
290        (f) Multiple uses of -f and --file-list are now supported.
291    
292        (g) In a Windows environment, the default for -d has been changed from
293            "read" (the GNU grep default) to "skip", because otherwise the presence
294            of a directory in the file list provokes an error.
295    
296        (h) The documentation has been revised and clarified in places.
297    
298    9.  Improve the matching speed of capturing brackets.
299    
300    10. Changed the meaning of \X so that it now matches a Unicode extended
301        grapheme cluster.
302    
303    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
304        out POSIX threads when JIT support is configured.
305    
306    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
307    
308    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
309        structure could go wrong in environments where size_t is not the same size
310        as int.
311    
312    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
313        set.
314    
315    15. The EBCDIC support had decayed; later updates to the code had included
316        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
317        general tidy up of EBCDIC-related issues, and the documentation was also
318        not quite right. There is now a test that can be run on ASCII systems to
319        check some of the EBCDIC-related things (but is it not a full test).
320    
321    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
322        in a small tidy to the code.
323    
324    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
325    
326    18. If the --only-matching (-o) option in pcregrep is specified multiple
327        times, each one causes appropriate output. For example, -o1 -o2 outputs the
328        substrings matched by the 1st and 2nd capturing parentheses. A separating
329        string can be specified by --om-separator (default empty).
330    
331    19. Improving the first n character searches.
332    
333    20. Turn case lists for horizontal and vertical white space into macros so that
334        they are defined only once.
335    
336    21. This set of changes together give more compatible Unicode case-folding
337        behaviour for characters that have more than one other case when UCP
338        support is available.
339    
340        (a) The Unicode property table now has offsets into a new table of sets of
341            three or more characters that are case-equivalent. The MultiStage2.py
342            script that generates these tables (the pcre_ucd.c file) now scans
343            CaseFolding.txt instead of UnicodeData.txt for character case
344            information.
345    
346        (b) The code for adding characters or ranges of characters to a character
347            class has been abstracted into a generalized function that also handles
348            case-independence. In UTF-mode with UCP support, this uses the new data
349            to handle characters with more than one other case.
350    
351        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
352            whose other case is greater than 256 are now correctly matched
353            caselessly. Previously, the high codepoint matched the low one, but not
354            vice versa.
355    
356        (d) The processing of \h, \H, \v, and \ in character classes now makes use
357            of the new class addition function, using character lists defined as
358            macros alongside the case definitions of 20 above.
359    
360        (e) Caseless back references now work with characters that have more than
361            one other case.
362    
363        (f) General caseless matching of characters with more than one other case
364            is supported.
365    
366    22. Unicode character properties were updated from Unicode 6.2.0
367    
368    23. Improved CMake support under Windows. Patch by Daniel Richard G.
369    
370    24. Add support for 32-bit character strings, and UTF-32
371    
372    25. Major JIT compiler update (code refactoring and bugfixing).
373        Experimental Sparc 32 support is added.
374    
375    26. Applied a modified version of Daniel Richard G's patch to create
376        pcre.h.generic and config.h.generic by "make" instead of in the
377        PrepareRelease script.
378    
379    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
380        pcre_compile.c when checking for a zero character.
381    
382    28. Introducing a native interface for JIT. Through this interface, the compiled
383        machine code can be directly executed. The purpose of this interface is to
384        provide fast pattern matching, so several sanity checks are not performed.
385        However, feature tests are still performed. The new interface provides
386        1.4x speedup compared to the old one.
387    
388    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
389        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
390        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
391    
392    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
393        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
394        the "old" RFC 2279). Instead, it ended up passing a negative length to
395        pcre_exec().
396    
397    31. Add support for GCC's visibility feature to hide internal functions.
398    
399    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
400        "unknown -C option" after outputting 0 or 1.
401    
402    33. There is now support for generating a code coverage report for the test
403        suite in environments where gcc is the compiler and lcov is installed. This
404        is mainly for the benefit of the developers.
405    
406    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
407        unaddressable using valgrind annotations, allowing valgrind to detect
408        invalid memory accesses. This is mainly for the benefit of the developers.
409    
410    25. (*UTF) can now be used to start a pattern in any of the three libraries.
411    
412    26. Give configure error if --enable-cpp but no C++ compiler found.
413    
414    
415    Version 8.31 06-July-2012
416    -------------------------
417    
418    1.  Fixing a wrong JIT test case and some compiler warnings.
419    
420    2.  Removed a bashism from the RunTest script.
421    
422    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
423        to unsigned type, result still unsigned" that was given by an MS compiler
424        on encountering the code "-sizeof(xxx)".
425    
426    4.  Partial matching support is added to the JIT compiler.
427    
428    5.  Fixed several bugs concerned with partial matching of items that consist
429        of more than one character:
430    
431        (a) /^(..)\1/ did not partially match "aba" because checking references was
432            done on an "all or nothing" basis. This also applied to repeated
433            references.
434    
435        (b) \R did not give a hard partial match if \r was found at the end of the
436            subject.
437    
438        (c) \X did not give a hard partial match after matching one or more
439            characters at the end of the subject.
440    
441        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
442            a partial match for the string "\r".
443    
444        (e) When newline was set to CRLF, the metacharacter "." did not recognize
445            a partial match for a CR character at the end of the subject string.
446    
447    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
448        running pcretest, the text "(JIT)" added to the output whenever JIT is
449        actually used to run the match.
450    
451    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
452        or /S+[+] with a digit between 1 and 7.
453    
454    8.  OP_NOT now supports any UTF character not just single-byte ones.
455    
456    9.  (*MARK) control verb is now supported by the JIT compiler.
457    
458    10. The command "./RunTest list" lists the available tests without actually
459        running any of them. (Because I keep forgetting what they all are.)
460    
461    11. Add PCRE_INFO_MAXLOOKBEHIND.
462    
463    12. Applied a (slightly modified) user-supplied patch that improves performance
464        when the heap is used for recursion (compiled with --disable-stack-for-
465        recursion). Instead of malloc and free for each heap frame each time a
466        logical recursion happens, frames are retained on a chain and re-used where
467        possible. This sometimes gives as much as 30% improvement.
468    
469    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
470        call.
471    
472    14. As documented, (*COMMIT) is now confined to within a positive assertion.
473    
474    15. It is now possible to link pcretest with libedit as an alternative to
475        libreadline.
476    
477    16. (*COMMIT) control verb is now supported by the JIT compiler.
478    
479    17. The Unicode data tables have been updated to Unicode 6.1.0.
480    
481    18. Added --file-list option to pcregrep.
482    
483    19. Added binary file support to pcregrep, including the -a, --binary-files,
484        -I, and --text options.
485    
486    20. The madvise function is renamed for posix_madvise for QNX compatibility
487        reasons. Fixed by Giuseppe D'Angelo.
488    
489    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
490    
491    22. Changed the option for creating symbolic links for 16-bit man pages from
492        -s to -sf so that re-installing does not cause issues.
493    
494    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
495    
496    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
497        matches in certain environments (the workspace was not being correctly
498        retained). Also added to pcre_dfa_exec() a simple plausibility check on
499        some of the workspace data at the beginning of a restart.
500    
501    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
502        was not doing so when it should - probably a typo introduced by SVN 528
503        (change 8.10/14).
504    
505    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
506        \w+ when the character tables indicated that \x{c4} was a word character.
507        There were several related cases, all because the tests for doing a table
508        lookup were testing for characters less than 127 instead of 255.
509    
510    27. If a pattern contains capturing parentheses that are not used in a match,
511        their slots in the ovector are set to -1. For those that are higher than
512        any matched groups, this happens at the end of processing. In the case when
513        there were back references that the ovector was too small to contain
514        (causing temporary malloc'd memory to be used during matching), and the
515        highest capturing number was not used, memory off the end of the ovector
516        was incorrectly being set to -1. (It was using the size of the temporary
517        memory instead of the true size.)
518    
519    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
520        ovector size, it uses memory at the end of the block that it has got.
521    
522    29. Check for an overlong MARK name and give an error at compile time. The
523        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
524    
525    30. JIT compiler update.
526    
527    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
528        Rill for the patch.
529    
530    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
531    
532    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
533    
534    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
535        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
536        was enabled.
537    
538    35. Improve JIT code generation for greedy plus quantifier.
539    
540    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
541        1 to "aa" instead of to an empty string. The bug affected repeated groups
542        that could potentially match an empty string.
543    
544    37. Optimizing single character iterators in JIT.
545    
546    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
547        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
548        codepoints that are too big for the mode are faulted, and in a UTF mode,
549        disallowed codepoints are also faulted.
550    
551    39. If PCRE was compiled with UTF support, in three places in the DFA
552        matcher there was code that should only have been obeyed in UTF mode, but
553        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
554        processing when bytes with values greater than 127 were present. In 16-bit
555        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
556        both cases the values are those that cannot be the first data item in a UTF
557        character. The three items that might have provoked this were recursions,
558        possessively repeated groups, and atomic groups.
559    
560    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
561        and pcregrep, because some OS require shared objects to be explicitly
562        passed to ld, causing the link step to fail if they are not.
563    
564    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
565        mode, patterns that started with \h* or \R* might be incorrectly matched.
566    
567    
568    Version 8.30 04-February-2012
569    -----------------------------
570    
571    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
572        name is defined in ctype.h.
573    
574    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
575        only in quite long subpatterns.
576    
577    3.  Removed the function pcre_info(), which has been obsolete and deprecated
578        since it was replaced by pcre_fullinfo() in February 2000.
579    
580    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
581        match a (*MARK), and the match failed at the start of the subject, a
582        reference to memory before the start of the subject could occur. This bug
583        was introduced by fix 17 of release 8.21.
584    
585    5.  A reference to an unset group with zero minimum repetition was giving
586        totally wrong answers (in non-JavaScript-compatibility mode). For example,
587        /(another)?(\1?)test/ matched against "hello world test". This bug was
588        introduced in release 8.13.
589    
590    6.  Add support for 16-bit character strings (a large amount of work involving
591        many changes and refactorings).
592    
593    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
594        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
595        from a file.
596    
597    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
598        rounding is not applied in this particular case).
599    
600    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
601        if they appear, or are escaped, in patterns.
602    
603    10. Get rid of a number of -Wunused-but-set-variable warnings.
604    
605    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
606        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
607        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
608        also returns the mark "x". This bug applied to capturing parentheses,
609        non-capturing parentheses, and atomic parentheses. It also applied to some
610        assertions.
611    
612    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
613        information out of configure.ac instead of relying on pcre.h.generic, which
614        is not stored in the repository.
615    
616    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
617        -lreadline.
618    
619    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
620    
621    15. Applied Graycode's patch to put the top-level frame on the stack rather
622        than the heap when not using the stack for recursion. This gives a
623        performance improvement in many cases when recursion is not deep.
624    
625    16. Experimental code added to "pcretest -C" to output the stack frame size.
626    
627    
628    Version 8.21 12-Dec-2011
629    ------------------------
630    
631    1.  Updating the JIT compiler.
632    
633    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
634        are added as well.
635    
636    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
637        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
638        calling _pcre_jit_exec. Some extra comments are added.
639    
640    4.  (*MARK) settings inside atomic groups that do not contain any capturing
641        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
642        was introduced by change 18 for 8.20.
643    
644    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
645        ECMA-262 standard.
646    
647    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
648        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
649        This bug was probably introduced by change 9 of 8.13.
650    
651    7.  While fixing 6 above, I noticed that a number of other items were being
652        incorrectly rejected as "not fixed length". This arose partly because newer
653        opcodes had not been added to the fixed-length checking code. I have (a)
654        corrected the bug and added tests for these items, and (b) arranged for an
655        error to occur if an unknown opcode is encountered while checking for fixed
656        length instead of just assuming "not fixed length". The items that were
657        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
658        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
659        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
660    
661    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
662        being incorrectly compiled and would have given unpredicatble results.
663    
664    9.  A possessively repeated subpattern with minimum repeat count greater than
665        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
666        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
667        the first (A) could occur when it should not.
668    
669    10. Add a cast and remove a redundant test from the code.
670    
671    11. JIT should use pcre_malloc/pcre_free for allocation.
672    
673    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
674        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
675        is anything other than /usr, -L is still shown).
676    
677    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
678    
679    14. Perl does not support \N without a following name in a [] class; PCRE now
680        also gives an error.
681    
682    15. If a forward reference was repeated with an upper limit of around 2000,
683        it caused the error "internal error: overran compiling workspace". The
684        maximum number of forward references (including repeats) was limited by the
685        internal workspace, and dependent on the LINK_SIZE. The code has been
686        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
687        the default depends on LINK_SIZE. There is a new upper limit (for safety)
688        of around 200,000 forward references. While doing this, I also speeded up
689        the filling in of repeated forward references.
690    
691    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
692        incorrectly expecting the subject to contain another "a" after the start.
693    
694    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
695        in the match, the SKIP should be ignored. This was not happening; instead
696        the SKIP was being treated as NOMATCH. For patterns such as
697        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
698        tested.
699    
700    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
701        now much more compatible with Perl, in particular in cases where the result
702        is a non-match for a non-anchored pattern. For example, if
703        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
704        "m", where previously it did not return a name. A side effect of this
705        change is that for partial matches, the last encountered mark name is
706        returned, as for non matches. A number of tests that were previously not
707        Perl-compatible have been moved into the Perl-compatible test files. The
708        refactoring has had the pleasing side effect of removing one argument from
709        the match() function, thus reducing its stack requirements.
710    
711    19. If the /S+ option was used in pcretest to study a pattern using JIT,
712        subsequent uses of /S (without +) incorrectly behaved like /S+.
713    
714    21. Retrieve executable code size support for the JIT compiler and fixing
715        some warnings.
716    
717    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
718        not work when the shorter character appeared right at the end of the
719        subject string.
720    
721    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
722        systems.
723    
724    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
725        output it when the /M option is used in pcretest.
726    
727    25. The CheckMan script was not being included in the distribution. Also, added
728        an explicit "perl" to run Perl scripts from the PrepareRelease script
729        because this is reportedly needed in Windows.
730    
731    26. If study data was being save in a file and studying had not found a set of
732        "starts with" bytes for the pattern, the data written to the file (though
733        never used) was taken from uninitialized memory and so caused valgrind to
734        complain.
735    
736    27. Updated RunTest.bat as provided by Sheri Pierce.
737    
738    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
739    
740    29. Computation of memory usage for the table of capturing group names was
741        giving an unnecessarily large value.
742    
743    
744    Version 8.20 21-Oct-2011
745    ------------------------
746    
747    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
748        a POSIX class. After further experiments with Perl, which convinced me that
749        Perl has bugs and confusions, a closing square bracket is no longer allowed
750        in a POSIX name. This bug also affected patterns with classes that started
751        with full stops.
752    
753    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
754        captured substring, but while checking the failing first alternative,
755        substring 1 is temporarily captured. If the output vector supplied to
756        pcre_exec() was not big enough for this capture, the yield of the function
757        was still zero ("insufficient space for captured substrings"). This cannot
758        be totally fixed without adding another stack variable, which seems a lot
759        of expense for a edge case. However, I have improved the situation in cases
760        such as /(a)(b)x|abc/ matched against "abc", where the return code
761        indicates that fewer than the maximum number of slots in the ovector have
762        been set.
763    
764    3.  Related to (2) above: when there are more back references in a pattern than
765        slots in the output vector, pcre_exec() uses temporary memory during
766        matching, and copies in the captures as far as possible afterwards. It was
767        using the entire output vector, but this conflicts with the specification
768        that only 2/3 is used for passing back captured substrings. Now it uses
769        only the first 2/3, for compatibility. This is, of course, another edge
770        case.
771    
772    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
773        main code base, and can be used by building with --enable-jit. When this is
774        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
775        runtime --no-jit option is given.
776    
777    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
778        ovector, the return from the function was zero, implying that there were
779        other matches that did not fit. The correct "exactly full" value is now
780        returned.
781    
782    6.  If a subpattern that was called recursively or as a subroutine contained
783        (*PRUNE) or any other control that caused it to give a non-standard return,
784        invalid errors such as "Error -26 (nested recursion at the same subject
785        position)" or even infinite loops could occur.
786    
787    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
788        computing the minimum length on reaching *ACCEPT, and so ended up with the
789        wrong value of 1 rather than 0. Further investigation indicates that
790        computing a minimum subject length in the presence of *ACCEPT is difficult
791        (think back references, subroutine calls), and so I have changed the code
792        so that no minimum is registered for a pattern that contains *ACCEPT.
793    
794    8.  If (*THEN) was present in the first (true) branch of a conditional group,
795        it was not handled as intended. [But see 16 below.]
796    
797    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
798        Sheri Pierce.
799    
800    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
801        the first byte in a match must be "a".
802    
803    11. Change 17 for 8.13 increased the recursion depth for patterns like
804        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
805        pattern contains any instances of (*THEN). If it does not, the old
806        optimizations are restored. It would be nice to do this on a per-group
807        basis, but at the moment that is not feasible.
808    
809    12. In some environments, the output of pcretest -C is CRLF terminated. This
810        broke RunTest's code that checks for the link size. A single white space
811        character after the value is now allowed for.
812    
813    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
814        For "fr", it uses the Windows-specific input and output files.
815    
816    14. If (*THEN) appeared in a group that was called recursively or as a
817        subroutine, it did not work as intended. [But see next item.]
818    
819    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
820        pattern fragments (but not containing any | characters). If A and B are
821        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
822        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
823        D. In other words, Perl considers parentheses that do not contain any |
824        characters to be part of a surrounding alternative, whereas PCRE was
825        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
826        differently. PCRE now behaves in the same way as Perl, except in the case
827        of subroutine/recursion calls such as (?1) which have in any case always
828        been different (but PCRE had them first :-).
829    
830    16. Related to 15 above: Perl does not treat the | in a conditional group as
831        creating alternatives. Such a group is treated in the same way as an
832        ordinary group without any | characters when processing (*THEN). PCRE has
833        been changed to match Perl's behaviour.
834    
835    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
836        RunGrepTest script failed.
837    
838    18. Change 22 for version 13 caused atomic groups to use more stack. This is
839        inevitable for groups that contain captures, but it can lead to a lot of
840        stack use in large patterns. The old behaviour has been restored for atomic
841        groups that do not contain any capturing parentheses.
842    
843    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
844        suppress the check for a minimum subject length at run time. (If it was
845        given to pcre_exec() or pcre_dfa_exec() it did work.)
846    
847    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
848        fail to work when decoding hex characters in data strings in EBCDIC
849        environments.
850    
851    21. It appears that in at least one Mac OS environment, the isxdigit() function
852        is implemented as a macro that evaluates to its argument more than once,
853        contravening the C 90 Standard (I haven't checked a later standard). There
854        was an instance in pcretest which caused it to go wrong when processing
855        \x{...} escapes in subject strings. The has been rewritten to avoid using
856        things like p++ in the argument of isxdigit().
857    
858    
859    Version 8.13 16-Aug-2011
860    ------------------------
861    
862    1.  The Unicode data tables have been updated to Unicode 6.0.0.
863    
864    2.  Two minor typos in pcre_internal.h have been fixed.
865    
866    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
867        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
868        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
869    
870    4.  There were a number of related bugs in the code for matching backrefences
871        caselessly in UTF-8 mode when codes for the characters concerned were
872        different numbers of bytes. For example, U+023A and U+2C65 are an upper
873        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
874        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
875        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
876        2-byte code at the end of the subject (it thought there wasn't enough data
877        left).
878    
879    5.  Comprehensive information about what went wrong is now returned by
880        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
881        as the output vector has at least 2 elements. The offset of the start of
882        the failing character and a reason code are placed in the vector.
883    
884    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
885        now returned is for the first byte of the failing character, instead of the
886        last byte inspected. This is an incompatible change, but I hope it is small
887        enough not to be a problem. It makes the returned offset consistent with
888        pcre_exec() and pcre_dfa_exec().
889    
890    7.  pcretest now gives a text phrase as well as the error number when
891        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
892        failure, the offset and reason code are output.
893    
894    8.  When \R was used with a maximizing quantifier it failed to skip backwards
895        over a \r\n pair if the subsequent match failed. Instead, it just skipped
896        back over a single character (\n). This seems wrong (because it treated the
897        two characters as a single entity when going forwards), conflicts with the
898        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
899        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
900        has been changed.
901    
902    9.  Some internal refactoring has changed the processing so that the handling
903        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
904        time (the PCRE_DOTALL option was changed this way some time ago: version
905        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
906        which was always a bit of a fudge. It also means that there is one less
907        argument for the match() function, which reduces its stack requirements
908        slightly. This change also fixes an incompatibility with Perl: the pattern
909        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
910    
911    10. More internal refactoring has drastically reduced the number of recursive
912        calls to match() for possessively repeated groups such as (abc)++ when
913        using pcre_exec().
914    
915    11. While implementing 10, a number of bugs in the handling of groups were
916        discovered and fixed:
917    
918        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
919        (a|)*(?1) gave a compile-time internal error.
920        ((a|)+)+  did not notice that the outer group could match an empty string.
921        (^a|^)+   was not marked as anchored.
922        (.*a|.*)+ was not marked as matching at start or after a newline.
923    
924    12. Yet more internal refactoring has removed another argument from the match()
925        function. Special calls to this function are now indicated by setting a
926        value in a variable in the "match data" data block.
927    
928    13. Be more explicit in pcre_study() instead of relying on "default" for
929        opcodes that mean there is no starting character; this means that when new
930        ones are added and accidentally left out of pcre_study(), testing should
931        pick them up.
932    
933    14. The -s option of pcretest has been documented for ages as being an old
934        synonym of -m (show memory usage). I have changed it to mean "force study
935        for every regex", that is, assume /S for every regex. This is similar to -i
936        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
937        using it. It makes it easier to run collections of tests with and without
938        study enabled, and thereby test pcre_study() more easily. All the standard
939        tests are now run with and without -s (but some patterns can be marked as
940        "never study" - see 20 below).
941    
942    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
943        restoration of the capturing data to the outer values was not happening
944        correctly.
945    
946    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
947        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
948        pattern had matched an empty string, and so incorrectly returned a no
949        match.
950    
951    17. There was optimizing code for the last branch of non-capturing parentheses,
952        and also for the obeyed branch of a conditional subexpression, which used
953        tail recursion to cut down on stack usage. Unfortunately, now that there is
954        the possibility of (*THEN) occurring in these branches, tail recursion is
955        no longer possible because the return has to be checked for (*THEN). These
956        two optimizations have therefore been removed. [But see 8.20/11 above.]
957    
958    18. If a pattern containing \R was studied, it was assumed that \R always
959        matched two bytes, thus causing the minimum subject length to be
960        incorrectly computed because \R can also match just one byte.
961    
962    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
963        was incorrectly computed.
964    
965    20. If /S is present twice on a test pattern in pcretest input, it now
966        *disables* studying, thereby overriding the use of -s on the command line
967        (see 14 above). This is necessary for one or two tests to keep the output
968        identical in both cases.
969    
970    21. When (*ACCEPT) was used in an assertion that matched an empty string and
971        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
972    
973    22. When an atomic group that contained a capturing parenthesis was
974        successfully matched, but the branch in which it appeared failed, the
975        capturing was not being forgotten if a higher numbered group was later
976        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
977        group 1 to "a", when in fact it should be unset. This applied to multi-
978        branched capturing and non-capturing groups, repeated or not, and also to
979        positive assertions (capturing in negative assertions does not happen
980        in PCRE) and also to nested atomic groups.
981    
982    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
983        subject after a captured substring, to make it easier to tell which of a
984        number of identical substrings has been captured.
985    
986    24. The way atomic groups are processed by pcre_exec() has been changed so that
987        if they are repeated, backtracking one repetition now resets captured
988        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
989        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
990        "aaa". Previously, it would have been "a". As part of this code
991        refactoring, the way recursive calls are handled has also been changed.
992    
993    25. If an assertion condition captured any substrings, they were not passed
994        back unless some other capturing happened later. For example, if
995        (?(?=(a))a) was matched against "a", no capturing was returned.
996    
997    26. When studying a pattern that contained subroutine calls or assertions,
998        the code for finding the minimum length of a possible match was handling
999        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1000        group 1 called group 2 while simultaneously a separate group 2 called group
1001        1). A stack overflow occurred in this case. I have fixed this by limiting
1002        the recursion depth to 10.
1003    
1004    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1005        Fortmann. This supports explicit test numbers on the command line, and has
1006        argument validation and error reporting.
1007    
1008    28. An instance of \X with an unlimited repeat could fail if at any point the
1009        first character it looked at was a mark character.
1010    
1011    29. Some minor code refactoring concerning Unicode properties and scripts
1012        should reduce the stack requirement of match() slightly.
1013    
1014    30. Added the '=' option to pcretest to check the setting of unused capturing
1015        slots at the end of the pattern, which are documented as being -1, but are
1016        not included in the return count.
1017    
1018    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1019        compiled something random. Now it gives a compile-time error (as does
1020        Perl).
1021    
1022    32. A *MARK encountered during the processing of a positive assertion is now
1023        recorded and passed back (compatible with Perl).
1024    
1025    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1026        had alternative anchored branches, the search for a second match in a line
1027        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1028        matched the line "0102" twice. The same bug affected patterns that started
1029        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1030        twice.
1031    
1032    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1033        does, and because of capturing effects, quantifying parenthesized
1034        assertions may at times be useful. Quantifiers are now allowed for
1035        parenthesized assertions.
1036    
1037    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1038    
1039    36. \g was being checked for fancy things in a character class, when it should
1040        just be a literal "g".
1041    
1042    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1043        appearance of a nested POSIX class supersedes an apparent external class.
1044        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1045        unescaped square brackets may also appear as part of class names. For
1046        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1047        more like Perl. (But see 8.20/1 above.)
1048    
1049    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1050        was because it thought it was \N{name}, which is not supported).
1051    
1052    39. Add minix to OS list not supporting the -S option in pcretest.
1053    
1054    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1055        cannot analyze patterns in sufficient detail to catch mutual recursions
1056        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1057        subgroup is called recursively as a subpattern for a second time at the
1058        same position in the subject string. In previous releases this might have
1059        been caught by the recursion limit, or it might have run out of stack.
1060    
1061    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1062        happen only once. PCRE was, however incorrectly giving a compile time error
1063        "recursive call could loop indefinitely" because it cannot analyze the
1064        pattern in sufficient detail. The compile time test no longer happens when
1065        PCRE is compiling a conditional subpattern, but actual runaway loops are
1066        now caught at runtime (see 40 above).
1067    
1068    42. It seems that Perl allows any characters other than a closing parenthesis
1069        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1070        has been changed to be the same.
1071    
1072    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1073        as not to get warnings when autogen.sh is called. Also changed
1074        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1075    
1076    44. To help people who use pcregrep to scan files containing exceedingly long
1077        lines, the following changes have been made:
1078    
1079        (a) The default value of the buffer size parameter has been increased from
1080            8K to 20K. (The actual buffer used is three times this size.)
1081    
1082        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1083            PCRE is built.
1084    
1085        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1086            to be set at run time.
1087    
1088        (d) Numerical values in pcregrep options can be followed by K or M, for
1089            example --buffer-size=50K.
1090    
1091        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1092            given and the return code is set to 2.
1093    
1094    45. Add a pointer to the latest mark to the callout data block.
1095    
1096    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1097        partial match of an empty string instead of no match. This was specific to
1098        the use of ".".
1099    
1100    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1101        complete match instead of a partial match. This bug was dependent on both
1102        the PCRE_UTF8 and PCRE_DOTALL options being set.
1103    
1104    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1105        starting byte set, because \b was not being ignored.
1106    
1107    
1108    Version 8.12 15-Jan-2011
1109    ------------------------
1110    
1111    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1112        checks for such things as part of the documentation building process.
1113    
1114    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1115        --match-limit and --recursion-limit options (added for 8.11). In
1116        particular, this made one of the standard tests fail. (The integer value
1117        went into the wrong half of a long int.)
1118    
1119    3.  If the --colour option was given to pcregrep with -v (invert match), it
1120        did strange things, either producing crazy output, or crashing. It should,
1121        of course, ignore a request for colour when reporting lines that do not
1122        match.
1123    
1124    4.  Another pcregrep bug caused similar problems if --colour was specified with
1125        -M (multiline) and the pattern match finished with a line ending.
1126    
1127    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1128        matched in multiline mode, the following line was shown as part of the
1129        match. This seems wrong, so I have changed it.
1130    
1131    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1132        the check for further matches in the same line (so they could be coloured)
1133        to overrun the end of the current line. If another match was found, it was
1134        incorrectly shown (and then shown again when found in the next line).
1135    
1136    7.  If pcregrep was compiled under Windows, there was a reference to the
1137        function pcregrep_exit() before it was defined. I am assuming this was
1138        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1139        reported by a user. I've moved the definition above the reference.
1140    
1141    
1142    Version 8.11 10-Dec-2010
1143    ------------------------
1144    
1145    1.  (*THEN) was not working properly if there were untried alternatives prior
1146        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1147        backtracked to try for "b" instead of moving to the next alternative branch
1148        at the same level (in this case, to look for "c"). The Perl documentation
1149        is clear that when (*THEN) is backtracked onto, it goes to the "next
1150        alternative in the innermost enclosing group".
1151    
1152    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1153        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1154        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1155        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1156        (*THEN).
1157    
1158    3.  If \s appeared in a character class, it removed the VT character from
1159        the class, even if it had been included by some previous item, for example
1160        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1161        of \s, but is part of the POSIX "space" class.)
1162    
1163    4.  A partial match never returns an empty string (because you can always
1164        match an empty string at the end of the subject); however the checking for
1165        an empty string was starting at the "start of match" point. This has been
1166        changed to the "earliest inspected character" point, because the returned
1167        data for a partial match starts at this character. This means that, for
1168        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1169        (previously it gave "no match").
1170    
1171    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1172        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1173        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1174        has an implication that the given string is incomplete (because a partial
1175        match is preferred over a full match). For this reason, these items now
1176        give a partial match in this situation. [Aside: previously, the one case
1177        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1178        match rather than a full match, which was wrong by the old rules, but is
1179        now correct.]
1180    
1181    6.  There was a bug in the handling of #-introduced comments, recognized when
1182        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1183        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1184        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1185        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1186        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1187        places in pcre_compile().
1188    
1189    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1190        comments when looking ahead for named forward references to subpatterns,
1191        the only newline sequence it recognized was NL. It now handles newlines
1192        according to the set newline convention.
1193    
1194    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1195        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1196        cater for a lack of strerror(). These oversights have been fixed.
1197    
1198    9.  Added --match-limit and --recursion-limit to pcregrep.
1199    
1200    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1201    
1202    11. When the -o option was used, pcregrep was setting a return code of 1, even
1203        when matches were found, and --line-buffered was not being honoured.
1204    
1205    12. Added an optional parentheses number to the -o and --only-matching options
1206        of pcregrep.
1207    
1208    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1209        can match an empty string. The code to do it in pcretest and pcredemo
1210        needed fixing:
1211    
1212        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1213            only one byte after an empty string match just before CRLF (this case
1214            just got forgotten; "any" and "anycrlf" were OK).
1215    
1216        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1217            mode when an empty string match preceded an ASCII character followed by
1218            a non-ASCII character. (The code for advancing by one character rather
1219            than one byte was nonsense.)
1220    
1221        (c) The pcredemo.c sample program did not have any code at all to handle
1222            the cases when CRLF is a valid newline sequence.
1223    
1224    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1225        as a starting offset was within the subject string. There is now a new
1226        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1227        negative or greater than the length of the string. In order to test this,
1228        pcretest is extended to allow the setting of negative starting offsets.
1229    
1230    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1231        starting offset points to the beginning of a UTF-8 character was
1232        unnecessarily clumsy. I tidied it up.
1233    
1234    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1235        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1236    
1237    17. Nobody had reported that the --include_dir option, which was added in
1238        release 7.7 should have been called --include-dir (hyphen, not underscore)
1239        for compatibility with GNU grep. I have changed it to --include-dir, but
1240        left --include_dir as an undocumented synonym, and the same for
1241        --exclude-dir, though that is not available in GNU grep, at least as of
1242        release 2.5.4.
1243    
1244    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1245        characters from a string of bytes) have been redefined so as not to use
1246        loops, in order to improve performance in some environments. At the same
1247        time, I abstracted some of the common code into auxiliary macros to save
1248        repetition (this should not affect the compiled code).
1249    
1250    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1251        compile-time error is now given if \c is not followed by an ASCII
1252        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1253        different, and any byte value is allowed.)
1254    
1255    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1256        START_OPTIMIZE option, which is now allowed at compile time - but just
1257        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1258        to pcregrep and other applications that have no direct access to PCRE
1259        options. The new /Y option in pcretest sets this option when calling
1260        pcre_compile().
1261    
1262    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1263        back references. Groups containing recursive back references were forced to
1264        be atomic by that change, but in the case of named groups, the amount of
1265        memory required was incorrectly computed, leading to "Failed: internal
1266        error: code overflow". This has been fixed.
1267    
1268    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1269        pcretest.c, to avoid build problems in some Borland environments.
1270    
1271    
1272    Version 8.10 25-Jun-2010
1273    ------------------------
1274    
1275    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1276        THEN.
1277    
1278    2.  (*ACCEPT) was not working when inside an atomic group.
1279    
1280    3.  Inside a character class, \B is treated as a literal by default, but
1281        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1282        causes the error). The code is unchanged, but I tidied the documentation.
1283    
1284    4.  Inside a character class, PCRE always treated \R and \X as literals,
1285        whereas Perl faults them if its -w option is set. I have changed PCRE so
1286        that it faults them when PCRE_EXTRA is set.
1287    
1288    5.  Added support for \N, which always matches any character other than
1289        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1290    
1291    6.  When compiling pcregrep with newer versions of gcc which may have
1292        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1293        declared with attribute warn_unused_result" were given. Just casting the
1294        result to (void) does not stop the warnings; a more elaborate fudge is
1295        needed. I've used a macro to implement this.
1296    
1297    7.  Minor change to pcretest.c to avoid a compiler warning.
1298    
1299    8.  Added four artifical Unicode properties to help with an option to make
1300        \s etc use properties (see next item). The new properties are: Xan
1301        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1302    
1303    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1304        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1305        this option. Modified pcretest to add /W to test this facility. Added
1306        REG_UCP to make it available via the POSIX interface.
1307    
1308    10. Added --line-buffered to pcregrep.
1309    
1310    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1311        studied, and the match started with a letter with a code point greater than
1312        127 whose first byte was different to the first byte of the other case of
1313        the letter, the other case of this starting letter was not recognized
1314        (#976).
1315    
1316    12. If a pattern that was studied started with a repeated Unicode property
1317        test, for example, \p{Nd}+, there was the theoretical possibility of
1318        setting up an incorrect bitmap of starting bytes, but fortunately it could
1319        not have actually happened in practice until change 8 above was made (it
1320        added property types that matched character-matching opcodes).
1321    
1322    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1323        possible starting bytes for non-anchored patterns.
1324    
1325    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1326        \R, and also a number of cases that involve Unicode properties, both
1327        explicit and implicit when PCRE_UCP is set.
1328    
1329    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1330        input, it could crash or give wrong results if characters with values
1331        greater than 0xc0 were present in the subject string. (Detail: it assumed
1332        UTF-8 input when processing these items.)
1333    
1334    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1335        size_t is 64-bit (#991).
1336    
1337    17. Added a check for running out of memory when PCRE is compiled with
1338        --disable-stack-for-recursion (#990).
1339    
1340    18. If the last data line in a file for pcretest does not have a newline on
1341        the end, a newline was missing in the output.
1342    
1343    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1344        less than 128) in its various bitmaps. However, there is a facility for
1345        generating tables according to the current locale when PCRE is compiled. It
1346        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1347        characters, are recognized by isspace() and therefore were getting set in
1348        these tables, and indeed these tables seem to approximate to ISO 8859. This
1349        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1350        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1351        which of course cannot start UTF-8 characters. I have changed the code so
1352        that only real ASCII characters (less than 128) and the correct starting
1353        bytes for UTF-8 encodings are set for characters greater than 127 when in
1354        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1355        altogether.)
1356    
1357    20. Added the /T option to pcretest so as to be able to run tests with non-
1358        standard character tables, thus making it possible to include the tests
1359        used for 19 above in the standard set of tests.
1360    
1361    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1362        reference to a subpattern the other side of a comment that contains an
1363        opening parenthesis caused either an internal compiling error, or a
1364        reference to the wrong subpattern.
1365    
1366    
1367    Version 8.02 19-Mar-2010
1368    ------------------------
1369    
1370    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1371    
1372    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1373        configured.
1374    
1375    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1376        original author of that file, following a query about its status.
1377    
1378    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1379        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1380    
1381    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1382        quantifier applied to a forward-referencing subroutine call, could compile
1383        incorrect code or give the error "internal error: previously-checked
1384        referenced subpattern not found".
1385    
1386    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1387        variables to point to external functions. For these systems, therefore,
1388        pcre_malloc etc. are now initialized to local functions that call the
1389        relevant global functions.
1390    
1391    7.  There were two entries missing in the vectors called coptable and poptable
1392        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1393        I've fixed the data, and added a kludgy way of testing at compile time that
1394        the lengths are correct (equal to the number of opcodes).
1395    
1396    8.  Following on from 7, I added a similar kludge to check the length of the
1397        eint vector in pcreposix.c.
1398    
1399    9.  Error texts for pcre_compile() are held as one long string to avoid too
1400        much relocation at load time. To find a text, the string is searched,
1401        counting zeros. There was no check for running off the end of the string,
1402        which could happen if a new error number was added without updating the
1403        string.
1404    
1405    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1406    
1407    11. \K was not working if it appeared in an atomic group or in a group that
1408        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1409        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1410        the assertion is positive, but not if it is negative.
1411    
1412    12. Change 11 fortuitously reduced the size of the stack frame used in the
1413        "match()" function of pcre_exec.c by one pointer. Forthcoming
1414        implementation of support for (*MARK) will need an extra pointer on the
1415        stack; I have reserved it now, so that the stack frame size does not
1416        decrease.
1417    
1418    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1419        item in branch that calls a recursion is a subroutine call - as in the
1420        second branch in the above example - was incorrectly given the compile-
1421        time error "recursive call could loop indefinitely" because pcre_compile()
1422        was not correctly checking the subroutine for matching a non-empty string.
1423    
1424    14. The checks for overrunning compiling workspace could trigger after an
1425        overrun had occurred. This is a "should never occur" error, but it can be
1426        triggered by pathological patterns such as hundreds of nested parentheses.
1427        The checks now trigger 100 bytes before the end of the workspace.
1428    
1429    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1430    
1431    
1432    Version 8.01 19-Jan-2010
1433    ------------------------
1434    
1435    1.  If a pattern contained a conditional subpattern with only one branch (in
1436        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1437        computed the wrong minimum data length (which is of course zero for such
1438        subpatterns). This could cause incorrect "no match" results.
1439    
1440    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1441        the pattern is reset in the first branch, pcre_compile() failed with
1442        "internal error: code overflow at offset...". This happened only when
1443        the reset was to the original external option setting. (An optimization
1444        abstracts leading options settings into an external setting, which was the
1445        cause of this.)
1446    
1447    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1448        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1449        assertion pattern did not match (meaning that the assertion was true), it
1450        was incorrectly treated as false if the SKIP had been reached during the
1451        matching. This also applied to assertions used as conditions.
1452    
1453    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1454        assertion subpattern, including such a pattern used as a condition,
1455        unpredictable results occurred, instead of the error return
1456        PCRE_ERROR_DFA_UITEM.
1457    
1458    5.  The C++ GlobalReplace function was not working like Perl for the special
1459        situation when an empty string is matched. It now does the fancy magic
1460        stuff that is necessary.
1461    
1462    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1463        removed. (These were left over from very, very early versions of PCRE.)
1464    
1465    7.  Some cosmetic changes to the code to make life easier when compiling it
1466        as part of something else:
1467    
1468        (a) Change DEBUG to PCRE_DEBUG.
1469    
1470        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1471            called "current" as "current_branch", to prevent a collision with the
1472            Linux macro when compiled as a kernel module.
1473    
1474        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1475            prevent a collision with the Linux macro when compiled as a kernel
1476            module.
1477    
1478    8.  In pcre_compile() there are some checks for integer overflows that used to
1479        cast potentially large values to (double). This has been changed to that
1480        when building, a check for int64_t is made, and if it is found, it is used
1481        instead, thus avoiding the use of floating point arithmetic. (There is no
1482        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1483        double.
1484    
1485    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1486        2005 (difference between two addresses compared to an unsigned value).
1487    
1488    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1489        custom one, because of the following reported problem in Windows:
1490    
1491          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1492              under Win32.
1493          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1494              therefore missing the function definition.
1495          - The compiler thus generates a "C" signature for the test function.
1496          - The linker fails to find the "C" function.
1497          - PCRE fails to configure if asked to do so against libbz2.
1498    
1499    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1500        messages were output:
1501    
1502          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1503          rerunning libtoolize, to keep the correct libtool macros in-tree.
1504          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1505    
1506        I have done both of these things.
1507    
1508    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1509        most of the time, it *can* run out if it is given a pattern that contains a
1510        runaway infinite recursion. I updated the discussion in the pcrestack man
1511        page.
1512    
1513    13. Now that we have gone to the x.xx style of version numbers, the minor
1514        version may start with zero. Using 08 or 09 is a bad idea because users
1515        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1516        interpreted as invalid octal numbers. I've updated the previous comment in
1517        configure.ac, and also added a check that gives an error if 08 or 09 are
1518        used.
1519    
1520    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1521        causing partial matching to fail when the end of the subject matched \W
1522        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1523    
1524    15. There were some discrepancies between the declarations in pcre_internal.h
1525        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1526        their definitions. The declarations used "const uschar *" and the
1527        definitions used USPTR. Even though USPTR is normally defined as "const
1528        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1529        reported that: "This difference in casting confuses some C++ compilers, for
1530        example, SunCC recognizes above declarations as different functions and
1531        generates broken code for hbpcre." I have changed the declarations to use
1532        USPTR.
1533    
1534    16. GNU libtool is named differently on some systems. The autogen.sh script now
1535        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1536        (FreeBSD).
1537    
1538    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1539        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1540        comment: "Figure out how to create a longlong from a string: strtoll and
1541        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1542        instance, but it only takes 2 args instead of 3!"
1543    
1544    18. A subtle bug concerned with back references has been fixed by a change of
1545        specification, with a corresponding code fix. A pattern such as
1546        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1547        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1548        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1549        same bug. Such groups have to be quantified to be useful, or contained
1550        inside another quantified group. (If there's no repetition, the reference
1551        can never match.) The problem arises because, having left the group and
1552        moved on to the rest of the pattern, a later failure that backtracks into
1553        the group uses the captured value from the final iteration of the group
1554        rather than the correct earlier one. I have fixed this in PCRE by forcing
1555        any group that contains a reference to itself to be an atomic group; that
1556        is, there cannot be any backtracking into it once it has completed. This is
1557        similar to recursive and subroutine calls.
1558    
1559    
1560    Version 8.00 19-Oct-09
1561    ----------------------
1562    
1563    1.  The table for translating pcre_compile() error codes into POSIX error codes
1564        was out-of-date, and there was no check on the pcre_compile() error code
1565        being within the table. This could lead to an OK return being given in
1566        error.
1567    
1568    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1569        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1570        in a Windows environment.
1571    
1572    3.  The pcregrep --count option prints the count for each file even when it is
1573        zero, as does GNU grep. However, pcregrep was also printing all files when
1574        --files-with-matches was added. Now, when both options are given, it prints
1575        counts only for those files that have at least one match. (GNU grep just
1576        prints the file name in this circumstance, but including the count seems
1577        more useful - otherwise, why use --count?) Also ensured that the
1578        combination -clh just lists non-zero counts, with no names.
1579    
1580    4.  The long form of the pcregrep -F option was incorrectly implemented as
1581        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1582        but it seems right to fix it, and I didn't think it was worth preserving
1583        the old behaviour.
1584    
1585    5.  The command line items --regex=pattern and --regexp=pattern were not
1586        recognized by pcregrep, which required --regex pattern or --regexp pattern
1587        (with a space rather than an '='). The man page documented the '=' forms,
1588        which are compatible with GNU grep; these now work.
1589    
1590    6.  No libpcreposix.pc file was created for pkg-config; there was just
1591        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1592    
1593    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1594        when UCP support is not needed, by modifying the Python script that
1595        generates it from Unicode data files. This should not matter if the module
1596        is correctly used as a library, but I received one complaint about 50K of
1597        unwanted data. My guess is that the person linked everything into his
1598        program rather than using a library. Anyway, it does no harm.
1599    
1600    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1601        was a minimum greater than 1 for a wide character in a possessive
1602        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1603        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1604        character. Chaos in the form of incorrect output or a compiling loop could
1605        result.
1606    
1607    9.  The restrictions on what a pattern can contain when partial matching is
1608        requested for pcre_exec() have been removed. All patterns can now be
1609        partially matched by this function. In addition, if there are at least two
1610        slots in the offset vector, the offset of the earliest inspected character
1611        for the match and the offset of the end of the subject are set in them when
1612        PCRE_ERROR_PARTIAL is returned.
1613    
1614    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1615        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1616        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1617        and may be more useful for multi-segment matching.
1618    
1619    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1620        used to be given if ever the end of the subject was reached; now it is
1621        given only if matching could not proceed because another character was
1622        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1623        string "Z", which now yields "no match" instead of "partial match". In the
1624        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1625        final character ended with (*FAIL).
1626    
1627    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1628        if the pattern had a "must contain" character that was already found in the
1629        earlier partial match, unless partial matching was again requested. For
1630        example, with the pattern /dog.(body)?/, the "must contain" character is
1631        "g". If the first part-match was for the string "dog", restarting with
1632        "sbody" failed. This bug has been fixed.
1633    
1634    13. The string returned by pcre_dfa_exec() after a partial match has been
1635        changed so that it starts at the first inspected character rather than the
1636        first character of the match. This makes a difference only if the pattern
1637        starts with a lookbehind assertion or \b or \B (\K is not supported by
1638        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1639        matching functions compatible, and I think it's the right thing to do.
1640    
1641    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1642        so that the demonstration program is easily available in environments where
1643        PCRE has not been installed from source.
1644    
1645    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1646        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1647        library.
1648    
1649    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1650        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1651        is not the first non-POSIX option to be added. Clearly some people find
1652        these options useful.
1653    
1654    17. If a caller to the POSIX matching function regexec() passes a non-zero
1655        value for nmatch with a NULL value for pmatch, the value of
1656        nmatch is forced to zero.
1657    
1658    18. RunGrepTest did not have a test for the availability of the -u option of
1659        the diff command, as RunTest does. It now checks in the same way as
1660        RunTest, and also checks for the -b option.
1661    
1662    19. If an odd number of negated classes containing just a single character
1663        interposed, within parentheses, between a forward reference to a named
1664        subpattern and the definition of the subpattern, compilation crashed with
1665        an internal error, complaining that it could not find the referenced
1666        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1667        [The bug was that it was starting one character too far in when skipping
1668        over the character class, thus treating the ] as data rather than
1669        terminating the class. This meant it could skip too much.]
1670    
1671    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1672        /g option in pcretest when the pattern contains \K, which makes it possible
1673        to have an empty string match not at the start, even when the pattern is
1674        anchored. Updated pcretest and pcredemo to use this option.
1675    
1676    21. If the maximum number of capturing subpatterns in a recursion was greater
1677        than the maximum at the outer level, the higher number was returned, but
1678        with unset values at the outer level. The correct (outer level) value is
1679        now given.
1680    
1681    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1682        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1683        make it do so. The string so far is captured, making this feature
1684        compatible with Perl.
1685    
1686    23. The tests have been re-organized, adding tests 11 and 12, to make it
1687        possible to check the Perl 5.10 features against Perl 5.10.
1688    
1689    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1690        pattern matches a fixed length string. PCRE did not allow this; now it
1691        does. Neither allows recursion.
1692    
1693    25. I finally figured out how to implement a request to provide the minimum
1694        length of subject string that was needed in order to match a given pattern.
1695        (It was back references and recursion that I had previously got hung up
1696        on.) This code has now been added to pcre_study(); it finds a lower bound
1697        to the length of subject needed. It is not necessarily the greatest lower
1698        bound, but using it to avoid searching strings that are too short does give
1699        some useful speed-ups. The value is available to calling programs via
1700        pcre_fullinfo().
1701    
1702    26. While implementing 25, I discovered to my embarrassment that pcretest had
1703        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1704        study optimizations had never been tested with that matching function.
1705        Oops. What is worse, even when it was passed study data, there was a bug in
1706        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1707        were also very few tests of studied patterns with pcre_dfa_exec().
1708    
1709    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1710        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1711        on the other side of the coin, they are no longer allowed to have different
1712        names, because these cannot be distinguished in PCRE, and this has caused
1713        confusion. (This is a difference from Perl.)
1714    
1715    28. When duplicate subpattern names are present (necessarily with different
1716        numbers, as required by 27 above), and a test is made by name in a
1717        conditional pattern, either for a subpattern having been matched, or for
1718        recursion in such a pattern, all the associated numbered subpatterns are
1719        tested, and the overall condition is true if the condition is true for any
1720        one of them. This is the way Perl works, and is also more like the way
1721        testing by number works.
1722    
1723    
1724    Version 7.9 11-Apr-09
1725    ---------------------
1726    
1727    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1728        (pcretest), all targets were linked against these libraries. This included
1729        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1730        libraries. This caused unwanted dependencies to be created. This problem
1731        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1732        pcretest is linked with readline.
1733    
1734    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1735        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1736        moved outside it again, because FALSE and TRUE are already defined in AIX,
1737        but BOOL is not.
1738    
1739    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1740        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1741    
1742    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1743        hyphens) following file names and line numbers when outputting matching
1744        lines. This is not true; no spaces are inserted. I have also clarified the
1745        wording for the --colour (or --color) option.
1746    
1747    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1748        was not coloured; this is different to GNU grep, so I have changed it to be
1749        the same.
1750    
1751    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1752        each matching line was coloured. Now it goes on to look for further matches
1753        of any of the test patterns, which is the same behaviour as GNU grep.
1754    
1755    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1756        doesn't make sense to accept an empty string match in pcregrep, so I have
1757        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1758        seems to be how GNU grep behaves. [But see later change 40 for release
1759        8.33.]
1760    
1761    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1762        start or after a newline", because the conditional assertion was not being
1763        correctly handled. The rule now is that both the assertion and what follows
1764        in the first alternative must satisfy the test.
1765    
1766    9.  If auto-callout was enabled in a pattern with a conditional group whose
1767        condition was an assertion, PCRE could crash during matching, both with
1768        pcre_exec() and pcre_dfa_exec().
1769    
1770    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1771        used for matching.
1772    
1773    11. Unicode property support in character classes was not working for
1774        characters (bytes) greater than 127 when not in UTF-8 mode.
1775    
1776    12. Added the -M command line option to pcretest.
1777    
1778    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1779    
1780    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1781    
1782    16. Added comments and documentation about mis-use of no_arg in the C++
1783        wrapper.
1784    
1785    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1786        from Martin Jerabek that uses macro names for all relevant character and
1787        string constants.
1788    
1789    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1790        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1791        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1792        these, but not everybody uses configure.
1793    
1794    19. A conditional group that had only one branch was not being correctly
1795        recognized as an item that could match an empty string. This meant that an
1796        enclosing group might also not be so recognized, causing infinite looping
1797        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1798        with the subject "ab", where knowledge that the repeated group can match
1799        nothing is needed in order to break the loop.
1800    
1801    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1802        exec(), but without supplying a callout function, matching went wrong.
1803    
1804    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1805        leak if the size of the offset vector was greater than 30. When the vector
1806        is smaller, the saved offsets during recursion go onto a local stack
1807        vector, but for larger vectors malloc() is used. It was failing to free
1808        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1809        error, in fact).
1810    
1811    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1812        heapframe that is used only when UTF-8 support is enabled. This caused no
1813        problem, but was untidy.
1814    
1815    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1816        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1817        included within another project.
1818    
1819    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1820        slightly modified by me:
1821    
1822          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1823              not building pcregrep.
1824    
1825          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1826              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1827    
1828    25. Forward references, both numeric and by name, in patterns that made use of
1829        duplicate group numbers, could behave incorrectly or give incorrect errors,
1830        because when scanning forward to find the reference group, PCRE was not
1831        taking into account the duplicate group numbers. A pattern such as
1832        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1833    
1834    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1835        the feature of a custom pointer more persuasive (as requested by a user).
1836    
1837    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1838        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1839        pre-defined.
1840    
1841    28. Added support for (*UTF8) at the start of a pattern.
1842    
1843    29. Arrange for flags added by the "release type" setting in CMake to be shown
1844        in the configuration summary.
1845    
1846    
1847    Version 7.8 05-Sep-08
1848    ---------------------
1849    
1850    1.  Replaced UCP searching code with optimized version as implemented for Ad
1851        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1852        stage table and inline lookup instead of a function, giving speed ups of 2
1853        to 5 times on some simple patterns that I tested. Permission was given to
1854        distribute the MultiStage2.py script that generates the tables (it's not in
1855        the tarball, but is in the Subversion repository).
1856    
1857    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1858        scripts.
1859    
1860    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1861        a group with a zero qualifier. The result of the study could be incorrect,
1862        or the function might crash, depending on the pattern.
1863    
1864    4.  Caseless matching was not working for non-ASCII characters in back
1865        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1866        It now works when Unicode Property Support is available.
1867    
1868    5.  In pcretest, an escape such as \x{de} in the data was always generating
1869        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1870        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1871        truncation.
1872    
1873    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1874    
1875    7.  Added two (int) casts to pcregrep when printing the difference of two
1876        pointers, in case they are 64-bit values.
1877    
1878    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1879        test 2 if it fails.
1880    
1881    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1882        and a #define of that name to empty if it is not externally set. This is to
1883        allow users of MSVC to set it if necessary.
1884    
1885    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1886        the convenience functions in the pcre_get.c source file.
1887    
1888    11. An option change at the start of a pattern that had top-level alternatives
1889        could cause overwriting and/or a crash. This command provoked a crash in
1890        some environments:
1891    
1892          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1893    
1894        This potential security problem was recorded as CVE-2008-2371.
1895    
1896    12. For a pattern where the match had to start at the beginning or immediately
1897        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1898        pcre_dfa_exec() could read past the end of the passed subject if there was
1899        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1900        pcretest so that it places the subject at the end of its malloc-ed buffer.
1901    
1902    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1903        exec() might read past the end of the data buffer in UTF-8 mode.
1904    
1905    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1906        the data contained the byte 0x85 as part of a UTF-8 character within its
1907        first line. This applied both to normal and DFA matching.
1908    
1909    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1910        /^[^d]*?$/8 failed to match "abc".
1911    
1912    16. Added a missing copyright notice to pcrecpp_internal.h.
1913    
1914    17. Make it more clear in the documentation that values returned from
1915        pcre_exec() in ovector are byte offsets, not character counts.
1916    
1917    18. Tidied a few places to stop certain compilers from issuing warnings.
1918    
1919    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1920        supplied by Stefan Weber. I made a further small update for 7.8 because
1921        there is a change of source arrangements: the pcre_searchfuncs.c module is
1922        replaced by pcre_ucd.c.
1923    
1924    
1925    Version 7.7 07-May-08
1926    ---------------------
1927    
1928    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1929        a string to a long long, pretend we don't even have a long long." This is
1930        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1931    
1932    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1933        pre-7.6 versions, which defined a global no_arg variable instead of putting
1934        it in the RE class. (See also #8 below.)
1935    
1936    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1937        Lopes.
1938    
1939    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1940    
1941        (1) The include/exclude patterns were being applied to the whole pathnames
1942            of files, instead of just to the final components.
1943    
1944        (2) If there was more than one level of directory, the subdirectories were
1945            skipped unless they satisfied the include/exclude conditions. This is
1946            inconsistent with GNU grep (and could even be seen as contrary to the
1947            pcregrep specification - which I improved to make it absolutely clear).
1948            The action now is always to scan all levels of directory, and just
1949            apply the include/exclude patterns to regular files.
1950    
1951    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1952        --exclude_dir in the tests to avoid scanning .svn directories.
1953    
1954    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1955        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1956        doesn't support NULs in patterns.
1957    
1958    7.  Added some missing "const"s to declarations of static tables in
1959        pcre_compile.c and pcre_dfa_exec.c.
1960    
1961    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1962        caused by fix #2  above. (Subsequently also a second patch to fix the
1963        first patch. And a third patch - this was a messy problem.)
1964    
1965    9.  Applied Craig's patch to remove the use of push_back().
1966    
1967    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1968        matching function regexec().
1969    
1970    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1971        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1972        references. PCRE supports relative numbers with this syntax (I don't think
1973        Oniguruma does).
1974    
1975    12. Previously, a group with a zero repeat such as (...){0} was completely
1976        omitted from the compiled regex. However, this means that if the group
1977        was called as a subroutine from elsewhere in the pattern, things went wrong
1978        (an internal error was given). Such groups are now left in the compiled
1979        pattern, with a new opcode that causes them to be skipped at execution
1980        time.
1981    
1982    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1983        to the way PCRE behaves:
1984    
1985        (a) A lone ] character is dis-allowed (Perl treats it as data).
1986    
1987        (b) A back reference to an unmatched subpattern matches an empty string
1988            (Perl fails the current match path).
1989    
1990        (c) A data ] in a character class must be notated as \] because if the
1991            first data character in a class is ], it defines an empty class. (In
1992            Perl it is not possible to have an empty class.) The empty class []
1993            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1994            The negative empty class [^] matches any one character, independently
1995            of the DOTALL setting.
1996    
1997    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1998        non-existent subpattern following a character class starting with ']' and
1999        containing () gave an internal compiling error instead of "reference to
2000        non-existent subpattern". Fortunately, when the pattern did exist, the
2001        compiled code was correct. (When scanning forwards to check for the
2002        existence of the subpattern, it was treating the data ']' as terminating
2003        the class, so got the count wrong. When actually compiling, the reference
2004        was subsequently set up correctly.)
2005    
2006    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2007        it was being rejected as not supported by pcre_dfa_exec(), even though
2008        other assertions are supported. I have made pcre_dfa_exec() support
2009        (*FAIL).
2010    
2011    16. The implementation of 13c above involved the invention of a new opcode,
2012        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2013        cannot be changed at match time, I realized I could make a small
2014        improvement to matching performance by compiling OP_ALLANY instead of
2015        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2016        on the OP_ANY path.
2017    
2018    17. Compiling pcretest on Windows with readline support failed without the
2019        following two fixes: (1) Make the unistd.h include conditional on
2020        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2021    
2022    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2023        ncurses library to be included for pcretest when ReadLine support is
2024        requested, but also to allow for it to be overridden. This patch came from
2025        Daniel Bergström.
2026    
2027    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2028        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2029        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2030        spotting this.
2031    
2032    
2033    Version 7.6 28-Jan-08
2034    ---------------------
2035    
2036    1.  A character class containing a very large number of characters with
2037        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2038        overflow.
2039    
2040    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2041        HAVE_LONG_LONG is not defined.
2042    
2043    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2044        bring them up to date and include new features. This patch includes:
2045    
2046        - Fixed PH's badly added libz and libbz2 support.
2047        - Fixed a problem with static linking.
2048        - Added pcredemo. [But later removed - see 7 below.]
2049        - Fixed dftables problem and added an option.
2050        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2051            HAVE_LONG_LONG.
2052        - Added readline support for pcretest.
2053        - Added an listing of the option settings after cmake has run.
2054    
2055    4.  A user submitted a patch to Makefile that makes it easy to create
2056        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2057        Makefile.am that cause it to include this special target, without
2058        affecting anything else. Note that the same mingw target plus all
2059        the other distribution libraries and programs are now supported
2060        when configuring with CMake (see 6 below) instead of with
2061        Configure/Make.
2062    
2063    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2064        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2065        exported in the Windows port". It has not yet been confirmed that the patch
2066        solves the problem, but it does no harm.
2067    
2068    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2069        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2070        with CMake, and also correct the comment about stack recursion.
2071    
2072    7.  Remove the automatic building of pcredemo from the ./configure system and
2073        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2074        of a program that users should build themselves after PCRE is installed, so
2075        building it automatically is not really right. What is more, it gave
2076        trouble in some build environments.
2077    
2078    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2079    
2080    
2081    Version 7.5 10-Jan-08
2082    ---------------------
2083    
2084    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2085        values in parens when parsing an RE using the C++ wrapper."
2086    
2087    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2088        Characters greater than 255 were excluded from the class instead of being
2089        included.
2090    
2091    3.  The same bug as (2) above applied to negated POSIX classes such as
2092        [:^space:].
2093    
2094    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2095        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2096        I have changed it.
2097    
2098    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2099        first named subpattern) and a construct such as (?&a) would reference the
2100        first named subpattern whose name started with "a" (in other words, the
2101        length check was missing). Both these problems are fixed. "Subpattern name
2102        expected" is now given for (?&) (a zero-length name), and this patch also
2103        makes it give the same error for \k'' (previously it complained that that
2104        was a reference to a non-existent subpattern).
2105    
2106    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2107        this is right because (?- can be followed by option settings as well as by
2108        digits. I have, however, made the messages clearer.
2109    
2110    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2111        than the number used in the conditional) now cause a compile-time error.
2112        This is actually not compatible with Perl, which accepts such patterns, but
2113        treats the conditional as always being FALSE (as PCRE used to), but it
2114        seems to me that giving a diagnostic is better.
2115    
2116    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2117        and messages.
2118    
2119    9.  Fix two occurrences of "backslash" in comments that should have been
2120        "backspace".
2121    
2122    10. Remove two redundant lines of code that can never be obeyed (their function
2123        was moved elsewhere).
2124    
2125    11. The program that makes PCRE's Unicode character property table had a bug
2126        which caused it to generate incorrect table entries for sequences of
2127        characters that have the same character type, but are in different scripts.
2128        It amalgamated them into a single range, with the script of the first of
2129        them. In other words, some characters were in the wrong script. There were
2130        thirteen such cases, affecting characters in the following ranges:
2131    
2132          U+002b0 - U+002c1
2133          U+0060c - U+0060d
2134          U+0061e - U+00612
2135          U+0064b - U+0065e
2136          U+0074d - U+0076d
2137          U+01800 - U+01805
2138          U+01d00 - U+01d77
2139          U+01d9b - U+01dbf
2140          U+0200b - U+0200f
2141          U+030fc - U+030fe
2142          U+03260 - U+0327f
2143          U+0fb46 - U+0fbb1
2144          U+10450 - U+1049d
2145    
2146    12. The -o option (show only the matching part of a line) for pcregrep was not
2147        compatible with GNU grep in that, if there was more than one match in a
2148        line, it showed only the first of them. It now behaves in the same way as
2149        GNU grep.
2150    
2151    13. If the -o and -v options were combined for pcregrep, it printed a blank
2152        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2153        does the same. The return code can be used to tell if there were any
2154        non-matching lines.
2155    
2156    14. Added --file-offsets and --line-offsets to pcregrep.
2157    
2158    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2159        infinitely looping recursion. The bug was that positive lookaheads were not
2160        being skipped when checking for a possible empty match (negative lookaheads
2161        and both kinds of lookbehind were skipped).
2162    
2163    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2164        inclusion of <windows.h> to before rather than after the definition of
2165        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2166    
2167    17. Specifying a possessive quantifier with a specific limit for a Unicode
2168        character property caused pcre_compile() to compile bad code, which led at
2169        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2170        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2171        caused the error; without that there was no problem.
2172    
2173    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2174    
2175    19. Added --enable-pcretest-libreadline.
2176    
2177    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2178        RE::GlobalReplace(). As a result, the number of replacements returned was
2179        double what it should be. I removed one of the increments, but Craig sent a
2180        later patch that removed the other one (the right fix) and added unit tests
2181        that check the return values (which was not done before).
2182    
2183    21. Several CMake things:
2184    
2185        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2186            the names libpcre and libpcreposix, not just pcre and pcreposix.
2187    
2188        (2) The above change means that pcretest and pcregrep are now correctly
2189            linked with the newly-built libraries, not previously installed ones.
2190    
2191        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2192    
2193    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2194        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2195        UTF-8 newline character). The key issue is that the pattern starts .*;
2196        this means that the match must be either at the beginning, or after a
2197        newline. The bug was in the code for advancing after a failed match and
2198        checking that the new position followed a newline. It was not taking
2199        account of UTF-8 characters correctly.
2200    
2201    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2202        character classes. PCRE was not treating the sequence [:...:] as a
2203        character class unless the ... were all letters. Perl, however, seems to
2204        allow any characters between [: and :], though of course it rejects as
2205        unknown any "names" that contain non-letters, because all the known class
2206        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2207        for example, whereas PCRE did not - it did not recognize a POSIX character
2208        class. This seemed a bit dangerous, so the code has been changed to be
2209        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2210        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2211        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2212        Perl does, and where it didn't before.
2213    
2214    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2215        Windows environments %n is disabled by default.
2216    
2217    
2218    Version 7.4 21-Sep-07
2219    ---------------------
2220    
2221    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2222        means that a class such as [\s] counted as "explicit reference to CR or
2223        LF". That isn't really right - the whole point of the change was to try to
2224        help when there was an actual mention of one of the two characters. So now
2225        the change happens only if \r or \n (or a literal CR or LF) character is
2226        encountered.
2227    
2228    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2229        of both had grown to the point where there were only 3 bits left.
2230        Fortunately, there was spare space in the data structure, and so I have
2231        moved the internal flags into a new 16-bit field to free up more option
2232        bits.
2233    
2234    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2235        but did not set the internal JCHANGED flag - either of these is enough to
2236        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2237        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2238        start sets both bits.
2239    
2240    4.  Added options (at build time, compile time, exec time) to change \R from
2241        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2242    
2243    5.  doc/pcresyntax.html was missing from the distribution.
2244    
2245    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2246        compatibility, even though it is no longer used.
2247    
2248    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2249        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2250        windows.h file is present (where different names are used). [This was
2251        reversed later after testing - see 16 below.]
2252    
2253    8.  Changed all #include <config.h> to #include "config.h". There were also
2254        some further <pcre.h> cases that I changed to "pcre.h".
2255    
2256    9.  When pcregrep was used with the --colour option, it missed the line ending
2257        sequence off the lines that it output.
2258    
2259    10. It was pointed out to me that arrays of string pointers cause lots of
2260        relocations when a shared library is dynamically loaded. A technique of
2261        using a single long string with a table of offsets can drastically reduce
2262        these. I have refactored PCRE in four places to do this. The result is
2263        dramatic:
2264    
2265          Originally:                          290
2266          After changing UCP table:            187
2267          After changing error message table:   43
2268          After changing table of "verbs"       36
2269          After changing table of Posix names   22
2270    
2271        Thanks to the folks working on Gregex for glib for this insight.
2272    
2273    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2274        unicode-properties was also set.
2275    
2276    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2277    
2278    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2279        checked only for CRLF.
2280    
2281    14. Added casts to pcretest.c to avoid compiler warnings.
2282    
2283    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2284    
2285    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2286        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2287        entirely. This removes changes made in 7 above.
2288    
2289    17. The CMake files have been updated, and there is now more information about
2290        building with CMake in the NON-UNIX-USE document.
2291    
2292    
2293    Version 7.3 28-Aug-07
2294    ---------------------
2295    
2296     1. In the rejigging of the build system that eventually resulted in 7.1, the
2297        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2298        brackets there is not right, since it causes compilers to look for an
2299        installed pcre.h, not the version that is in the source that is being
2300        compiled (which of course may be different). I have changed it back to:
2301    
2302          #include "pcre.h"
2303    
2304        I have a vague recollection that the change was concerned with compiling in
2305        different directories, but in the new build system, that is taken care of
2306        by the VPATH setting the Makefile.
2307    
2308     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2309        when the subject happened to end in the byte 0x85 (e.g. if the last
2310        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2311        characters but of course it shouldn't be taken as a newline when it is part
2312        of another character. The bug was that, for an unlimited repeat of . in
2313        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2314        characters when looking for a newline.
2315    
2316     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2317    
2318     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2319        in debug output.
2320    
2321     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2322        long printing in the pcrecpp unittest when running under MinGW.
2323    
2324     6. ESC_K was left out of the EBCDIC table.
2325    
2326     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2327        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2328        limit also applies to "virtual nesting" when a pattern is recursive, and in
2329        this case 1000 isn't so big. I have been able to remove this limit at the
2330        expense of backing off one optimization in certain circumstances. Normally,
2331        when pcre_exec() would call its internal match() function recursively and
2332        immediately return the result unconditionally, it uses a "tail recursion"
2333        feature to save stack. However, when a subpattern that can match an empty
2334        string has an unlimited repetition quantifier, it no longer makes this
2335        optimization. That gives it a stack frame in which to save the data for
2336        checking that an empty string has been matched. Previously this was taken
2337        from the 1000-entry workspace that had been reserved. So now there is no
2338        explicit limit, but more stack is used.
2339    
2340     8. Applied Daniel's patches to solve problems with the import/export magic
2341        syntax that is required for Windows, and which was going wrong for the
2342        pcreposix and pcrecpp parts of the library. These were overlooked when this
2343        problem was solved for the main library.
2344    
2345     9. There were some crude static tests to avoid integer overflow when computing
2346        the size of patterns that contain repeated groups with explicit upper
2347        limits. As the maximum quantifier is 65535, the maximum group length was
2348        set at 30,000 so that the product of these two numbers did not overflow a
2349        32-bit integer. However, it turns out that people want to use groups that
2350        are longer than 30,000 bytes (though not repeat them that many times).
2351        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2352        made it possible to implement the integer overflow checks in a much more
2353        dynamic way, which I have now done. The artificial limitation on group
2354        length has been removed - we now have only the limit on the total length of
2355        the compiled pattern, which depends on the LINK_SIZE setting.
2356    
2357    10. Fixed a bug in the documentation for get/copy named substring when
2358        duplicate names are permitted. If none of the named substrings are set, the
2359        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2360        empty string.
2361    
2362    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2363        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2364        because the ] is interpreted as the first data character and the
2365        terminating ] is not found. PCRE has been made compatible with Perl in this
2366        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2367        cause memory overwriting.
2368    
2369    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2370        string has been matched (to stop an infinite loop). It was not recognizing
2371        a conditional subpattern that could match an empty string if that
2372        subpattern was within another subpattern. For example, it looped when
2373        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2374        condition was not nested. This bug has been fixed.
2375    
2376    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2377        past the start of the subject in the presence of bytes with the top bit
2378        set, for example "\x8aBCD".
2379    
2380    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2381        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2382    
2383    14. Optimized (?!) to (*FAIL).
2384    
2385    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2386        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2387        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2388        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2389        does: it's just the validity check that is more restrictive.
2390    
2391    16. Inserted checks for integer overflows during escape sequence (backslash)
2392        processing, and also fixed erroneous offset values for syntax errors during
2393        backslash processing.
2394    
2395    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2396        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2397    
2398    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2399        caused an overrun.
2400    
2401    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2402        something other than just ASCII characters) inside a group that had an
2403        unlimited repeat caused a loop at compile time (while checking to see
2404        whether the group could match an empty string).
2405    
2406    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2407        [\P{Any}] did so. (Error in the code for printing property names.)
2408    
2409    21. An orphan \E inside a character class could cause a crash.
2410    
2411    22. A repeated capturing bracket such as (A)? could cause a wild memory
2412        reference during compilation.
2413    
2414    23. There are several functions in pcre_compile() that scan along a compiled
2415        expression for various reasons (e.g. to see if it's fixed length for look
2416        behind). There were bugs in these functions when a repeated \p or \P was
2417        present in the pattern. These operators have additional parameters compared
2418        with \d, etc, and these were not being taken into account when moving along
2419        the compiled data. Specifically:
2420    
2421        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2422            length.
2423    
2424        (b) An item such as \pL+ within a repeated group could cause crashes or
2425            loops.
2426    
2427        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2428            "reference to non-existent subpattern" error.
2429    
2430        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2431    
2432    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2433        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2434    
2435    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2436    
2437    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2438        character were causing crashes (broken optimization).
2439    
2440    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2441        \p or \P) caused a compile-time loop.
2442    
2443    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2444        break. For example, the unstudied pattern [\r\n]A does not match the string
2445        "\r\nA" because change 7.0/46 below moves the current point on by two
2446        characters after failing to match at the start. However, the pattern \nA
2447        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2448        the same is true. There doesn't seem any very clean way out of this, but
2449        what I have chosen to do makes the common cases work: PCRE now takes note
2450        of whether there can be an explicit match for \r or \n anywhere in the
2451        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2452        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2453        pattern has explicit CR or LF references.
2454    
2455    29. Added (*CR) etc for changing newline setting at start of pattern.
2456    
2457    
2458    Version 7.2 19-Jun-07
2459    ---------------------
2460    
2461     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2462        which is apparently normally available under Windows.
2463    
2464     2. Re-jig the pcregrep tests with different newline settings in an attempt
2465        to make them independent of the local environment's newline setting.
2466    
2467     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2468    
2469     4. Some of the "internals" tests were previously cut out when the link size
2470        was not 2, because the output contained actual offsets. The recent new
2471        "Z" feature of pcretest means that these can be cut out, making the tests
2472        usable with all link sizes.
2473    
2474     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2475        stack recursion. This gives a massive performance boost under BSD, but just
2476        a small improvement under Linux. However, it saves one field in the frame
2477        in all cases.
2478    
2479     6. Added more features from the forthcoming Perl 5.10:
2480    
2481        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2482            recursion call. It refers to the nth most recently opened parentheses.
2483    
2484        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2485            to be opened parentheses.
2486    
2487        (c) Conditions that refer to capturing parentheses can be specified
2488            relatively, for example, (?(-2)... or (?(+3)...
2489    
2490        (d) \K resets the start of the current match so that everything before
2491            is not part of it.
2492    
2493        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2494    
2495        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2496            reference syntax.
2497    
2498        (g) (?| introduces a group in which the numbering of parentheses in each
2499            alternative starts with the same number.
2500    
2501        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2502    
2503     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2504        PCRE_INFO_JCHANGED.
2505    
2506     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2507        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2508        for detecting groups that can match an empty string.
2509    
2510     9. A pattern with a very large number of alternatives (more than several
2511        hundred) was running out of internal workspace during the pre-compile
2512        phase, where pcre_compile() figures out how much memory will be needed. A
2513        bit of new cunning has reduced the workspace needed for groups with
2514        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2515        workspace instead of running out of the 4096 that are available.
2516    
2517    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2518    
2519    11. Applied patch from Google to remove an optimization that didn't quite work.
2520        The report of the bug said:
2521    
2522          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2523          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2524          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2525    
2526    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2527        it matched the wrong number of bytes.
2528    
2529    
2530    Version 7.1 24-Apr-07
2531  ---------------------  ---------------------
2532    
2533   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2534      that is more "standard", making use of automake and other autotools.      that is more "standard", making use of automake and other Autotools. There
2535        is some re-arrangement of the files and adjustment of comments consequent
2536        on this.
2537    
2538     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2539        for recursive directory scanning broke on some systems because the files
2540        are not scanned in any specific order and on different systems the order
2541        was different. A call to "sort" has been inserted into RunGrepTest for the
2542        approprate test as a short-term fix. In the longer term there may be an
2543        alternative.
2544    
2545     3. I had an email from Eric Raymond about problems translating some of PCRE's
2546        man pages to HTML (despite the fact that I distribute HTML pages, some
2547        people do their own conversions for various reasons). The problems
2548        concerned the use of low-level troff macros .br and .in. I have therefore
2549        removed all such uses from the man pages (some were redundant, some could
2550        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2551        HTML has been updated to handle .nf/.fi and to complain if it encounters
2552        .br or .in.
2553    
2554     4. Updated comments in configure.ac that get placed in config.h.in and also
2555        arranged for config.h to be included in the distribution, with the name
2556        config.h.generic, for the benefit of those who have to compile without
2557        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2558    
2559     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2560        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2561        makevp.bat for the current PCRE, using the additional files
2562        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2563    
2564     6. A Windows user reported a minor discrepancy with test 2, which turned out
2565        to be caused by a trailing space on an input line that had got lost in his
2566        copy. The trailing space was an accident, so I've just removed it.
2567    
2568     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2569        that is needed.
2570    
2571     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2572        as "const" (a) because they are and (b) because it helps the PHP
2573        maintainers who have recently made a script to detect big data structures
2574        in the php code that should be moved to the .rodata section. I remembered
2575        to update Builducptable as well, so it won't revert if ucptable.h is ever
2576        re-created.
2577    
2578     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2579        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2580        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2581        support is not required. This saves 1.5-2K of code, which is important in
2582        some applications.
2583    
2584        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2585        so as not to refer to the tables, even though these functions will never be
2586        called when UTF-8 support is disabled. Otherwise there are problems with a
2587        shared library.
2588    
2589    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2590    
2591        (a) It was defining its arguments as char * instead of void *.
2592    
2593        (b) It was assuming that all moves were upwards in memory; this was true
2594            a long time ago when I wrote it, but is no longer the case.
2595    
2596        The emulated memove() is provided for those environments that have neither
2597        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2598        is clearly not the case, as these two bugs were recently reported.
2599    
2600    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2601        and Detrail to create the HTML documentation, the .txt form of the man
2602        pages, and it removes trailing spaces from listed files. It also creates
2603        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2604        case, it wraps all the #defines with #ifndefs. This script should be run
2605        before "make dist".
2606    
2607    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2608        with Unicode property support.
2609    
2610        (a) For a maximizing quantifier, if the two different cases of the
2611            character were of different lengths in their UTF-8 codings (there are
2612            some cases like this - I found 11), and the matching function had to
2613            back up over a mixture of the two cases, it incorrectly assumed they
2614            were both the same length.
2615    
2616        (b) When PCRE was configured to use the heap rather than the stack for
2617            recursion during matching, it was not correctly preserving the data for
2618            the other case of a UTF-8 character when checking ahead for a match
2619            while processing a minimizing repeat. If the check also involved
2620            matching a wide character, but failed, corruption could cause an
2621            erroneous result when trying to check for a repeat of the original
2622            character.
2623    
2624    13. Some tidying changes to the testing mechanism:
2625    
2626        (a) The RunTest script now detects the internal link size and whether there
2627            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2628            values substituted by "configure". (The RunGrepTest script already did
2629            this for UTF-8.) The configure.ac script no longer substitutes the
2630            relevant variables.
2631    
2632        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2633            with length and offset values. This means that the output is different
2634            for different internal link sizes. Test 2 is skipped for link sizes
2635            other than 2 because of this, bypassing the problem. Unfortunately,
2636            there was also a test in test 3 (the locale tests) that used /B and
2637            failed for link sizes other than 2. Rather than cut the whole test out,
2638            I have added a new /Z option to pcretest that replaces the length and
2639            offset values with spaces. This is now used to make test 3 independent
2640            of link size. (Test 2 will be tidied up later.)
2641    
2642    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2643        segmentation fault instead of returning the appropriate error message.
2644    
2645    15. In multiline mode when the newline sequence was set to "any", the pattern
2646        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2647        This doesn't seem right; it now treats the CRLF combination as the line
2648        ending, and so does not match in that case. It's only a pattern such as ^$
2649        that would hit this one: something like ^ABC$ would have failed after \r
2650        and then tried again after \r\n.
2651    
2652    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2653        in an attempt to make files that differ only in their line terminators
2654        compare equal. This works on Linux.
2655    
2656    17. Under certain error circumstances pcregrep might try to free random memory
2657        as it exited. This is now fixed, thanks to valgrind.
2658    
2659    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2660        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2661        was because its rules for how to advance for /g after matching an empty
2662        string at the end of a line did not allow for this case. They now check for
2663        it specially.
2664    
2665    20. pcretest is supposed to handle patterns and data of any length, by
2666        extending its buffers when necessary. It was getting this wrong when the
2667        buffer for a data line had to be extended.
2668    
2669    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2670        CRLF as a newline sequence.
2671    
2672    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2673        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2674        I have nevertheless tidied it up.
2675    
2676    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2677    
2678    24. Added a man page for pcre-config.
2679    
2680    
2681  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.98  
changed lines
  Added in v.1361

  ViewVC Help
Powered by ViewVC 1.1.5