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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.153  
changed lines
  Added in v.1053

  ViewVC Help
Powered by ViewVC 1.1.5