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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.108  
changed lines
  Added in v.1196

  ViewVC Help
Powered by ViewVC 1.1.5