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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.98  
changed lines
  Added in v.1115

  ViewVC Help
Powered by ViewVC 1.1.5