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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.200  
changed lines
  Added in v.1327

  ViewVC Help
Powered by ViewVC 1.1.5