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

Diff of /code/trunk/ChangeLog

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

revision 65 by nigel, Sat Feb 24 21:40:08 2007 UTC revision 1322 by zherczeg, Wed May 1 19:42:31 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    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    
151    Version 8.32 30-November-2012
152    -----------------------------
153    
154    1.  Improved JIT compiler optimizations for first character search and single
155        character iterators.
156    
157    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
158        Patch by Daniel Richard G.
159    
160    3.  Single character iterator optimizations in the JIT compiler.
161    
162    4.  Improved JIT compiler optimizations for character ranges.
163    
164    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
165        Reported by Giuseppe D'Angelo.
166    
167    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
168        of a line, was being set incorrectly in cases where .* appeared inside
169        atomic brackets at the start of a pattern, or where there was a subsequent
170        *PRUNE or *SKIP.
171    
172    7.  Improved instruction cache flush for POWER/PowerPC.
173        Patch by Daniel Richard G.
174    
175    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
176        grep:
177    
178        (a) There is now no limit to the number of patterns to be matched.
179    
180        (b) An error is given if a pattern is too long.
181    
182        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
183            are now supported.
184    
185        (d) --exclude-from and --include-from (multiple use) have been added.
186    
187        (e) Exclusions and inclusions now apply to all files and directories, not
188            just to those obtained from scanning a directory recursively.
189    
190        (f) Multiple uses of -f and --file-list are now supported.
191    
192        (g) In a Windows environment, the default for -d has been changed from
193            "read" (the GNU grep default) to "skip", because otherwise the presence
194            of a directory in the file list provokes an error.
195    
196        (h) The documentation has been revised and clarified in places.
197    
198    9.  Improve the matching speed of capturing brackets.
199    
200    10. Changed the meaning of \X so that it now matches a Unicode extended
201        grapheme cluster.
202    
203    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
204        out POSIX threads when JIT support is configured.
205    
206    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
207    
208    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
209        structure could go wrong in environments where size_t is not the same size
210        as int.
211    
212    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
213        set.
214    
215    15. The EBCDIC support had decayed; later updates to the code had included
216        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
217        general tidy up of EBCDIC-related issues, and the documentation was also
218        not quite right. There is now a test that can be run on ASCII systems to
219        check some of the EBCDIC-related things (but is it not a full test).
220    
221    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
222        in a small tidy to the code.
223    
224    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
225    
226    18. If the --only-matching (-o) option in pcregrep is specified multiple
227        times, each one causes appropriate output. For example, -o1 -o2 outputs the
228        substrings matched by the 1st and 2nd capturing parentheses. A separating
229        string can be specified by --om-separator (default empty).
230    
231    19. Improving the first n character searches.
232    
233    20. Turn case lists for horizontal and vertical white space into macros so that
234        they are defined only once.
235    
236    21. This set of changes together give more compatible Unicode case-folding
237        behaviour for characters that have more than one other case when UCP
238        support is available.
239    
240        (a) The Unicode property table now has offsets into a new table of sets of
241            three or more characters that are case-equivalent. The MultiStage2.py
242            script that generates these tables (the pcre_ucd.c file) now scans
243            CaseFolding.txt instead of UnicodeData.txt for character case
244            information.
245    
246        (b) The code for adding characters or ranges of characters to a character
247            class has been abstracted into a generalized function that also handles
248            case-independence. In UTF-mode with UCP support, this uses the new data
249            to handle characters with more than one other case.
250    
251        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
252            whose other case is greater than 256 are now correctly matched
253            caselessly. Previously, the high codepoint matched the low one, but not
254            vice versa.
255    
256        (d) The processing of \h, \H, \v, and \ in character classes now makes use
257            of the new class addition function, using character lists defined as
258            macros alongside the case definitions of 20 above.
259    
260        (e) Caseless back references now work with characters that have more than
261            one other case.
262    
263        (f) General caseless matching of characters with more than one other case
264            is supported.
265    
266    22. Unicode character properties were updated from Unicode 6.2.0
267    
268    23. Improved CMake support under Windows. Patch by Daniel Richard G.
269    
270    24. Add support for 32-bit character strings, and UTF-32
271    
272    25. Major JIT compiler update (code refactoring and bugfixing).
273        Experimental Sparc 32 support is added.
274    
275    26. Applied a modified version of Daniel Richard G's patch to create
276        pcre.h.generic and config.h.generic by "make" instead of in the
277        PrepareRelease script.
278    
279    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
280        pcre_compile.c when checking for a zero character.
281    
282    28. Introducing a native interface for JIT. Through this interface, the compiled
283        machine code can be directly executed. The purpose of this interface is to
284        provide fast pattern matching, so several sanity checks are not performed.
285        However, feature tests are still performed. The new interface provides
286        1.4x speedup compared to the old one.
287    
288    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
289        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
290        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
291    
292    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
293        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
294        the "old" RFC 2279). Instead, it ended up passing a negative length to
295        pcre_exec().
296    
297    31. Add support for GCC's visibility feature to hide internal functions.
298    
299    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
300        "unknown -C option" after outputting 0 or 1.
301    
302    33. There is now support for generating a code coverage report for the test
303        suite in environments where gcc is the compiler and lcov is installed. This
304        is mainly for the benefit of the developers.
305    
306    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
307        unaddressable using valgrind annotations, allowing valgrind to detect
308        invalid memory accesses. This is mainly for the benefit of the developers.
309    
310    25. (*UTF) can now be used to start a pattern in any of the three libraries.
311    
312    26. Give configure error if --enable-cpp but no C++ compiler found.
313    
314    
315    Version 8.31 06-July-2012
316    -------------------------
317    
318    1.  Fixing a wrong JIT test case and some compiler warnings.
319    
320    2.  Removed a bashism from the RunTest script.
321    
322    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
323        to unsigned type, result still unsigned" that was given by an MS compiler
324        on encountering the code "-sizeof(xxx)".
325    
326    4.  Partial matching support is added to the JIT compiler.
327    
328    5.  Fixed several bugs concerned with partial matching of items that consist
329        of more than one character:
330    
331        (a) /^(..)\1/ did not partially match "aba" because checking references was
332            done on an "all or nothing" basis. This also applied to repeated
333            references.
334    
335        (b) \R did not give a hard partial match if \r was found at the end of the
336            subject.
337    
338        (c) \X did not give a hard partial match after matching one or more
339            characters at the end of the subject.
340    
341        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
342            a partial match for the string "\r".
343    
344        (e) When newline was set to CRLF, the metacharacter "." did not recognize
345            a partial match for a CR character at the end of the subject string.
346    
347    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
348        running pcretest, the text "(JIT)" added to the output whenever JIT is
349        actually used to run the match.
350    
351    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
352        or /S+[+] with a digit between 1 and 7.
353    
354    8.  OP_NOT now supports any UTF character not just single-byte ones.
355    
356    9.  (*MARK) control verb is now supported by the JIT compiler.
357    
358    10. The command "./RunTest list" lists the available tests without actually
359        running any of them. (Because I keep forgetting what they all are.)
360    
361    11. Add PCRE_INFO_MAXLOOKBEHIND.
362    
363    12. Applied a (slightly modified) user-supplied patch that improves performance
364        when the heap is used for recursion (compiled with --disable-stack-for-
365        recursion). Instead of malloc and free for each heap frame each time a
366        logical recursion happens, frames are retained on a chain and re-used where
367        possible. This sometimes gives as much as 30% improvement.
368    
369    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
370        call.
371    
372    14. As documented, (*COMMIT) is now confined to within a positive assertion.
373    
374    15. It is now possible to link pcretest with libedit as an alternative to
375        libreadline.
376    
377    16. (*COMMIT) control verb is now supported by the JIT compiler.
378    
379    17. The Unicode data tables have been updated to Unicode 6.1.0.
380    
381    18. Added --file-list option to pcregrep.
382    
383    19. Added binary file support to pcregrep, including the -a, --binary-files,
384        -I, and --text options.
385    
386    20. The madvise function is renamed for posix_madvise for QNX compatibility
387        reasons. Fixed by Giuseppe D'Angelo.
388    
389    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
390    
391    22. Changed the option for creating symbolic links for 16-bit man pages from
392        -s to -sf so that re-installing does not cause issues.
393    
394    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
395    
396    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
397        matches in certain environments (the workspace was not being correctly
398        retained). Also added to pcre_dfa_exec() a simple plausibility check on
399        some of the workspace data at the beginning of a restart.
400    
401    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
402        was not doing so when it should - probably a typo introduced by SVN 528
403        (change 8.10/14).
404    
405    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
406        \w+ when the character tables indicated that \x{c4} was a word character.
407        There were several related cases, all because the tests for doing a table
408        lookup were testing for characters less than 127 instead of 255.
409    
410    27. If a pattern contains capturing parentheses that are not used in a match,
411        their slots in the ovector are set to -1. For those that are higher than
412        any matched groups, this happens at the end of processing. In the case when
413        there were back references that the ovector was too small to contain
414        (causing temporary malloc'd memory to be used during matching), and the
415        highest capturing number was not used, memory off the end of the ovector
416        was incorrectly being set to -1. (It was using the size of the temporary
417        memory instead of the true size.)
418    
419    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
420        ovector size, it uses memory at the end of the block that it has got.
421    
422    29. Check for an overlong MARK name and give an error at compile time. The
423        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
424    
425    30. JIT compiler update.
426    
427    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
428        Rill for the patch.
429    
430    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
431    
432    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
433    
434    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
435        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
436        was enabled.
437    
438    35. Improve JIT code generation for greedy plus quantifier.
439    
440    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
441        1 to "aa" instead of to an empty string. The bug affected repeated groups
442        that could potentially match an empty string.
443    
444    37. Optimizing single character iterators in JIT.
445    
446    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
447        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
448        codepoints that are too big for the mode are faulted, and in a UTF mode,
449        disallowed codepoints are also faulted.
450    
451    39. If PCRE was compiled with UTF support, in three places in the DFA
452        matcher there was code that should only have been obeyed in UTF mode, but
453        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
454        processing when bytes with values greater than 127 were present. In 16-bit
455        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
456        both cases the values are those that cannot be the first data item in a UTF
457        character. The three items that might have provoked this were recursions,
458        possessively repeated groups, and atomic groups.
459    
460    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
461        and pcregrep, because some OS require shared objects to be explicitly
462        passed to ld, causing the link step to fail if they are not.
463    
464    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
465        mode, patterns that started with \h* or \R* might be incorrectly matched.
466    
467    
468    Version 8.30 04-February-2012
469    -----------------------------
470    
471    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
472        name is defined in ctype.h.
473    
474    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
475        only in quite long subpatterns.
476    
477    3.  Removed the function pcre_info(), which has been obsolete and deprecated
478        since it was replaced by pcre_fullinfo() in February 2000.
479    
480    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
481        match a (*MARK), and the match failed at the start of the subject, a
482        reference to memory before the start of the subject could occur. This bug
483        was introduced by fix 17 of release 8.21.
484    
485    5.  A reference to an unset group with zero minimum repetition was giving
486        totally wrong answers (in non-JavaScript-compatibility mode). For example,
487        /(another)?(\1?)test/ matched against "hello world test". This bug was
488        introduced in release 8.13.
489    
490    6.  Add support for 16-bit character strings (a large amount of work involving
491        many changes and refactorings).
492    
493    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
494        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
495        from a file.
496    
497    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
498        rounding is not applied in this particular case).
499    
500    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
501        if they appear, or are escaped, in patterns.
502    
503    10. Get rid of a number of -Wunused-but-set-variable warnings.
504    
505    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
506        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
507        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
508        also returns the mark "x". This bug applied to capturing parentheses,
509        non-capturing parentheses, and atomic parentheses. It also applied to some
510        assertions.
511    
512    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
513        information out of configure.ac instead of relying on pcre.h.generic, which
514        is not stored in the repository.
515    
516    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
517        -lreadline.
518    
519    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
520    
521    15. Applied Graycode's patch to put the top-level frame on the stack rather
522        than the heap when not using the stack for recursion. This gives a
523        performance improvement in many cases when recursion is not deep.
524    
525    16. Experimental code added to "pcretest -C" to output the stack frame size.
526    
527    
528    Version 8.21 12-Dec-2011
529    ------------------------
530    
531    1.  Updating the JIT compiler.
532    
533    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
534        are added as well.
535    
536    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
537        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
538        calling _pcre_jit_exec. Some extra comments are added.
539    
540    4.  (*MARK) settings inside atomic groups that do not contain any capturing
541        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
542        was introduced by change 18 for 8.20.
543    
544    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
545        ECMA-262 standard.
546    
547    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
548        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
549        This bug was probably introduced by change 9 of 8.13.
550    
551    7.  While fixing 6 above, I noticed that a number of other items were being
552        incorrectly rejected as "not fixed length". This arose partly because newer
553        opcodes had not been added to the fixed-length checking code. I have (a)
554        corrected the bug and added tests for these items, and (b) arranged for an
555        error to occur if an unknown opcode is encountered while checking for fixed
556        length instead of just assuming "not fixed length". The items that were
557        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
558        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
559        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
560    
561    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
562        being incorrectly compiled and would have given unpredicatble results.
563    
564    9.  A possessively repeated subpattern with minimum repeat count greater than
565        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
566        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
567        the first (A) could occur when it should not.
568    
569    10. Add a cast and remove a redundant test from the code.
570    
571    11. JIT should use pcre_malloc/pcre_free for allocation.
572    
573    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
574        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
575        is anything other than /usr, -L is still shown).
576    
577    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
578    
579    14. Perl does not support \N without a following name in a [] class; PCRE now
580        also gives an error.
581    
582    15. If a forward reference was repeated with an upper limit of around 2000,
583        it caused the error "internal error: overran compiling workspace". The
584        maximum number of forward references (including repeats) was limited by the
585        internal workspace, and dependent on the LINK_SIZE. The code has been
586        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
587        the default depends on LINK_SIZE. There is a new upper limit (for safety)
588        of around 200,000 forward references. While doing this, I also speeded up
589        the filling in of repeated forward references.
590    
591    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
592        incorrectly expecting the subject to contain another "a" after the start.
593    
594    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
595        in the match, the SKIP should be ignored. This was not happening; instead
596        the SKIP was being treated as NOMATCH. For patterns such as
597        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
598        tested.
599    
600    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
601        now much more compatible with Perl, in particular in cases where the result
602        is a non-match for a non-anchored pattern. For example, if
603        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
604        "m", where previously it did not return a name. A side effect of this
605        change is that for partial matches, the last encountered mark name is
606        returned, as for non matches. A number of tests that were previously not
607        Perl-compatible have been moved into the Perl-compatible test files. The
608        refactoring has had the pleasing side effect of removing one argument from
609        the match() function, thus reducing its stack requirements.
610    
611    19. If the /S+ option was used in pcretest to study a pattern using JIT,
612        subsequent uses of /S (without +) incorrectly behaved like /S+.
613    
614    21. Retrieve executable code size support for the JIT compiler and fixing
615        some warnings.
616    
617    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
618        not work when the shorter character appeared right at the end of the
619        subject string.
620    
621    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
622        systems.
623    
624    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
625        output it when the /M option is used in pcretest.
626    
627    25. The CheckMan script was not being included in the distribution. Also, added
628        an explicit "perl" to run Perl scripts from the PrepareRelease script
629        because this is reportedly needed in Windows.
630    
631    26. If study data was being save in a file and studying had not found a set of
632        "starts with" bytes for the pattern, the data written to the file (though
633        never used) was taken from uninitialized memory and so caused valgrind to
634        complain.
635    
636    27. Updated RunTest.bat as provided by Sheri Pierce.
637    
638    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
639    
640    29. Computation of memory usage for the table of capturing group names was
641        giving an unnecessarily large value.
642    
643    
644    Version 8.20 21-Oct-2011
645    ------------------------
646    
647    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
648        a POSIX class. After further experiments with Perl, which convinced me that
649        Perl has bugs and confusions, a closing square bracket is no longer allowed
650        in a POSIX name. This bug also affected patterns with classes that started
651        with full stops.
652    
653    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
654        captured substring, but while checking the failing first alternative,
655        substring 1 is temporarily captured. If the output vector supplied to
656        pcre_exec() was not big enough for this capture, the yield of the function
657        was still zero ("insufficient space for captured substrings"). This cannot
658        be totally fixed without adding another stack variable, which seems a lot
659        of expense for a edge case. However, I have improved the situation in cases
660        such as /(a)(b)x|abc/ matched against "abc", where the return code
661        indicates that fewer than the maximum number of slots in the ovector have
662        been set.
663    
664    3.  Related to (2) above: when there are more back references in a pattern than
665        slots in the output vector, pcre_exec() uses temporary memory during
666        matching, and copies in the captures as far as possible afterwards. It was
667        using the entire output vector, but this conflicts with the specification
668        that only 2/3 is used for passing back captured substrings. Now it uses
669        only the first 2/3, for compatibility. This is, of course, another edge
670        case.
671    
672    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
673        main code base, and can be used by building with --enable-jit. When this is
674        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
675        runtime --no-jit option is given.
676    
677    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
678        ovector, the return from the function was zero, implying that there were
679        other matches that did not fit. The correct "exactly full" value is now
680        returned.
681    
682    6.  If a subpattern that was called recursively or as a subroutine contained
683        (*PRUNE) or any other control that caused it to give a non-standard return,
684        invalid errors such as "Error -26 (nested recursion at the same subject
685        position)" or even infinite loops could occur.
686    
687    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
688        computing the minimum length on reaching *ACCEPT, and so ended up with the
689        wrong value of 1 rather than 0. Further investigation indicates that
690        computing a minimum subject length in the presence of *ACCEPT is difficult
691        (think back references, subroutine calls), and so I have changed the code
692        so that no minimum is registered for a pattern that contains *ACCEPT.
693    
694    8.  If (*THEN) was present in the first (true) branch of a conditional group,
695        it was not handled as intended. [But see 16 below.]
696    
697    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
698        Sheri Pierce.
699    
700    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
701        the first byte in a match must be "a".
702    
703    11. Change 17 for 8.13 increased the recursion depth for patterns like
704        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
705        pattern contains any instances of (*THEN). If it does not, the old
706        optimizations are restored. It would be nice to do this on a per-group
707        basis, but at the moment that is not feasible.
708    
709    12. In some environments, the output of pcretest -C is CRLF terminated. This
710        broke RunTest's code that checks for the link size. A single white space
711        character after the value is now allowed for.
712    
713    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
714        For "fr", it uses the Windows-specific input and output files.
715    
716    14. If (*THEN) appeared in a group that was called recursively or as a
717        subroutine, it did not work as intended. [But see next item.]
718    
719    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
720        pattern fragments (but not containing any | characters). If A and B are
721        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
722        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
723        D. In other words, Perl considers parentheses that do not contain any |
724        characters to be part of a surrounding alternative, whereas PCRE was
725        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
726        differently. PCRE now behaves in the same way as Perl, except in the case
727        of subroutine/recursion calls such as (?1) which have in any case always
728        been different (but PCRE had them first :-).
729    
730    16. Related to 15 above: Perl does not treat the | in a conditional group as
731        creating alternatives. Such a group is treated in the same way as an
732        ordinary group without any | characters when processing (*THEN). PCRE has
733        been changed to match Perl's behaviour.
734    
735    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
736        RunGrepTest script failed.
737    
738    18. Change 22 for version 13 caused atomic groups to use more stack. This is
739        inevitable for groups that contain captures, but it can lead to a lot of
740        stack use in large patterns. The old behaviour has been restored for atomic
741        groups that do not contain any capturing parentheses.
742    
743    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
744        suppress the check for a minimum subject length at run time. (If it was
745        given to pcre_exec() or pcre_dfa_exec() it did work.)
746    
747    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
748        fail to work when decoding hex characters in data strings in EBCDIC
749        environments.
750    
751    21. It appears that in at least one Mac OS environment, the isxdigit() function
752        is implemented as a macro that evaluates to its argument more than once,
753        contravening the C 90 Standard (I haven't checked a later standard). There
754        was an instance in pcretest which caused it to go wrong when processing
755        \x{...} escapes in subject strings. The has been rewritten to avoid using
756        things like p++ in the argument of isxdigit().
757    
758    
759    Version 8.13 16-Aug-2011
760    ------------------------
761    
762    1.  The Unicode data tables have been updated to Unicode 6.0.0.
763    
764    2.  Two minor typos in pcre_internal.h have been fixed.
765    
766    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
767        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
768        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
769    
770    4.  There were a number of related bugs in the code for matching backrefences
771        caselessly in UTF-8 mode when codes for the characters concerned were
772        different numbers of bytes. For example, U+023A and U+2C65 are an upper
773        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
774        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
775        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
776        2-byte code at the end of the subject (it thought there wasn't enough data
777        left).
778    
779    5.  Comprehensive information about what went wrong is now returned by
780        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
781        as the output vector has at least 2 elements. The offset of the start of
782        the failing character and a reason code are placed in the vector.
783    
784    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
785        now returned is for the first byte of the failing character, instead of the
786        last byte inspected. This is an incompatible change, but I hope it is small
787        enough not to be a problem. It makes the returned offset consistent with
788        pcre_exec() and pcre_dfa_exec().
789    
790    7.  pcretest now gives a text phrase as well as the error number when
791        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
792        failure, the offset and reason code are output.
793    
794    8.  When \R was used with a maximizing quantifier it failed to skip backwards
795        over a \r\n pair if the subsequent match failed. Instead, it just skipped
796        back over a single character (\n). This seems wrong (because it treated the
797        two characters as a single entity when going forwards), conflicts with the
798        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
799        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
800        has been changed.
801    
802    9.  Some internal refactoring has changed the processing so that the handling
803        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
804        time (the PCRE_DOTALL option was changed this way some time ago: version
805        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
806        which was always a bit of a fudge. It also means that there is one less
807        argument for the match() function, which reduces its stack requirements
808        slightly. This change also fixes an incompatibility with Perl: the pattern
809        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
810    
811    10. More internal refactoring has drastically reduced the number of recursive
812        calls to match() for possessively repeated groups such as (abc)++ when
813        using pcre_exec().
814    
815    11. While implementing 10, a number of bugs in the handling of groups were
816        discovered and fixed:
817    
818        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
819        (a|)*(?1) gave a compile-time internal error.
820        ((a|)+)+  did not notice that the outer group could match an empty string.
821        (^a|^)+   was not marked as anchored.
822        (.*a|.*)+ was not marked as matching at start or after a newline.
823    
824    12. Yet more internal refactoring has removed another argument from the match()
825        function. Special calls to this function are now indicated by setting a
826        value in a variable in the "match data" data block.
827    
828    13. Be more explicit in pcre_study() instead of relying on "default" for
829        opcodes that mean there is no starting character; this means that when new
830        ones are added and accidentally left out of pcre_study(), testing should
831        pick them up.
832    
833    14. The -s option of pcretest has been documented for ages as being an old
834        synonym of -m (show memory usage). I have changed it to mean "force study
835        for every regex", that is, assume /S for every regex. This is similar to -i
836        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
837        using it. It makes it easier to run collections of tests with and without
838        study enabled, and thereby test pcre_study() more easily. All the standard
839        tests are now run with and without -s (but some patterns can be marked as
840        "never study" - see 20 below).
841    
842    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
843        restoration of the capturing data to the outer values was not happening
844        correctly.
845    
846    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
847        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
848        pattern had matched an empty string, and so incorrectly returned a no
849        match.
850    
851    17. There was optimizing code for the last branch of non-capturing parentheses,
852        and also for the obeyed branch of a conditional subexpression, which used
853        tail recursion to cut down on stack usage. Unfortunately, now that there is
854        the possibility of (*THEN) occurring in these branches, tail recursion is
855        no longer possible because the return has to be checked for (*THEN). These
856        two optimizations have therefore been removed. [But see 8.20/11 above.]
857    
858    18. If a pattern containing \R was studied, it was assumed that \R always
859        matched two bytes, thus causing the minimum subject length to be
860        incorrectly computed because \R can also match just one byte.
861    
862    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
863        was incorrectly computed.
864    
865    20. If /S is present twice on a test pattern in pcretest input, it now
866        *disables* studying, thereby overriding the use of -s on the command line
867        (see 14 above). This is necessary for one or two tests to keep the output
868        identical in both cases.
869    
870    21. When (*ACCEPT) was used in an assertion that matched an empty string and
871        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
872    
873    22. When an atomic group that contained a capturing parenthesis was
874        successfully matched, but the branch in which it appeared failed, the
875        capturing was not being forgotten if a higher numbered group was later
876        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
877        group 1 to "a", when in fact it should be unset. This applied to multi-
878        branched capturing and non-capturing groups, repeated or not, and also to
879        positive assertions (capturing in negative assertions does not happen
880        in PCRE) and also to nested atomic groups.
881    
882    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
883        subject after a captured substring, to make it easier to tell which of a
884        number of identical substrings has been captured.
885    
886    24. The way atomic groups are processed by pcre_exec() has been changed so that
887        if they are repeated, backtracking one repetition now resets captured
888        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
889        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
890        "aaa". Previously, it would have been "a". As part of this code
891        refactoring, the way recursive calls are handled has also been changed.
892    
893    25. If an assertion condition captured any substrings, they were not passed
894        back unless some other capturing happened later. For example, if
895        (?(?=(a))a) was matched against "a", no capturing was returned.
896    
897    26. When studying a pattern that contained subroutine calls or assertions,
898        the code for finding the minimum length of a possible match was handling
899        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
900        group 1 called group 2 while simultaneously a separate group 2 called group
901        1). A stack overflow occurred in this case. I have fixed this by limiting
902        the recursion depth to 10.
903    
904    27. Updated RunTest.bat in the distribution to the version supplied by Tom
905        Fortmann. This supports explicit test numbers on the command line, and has
906        argument validation and error reporting.
907    
908    28. An instance of \X with an unlimited repeat could fail if at any point the
909        first character it looked at was a mark character.
910    
911    29. Some minor code refactoring concerning Unicode properties and scripts
912        should reduce the stack requirement of match() slightly.
913    
914    30. Added the '=' option to pcretest to check the setting of unused capturing
915        slots at the end of the pattern, which are documented as being -1, but are
916        not included in the return count.
917    
918    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
919        compiled something random. Now it gives a compile-time error (as does
920        Perl).
921    
922    32. A *MARK encountered during the processing of a positive assertion is now
923        recorded and passed back (compatible with Perl).
924    
925    33. If --only-matching or --colour was set on a pcregrep call whose pattern
926        had alternative anchored branches, the search for a second match in a line
927        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
928        matched the line "0102" twice. The same bug affected patterns that started
929        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
930        twice.
931    
932    34. Previously, PCRE did not allow quantification of assertions. However, Perl
933        does, and because of capturing effects, quantifying parenthesized
934        assertions may at times be useful. Quantifiers are now allowed for
935        parenthesized assertions.
936    
937    35. A minor code tidy in pcre_compile() when checking options for \R usage.
938    
939    36. \g was being checked for fancy things in a character class, when it should
940        just be a literal "g".
941    
942    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
943        appearance of a nested POSIX class supersedes an apparent external class.
944        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
945        unescaped square brackets may also appear as part of class names. For
946        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
947        more like Perl. (But see 8.20/1 above.)
948    
949    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
950        was because it thought it was \N{name}, which is not supported).
951    
952    39. Add minix to OS list not supporting the -S option in pcretest.
953    
954    40. PCRE tries to detect cases of infinite recursion at compile time, but it
955        cannot analyze patterns in sufficient detail to catch mutual recursions
956        such as ((?1))((?2)). There is now a runtime test that gives an error if a
957        subgroup is called recursively as a subpattern for a second time at the
958        same position in the subject string. In previous releases this might have
959        been caught by the recursion limit, or it might have run out of stack.
960    
961    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
962        happen only once. PCRE was, however incorrectly giving a compile time error
963        "recursive call could loop indefinitely" because it cannot analyze the
964        pattern in sufficient detail. The compile time test no longer happens when
965        PCRE is compiling a conditional subpattern, but actual runaway loops are
966        now caught at runtime (see 40 above).
967    
968    42. It seems that Perl allows any characters other than a closing parenthesis
969        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
970        has been changed to be the same.
971    
972    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
973        as not to get warnings when autogen.sh is called. Also changed
974        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
975    
976    44. To help people who use pcregrep to scan files containing exceedingly long
977        lines, the following changes have been made:
978    
979        (a) The default value of the buffer size parameter has been increased from
980            8K to 20K. (The actual buffer used is three times this size.)
981    
982        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
983            PCRE is built.
984    
985        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
986            to be set at run time.
987    
988        (d) Numerical values in pcregrep options can be followed by K or M, for
989            example --buffer-size=50K.
990    
991        (e) If a line being scanned overflows pcregrep's buffer, an error is now
992            given and the return code is set to 2.
993    
994    45. Add a pointer to the latest mark to the callout data block.
995    
996    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
997        partial match of an empty string instead of no match. This was specific to
998        the use of ".".
999    
1000    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1001        complete match instead of a partial match. This bug was dependent on both
1002        the PCRE_UTF8 and PCRE_DOTALL options being set.
1003    
1004    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1005        starting byte set, because \b was not being ignored.
1006    
1007    
1008    Version 8.12 15-Jan-2011
1009    ------------------------
1010    
1011    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1012        checks for such things as part of the documentation building process.
1013    
1014    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1015        --match-limit and --recursion-limit options (added for 8.11). In
1016        particular, this made one of the standard tests fail. (The integer value
1017        went into the wrong half of a long int.)
1018    
1019    3.  If the --colour option was given to pcregrep with -v (invert match), it
1020        did strange things, either producing crazy output, or crashing. It should,
1021        of course, ignore a request for colour when reporting lines that do not
1022        match.
1023    
1024    4.  Another pcregrep bug caused similar problems if --colour was specified with
1025        -M (multiline) and the pattern match finished with a line ending.
1026    
1027    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1028        matched in multiline mode, the following line was shown as part of the
1029        match. This seems wrong, so I have changed it.
1030    
1031    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1032        the check for further matches in the same line (so they could be coloured)
1033        to overrun the end of the current line. If another match was found, it was
1034        incorrectly shown (and then shown again when found in the next line).
1035    
1036    7.  If pcregrep was compiled under Windows, there was a reference to the
1037        function pcregrep_exit() before it was defined. I am assuming this was
1038        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1039        reported by a user. I've moved the definition above the reference.
1040    
1041    
1042    Version 8.11 10-Dec-2010
1043    ------------------------
1044    
1045    1.  (*THEN) was not working properly if there were untried alternatives prior
1046        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1047        backtracked to try for "b" instead of moving to the next alternative branch
1048        at the same level (in this case, to look for "c"). The Perl documentation
1049        is clear that when (*THEN) is backtracked onto, it goes to the "next
1050        alternative in the innermost enclosing group".
1051    
1052    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1053        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1054        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1055        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1056        (*THEN).
1057    
1058    3.  If \s appeared in a character class, it removed the VT character from
1059        the class, even if it had been included by some previous item, for example
1060        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1061        of \s, but is part of the POSIX "space" class.)
1062    
1063    4.  A partial match never returns an empty string (because you can always
1064        match an empty string at the end of the subject); however the checking for
1065        an empty string was starting at the "start of match" point. This has been
1066        changed to the "earliest inspected character" point, because the returned
1067        data for a partial match starts at this character. This means that, for
1068        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1069        (previously it gave "no match").
1070    
1071    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1072        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1073        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1074        has an implication that the given string is incomplete (because a partial
1075        match is preferred over a full match). For this reason, these items now
1076        give a partial match in this situation. [Aside: previously, the one case
1077        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1078        match rather than a full match, which was wrong by the old rules, but is
1079        now correct.]
1080    
1081    6.  There was a bug in the handling of #-introduced comments, recognized when
1082        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1083        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1084        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1085        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1086        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1087        places in pcre_compile().
1088    
1089    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1090        comments when looking ahead for named forward references to subpatterns,
1091        the only newline sequence it recognized was NL. It now handles newlines
1092        according to the set newline convention.
1093    
1094    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1095        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1096        cater for a lack of strerror(). These oversights have been fixed.
1097    
1098    9.  Added --match-limit and --recursion-limit to pcregrep.
1099    
1100    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1101    
1102    11. When the -o option was used, pcregrep was setting a return code of 1, even
1103        when matches were found, and --line-buffered was not being honoured.
1104    
1105    12. Added an optional parentheses number to the -o and --only-matching options
1106        of pcregrep.
1107    
1108    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1109        can match an empty string. The code to do it in pcretest and pcredemo
1110        needed fixing:
1111    
1112        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1113            only one byte after an empty string match just before CRLF (this case
1114            just got forgotten; "any" and "anycrlf" were OK).
1115    
1116        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1117            mode when an empty string match preceded an ASCII character followed by
1118            a non-ASCII character. (The code for advancing by one character rather
1119            than one byte was nonsense.)
1120    
1121        (c) The pcredemo.c sample program did not have any code at all to handle
1122            the cases when CRLF is a valid newline sequence.
1123    
1124    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1125        as a starting offset was within the subject string. There is now a new
1126        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1127        negative or greater than the length of the string. In order to test this,
1128        pcretest is extended to allow the setting of negative starting offsets.
1129    
1130    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1131        starting offset points to the beginning of a UTF-8 character was
1132        unnecessarily clumsy. I tidied it up.
1133    
1134    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1135        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1136    
1137    17. Nobody had reported that the --include_dir option, which was added in
1138        release 7.7 should have been called --include-dir (hyphen, not underscore)
1139        for compatibility with GNU grep. I have changed it to --include-dir, but
1140        left --include_dir as an undocumented synonym, and the same for
1141        --exclude-dir, though that is not available in GNU grep, at least as of
1142        release 2.5.4.
1143    
1144    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1145        characters from a string of bytes) have been redefined so as not to use
1146        loops, in order to improve performance in some environments. At the same
1147        time, I abstracted some of the common code into auxiliary macros to save
1148        repetition (this should not affect the compiled code).
1149    
1150    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1151        compile-time error is now given if \c is not followed by an ASCII
1152        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1153        different, and any byte value is allowed.)
1154    
1155    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1156        START_OPTIMIZE option, which is now allowed at compile time - but just
1157        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1158        to pcregrep and other applications that have no direct access to PCRE
1159        options. The new /Y option in pcretest sets this option when calling
1160        pcre_compile().
1161    
1162    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1163        back references. Groups containing recursive back references were forced to
1164        be atomic by that change, but in the case of named groups, the amount of
1165        memory required was incorrectly computed, leading to "Failed: internal
1166        error: code overflow". This has been fixed.
1167    
1168    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1169        pcretest.c, to avoid build problems in some Borland environments.
1170    
1171    
1172    Version 8.10 25-Jun-2010
1173    ------------------------
1174    
1175    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1176        THEN.
1177    
1178    2.  (*ACCEPT) was not working when inside an atomic group.
1179    
1180    3.  Inside a character class, \B is treated as a literal by default, but
1181        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1182        causes the error). The code is unchanged, but I tidied the documentation.
1183    
1184    4.  Inside a character class, PCRE always treated \R and \X as literals,
1185        whereas Perl faults them if its -w option is set. I have changed PCRE so
1186        that it faults them when PCRE_EXTRA is set.
1187    
1188    5.  Added support for \N, which always matches any character other than
1189        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1190    
1191    6.  When compiling pcregrep with newer versions of gcc which may have
1192        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1193        declared with attribute warn_unused_result" were given. Just casting the
1194        result to (void) does not stop the warnings; a more elaborate fudge is
1195        needed. I've used a macro to implement this.
1196    
1197    7.  Minor change to pcretest.c to avoid a compiler warning.
1198    
1199    8.  Added four artifical Unicode properties to help with an option to make
1200        \s etc use properties (see next item). The new properties are: Xan
1201        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1202    
1203    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1204        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1205        this option. Modified pcretest to add /W to test this facility. Added
1206        REG_UCP to make it available via the POSIX interface.
1207    
1208    10. Added --line-buffered to pcregrep.
1209    
1210    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1211        studied, and the match started with a letter with a code point greater than
1212        127 whose first byte was different to the first byte of the other case of
1213        the letter, the other case of this starting letter was not recognized
1214        (#976).
1215    
1216    12. If a pattern that was studied started with a repeated Unicode property
1217        test, for example, \p{Nd}+, there was the theoretical possibility of
1218        setting up an incorrect bitmap of starting bytes, but fortunately it could
1219        not have actually happened in practice until change 8 above was made (it
1220        added property types that matched character-matching opcodes).
1221    
1222    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1223        possible starting bytes for non-anchored patterns.
1224    
1225    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1226        \R, and also a number of cases that involve Unicode properties, both
1227        explicit and implicit when PCRE_UCP is set.
1228    
1229    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1230        input, it could crash or give wrong results if characters with values
1231        greater than 0xc0 were present in the subject string. (Detail: it assumed
1232        UTF-8 input when processing these items.)
1233    
1234    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1235        size_t is 64-bit (#991).
1236    
1237    17. Added a check for running out of memory when PCRE is compiled with
1238        --disable-stack-for-recursion (#990).
1239    
1240    18. If the last data line in a file for pcretest does not have a newline on
1241        the end, a newline was missing in the output.
1242    
1243    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1244        less than 128) in its various bitmaps. However, there is a facility for
1245        generating tables according to the current locale when PCRE is compiled. It
1246        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1247        characters, are recognized by isspace() and therefore were getting set in
1248        these tables, and indeed these tables seem to approximate to ISO 8859. This
1249        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1250        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1251        which of course cannot start UTF-8 characters. I have changed the code so
1252        that only real ASCII characters (less than 128) and the correct starting
1253        bytes for UTF-8 encodings are set for characters greater than 127 when in
1254        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1255        altogether.)
1256    
1257    20. Added the /T option to pcretest so as to be able to run tests with non-
1258        standard character tables, thus making it possible to include the tests
1259        used for 19 above in the standard set of tests.
1260    
1261    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1262        reference to a subpattern the other side of a comment that contains an
1263        opening parenthesis caused either an internal compiling error, or a
1264        reference to the wrong subpattern.
1265    
1266    
1267    Version 8.02 19-Mar-2010
1268    ------------------------
1269    
1270    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1271    
1272    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1273        configured.
1274    
1275    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1276        original author of that file, following a query about its status.
1277    
1278    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1279        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1280    
1281    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1282        quantifier applied to a forward-referencing subroutine call, could compile
1283        incorrect code or give the error "internal error: previously-checked
1284        referenced subpattern not found".
1285    
1286    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1287        variables to point to external functions. For these systems, therefore,
1288        pcre_malloc etc. are now initialized to local functions that call the
1289        relevant global functions.
1290    
1291    7.  There were two entries missing in the vectors called coptable and poptable
1292        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1293        I've fixed the data, and added a kludgy way of testing at compile time that
1294        the lengths are correct (equal to the number of opcodes).
1295    
1296    8.  Following on from 7, I added a similar kludge to check the length of the
1297        eint vector in pcreposix.c.
1298    
1299    9.  Error texts for pcre_compile() are held as one long string to avoid too
1300        much relocation at load time. To find a text, the string is searched,
1301        counting zeros. There was no check for running off the end of the string,
1302        which could happen if a new error number was added without updating the
1303        string.
1304    
1305    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1306    
1307    11. \K was not working if it appeared in an atomic group or in a group that
1308        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1309        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1310        the assertion is positive, but not if it is negative.
1311    
1312    12. Change 11 fortuitously reduced the size of the stack frame used in the
1313        "match()" function of pcre_exec.c by one pointer. Forthcoming
1314        implementation of support for (*MARK) will need an extra pointer on the
1315        stack; I have reserved it now, so that the stack frame size does not
1316        decrease.
1317    
1318    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1319        item in branch that calls a recursion is a subroutine call - as in the
1320        second branch in the above example - was incorrectly given the compile-
1321        time error "recursive call could loop indefinitely" because pcre_compile()
1322        was not correctly checking the subroutine for matching a non-empty string.
1323    
1324    14. The checks for overrunning compiling workspace could trigger after an
1325        overrun had occurred. This is a "should never occur" error, but it can be
1326        triggered by pathological patterns such as hundreds of nested parentheses.
1327        The checks now trigger 100 bytes before the end of the workspace.
1328    
1329    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1330    
1331    
1332    Version 8.01 19-Jan-2010
1333    ------------------------
1334    
1335    1.  If a pattern contained a conditional subpattern with only one branch (in
1336        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1337        computed the wrong minimum data length (which is of course zero for such
1338        subpatterns). This could cause incorrect "no match" results.
1339    
1340    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1341        the pattern is reset in the first branch, pcre_compile() failed with
1342        "internal error: code overflow at offset...". This happened only when
1343        the reset was to the original external option setting. (An optimization
1344        abstracts leading options settings into an external setting, which was the
1345        cause of this.)
1346    
1347    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1348        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1349        assertion pattern did not match (meaning that the assertion was true), it
1350        was incorrectly treated as false if the SKIP had been reached during the
1351        matching. This also applied to assertions used as conditions.
1352    
1353    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1354        assertion subpattern, including such a pattern used as a condition,
1355        unpredictable results occurred, instead of the error return
1356        PCRE_ERROR_DFA_UITEM.
1357    
1358    5.  The C++ GlobalReplace function was not working like Perl for the special
1359        situation when an empty string is matched. It now does the fancy magic
1360        stuff that is necessary.
1361    
1362    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1363        removed. (These were left over from very, very early versions of PCRE.)
1364    
1365    7.  Some cosmetic changes to the code to make life easier when compiling it
1366        as part of something else:
1367    
1368        (a) Change DEBUG to PCRE_DEBUG.
1369    
1370        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1371            called "current" as "current_branch", to prevent a collision with the
1372            Linux macro when compiled as a kernel module.
1373    
1374        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1375            prevent a collision with the Linux macro when compiled as a kernel
1376            module.
1377    
1378    8.  In pcre_compile() there are some checks for integer overflows that used to
1379        cast potentially large values to (double). This has been changed to that
1380        when building, a check for int64_t is made, and if it is found, it is used
1381        instead, thus avoiding the use of floating point arithmetic. (There is no
1382        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1383        double.
1384    
1385    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1386        2005 (difference between two addresses compared to an unsigned value).
1387    
1388    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1389        custom one, because of the following reported problem in Windows:
1390    
1391          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1392              under Win32.
1393          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1394              therefore missing the function definition.
1395          - The compiler thus generates a "C" signature for the test function.
1396          - The linker fails to find the "C" function.
1397          - PCRE fails to configure if asked to do so against libbz2.
1398    
1399    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1400        messages were output:
1401    
1402          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1403          rerunning libtoolize, to keep the correct libtool macros in-tree.
1404          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1405    
1406        I have done both of these things.
1407    
1408    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1409        most of the time, it *can* run out if it is given a pattern that contains a
1410        runaway infinite recursion. I updated the discussion in the pcrestack man
1411        page.
1412    
1413    13. Now that we have gone to the x.xx style of version numbers, the minor
1414        version may start with zero. Using 08 or 09 is a bad idea because users
1415        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1416        interpreted as invalid octal numbers. I've updated the previous comment in
1417        configure.ac, and also added a check that gives an error if 08 or 09 are
1418        used.
1419    
1420    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1421        causing partial matching to fail when the end of the subject matched \W
1422        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1423    
1424    15. There were some discrepancies between the declarations in pcre_internal.h
1425        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1426        their definitions. The declarations used "const uschar *" and the
1427        definitions used USPTR. Even though USPTR is normally defined as "const
1428        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1429        reported that: "This difference in casting confuses some C++ compilers, for
1430        example, SunCC recognizes above declarations as different functions and
1431        generates broken code for hbpcre." I have changed the declarations to use
1432        USPTR.
1433    
1434    16. GNU libtool is named differently on some systems. The autogen.sh script now
1435        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1436        (FreeBSD).
1437    
1438    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1439        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1440        comment: "Figure out how to create a longlong from a string: strtoll and
1441        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1442        instance, but it only takes 2 args instead of 3!"
1443    
1444    18. A subtle bug concerned with back references has been fixed by a change of
1445        specification, with a corresponding code fix. A pattern such as
1446        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1447        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1448        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1449        same bug. Such groups have to be quantified to be useful, or contained
1450        inside another quantified group. (If there's no repetition, the reference
1451        can never match.) The problem arises because, having left the group and
1452        moved on to the rest of the pattern, a later failure that backtracks into
1453        the group uses the captured value from the final iteration of the group
1454        rather than the correct earlier one. I have fixed this in PCRE by forcing
1455        any group that contains a reference to itself to be an atomic group; that
1456        is, there cannot be any backtracking into it once it has completed. This is
1457        similar to recursive and subroutine calls.
1458    
1459    
1460    Version 8.00 19-Oct-09
1461    ----------------------
1462    
1463    1.  The table for translating pcre_compile() error codes into POSIX error codes
1464        was out-of-date, and there was no check on the pcre_compile() error code
1465        being within the table. This could lead to an OK return being given in
1466        error.
1467    
1468    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1469        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1470        in a Windows environment.
1471    
1472    3.  The pcregrep --count option prints the count for each file even when it is
1473        zero, as does GNU grep. However, pcregrep was also printing all files when
1474        --files-with-matches was added. Now, when both options are given, it prints
1475        counts only for those files that have at least one match. (GNU grep just
1476        prints the file name in this circumstance, but including the count seems
1477        more useful - otherwise, why use --count?) Also ensured that the
1478        combination -clh just lists non-zero counts, with no names.
1479    
1480    4.  The long form of the pcregrep -F option was incorrectly implemented as
1481        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1482        but it seems right to fix it, and I didn't think it was worth preserving
1483        the old behaviour.
1484    
1485    5.  The command line items --regex=pattern and --regexp=pattern were not
1486        recognized by pcregrep, which required --regex pattern or --regexp pattern
1487        (with a space rather than an '='). The man page documented the '=' forms,
1488        which are compatible with GNU grep; these now work.
1489    
1490    6.  No libpcreposix.pc file was created for pkg-config; there was just
1491        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1492    
1493    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1494        when UCP support is not needed, by modifying the Python script that
1495        generates it from Unicode data files. This should not matter if the module
1496        is correctly used as a library, but I received one complaint about 50K of
1497        unwanted data. My guess is that the person linked everything into his
1498        program rather than using a library. Anyway, it does no harm.
1499    
1500    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1501        was a minimum greater than 1 for a wide character in a possessive
1502        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1503        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1504        character. Chaos in the form of incorrect output or a compiling loop could
1505        result.
1506    
1507    9.  The restrictions on what a pattern can contain when partial matching is
1508        requested for pcre_exec() have been removed. All patterns can now be
1509        partially matched by this function. In addition, if there are at least two
1510        slots in the offset vector, the offset of the earliest inspected character
1511        for the match and the offset of the end of the subject are set in them when
1512        PCRE_ERROR_PARTIAL is returned.
1513    
1514    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1515        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1516        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1517        and may be more useful for multi-segment matching.
1518    
1519    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1520        used to be given if ever the end of the subject was reached; now it is
1521        given only if matching could not proceed because another character was
1522        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1523        string "Z", which now yields "no match" instead of "partial match". In the
1524        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1525        final character ended with (*FAIL).
1526    
1527    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1528        if the pattern had a "must contain" character that was already found in the
1529        earlier partial match, unless partial matching was again requested. For
1530        example, with the pattern /dog.(body)?/, the "must contain" character is
1531        "g". If the first part-match was for the string "dog", restarting with
1532        "sbody" failed. This bug has been fixed.
1533    
1534    13. The string returned by pcre_dfa_exec() after a partial match has been
1535        changed so that it starts at the first inspected character rather than the
1536        first character of the match. This makes a difference only if the pattern
1537        starts with a lookbehind assertion or \b or \B (\K is not supported by
1538        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1539        matching functions compatible, and I think it's the right thing to do.
1540    
1541    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1542        so that the demonstration program is easily available in environments where
1543        PCRE has not been installed from source.
1544    
1545    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1546        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1547        library.
1548    
1549    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1550        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1551        is not the first non-POSIX option to be added. Clearly some people find
1552        these options useful.
1553    
1554    17. If a caller to the POSIX matching function regexec() passes a non-zero
1555        value for nmatch with a NULL value for pmatch, the value of
1556        nmatch is forced to zero.
1557    
1558    18. RunGrepTest did not have a test for the availability of the -u option of
1559        the diff command, as RunTest does. It now checks in the same way as
1560        RunTest, and also checks for the -b option.
1561    
1562    19. If an odd number of negated classes containing just a single character
1563        interposed, within parentheses, between a forward reference to a named
1564        subpattern and the definition of the subpattern, compilation crashed with
1565        an internal error, complaining that it could not find the referenced
1566        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1567        [The bug was that it was starting one character too far in when skipping
1568        over the character class, thus treating the ] as data rather than
1569        terminating the class. This meant it could skip too much.]
1570    
1571    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1572        /g option in pcretest when the pattern contains \K, which makes it possible
1573        to have an empty string match not at the start, even when the pattern is
1574        anchored. Updated pcretest and pcredemo to use this option.
1575    
1576    21. If the maximum number of capturing subpatterns in a recursion was greater
1577        than the maximum at the outer level, the higher number was returned, but
1578        with unset values at the outer level. The correct (outer level) value is
1579        now given.
1580    
1581    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1582        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1583        make it do so. The string so far is captured, making this feature
1584        compatible with Perl.
1585    
1586    23. The tests have been re-organized, adding tests 11 and 12, to make it
1587        possible to check the Perl 5.10 features against Perl 5.10.
1588    
1589    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1590        pattern matches a fixed length string. PCRE did not allow this; now it
1591        does. Neither allows recursion.
1592    
1593    25. I finally figured out how to implement a request to provide the minimum
1594        length of subject string that was needed in order to match a given pattern.
1595        (It was back references and recursion that I had previously got hung up
1596        on.) This code has now been added to pcre_study(); it finds a lower bound
1597        to the length of subject needed. It is not necessarily the greatest lower
1598        bound, but using it to avoid searching strings that are too short does give
1599        some useful speed-ups. The value is available to calling programs via
1600        pcre_fullinfo().
1601    
1602    26. While implementing 25, I discovered to my embarrassment that pcretest had
1603        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1604        study optimizations had never been tested with that matching function.
1605        Oops. What is worse, even when it was passed study data, there was a bug in
1606        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1607        were also very few tests of studied patterns with pcre_dfa_exec().
1608    
1609    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1610        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1611        on the other side of the coin, they are no longer allowed to have different
1612        names, because these cannot be distinguished in PCRE, and this has caused
1613        confusion. (This is a difference from Perl.)
1614    
1615    28. When duplicate subpattern names are present (necessarily with different
1616        numbers, as required by 27 above), and a test is made by name in a
1617        conditional pattern, either for a subpattern having been matched, or for
1618        recursion in such a pattern, all the associated numbered subpatterns are
1619        tested, and the overall condition is true if the condition is true for any
1620        one of them. This is the way Perl works, and is also more like the way
1621        testing by number works.
1622    
1623    
1624    Version 7.9 11-Apr-09
1625    ---------------------
1626    
1627    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1628        (pcretest), all targets were linked against these libraries. This included
1629        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1630        libraries. This caused unwanted dependencies to be created. This problem
1631        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1632        pcretest is linked with readline.
1633    
1634    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1635        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1636        moved outside it again, because FALSE and TRUE are already defined in AIX,
1637        but BOOL is not.
1638    
1639    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1640        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1641    
1642    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1643        hyphens) following file names and line numbers when outputting matching
1644        lines. This is not true; no spaces are inserted. I have also clarified the
1645        wording for the --colour (or --color) option.
1646    
1647    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1648        was not coloured; this is different to GNU grep, so I have changed it to be
1649        the same.
1650    
1651    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1652        each matching line was coloured. Now it goes on to look for further matches
1653        of any of the test patterns, which is the same behaviour as GNU grep.
1654    
1655    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1656        doesn't make sense to accept an empty string match in pcregrep, so I have
1657        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1658        seems to be how GNU grep behaves.
1659    
1660    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1661        start or after a newline", because the conditional assertion was not being
1662        correctly handled. The rule now is that both the assertion and what follows
1663        in the first alternative must satisfy the test.
1664    
1665    9.  If auto-callout was enabled in a pattern with a conditional group whose
1666        condition was an assertion, PCRE could crash during matching, both with
1667        pcre_exec() and pcre_dfa_exec().
1668    
1669    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1670        used for matching.
1671    
1672    11. Unicode property support in character classes was not working for
1673        characters (bytes) greater than 127 when not in UTF-8 mode.
1674    
1675    12. Added the -M command line option to pcretest.
1676    
1677    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1678    
1679    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1680    
1681    16. Added comments and documentation about mis-use of no_arg in the C++
1682        wrapper.
1683    
1684    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1685        from Martin Jerabek that uses macro names for all relevant character and
1686        string constants.
1687    
1688    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1689        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1690        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1691        these, but not everybody uses configure.
1692    
1693    19. A conditional group that had only one branch was not being correctly
1694        recognized as an item that could match an empty string. This meant that an
1695        enclosing group might also not be so recognized, causing infinite looping
1696        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1697        with the subject "ab", where knowledge that the repeated group can match
1698        nothing is needed in order to break the loop.
1699    
1700    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1701        exec(), but without supplying a callout function, matching went wrong.
1702    
1703    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1704        leak if the size of the offset vector was greater than 30. When the vector
1705        is smaller, the saved offsets during recursion go onto a local stack
1706        vector, but for larger vectors malloc() is used. It was failing to free
1707        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1708        error, in fact).
1709    
1710    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1711        heapframe that is used only when UTF-8 support is enabled. This caused no
1712        problem, but was untidy.
1713    
1714    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1715        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1716        included within another project.
1717    
1718    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1719        slightly modified by me:
1720    
1721          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1722              not building pcregrep.
1723    
1724          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1725              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1726    
1727    25. Forward references, both numeric and by name, in patterns that made use of
1728        duplicate group numbers, could behave incorrectly or give incorrect errors,
1729        because when scanning forward to find the reference group, PCRE was not
1730        taking into account the duplicate group numbers. A pattern such as
1731        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1732    
1733    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1734        the feature of a custom pointer more persuasive (as requested by a user).
1735    
1736    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1737        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1738        pre-defined.
1739    
1740    28. Added support for (*UTF8) at the start of a pattern.
1741    
1742    29. Arrange for flags added by the "release type" setting in CMake to be shown
1743        in the configuration summary.
1744    
1745    
1746    Version 7.8 05-Sep-08
1747    ---------------------
1748    
1749    1.  Replaced UCP searching code with optimized version as implemented for Ad
1750        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1751        stage table and inline lookup instead of a function, giving speed ups of 2
1752        to 5 times on some simple patterns that I tested. Permission was given to
1753        distribute the MultiStage2.py script that generates the tables (it's not in
1754        the tarball, but is in the Subversion repository).
1755    
1756    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1757        scripts.
1758    
1759    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1760        a group with a zero qualifier. The result of the study could be incorrect,
1761        or the function might crash, depending on the pattern.
1762    
1763    4.  Caseless matching was not working for non-ASCII characters in back
1764        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1765        It now works when Unicode Property Support is available.
1766    
1767    5.  In pcretest, an escape such as \x{de} in the data was always generating
1768        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1769        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1770        truncation.
1771    
1772    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1773    
1774    7.  Added two (int) casts to pcregrep when printing the difference of two
1775        pointers, in case they are 64-bit values.
1776    
1777    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1778        test 2 if it fails.
1779    
1780    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1781        and a #define of that name to empty if it is not externally set. This is to
1782        allow users of MSVC to set it if necessary.
1783    
1784    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1785        the convenience functions in the pcre_get.c source file.
1786    
1787    11. An option change at the start of a pattern that had top-level alternatives
1788        could cause overwriting and/or a crash. This command provoked a crash in
1789        some environments:
1790    
1791          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1792    
1793        This potential security problem was recorded as CVE-2008-2371.
1794    
1795    12. For a pattern where the match had to start at the beginning or immediately
1796        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1797        pcre_dfa_exec() could read past the end of the passed subject if there was
1798        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1799        pcretest so that it places the subject at the end of its malloc-ed buffer.
1800    
1801    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1802        exec() might read past the end of the data buffer in UTF-8 mode.
1803    
1804    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1805        the data contained the byte 0x85 as part of a UTF-8 character within its
1806        first line. This applied both to normal and DFA matching.
1807    
1808    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1809        /^[^d]*?$/8 failed to match "abc".
1810    
1811    16. Added a missing copyright notice to pcrecpp_internal.h.
1812    
1813    17. Make it more clear in the documentation that values returned from
1814        pcre_exec() in ovector are byte offsets, not character counts.
1815    
1816    18. Tidied a few places to stop certain compilers from issuing warnings.
1817    
1818    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1819        supplied by Stefan Weber. I made a further small update for 7.8 because
1820        there is a change of source arrangements: the pcre_searchfuncs.c module is
1821        replaced by pcre_ucd.c.
1822    
1823    
1824    Version 7.7 07-May-08
1825    ---------------------
1826    
1827    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1828        a string to a long long, pretend we don't even have a long long." This is
1829        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1830    
1831    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1832        pre-7.6 versions, which defined a global no_arg variable instead of putting
1833        it in the RE class. (See also #8 below.)
1834    
1835    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1836        Lopes.
1837    
1838    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1839    
1840        (1) The include/exclude patterns were being applied to the whole pathnames
1841            of files, instead of just to the final components.
1842    
1843        (2) If there was more than one level of directory, the subdirectories were
1844            skipped unless they satisfied the include/exclude conditions. This is
1845            inconsistent with GNU grep (and could even be seen as contrary to the
1846            pcregrep specification - which I improved to make it absolutely clear).
1847            The action now is always to scan all levels of directory, and just
1848            apply the include/exclude patterns to regular files.
1849    
1850    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1851        --exclude_dir in the tests to avoid scanning .svn directories.
1852    
1853    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1854        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1855        doesn't support NULs in patterns.
1856    
1857    7.  Added some missing "const"s to declarations of static tables in
1858        pcre_compile.c and pcre_dfa_exec.c.
1859    
1860    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1861        caused by fix #2  above. (Subsequently also a second patch to fix the
1862        first patch. And a third patch - this was a messy problem.)
1863    
1864    9.  Applied Craig's patch to remove the use of push_back().
1865    
1866    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1867        matching function regexec().
1868    
1869    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1870        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1871        references. PCRE supports relative numbers with this syntax (I don't think
1872        Oniguruma does).
1873    
1874    12. Previously, a group with a zero repeat such as (...){0} was completely
1875        omitted from the compiled regex. However, this means that if the group
1876        was called as a subroutine from elsewhere in the pattern, things went wrong
1877        (an internal error was given). Such groups are now left in the compiled
1878        pattern, with a new opcode that causes them to be skipped at execution
1879        time.
1880    
1881    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1882        to the way PCRE behaves:
1883    
1884        (a) A lone ] character is dis-allowed (Perl treats it as data).
1885    
1886        (b) A back reference to an unmatched subpattern matches an empty string
1887            (Perl fails the current match path).
1888    
1889        (c) A data ] in a character class must be notated as \] because if the
1890            first data character in a class is ], it defines an empty class. (In
1891            Perl it is not possible to have an empty class.) The empty class []
1892            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1893            The negative empty class [^] matches any one character, independently
1894            of the DOTALL setting.
1895    
1896    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1897        non-existent subpattern following a character class starting with ']' and
1898        containing () gave an internal compiling error instead of "reference to
1899        non-existent subpattern". Fortunately, when the pattern did exist, the
1900        compiled code was correct. (When scanning forwards to check for the
1901        existencd of the subpattern, it was treating the data ']' as terminating
1902        the class, so got the count wrong. When actually compiling, the reference
1903        was subsequently set up correctly.)
1904    
1905    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1906        it was being rejected as not supported by pcre_dfa_exec(), even though
1907        other assertions are supported. I have made pcre_dfa_exec() support
1908        (*FAIL).
1909    
1910    16. The implementation of 13c above involved the invention of a new opcode,
1911        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1912        cannot be changed at match time, I realized I could make a small
1913        improvement to matching performance by compiling OP_ALLANY instead of
1914        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1915        on the OP_ANY path.
1916    
1917    17. Compiling pcretest on Windows with readline support failed without the
1918        following two fixes: (1) Make the unistd.h include conditional on
1919        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1920    
1921    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1922        ncurses library to be included for pcretest when ReadLine support is
1923        requested, but also to allow for it to be overridden. This patch came from
1924        Daniel Bergström.
1925    
1926    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1927        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1928        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1929        spotting this.
1930    
1931    
1932    Version 7.6 28-Jan-08
1933    ---------------------
1934    
1935    1.  A character class containing a very large number of characters with
1936        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1937        overflow.
1938    
1939    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1940        HAVE_LONG_LONG is not defined.
1941    
1942    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1943        bring them up to date and include new features. This patch includes:
1944    
1945        - Fixed PH's badly added libz and libbz2 support.
1946        - Fixed a problem with static linking.
1947        - Added pcredemo. [But later removed - see 7 below.]
1948        - Fixed dftables problem and added an option.
1949        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1950            HAVE_LONG_LONG.
1951        - Added readline support for pcretest.
1952        - Added an listing of the option settings after cmake has run.
1953    
1954    4.  A user submitted a patch to Makefile that makes it easy to create
1955        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1956        Makefile.am that cause it to include this special target, without
1957        affecting anything else. Note that the same mingw target plus all
1958        the other distribution libraries and programs are now supported
1959        when configuring with CMake (see 6 below) instead of with
1960        Configure/Make.
1961    
1962    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1963        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1964        exported in the Windows port". It has not yet been confirmed that the patch
1965        solves the problem, but it does no harm.
1966    
1967    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1968        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1969        with CMake, and also correct the comment about stack recursion.
1970    
1971    7.  Remove the automatic building of pcredemo from the ./configure system and
1972        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1973        of a program that users should build themselves after PCRE is installed, so
1974        building it automatically is not really right. What is more, it gave
1975        trouble in some build environments.
1976    
1977    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1978    
1979    
1980    Version 7.5 10-Jan-08
1981    ---------------------
1982    
1983    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1984        values in parens when parsing an RE using the C++ wrapper."
1985    
1986    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1987        Characters greater than 255 were excluded from the class instead of being
1988        included.
1989    
1990    3.  The same bug as (2) above applied to negated POSIX classes such as
1991        [:^space:].
1992    
1993    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1994        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1995        I have changed it.
1996    
1997    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1998        first named subpattern) and a construct such as (?&a) would reference the
1999        first named subpattern whose name started with "a" (in other words, the
2000        length check was missing). Both these problems are fixed. "Subpattern name
2001        expected" is now given for (?&) (a zero-length name), and this patch also
2002        makes it give the same error for \k'' (previously it complained that that
2003        was a reference to a non-existent subpattern).
2004    
2005    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2006        this is right because (?- can be followed by option settings as well as by
2007        digits. I have, however, made the messages clearer.
2008    
2009    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2010        than the number used in the conditional) now cause a compile-time error.
2011        This is actually not compatible with Perl, which accepts such patterns, but
2012        treats the conditional as always being FALSE (as PCRE used to), but it
2013        seems to me that giving a diagnostic is better.
2014    
2015    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2016        and messages.
2017    
2018    9.  Fix two occurrences of "backslash" in comments that should have been
2019        "backspace".
2020    
2021    10. Remove two redundant lines of code that can never be obeyed (their function
2022        was moved elsewhere).
2023    
2024    11. The program that makes PCRE's Unicode character property table had a bug
2025        which caused it to generate incorrect table entries for sequences of
2026        characters that have the same character type, but are in different scripts.
2027        It amalgamated them into a single range, with the script of the first of
2028        them. In other words, some characters were in the wrong script. There were
2029        thirteen such cases, affecting characters in the following ranges:
2030    
2031          U+002b0 - U+002c1
2032          U+0060c - U+0060d
2033          U+0061e - U+00612
2034          U+0064b - U+0065e
2035          U+0074d - U+0076d
2036          U+01800 - U+01805
2037          U+01d00 - U+01d77
2038          U+01d9b - U+01dbf
2039          U+0200b - U+0200f
2040          U+030fc - U+030fe
2041          U+03260 - U+0327f
2042          U+0fb46 - U+0fbb1
2043          U+10450 - U+1049d
2044    
2045    12. The -o option (show only the matching part of a line) for pcregrep was not
2046        compatible with GNU grep in that, if there was more than one match in a
2047        line, it showed only the first of them. It now behaves in the same way as
2048        GNU grep.
2049    
2050    13. If the -o and -v options were combined for pcregrep, it printed a blank
2051        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2052        does the same. The return code can be used to tell if there were any
2053        non-matching lines.
2054    
2055    14. Added --file-offsets and --line-offsets to pcregrep.
2056    
2057    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2058        infinitely looping recursion. The bug was that positive lookaheads were not
2059        being skipped when checking for a possible empty match (negative lookaheads
2060        and both kinds of lookbehind were skipped).
2061    
2062    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2063        inclusion of <windows.h> to before rather than after the definition of
2064        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2065    
2066    17. Specifying a possessive quantifier with a specific limit for a Unicode
2067        character property caused pcre_compile() to compile bad code, which led at
2068        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2069        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2070        caused the error; without that there was no problem.
2071    
2072    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2073    
2074    19. Added --enable-pcretest-libreadline.
2075    
2076    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2077        RE::GlobalReplace(). As a result, the number of replacements returned was
2078        double what it should be. I removed one of the increments, but Craig sent a
2079        later patch that removed the other one (the right fix) and added unit tests
2080        that check the return values (which was not done before).
2081    
2082    21. Several CMake things:
2083    
2084        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2085            the names libpcre and libpcreposix, not just pcre and pcreposix.
2086    
2087        (2) The above change means that pcretest and pcregrep are now correctly
2088            linked with the newly-built libraries, not previously installed ones.
2089    
2090        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2091    
2092    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2093        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2094        UTF-8 newline character). The key issue is that the pattern starts .*;
2095        this means that the match must be either at the beginning, or after a
2096        newline. The bug was in the code for advancing after a failed match and
2097        checking that the new position followed a newline. It was not taking
2098        account of UTF-8 characters correctly.
2099    
2100    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2101        character classes. PCRE was not treating the sequence [:...:] as a
2102        character class unless the ... were all letters. Perl, however, seems to
2103        allow any characters between [: and :], though of course it rejects as
2104        unknown any "names" that contain non-letters, because all the known class
2105        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2106        for example, whereas PCRE did not - it did not recognize a POSIX character
2107        class. This seemed a bit dangerous, so the code has been changed to be
2108        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2109        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2110        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2111        Perl does, and where it didn't before.
2112    
2113    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2114        Windows environments %n is disabled by default.
2115    
2116    
2117    Version 7.4 21-Sep-07
2118    ---------------------
2119    
2120    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2121        means that a class such as [\s] counted as "explicit reference to CR or
2122        LF". That isn't really right - the whole point of the change was to try to
2123        help when there was an actual mention of one of the two characters. So now
2124        the change happens only if \r or \n (or a literal CR or LF) character is
2125        encountered.
2126    
2127    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2128        of both had grown to the point where there were only 3 bits left.
2129        Fortunately, there was spare space in the data structure, and so I have
2130        moved the internal flags into a new 16-bit field to free up more option
2131        bits.
2132    
2133    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2134        but did not set the internal JCHANGED flag - either of these is enough to
2135        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2136        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2137        start sets both bits.
2138    
2139    4.  Added options (at build time, compile time, exec time) to change \R from
2140        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2141    
2142    5.  doc/pcresyntax.html was missing from the distribution.
2143    
2144    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2145        compatibility, even though it is no longer used.
2146    
2147    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2148        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2149        windows.h file is present (where different names are used). [This was
2150        reversed later after testing - see 16 below.]
2151    
2152    8.  Changed all #include <config.h> to #include "config.h". There were also
2153        some further <pcre.h> cases that I changed to "pcre.h".
2154    
2155    9.  When pcregrep was used with the --colour option, it missed the line ending
2156        sequence off the lines that it output.
2157    
2158    10. It was pointed out to me that arrays of string pointers cause lots of
2159        relocations when a shared library is dynamically loaded. A technique of
2160        using a single long string with a table of offsets can drastically reduce
2161        these. I have refactored PCRE in four places to do this. The result is
2162        dramatic:
2163    
2164          Originally:                          290
2165          After changing UCP table:            187
2166          After changing error message table:   43
2167          After changing table of "verbs"       36
2168          After changing table of Posix names   22
2169    
2170        Thanks to the folks working on Gregex for glib for this insight.
2171    
2172    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2173        unicode-properties was also set.
2174    
2175    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2176    
2177    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2178        checked only for CRLF.
2179    
2180    14. Added casts to pcretest.c to avoid compiler warnings.
2181    
2182    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2183    
2184    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2185        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2186        entirely. This removes changes made in 7 above.
2187    
2188    17. The CMake files have been updated, and there is now more information about
2189        building with CMake in the NON-UNIX-USE document.
2190    
2191    
2192    Version 7.3 28-Aug-07
2193    ---------------------
2194    
2195     1. In the rejigging of the build system that eventually resulted in 7.1, the
2196        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2197        brackets there is not right, since it causes compilers to look for an
2198        installed pcre.h, not the version that is in the source that is being
2199        compiled (which of course may be different). I have changed it back to:
2200    
2201          #include "pcre.h"
2202    
2203        I have a vague recollection that the change was concerned with compiling in
2204        different directories, but in the new build system, that is taken care of
2205        by the VPATH setting the Makefile.
2206    
2207     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2208        when the subject happened to end in the byte 0x85 (e.g. if the last
2209        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2210        characters but of course it shouldn't be taken as a newline when it is part
2211        of another character. The bug was that, for an unlimited repeat of . in
2212        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2213        characters when looking for a newline.
2214    
2215     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2216    
2217     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2218        in debug output.
2219    
2220     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2221        long printing in the pcrecpp unittest when running under MinGW.
2222    
2223     6. ESC_K was left out of the EBCDIC table.
2224    
2225     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2226        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2227        limit also applies to "virtual nesting" when a pattern is recursive, and in
2228        this case 1000 isn't so big. I have been able to remove this limit at the
2229        expense of backing off one optimization in certain circumstances. Normally,
2230        when pcre_exec() would call its internal match() function recursively and
2231        immediately return the result unconditionally, it uses a "tail recursion"
2232        feature to save stack. However, when a subpattern that can match an empty
2233        string has an unlimited repetition quantifier, it no longer makes this
2234        optimization. That gives it a stack frame in which to save the data for
2235        checking that an empty string has been matched. Previously this was taken
2236        from the 1000-entry workspace that had been reserved. So now there is no
2237        explicit limit, but more stack is used.
2238    
2239     8. Applied Daniel's patches to solve problems with the import/export magic
2240        syntax that is required for Windows, and which was going wrong for the
2241        pcreposix and pcrecpp parts of the library. These were overlooked when this
2242        problem was solved for the main library.
2243    
2244     9. There were some crude static tests to avoid integer overflow when computing
2245        the size of patterns that contain repeated groups with explicit upper
2246        limits. As the maximum quantifier is 65535, the maximum group length was
2247        set at 30,000 so that the product of these two numbers did not overflow a
2248        32-bit integer. However, it turns out that people want to use groups that
2249        are longer than 30,000 bytes (though not repeat them that many times).
2250        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2251        made it possible to implement the integer overflow checks in a much more
2252        dynamic way, which I have now done. The artificial limitation on group
2253        length has been removed - we now have only the limit on the total length of
2254        the compiled pattern, which depends on the LINK_SIZE setting.
2255    
2256    10. Fixed a bug in the documentation for get/copy named substring when
2257        duplicate names are permitted. If none of the named substrings are set, the
2258        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2259        empty string.
2260    
2261    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2262        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2263        because the ] is interpreted as the first data character and the
2264        terminating ] is not found. PCRE has been made compatible with Perl in this
2265        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2266        cause memory overwriting.
2267    
2268    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2269        string has been matched (to stop an infinite loop). It was not recognizing
2270        a conditional subpattern that could match an empty string if that
2271        subpattern was within another subpattern. For example, it looped when
2272        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2273        condition was not nested. This bug has been fixed.
2274    
2275    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2276        past the start of the subject in the presence of bytes with the top bit
2277        set, for example "\x8aBCD".
2278    
2279    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2280        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2281    
2282    14. Optimized (?!) to (*FAIL).
2283    
2284    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2285        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2286        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2287        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2288        does: it's just the validity check that is more restrictive.
2289    
2290    16. Inserted checks for integer overflows during escape sequence (backslash)
2291        processing, and also fixed erroneous offset values for syntax errors during
2292        backslash processing.
2293    
2294    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2295        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2296    
2297    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2298        caused an overrun.
2299    
2300    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2301        something other than just ASCII characters) inside a group that had an
2302        unlimited repeat caused a loop at compile time (while checking to see
2303        whether the group could match an empty string).
2304    
2305    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2306        [\P{Any}] did so. (Error in the code for printing property names.)
2307    
2308    21. An orphan \E inside a character class could cause a crash.
2309    
2310    22. A repeated capturing bracket such as (A)? could cause a wild memory
2311        reference during compilation.
2312    
2313    23. There are several functions in pcre_compile() that scan along a compiled
2314        expression for various reasons (e.g. to see if it's fixed length for look
2315        behind). There were bugs in these functions when a repeated \p or \P was
2316        present in the pattern. These operators have additional parameters compared
2317        with \d, etc, and these were not being taken into account when moving along
2318        the compiled data. Specifically:
2319    
2320        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2321            length.
2322    
2323        (b) An item such as \pL+ within a repeated group could cause crashes or
2324            loops.
2325    
2326        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2327            "reference to non-existent subpattern" error.
2328    
2329        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2330    
2331    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2332        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2333    
2334    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2335    
2336    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2337        character were causing crashes (broken optimization).
2338    
2339    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2340        \p or \P) caused a compile-time loop.
2341    
2342    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2343        break. For example, the unstudied pattern [\r\n]A does not match the string
2344        "\r\nA" because change 7.0/46 below moves the current point on by two
2345        characters after failing to match at the start. However, the pattern \nA
2346        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2347        the same is true. There doesn't seem any very clean way out of this, but
2348        what I have chosen to do makes the common cases work: PCRE now takes note
2349        of whether there can be an explicit match for \r or \n anywhere in the
2350        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2351        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2352        pattern has explicit CR or LF references.
2353    
2354    29. Added (*CR) etc for changing newline setting at start of pattern.
2355    
2356    
2357    Version 7.2 19-Jun-07
2358    ---------------------
2359    
2360     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2361        which is apparently normally available under Windows.
2362    
2363     2. Re-jig the pcregrep tests with different newline settings in an attempt
2364        to make them independent of the local environment's newline setting.
2365    
2366     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2367    
2368     4. Some of the "internals" tests were previously cut out when the link size
2369        was not 2, because the output contained actual offsets. The recent new
2370        "Z" feature of pcretest means that these can be cut out, making the tests
2371        usable with all link sizes.
2372    
2373     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2374        stack recursion. This gives a massive performance boost under BSD, but just
2375        a small improvement under Linux. However, it saves one field in the frame
2376        in all cases.
2377    
2378     6. Added more features from the forthcoming Perl 5.10:
2379    
2380        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2381            recursion call. It refers to the nth most recently opened parentheses.
2382    
2383        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2384            to be opened parentheses.
2385    
2386        (c) Conditions that refer to capturing parentheses can be specified
2387            relatively, for example, (?(-2)... or (?(+3)...
2388    
2389        (d) \K resets the start of the current match so that everything before
2390            is not part of it.
2391    
2392        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2393    
2394        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2395            reference syntax.
2396    
2397        (g) (?| introduces a group in which the numbering of parentheses in each
2398            alternative starts with the same number.
2399    
2400        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2401    
2402     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2403        PCRE_INFO_JCHANGED.
2404    
2405     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2406        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2407        for detecting groups that can match an empty string.
2408    
2409     9. A pattern with a very large number of alternatives (more than several
2410        hundred) was running out of internal workspace during the pre-compile
2411        phase, where pcre_compile() figures out how much memory will be needed. A
2412        bit of new cunning has reduced the workspace needed for groups with
2413        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2414        workspace instead of running out of the 4096 that are available.
2415    
2416    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2417    
2418    11. Applied patch from Google to remove an optimization that didn't quite work.
2419        The report of the bug said:
2420    
2421          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2422          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2423          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2424    
2425    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2426        it matched the wrong number of bytes.
2427    
2428    
2429    Version 7.1 24-Apr-07
2430    ---------------------
2431    
2432     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2433        that is more "standard", making use of automake and other Autotools. There
2434        is some re-arrangement of the files and adjustment of comments consequent
2435        on this.
2436    
2437     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2438        for recursive directory scanning broke on some systems because the files
2439        are not scanned in any specific order and on different systems the order
2440        was different. A call to "sort" has been inserted into RunGrepTest for the
2441        approprate test as a short-term fix. In the longer term there may be an
2442        alternative.
2443    
2444     3. I had an email from Eric Raymond about problems translating some of PCRE's
2445        man pages to HTML (despite the fact that I distribute HTML pages, some
2446        people do their own conversions for various reasons). The problems
2447        concerned the use of low-level troff macros .br and .in. I have therefore
2448        removed all such uses from the man pages (some were redundant, some could
2449        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2450        HTML has been updated to handle .nf/.fi and to complain if it encounters
2451        .br or .in.
2452    
2453     4. Updated comments in configure.ac that get placed in config.h.in and also
2454        arranged for config.h to be included in the distribution, with the name
2455        config.h.generic, for the benefit of those who have to compile without
2456        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2457    
2458     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2459        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2460        makevp.bat for the current PCRE, using the additional files
2461        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2462    
2463     6. A Windows user reported a minor discrepancy with test 2, which turned out
2464        to be caused by a trailing space on an input line that had got lost in his
2465        copy. The trailing space was an accident, so I've just removed it.
2466    
2467     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2468        that is needed.
2469    
2470     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2471        as "const" (a) because they are and (b) because it helps the PHP
2472        maintainers who have recently made a script to detect big data structures
2473        in the php code that should be moved to the .rodata section. I remembered
2474        to update Builducptable as well, so it won't revert if ucptable.h is ever
2475        re-created.
2476    
2477     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2478        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2479        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2480        support is not required. This saves 1.5-2K of code, which is important in
2481        some applications.
2482    
2483        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2484        so as not to refer to the tables, even though these functions will never be
2485        called when UTF-8 support is disabled. Otherwise there are problems with a
2486        shared library.
2487    
2488    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2489    
2490        (a) It was defining its arguments as char * instead of void *.
2491    
2492        (b) It was assuming that all moves were upwards in memory; this was true
2493            a long time ago when I wrote it, but is no longer the case.
2494    
2495        The emulated memove() is provided for those environments that have neither
2496        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2497        is clearly not the case, as these two bugs were recently reported.
2498    
2499    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2500        and Detrail to create the HTML documentation, the .txt form of the man
2501        pages, and it removes trailing spaces from listed files. It also creates
2502        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2503        case, it wraps all the #defines with #ifndefs. This script should be run
2504        before "make dist".
2505    
2506    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2507        with Unicode property support.
2508    
2509        (a) For a maximizing quantifier, if the two different cases of the
2510            character were of different lengths in their UTF-8 codings (there are
2511            some cases like this - I found 11), and the matching function had to
2512            back up over a mixture of the two cases, it incorrectly assumed they
2513            were both the same length.
2514    
2515        (b) When PCRE was configured to use the heap rather than the stack for
2516            recursion during matching, it was not correctly preserving the data for
2517            the other case of a UTF-8 character when checking ahead for a match
2518            while processing a minimizing repeat. If the check also involved
2519            matching a wide character, but failed, corruption could cause an
2520            erroneous result when trying to check for a repeat of the original
2521            character.
2522    
2523    13. Some tidying changes to the testing mechanism:
2524    
2525        (a) The RunTest script now detects the internal link size and whether there
2526            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2527            values substituted by "configure". (The RunGrepTest script already did
2528            this for UTF-8.) The configure.ac script no longer substitutes the
2529            relevant variables.
2530    
2531        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2532            with length and offset values. This means that the output is different
2533            for different internal link sizes. Test 2 is skipped for link sizes
2534            other than 2 because of this, bypassing the problem. Unfortunately,
2535            there was also a test in test 3 (the locale tests) that used /B and
2536            failed for link sizes other than 2. Rather than cut the whole test out,
2537            I have added a new /Z option to pcretest that replaces the length and
2538            offset values with spaces. This is now used to make test 3 independent
2539            of link size. (Test 2 will be tidied up later.)
2540    
2541    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2542        segmentation fault instead of returning the appropriate error message.
2543    
2544    15. In multiline mode when the newline sequence was set to "any", the pattern
2545        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2546        This doesn't seem right; it now treats the CRLF combination as the line
2547        ending, and so does not match in that case. It's only a pattern such as ^$
2548        that would hit this one: something like ^ABC$ would have failed after \r
2549        and then tried again after \r\n.
2550    
2551    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2552        in an attempt to make files that differ only in their line terminators
2553        compare equal. This works on Linux.
2554    
2555    17. Under certain error circumstances pcregrep might try to free random memory
2556        as it exited. This is now fixed, thanks to valgrind.
2557    
2558    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2559        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2560        was because its rules for how to advance for /g after matching an empty
2561        string at the end of a line did not allow for this case. They now check for
2562        it specially.
2563    
2564    20. pcretest is supposed to handle patterns and data of any length, by
2565        extending its buffers when necessary. It was getting this wrong when the
2566        buffer for a data line had to be extended.
2567    
2568    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2569        CRLF as a newline sequence.
2570    
2571    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2572        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2573        I have nevertheless tidied it up.
2574    
2575    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2576    
2577    24. Added a man page for pcre-config.
2578    
2579    
2580    Version 7.0 19-Dec-06
2581    ---------------------
2582    
2583     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2584        moving to gcc 4.1.1.
2585    
2586     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2587        sys/time.h, which is documented as needed for this function. It doesn't
2588        seem to matter on Linux, but it showed up on some releases of OS X.
2589    
2590     3. It seems that there are systems where bytes whose values are greater than
2591        127 match isprint() in the "C" locale. The "C" locale should be the
2592        default when a C program starts up. In most systems, only ASCII printing
2593        characters match isprint(). This difference caused the output from pcretest
2594        to vary, making some of the tests fail. I have changed pcretest so that:
2595    
2596        (a) When it is outputting text in the compiled version of a pattern, bytes
2597            other than 32-126 are always shown as hex escapes.
2598    
2599        (b) When it is outputting text that is a matched part of a subject string,
2600            it does the same, unless a different locale has been set for the match
2601            (using the /L modifier). In this case, it uses isprint() to decide.
2602    
2603     4. Fixed a major bug that caused incorrect computation of the amount of memory
2604        required for a compiled pattern when options that changed within the
2605        pattern affected the logic of the preliminary scan that determines the
2606        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2607        that the computed length was too small. The symptoms of this bug were
2608        either the PCRE error "internal error: code overflow" from pcre_compile(),
2609        or a glibc crash with a message such as "pcretest: free(): invalid next
2610        size (fast)". Examples of patterns that provoked this bug (shown in
2611        pcretest format) are:
2612    
2613          /(?-x: )/x
2614          /(?x)(?-x: \s*#\s*)/
2615          /((?i)[\x{c0}])/8
2616          /(?i:[\x{c0}])/8
2617    
2618        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2619        is now done differently.
2620    
2621     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2622        wrapper classes; (b) implement a new function in the C++ scanner that is
2623        more efficient than the old way of doing things because it avoids levels of
2624        recursion in the regex matching; (c) add a paragraph to the documentation
2625        for the FullMatch() function.
2626    
2627     6. The escape sequence \n was being treated as whatever was defined as
2628        "newline". Not only was this contrary to the documentation, which states
2629        that \n is character 10 (hex 0A), but it also went horribly wrong when
2630        "newline" was defined as CRLF. This has been fixed.
2631    
2632     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2633        was being set to -1 for the "end of line" case (supposedly a value that no
2634        character can have). Though this value is never used (the check for end of
2635        line is "zero bytes in current character"), it caused compiler complaints.
2636        I've changed it to 0xffffffff.
2637    
2638     8. In pcre_version.c, the version string was being built by a sequence of
2639        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2640        string (as it is for production releases) called a macro with an empty
2641        argument. The C standard says the result of this is undefined. The gcc
2642        compiler treats it as an empty string (which was what was wanted) but it is
2643        reported that Visual C gives an error. The source has been hacked around to
2644        avoid this problem.
2645    
2646     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2647        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2648        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2649        of them did).
2650    
2651    10. Originally, pcretest opened its input and output without "b"; then I was
2652        told that "b" was needed in some environments, so it was added for release
2653        5.0 to both the input and output. (It makes no difference on Unix-like
2654        systems.) Later I was told that it is wrong for the input on Windows. I've
2655        now abstracted the modes into two macros, to make it easier to fiddle with
2656        them, and removed "b" from the input mode under Windows.
2657    
2658    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2659    
2660    12. Added -help and --help to pcretest as an official way of being reminded
2661        of the options.
2662    
2663    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2664        and pcrecpp.cc because they annoy compilers at high warning levels.
2665    
2666    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2667    
2668    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2669        scripts are not C programs :-) and which was not noticed because it works
2670        on Linux.
2671    
2672    16. pcretest is supposed to handle any length of pattern and data line (as one
2673        line or as a continued sequence of lines) by extending its input buffer if
2674        necessary. This feature was broken for very long pattern lines, leading to
2675        a string of junk being passed to pcre_compile() if the pattern was longer
2676        than about 50K.
2677    
2678    17. I have done a major re-factoring of the way pcre_compile() computes the
2679        amount of memory needed for a compiled pattern. Previously, there was code
2680        that made a preliminary scan of the pattern in order to do this. That was
2681        OK when PCRE was new, but as the facilities have expanded, it has become
2682        harder and harder to keep it in step with the real compile phase, and there
2683        have been a number of bugs (see for example, 4 above). I have now found a
2684        cunning way of running the real compile function in a "fake" mode that
2685        enables it to compute how much memory it would need, while actually only
2686        ever using a few hundred bytes of working memory and without too many
2687        tests of the mode. This should make future maintenance and development
2688        easier. A side effect of this work is that the limit of 200 on the nesting
2689        depth of parentheses has been removed (though this was never a serious
2690        limitation, I suspect). However, there is a downside: pcre_compile() now
2691        runs more slowly than before (30% or more, depending on the pattern). I
2692        hope this isn't a big issue. There is no effect on runtime performance.
2693    
2694    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2695        newline (only possible for the last line of a file) and it was a
2696        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2697    
2698    19. Added additional timing features to pcretest. (1) The -tm option now times
2699        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2700        separate command line item, by a number that specifies the number of
2701        repeats to use when timing. The default is 50000; this gives better
2702        precision, but takes uncomfortably long for very large patterns.
2703    
2704    20. Extended pcre_study() to be more clever in cases where a branch of a
2705        subpattern has no definite first character. For example, (a*|b*)[cd] would
2706        previously give no result from pcre_study(). Now it recognizes that the
2707        first character must be a, b, c, or d.
2708    
2709    21. There was an incorrect error "recursive call could loop indefinitely" if
2710        a subpattern (or the entire pattern) that was being tested for matching an
2711        empty string contained only one non-empty item after a nested subpattern.
2712        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2713        incorrectly, because the \d was being skipped in the check.
2714    
2715    22. The pcretest program now has a new pattern option /B and a command line
2716        option -b, which is equivalent to adding /B to every pattern. This causes
2717        it to show the compiled bytecode, without the additional information that
2718        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2719        is the same as /B/I).
2720    
2721    23. A new optimization is now able automatically to treat some sequences such
2722        as a*b as a*+b. More specifically, if something simple (such as a character
2723        or a simple class like \d) has an unlimited quantifier, and is followed by
2724        something that cannot possibly match the quantified thing, the quantifier
2725        is automatically "possessified".
2726    
2727    24. A recursive reference to a subpattern whose number was greater than 39
2728        went wrong under certain circumstances in UTF-8 mode. This bug could also
2729        have affected the operation of pcre_study().
2730    
2731    25. Realized that a little bit of performance could be had by replacing
2732        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2733    
2734    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2735    
2736    27. Possessive quantifiers such as a++ were previously implemented by turning
2737        them into atomic groups such as ($>a+). Now they have their own opcodes,
2738        which improves performance. This includes the automatically created ones
2739        from 23 above.
2740    
2741    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2742        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2743        the first matched character to be a colon. This applied both to named and
2744        numbered groups.
2745    
2746    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2747    
2748    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2749        building PCRE on Windows easier, so I have included it in the distribution.
2750    
2751    31. There is now a check in pcretest against a ridiculously large number being
2752        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2753        loop, the loop is abandoned.
2754    
2755    32. Forward references to subpatterns in conditions such as (?(2)...) where
2756        subpattern 2 is defined later cause pcre_compile() to search forwards in
2757        the pattern for the relevant set of parentheses. This search went wrong
2758        when there were unescaped parentheses in a character class, parentheses
2759        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2760    
2761    33. "Subroutine" calls and backreferences were previously restricted to
2762        referencing subpatterns earlier in the regex. This restriction has now
2763        been removed.
2764    
2765    34. Added a number of extra features that are going to be in Perl 5.10. On the
2766        whole, these are just syntactic alternatives for features that PCRE had
2767        previously implemented using the Python syntax or my own invention. The
2768        other formats are all retained for compatibility.
2769    
2770        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2771            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2772            also .NET compatible.
2773    
2774        (b) A recursion or subroutine call to a named group can now be defined as
2775            (?&name) as well as (?P>name).
2776    
2777        (c) A backreference to a named group can now be defined as \k<name> or
2778            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2779            5.10, are also .NET compatible.
2780    
2781        (d) A conditional reference to a named group can now use the syntax
2782            (?(<name>) or (?('name') as well as (?(name).
2783    
2784        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2785            groups (named and numbered) that are never evaluated inline, but can be
2786            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2787            is always false. There may be only one alternative in such a group.
2788    
2789        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2790            as the simple (?(R). The condition is true only if the most recent
2791            recursion is that of the given number or name. It does not search out
2792            through the entire recursion stack.
2793    
2794        (g) The escape \gN or \g{N} has been added, where N is a positive or
2795            negative number, specifying an absolute or relative reference.
2796    
2797    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2798        some "unreachable code" warnings.
2799    
2800    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2801        things, this adds five new scripts.
2802    
2803    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2804        There were also incompatibilities regarding the handling of \Q..\E inside
2805        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2806        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2807    
2808    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2809        matches an empty string, and forcibly breaks the loop. There were bugs in
2810        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2811        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2812        separate and independent bugs (that affected different cases) have been
2813        fixed.
2814    
2815    39. Refactored the code to abolish the use of different opcodes for small
2816        capturing bracket numbers. This is a tidy that I avoided doing when I
2817        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2818        The new approach is not only tidier, it makes it possible to reduce the
2819        memory needed to fix the previous bug (38).
2820    
2821    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2822        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2823        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2824        mode.
2825    
2826    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2827        report.
2828    
2829    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2830        copy construction and assignment in the C++ wrapper.
2831    
2832    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2833        couple of bugs that could have given wrong results in the "--newline=crlf"
2834        case.
2835    
2836    44. Added a number of casts and did some reorganization of signed/unsigned int
2837        variables following suggestions from Dair Grant. Also renamed the variable
2838        "this" as "item" because it is a C++ keyword.
2839    
2840    45. Arranged for dftables to add
2841    
2842          #include "pcre_internal.h"
2843    
2844        to pcre_chartables.c because without it, gcc 4.x may remove the array
2845        definition from the final binary if PCRE is built into a static library and
2846        dead code stripping is activated.
2847    
2848    46. For an unanchored pattern, if a match attempt fails at the start of a
2849        newline sequence, and the newline setting is CRLF or ANY, and the next two
2850        characters are CRLF, advance by two characters instead of one.
2851    
2852    
2853    Version 6.7 04-Jul-06
2854    ---------------------
2855    
2856     1. In order to handle tests when input lines are enormously long, pcretest has
2857        been re-factored so that it automatically extends its buffers when
2858        necessary. The code is crude, but this _is_ just a test program. The
2859        default size has been increased from 32K to 50K.
2860    
2861     2. The code in pcre_study() was using the value of the re argument before
2862        testing it for NULL. (Of course, in any sensible call of the function, it
2863        won't be NULL.)
2864    
2865     3. The memmove() emulation function in pcre_internal.h, which is used on
2866        systems that lack both memmove() and bcopy() - that is, hardly ever -
2867        was missing a "static" storage class specifier.
2868    
2869     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2870        containing an extended class (one that cannot be represented by a bitmap
2871        because it contains high-valued characters or Unicode property items, e.g.
2872        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2873        pattern, but PCRE should not loop if you do not (it no longer does).
2874        [Detail: two cases were found: (a) a repeated subpattern containing an
2875        extended class; (b) a recursive reference to a subpattern that followed a
2876        previous extended class. It wasn't skipping over the extended class
2877        correctly when UTF-8 mode was not set.]
2878    
2879     5. A negated single-character class was not being recognized as fixed-length
2880        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2881        compile error "lookbehind assertion is not fixed length".
2882    
2883     6. The RunPerlTest auxiliary script was showing an unexpected difference
2884        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2885        write a Perl script that can interpret lines of an input file either as
2886        byte characters or as UTF-8, which is what "perltest" was being required to
2887        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2888        can't do is switch easily at run time between having the "use utf8;" pragma
2889        or not. In the end, I fudged it by using the RunPerlTest script to insert
2890        "use utf8;" explicitly for the UTF-8 tests.
2891    
2892     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2893        the end of the subject string, contrary to the documentation and to what
2894        Perl does. This was true of both matching functions. Now it matches only at
2895        the start of the subject and immediately after *internal* newlines.
2896    
2897     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2898        a pointer to an int instead of a pointer to an unsigned long int. This
2899        caused problems on 64-bit systems.
2900    
2901     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2902        instance of the 'standard' template library not being so standard".
2903    
2904    10. There was no check on the number of named subpatterns nor the maximum
2905        length of a subpattern name. The product of these values is used to compute
2906        the size of the memory block for a compiled pattern. By supplying a very
2907        long subpattern name and a large number of named subpatterns, the size
2908        computation could be caused to overflow. This is now prevented by limiting
2909        the length of names to 32 characters, and the number of named subpatterns
2910        to 10,000.
2911    
2912    11. Subpatterns that are repeated with specific counts have to be replicated in
2913        the compiled pattern. The size of memory for this was computed from the
2914        length of the subpattern and the repeat count. The latter is limited to
2915        65535, but there was no limit on the former, meaning that integer overflow
2916        could in principle occur. The compiled length of a repeated subpattern is
2917        now limited to 30,000 bytes in order to prevent this.
2918    
2919    12. Added the optional facility to have named substrings with the same name.
2920    
2921    13. Added the ability to use a named substring as a condition, using the
2922        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2923        are numbers (not recommended). Forward references are permitted.
2924    
2925    14. Added forward references in named backreferences (if you see what I mean).
2926    
2927    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2928        pattern could run off the end of the subject. For example, the pattern
2929        "(?s)(.{1,5})"8 did this with the subject "ab".
2930    
2931    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2932        PCRE_CASELESS was set when matching characters that were quantified with ?
2933        or *.
2934    
2935    17. A character class other than a single negated character that had a minimum
2936        but no maximum quantifier - for example [ab]{6,} - was not handled
2937        correctly by pce_dfa_exec(). It would match only one character.
2938    
2939    18. A valid (though odd) pattern that looked like a POSIX character
2940        class but used an invalid character after [ (for example [[,abc,]]) caused
2941        pcre_compile() to give the error "Failed: internal error: code overflow" or
2942        in some cases to crash with a glibc free() error. This could even happen if
2943        the pattern terminated after [[ but there just happened to be a sequence of
2944        letters, a binary zero, and a closing ] in the memory that followed.
2945    
2946    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2947        over the years. Originally (before any Unicode support), just the bottom 8
2948        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2949        output from "man perlunicode" includes this:
2950    
2951          The regular expression compiler produces polymorphic opcodes.  That
2952          is, the pattern adapts to the data and automatically switches to
2953          the Unicode character scheme when presented with Unicode data--or
2954          instead uses a traditional byte scheme when presented with byte
2955          data.
2956    
2957        Sadly, a wide octal escape does not cause a switch, and in a string with
2958        no other multibyte characters, these octal escapes are treated as before.
2959        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2960        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2961        Unicode string.
2962    
2963        I have not perpetrated such confusion in PCRE. Up till now, it took just
2964        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2965        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2966        translate to the appropriate multibyte character.
2967    
2968    29. Applied some refactoring to reduce the number of warnings from Microsoft
2969        and Borland compilers. This has included removing the fudge introduced
2970        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2971        a warning about an unused variable.
2972    
2973    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2974        characters since release 4.0, because Perl (from release 5.004) does not.
2975        [Or at least, is documented not to: some releases seem to be in conflict
2976        with the documentation.] However, when a pattern was studied with
2977        pcre_study() and all its branches started with \s, PCRE still included VT
2978        as a possible starting character. Of course, this did no harm; it just
2979        caused an unnecessary match attempt.
2980    
2981    22. Removed a now-redundant internal flag bit that recorded the fact that case
2982        dependency changed within the pattern. This was once needed for "required
2983        byte" processing, but is no longer used. This recovers a now-scarce options
2984        bit. Also moved the least significant internal flag bit to the most-
2985        significant bit of the word, which was not previously used (hangover from
2986        the days when it was an int rather than a uint) to free up another bit for
2987        the future.
2988    
2989    23. Added support for CRLF line endings as well as CR and LF. As well as the
2990        default being selectable at build time, it can now be changed at runtime
2991        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2992        specify that it is scanning data with non-default line endings.
2993    
2994    24. Changed the definition of CXXLINK to make it agree with the definition of
2995        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2996    
2997    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2998        recursions. This makes a big different to stack usage for some patterns.
2999    
3000    26. If a subpattern containing a named recursion or subroutine reference such
3001        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3002        the space required for the compiled pattern went wrong and gave too small a
3003        value. Depending on the environment, this could lead to "Failed: internal
3004        error: code overflow at offset 49" or "glibc detected double free or
3005        corruption" errors.
3006    
3007    27. Applied patches from Google (a) to support the new newline modes and (b) to
3008        advance over multibyte UTF-8 characters in GlobalReplace.
3009    
3010    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3011        difference for some implementation of PCRE in some Windows version.
3012    
3013    29. Added some extra testing facilities to pcretest:
3014    
3015        \q<number>   in a data line sets the "match limit" value
3016        \Q<number>   in a data line sets the "match recursion limt" value
3017        -S <number>  sets the stack size, where <number> is in megabytes
3018    
3019        The -S option isn't available for Windows.
3020    
3021    
3022    Version 6.6 06-Feb-06
3023    ---------------------
3024    
3025     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3026        in pcreposix.h. I have copied the definition from pcre.h.
3027    
3028     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3029        because pcre.h is no longer a built file.
3030    
3031     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3032        not normally included in the compiled code.
3033    
3034    
3035    Version 6.5 01-Feb-06
3036    ---------------------
3037    
3038     1. When using the partial match feature with pcre_dfa_exec(), it was not
3039        anchoring the second and subsequent partial matches at the new starting
3040        point. This could lead to incorrect results. For example, with the pattern
3041        /1234/, partially matching against "123" and then "a4" gave a match.
3042    
3043     2. Changes to pcregrep:
3044    
3045        (a) All non-match returns from pcre_exec() were being treated as failures
3046            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3047            error message is output. Some extra information is given for the
3048            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3049            probably the only errors that are likely to be caused by users (by
3050            specifying a regex that has nested indefinite repeats, for instance).
3051            If there are more than 20 of these errors, pcregrep is abandoned.
3052    
3053        (b) A binary zero was treated as data while matching, but terminated the
3054            output line if it was written out. This has been fixed: binary zeroes
3055            are now no different to any other data bytes.
3056    
3057        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3058            used to set a locale for matching. The --locale=xxxx long option has
3059            been added (no short equivalent) to specify a locale explicitly on the
3060            pcregrep command, overriding the environment variables.
3061    
3062        (d) When -B was used with -n, some line numbers in the output were one less
3063            than they should have been.
3064    
3065        (e) Added the -o (--only-matching) option.
3066    
3067        (f) If -A or -C was used with -c (count only), some lines of context were
3068            accidentally printed for the final match.
3069    
3070        (g) Added the -H (--with-filename) option.
3071    
3072        (h) The combination of options -rh failed to suppress file names for files
3073            that were found from directory arguments.
3074    
3075        (i) Added the -D (--devices) and -d (--directories) options.
3076    
3077        (j) Added the -F (--fixed-strings) option.
3078    
3079        (k) Allow "-" to be used as a file name for -f as well as for a data file.
3080    
3081        (l) Added the --colo(u)r option.
3082    
3083        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3084            is not present by default.
3085    
3086     3. A nasty bug was discovered in the handling of recursive patterns, that is,
3087        items such as (?R) or (?1), when the recursion could match a number of
3088        alternatives. If it matched one of the alternatives, but subsequently,
3089        outside the recursion, there was a failure, the code tried to back up into
3090        the recursion. However, because of the way PCRE is implemented, this is not
3091        possible, and the result was an incorrect result from the match.
3092    
3093        In order to prevent this happening, the specification of recursion has
3094        been changed so that all such subpatterns are automatically treated as
3095        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3096    
3097     4. I had overlooked the fact that, in some locales, there are characters for
3098        which isalpha() is true but neither isupper() nor islower() are true. In
3099        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3100        and ordfeminine) are like this. This affected the treatment of \w and \W
3101        when they appeared in character classes, but not when they appeared outside
3102        a character class. The bit map for "word" characters is now created
3103        separately from the results of isalnum() instead of just taking it from the
3104        upper, lower, and digit maps. (Plus the underscore character, of course.)
3105    
3106     5. The above bug also affected the handling of POSIX character classes such as
3107        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3108        permanent tables. Instead, the bit maps for such a class were previously
3109        created as the appropriate unions of the upper, lower, and digit bitmaps.
3110        Now they are created by subtraction from the [[:word:]] class, which has
3111        its own bitmap.
3112    
3113     6. The [[:blank:]] character class matches horizontal, but not vertical space.
3114        It is created by subtracting the vertical space characters (\x09, \x0a,
3115        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3116        subtraction was done in the overall bitmap for a character class, meaning
3117        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3118        be recognized. This bug has been fixed.
3119    
3120     7. Patches from the folks at Google:
3121    
3122          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3123          real life, but is still worth protecting against".
3124    
3125          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3126          regular expressions".
3127    
3128          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3129          have it.
3130    
3131          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3132          "configure" and the latter not, in order to fix a problem somebody had
3133          with compiling the Arg class on HP-UX.
3134    
3135          (e) Improve the error-handling of the C++ wrapper a little bit.
3136    
3137          (f) New tests for checking recursion limiting.
3138    
3139     8. The pcre_memmove() function, which is used only if the environment does not
3140        have a standard memmove() function (and is therefore rarely compiled),
3141        contained two bugs: (a) use of int instead of size_t, and (b) it was not
3142        returning a result (though PCRE never actually uses the result).
3143    
3144     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3145        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3146        returned instead of calling malloc() with an overflowing number that would
3147        most likely cause subsequent chaos.
3148    
3149    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3150    
3151    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3152        with this option is matched, the nmatch and pmatch options of regexec() are
3153        ignored.
3154    
3155    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3156        provided in case anyone wants to the the POSIX interface with UTF-8
3157        strings.
3158    
3159    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3160        C++ linking (needed for some HP-UX environments).
3161    
3162    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3163        (unused parameter) and in the pcre_printint() function (omitted "default"
3164        switch label when the default is to do nothing).
3165    
3166    15. Added some code to make it possible, when PCRE is compiled as a C++
3167        library, to replace subject pointers for pcre_exec() with a smart pointer
3168        class, thus making it possible to process discontinuous strings.
3169    
3170    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3171        much the same function. They were added by different people who were trying
3172        to make PCRE easy to compile on non-Unix systems. It has been suggested
3173        that PCRE_EXPORT be abolished now that there is more automatic apparatus
3174        for compiling on Windows systems. I have therefore replaced it with
3175        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3176        defaults to "extern" for C or "extern C" for C++, which works fine on
3177        Unix-like systems. It is now possible to override the value of PCRE_DATA_
3178        SCOPE with something explicit in config.h. In addition:
3179    
3180        (a) pcreposix.h still had just "extern" instead of either of these macros;
3181            I have replaced it with PCRE_DATA_SCOPE.
3182    
3183        (b) Functions such as _pcre_xclass(), which are internal to the library,
3184            but external in the C sense, all had PCRE_EXPORT in their definitions.
3185            This is apparently wrong for the Windows case, so I have removed it.
3186            (It makes no difference on Unix-like systems.)
3187    
3188    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3189        of recursive calls to match(). This is different to MATCH_LIMIT because
3190        that limits the total number of calls to match(), not all of which increase
3191        the depth of recursion. Limiting the recursion depth limits the amount of
3192        stack (or heap if NO_RECURSE is set) that is used. The default can be set
3193        when PCRE is compiled, and changed at run time. A patch from Google adds
3194        this functionality to the C++ interface.
3195    
3196    18. Changes to the handling of Unicode character properties:
3197    
3198        (a) Updated the table to Unicode 4.1.0.
3199    
3200        (b) Recognize characters that are not in the table as "Cn" (undefined).
3201    
3202        (c) I revised the way the table is implemented to a much improved format
3203            which includes recognition of ranges. It now supports the ranges that
3204            are defined in UnicodeData.txt, and it also amalgamates other
3205            characters into ranges. This has reduced the number of entries in the
3206            table from around 16,000 to around 3,000, thus reducing its size
3207            considerably. I realized I did not need to use a tree structure after
3208            all - a binary chop search is just as efficient. Having reduced the
3209            number of entries, I extended their size from 6 bytes to 8 bytes to
3210            allow for more data.
3211    
3212        (d) Added support for Unicode script names via properties such as \p{Han}.
3213    
3214    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3215        matching that character.
3216    
3217    20. When matching a repeated Unicode property with a minimum greater than zero,
3218        (for example \pL{2,}), PCRE could look past the end of the subject if it
3219        reached it while seeking the minimum number of characters. This could
3220        happen only if some of the characters were more than one byte long, because
3221        there is a check for at least the minimum number of bytes.
3222    
3223    21. Refactored the implementation of \p and \P so as to be more general, to
3224        allow for more different types of property in future. This has changed the
3225        compiled form incompatibly. Anybody with saved compiled patterns that use
3226        \p or \P will have to recompile them.
3227    
3228    22. Added "Any" and "L&" to the supported property types.
3229    
3230    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3231        but give a compile time error if the value is greater than 0xff.
3232    
3233    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3234        accidentally not being installed or uninstalled.
3235    
3236    25. The pcre.h file was built from pcre.h.in, but the only changes that were
3237        made were to insert the current release number. This seemed silly, because
3238        it made things harder for people building PCRE on systems that don't run
3239        "configure". I have turned pcre.h into a distributed file, no longer built
3240        by "configure", with the version identification directly included. There is
3241        no longer a pcre.h.in file.
3242    
3243        However, this change necessitated a change to the pcre-config script as
3244        well. It is built from pcre-config.in, and one of the substitutions was the
3245        release number. I have updated configure.ac so that ./configure now finds
3246        the release number by grepping pcre.h.
3247    
3248    26. Added the ability to run the tests under valgrind.
3249    
3250    
3251    Version 6.4 05-Sep-05
3252    ---------------------
3253    
3254     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3255        "--" to be printed when multiple files were scanned, even when none of the
3256        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3257        consider it to be a bug, and have restored the previous behaviour.
3258    
3259     2. A couple of code tidies to get rid of compiler warnings.
3260    
3261     3. The pcretest program used to cheat by referring to symbols in the library
3262        whose names begin with _pcre_. These are internal symbols that are not
3263        really supposed to be visible externally, and in some environments it is
3264        possible to suppress them. The cheating is now confined to including
3265        certain files from the library's source, which is a bit cleaner.
3266    
3267     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3268        file's purpose clearer.
3269    
3270     5. Reorganized pcre_ucp_findchar().
3271    
3272    
3273    Version 6.3 15-Aug-05
3274    ---------------------
3275    
3276     1. The file libpcre.pc.in did not have general read permission in the tarball.
3277    
3278     2. There were some problems when building without C++ support:
3279    
3280        (a) If C++ support was not built, "make install" and "make test" still
3281            tried to test it.
3282    
3283        (b) There were problems when the value of CXX was explicitly set. Some
3284            changes have been made to try to fix these, and ...
3285    
3286        (c) --disable-cpp can now be used to explicitly disable C++ support.
3287    
3288        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3289            backslash in a target when C++ was disabled. This confuses some
3290            versions of "make", apparently. Using an intermediate variable solves
3291            this. (Same for CPP_LOBJ.)
3292    
3293     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3294        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3295        necessary on certain architectures.
3296    
3297     4. Added a setting of -export-symbols-regex to the link command to remove
3298        those symbols that are exported in the C sense, but actually are local
3299        within the library, and not documented. Their names all begin with
3300        "_pcre_". This is not a perfect job, because (a) we have to except some
3301        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3302        available (and never for static libraries). I have made a note to try to
3303        find a way round (a) in the future.
3304    
3305    
3306    Version 6.2 01-Aug-05
3307    ---------------------
3308    
3309     1. There was no test for integer overflow of quantifier values. A construction
3310        such as {1111111111111111} would give undefined results. What is worse, if
3311        a minimum quantifier for a parenthesized subpattern overflowed and became
3312        negative, the calculation of the memory size went wrong. This could have
3313        led to memory overwriting.
3314    
3315     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3316    
3317     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3318        operating environments where this matters.
3319    
3320     4. Applied Giuseppe Maxia's patch to add additional features for controlling
3321        PCRE options from within the C++ wrapper.
3322    
3323     5. Named capturing subpatterns were not being correctly counted when a pattern
3324        was compiled. This caused two problems: (a) If there were more than 100
3325        such subpatterns, the calculation of the memory needed for the whole
3326        compiled pattern went wrong, leading to an overflow error. (b) Numerical
3327        back references of the form \12, where the number was greater than 9, were
3328        not recognized as back references, even though there were sufficient
3329        previous subpatterns.
3330    
3331     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3332        versions of gcc, e.g. 2.95.4.
3333    
3334    
3335    Version 6.1 21-Jun-05
3336    ---------------------
3337    
3338     1. There was one reference to the variable "posix" in pcretest.c that was not
3339        surrounded by "#if !defined NOPOSIX".
3340    
3341     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3342        the cross-check on the old pcre_info() function, for the benefit of the
3343        cut-down version of PCRE that is currently imported into Exim.
3344    
3345     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3346        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3347        patterns that start (?i) but I don't think that matters. The use of (?i) is
3348        just an example; this all applies to the other options as well.
3349    
3350     4. Since libtool seems to echo the compile commands it is issuing, the output
3351        from "make" can be reduced a bit by putting "@" in front of each libtool
3352        compile command.
3353    
3354     5. Patch from the folks at Google for configure.in to be a bit more thorough
3355        in checking for a suitable C++ installation before trying to compile the
3356        C++ stuff. This should fix a reported problem when a compiler was present,
3357        but no suitable headers.
3358    
3359     6. The man pages all had just "PCRE" as their title. I have changed them to
3360        be the relevant file name. I have also arranged that these names are
3361        retained in the file doc/pcre.txt, which is a concatenation in text format
3362        of all the man pages except the little individual ones for each function.
3363    
3364     7. The NON-UNIX-USE file had not been updated for the different set of source
3365        files that come with release 6. I also added a few comments about the C++
3366        wrapper.
3367    
3368    
3369    Version 6.0 07-Jun-05
3370    ---------------------
3371    
3372     1. Some minor internal re-organization to help with my DFA experiments.
3373    
3374     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3375        didn't matter for the library itself when fully configured, but did matter
3376        when compiling without UCP support, or within Exim, where the ucp files are
3377        not imported.
3378    
3379     3. Refactoring of the library code to split up the various functions into
3380        different source modules. The addition of the new DFA matching code (see
3381        below) to a single monolithic source would have made it really too
3382        unwieldy, quite apart from causing all the code to be include in a
3383        statically linked application, when only some functions are used. This is
3384        relevant even without the DFA addition now that patterns can be compiled in
3385        one application and matched in another.
3386    
3387        The downside of splitting up is that there have to be some external
3388        functions and data tables that are used internally in different modules of
3389        the library but which are not part of the API. These have all had their
3390        names changed to start with "_pcre_" so that they are unlikely to clash
3391        with other external names.
3392    
3393     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3394        a different (DFA) algorithm. Although it is slower than the original
3395        function, it does have some advantages for certain types of matching
3396        problem.
3397    
3398     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3399        including restarting after a partial match.
3400    
3401     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3402        defined when compiling for Windows was sent to me. I have put it into the
3403        code, though I have no means of testing or verifying it.
3404    
3405     7. Added the pcre_refcount() auxiliary function.
3406    
3407     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3408        match before or at the first newline in the subject string. In pcretest,
3409        the /f option on a pattern can be used to set this.
3410    
3411     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3412        would behave wrongly. This has been present in PCRE since release 4.0.
3413    
3414    10. A number of changes to the pcregrep command:
3415    
3416        (a) Refactored how -x works; insert ^(...)$ instead of setting
3417            PCRE_ANCHORED and checking the length, in preparation for adding
3418            something similar for -w.
3419    
3420        (b) Added the -w (match as a word) option.
3421    
3422        (c) Refactored the way lines are read and buffered so as to have more
3423            than one at a time available.
3424    
3425        (d) Implemented a pcregrep test script.
3426    
3427        (e) Added the -M (multiline match) option. This allows patterns to match
3428            over several lines of the subject. The buffering ensures that at least
3429            8K, or the rest of the document (whichever is the shorter) is available
3430            for matching (and similarly the previous 8K for lookbehind assertions).
3431    
3432        (f) Changed the --help output so that it now says
3433    
3434              -w, --word-regex(p)
3435    
3436            instead of two lines, one with "regex" and the other with "regexp"
3437            because that confused at least one person since the short forms are the
3438            same. (This required a bit of code, as the output is generated
3439            automatically from a table. It wasn't just a text change.)
3440    
3441        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3442            option but starts with a hyphen. Could be a pattern or a path name
3443            starting with a hyphen, for instance.
3444    
3445        (h) "-" can be given as a file name to represent stdin.
3446    
3447        (i) When file names are being printed, "(standard input)" is used for
3448            the standard input, for compatibility with GNU grep. Previously
3449            "<stdin>" was used.
3450    
3451        (j) The option --label=xxx can be used to supply a name to be used for
3452            stdin when file names are being printed. There is no short form.
3453    
3454        (k) Re-factored the options decoding logic because we are going to add
3455            two more options that take data. Such options can now be given in four
3456            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3457    
3458        (l) Added the -A, -B, and -C options for requesting that lines of context
3459            around matches be printed.
3460    
3461        (m) Added the -L option to print the names of files that do not contain
3462            any matching lines, that is, the complement of -l.
3463    
3464        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3465            continue to scan other files.
3466    
3467        (o) The -s option was incorrectly implemented. For compatibility with other
3468            greps, it now suppresses the error message for a non-existent or non-
3469            accessible file (but not the return code). There is a new option called
3470            -q that suppresses the output of matching lines, which was what -s was
3471            previously doing.
3472    
3473        (p) Added --include and --exclude options to specify files for inclusion
3474            and exclusion when recursing.
3475    
3476    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3477        Hopefully, it now does.
3478    
3479    12. Missing cast in pcre_study().
3480    
3481    13. Added an "uninstall" target to the makefile.
3482    
3483    14. Replaced "extern" in the function prototypes in Makefile.in with
3484        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3485        world, but is set differently for Windows.
3486    
3487    15. Added a second compiling function called pcre_compile2(). The only
3488        difference is that it has an extra argument, which is a pointer to an
3489        integer error code. When there is a compile-time failure, this is set
3490        non-zero, in addition to the error test pointer being set to point to an
3491        error message. The new argument may be NULL if no error number is required
3492        (but then you may as well call pcre_compile(), which is now just a
3493        wrapper). This facility is provided because some applications need a
3494        numeric error indication, but it has also enabled me to tidy up the way
3495        compile-time errors are handled in the POSIX wrapper.
3496    
3497    16. Added VPATH=.libs to the makefile; this should help when building with one
3498        prefix path and installing with another. (Or so I'm told by someone who
3499        knows more about this stuff than I do.)
3500    
3501    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3502        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3503        match everything, including newlines. This is not POSIX-compatible, but
3504        somebody wanted the feature. From pcretest it can be activated by using
3505        both the P and the s flags.
3506    
3507    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3508    
3509    19. libpcre.pc was being incorrectly installed as executable.
3510    
3511    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3512        it now also looks for '\r' so that it will work unmodified on Windows.
3513    
3514    21. Added Google's contributed C++ wrapper to the distribution.
3515    
3516    22. Added some untidy missing memory free() calls in pcretest, to keep
3517        Electric Fence happy when testing.
3518    
3519    
3520    
3521    Version 5.0 13-Sep-04
3522    ---------------------
3523    
3524     1. Internal change: literal characters are no longer packed up into items
3525        containing multiple characters in a single byte-string. Each character
3526        is now matched using a separate opcode. However, there may be more than one
3527        byte in the character in UTF-8 mode.
3528    
3529     2. The pcre_callout_block structure has two new fields: pattern_position and
3530        next_item_length. These contain the offset in the pattern to the next match
3531        item, and its length, respectively.
3532    
3533     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3534        insertion of callouts before each pattern item. Added the /C option to
3535        pcretest to make use of this.
3536    
3537     4. On the advice of a Windows user, the lines
3538    
3539          #if defined(_WIN32) || defined(WIN32)
3540          _setmode( _fileno( stdout ), 0x8000 );
3541          #endif  /* defined(_WIN32) || defined(WIN32) */
3542    
3543        have been added to the source of pcretest. This apparently does useful
3544        magic in relation to line terminators.
3545    
3546     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3547        for the benefit of those environments where the "b" makes a difference.
3548    
3549     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3550        to know about it. I have put a hack into configure.in that adds in code
3551        to set GCC=yes if CC=icc. This seems to end up at a point in the
3552        generated configure script that is early enough to affect the setting of
3553        compiler options, which is what is needed, but I have no means of testing
3554        whether it really works. (The user who reported this had patched the
3555        generated configure script, which of course I cannot do.)
3556    
3557        LATER: After change 22 below (new libtool files), the configure script
3558        seems to know about icc (and also ecc). Therefore, I have commented out
3559        this hack in configure.in.
3560    
3561     7. Added support for pkg-config (2 patches were sent in).
3562    
3563     8. Negated POSIX character classes that used a combination of internal tables
3564        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3565        [[:^ascii]]. Typically, they would match almost any characters. The other
3566        POSIX classes were not broken in this way.
3567    
3568     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3569        to find the match, as PCRE was deluded into thinking that the match had to
3570        start at the start point or following a newline. The same bug applied to
3571        patterns with negative forward assertions or any backward assertions
3572        preceding ".*" at the start, unless the pattern required a fixed first
3573        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3574    
3575    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3576        starting at the last subject character, bytes beyond the end of the subject
3577        string were read.
3578    
3579    11. Renamed the variable "class" as "classbits" to make life easier for C++
3580        users. (Previously there was a macro definition, but it apparently wasn't
3581        enough.)
3582    
3583    12. Added the new field "tables" to the extra data so that tables can be passed
3584        in at exec time, or the internal tables can be re-selected. This allows
3585        a compiled regex to be saved and re-used at a later time by a different
3586        program that might have everything at different addresses.
3587    
3588    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3589        -R library as well as a -L library.
3590    
3591    14. The debugging options of pcretest (-d on the command line or D on a
3592        pattern) showed incorrect output for anything following an extended class
3593        that contained multibyte characters and which was followed by a quantifier.
3594    
3595    15. Added optional support for general category Unicode character properties
3596        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3597        support. It adds about 90K to the size of the library. The meanings of the
3598        inbuilt class escapes such as \d and \s have NOT been changed.
3599    
3600    16. Updated pcredemo.c to include calls to free() to release the memory for the
3601        compiled pattern.
3602    
3603    17. The generated file chartables.c was being created in the source directory
3604        instead of in the building directory. This caused the build to fail if the
3605        source directory was different from the building directory, and was
3606        read-only.
3607    
3608    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3609        file. No doubt somebody will tell me if they don't make sense... Also added
3610        Dan Mooney's comments about building on OpenVMS.
3611    
3612    19. Added support for partial matching via the PCRE_PARTIAL option for
3613        pcre_exec() and the \P data escape in pcretest.
3614    
3615    20. Extended pcretest with 3 new pattern features:
3616    
3617        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3618              write the compiled pattern to the file whose name is "rest-of-line".
3619              This is a straight binary dump of the data, with the saved pointer to
3620              the character tables forced to be NULL. The study data, if any, is
3621              written too. After writing, pcretest reads a new pattern.
3622    
3623        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3624              compiled pattern from the given file. There must not be any
3625              occurrences of "<" in the file name (pretty unlikely); if there are,
3626              pcretest will instead treat the initial "<" as a pattern delimiter.
3627              After reading in the pattern, pcretest goes on to read data lines as
3628              usual.
3629    
3630        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3631              and 16-bit fields in a compiled pattern, to simulate a pattern that
3632              was compiled on a host of opposite endianness.
3633    
3634    21. The pcre-exec() function can now cope with patterns that were compiled on
3635        hosts of opposite endianness, with this restriction:
3636    
3637          As for any compiled expression that is saved and used later, the tables
3638          pointer field cannot be preserved; the extra_data field in the arguments
3639          to pcre_exec() should be used to pass in a tables address if a value
3640          other than the default internal tables were used at compile time.
3641    
3642    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3643        now diagnosed as an error. Previously, most of the time, a negative number
3644        would have been treated as zero, but if in addition "ovector" was passed as
3645        NULL, a crash could occur.
3646    
3647    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3648        new versions from the libtool 1.5 distribution (the last one is a copy of
3649        a file called libtool.m4). This seems to have fixed the need to patch
3650        "configure" to support Darwin 1.3 (which I used to do). However, I still
3651        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3652        workstation).
3653    
3654    24. Changed the PCRE licence to be the more standard "BSD" licence.
3655    
3656    
3657    Version 4.5 01-Dec-03
3658    ---------------------
3659    
3660     1. There has been some re-arrangement of the code for the match() function so
3661        that it can be compiled in a version that does not call itself recursively.
3662        Instead, it keeps those local variables that need separate instances for
3663        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3664        needs to "recurse". Keeping track of where control must go is done by means
3665        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3666        hide most of the details from the main code, and operates only if
3667        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3668        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3669        operating.
3670    
3671        To make it easier for callers to provide specially tailored get/free
3672        functions for this usage, two new functions, pcre_stack_malloc, and
3673        pcre_stack_free, are used. They are always called in strict stacking order,
3674        and the size of block requested is always the same.
3675    
3676        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3677        PCRE has been compiled to use the stack or the heap for recursion. The
3678        -C option of pcretest uses this to show which version is compiled.
3679    
3680        A new data escape \S, is added to pcretest; it causes the amounts of store
3681        obtained and freed by both kinds of malloc/free at match time to be added
3682        to the output.
3683    
3684     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3685        what's available on my current Linux desktop machine.
3686    
3687     3. When matching a UTF-8 string, the test for a valid string at the start has
3688        been extended. If start_offset is not zero, PCRE now checks that it points
3689        to a byte that is the start of a UTF-8 character. If not, it returns
3690        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3691        this is necessary because there may be backward assertions in the pattern.
3692        When matching the same subject several times, it may save resources to use
3693        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3694    
3695     4. The code for checking the validity of UTF-8 strings has been tightened so
3696        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3697        containing "overlong sequences".
3698    
3699     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3700        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3701        should have been "&", but it just so happened that all the cases this let
3702        through by mistake were picked up later in the function.
3703    
3704     6. I had used a variable called "isblank" - this is a C99 function, causing
3705        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3706    
3707     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3708        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3709        so the version is shown; (c) stop "make test" echoing "./RunTest".
3710    
3711     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3712    
3713     9. The prototype for memmove() for systems that don't have it was using
3714        size_t, but the inclusion of the header that defines size_t was later. I've
3715        moved the #includes for the C headers earlier to avoid this.
3716    
3717    10. Added some adjustments to the code to make it easier to compiler on certain
3718        special systems:
3719    
3720          (a) Some "const" qualifiers were missing.
3721          (b) Added the macro EXPORT before all exported functions; by default this
3722              is defined to be empty.
3723          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3724              that it reads its output file name as an argument instead of writing
3725              to the standard output and assuming this can be redirected.
3726    
3727    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3728        class containing characters with values greater than 255, PCRE compilation
3729        went into a loop.
3730    
3731    12. A recursive reference to a subpattern that was within another subpattern
3732        that had a minimum quantifier of zero caused PCRE to crash. For example,
3733        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3734        recursion. If the recursively-called subpattern itself had a zero repeat,
3735        that was OK.
3736    
3737    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3738        buffer into which it was copied (for escape processing) was still set at
3739        1024, so long lines caused crashes.
3740    
3741    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3742        "internal error: code overflow...". This applied to any character class
3743        that was followed by a possessive quantifier.
3744    
3745    15. Modified the Makefile to add libpcre.la as a prerequisite for
3746        libpcreposix.la because I was told this is needed for a parallel build to
3747        work.
3748    
3749    16. If a pattern that contained .* following optional items at the start was
3750        studied, the wrong optimizing data was generated, leading to matching
3751        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3752        matching string must start with a or b or c. The correct conclusion for
3753        this pattern is that a match can start with any character.
3754    
3755    
3756    Version 4.4 13-Aug-03
3757    ---------------------
3758    
3759     1. In UTF-8 mode, a character class containing characters with values between
3760        127 and 255 was not handled correctly if the compiled pattern was studied.
3761        In fixing this, I have also improved the studying algorithm for such
3762        classes (slightly).
3763    
3764     2. Three internal functions had redundant arguments passed to them. Removal
3765        might give a very teeny performance improvement.
3766    
3767     3. Documentation bug: the value of the capture_top field in a callout is *one
3768        more than* the number of the hightest numbered captured substring.
3769    
3770     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3771        in incorrectly linking with a previously installed version. They now link
3772        explicitly with libpcre.la.
3773    
3774     5. configure.in no longer needs to recognize Cygwin specially.
3775    
3776     6. A problem in pcre.in for Windows platforms is fixed.
3777    
3778     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3779        pcretest, it used to include the size of the study block as part of its
3780        output. Unfortunately, the structure contains a field that has a different
3781        size on different hardware architectures. This meant that the tests that
3782        showed this size failed. As the block is currently always of a fixed size,
3783        this information isn't actually particularly useful in pcretest output, so
3784        I have just removed it.
3785    
3786     8. Three pre-processor statements accidentally did not start in column 1.
3787        Sadly, there are *still* compilers around that complain, even though
3788        standard C has not required this for well over a decade. Sigh.
3789<