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

Diff of /code/trunk/ChangeLog

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

revision 61 by nigel, Sat Feb 24 21:39:58 2007 UTC revision 963 by ph10, Sat Apr 21 18:06:31 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 3.0 02-Jan-02  Version 8.31
5    -----------------------------
6    
7    1.  Fixing a wrong JIT test case and some compiler warnings.
8    
9    2.  Removed a bashism from the RunTest script.
10    
11    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12        to unsigned type, result still unsigned" that was given by an MS compiler
13        on encountering the code "-sizeof(xxx)".
14    
15    4.  Partial matching support is added to the JIT compiler.
16    
17    5.  Fixed several bugs concerned with partial matching of items that consist
18        of more than one character:
19    
20        (a) /^(..)\1/ did not partially match "aba" because checking references was
21            done on an "all or nothing" basis. This also applied to repeated
22            references.
23    
24        (b) \R did not give a hard partial match if \r was found at the end of the
25            subject.
26    
27        (c) \X did not give a hard partial match after matching one or more
28            characters at the end of the subject.
29    
30        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
31            a partial match for the string "\r".
32    
33        (e) When newline was set to CRLF, the metacharacter "." did not recognize
34            a partial match for a CR character at the end of the subject string.
35    
36    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
37        running pcretest, the text "(JIT)" added to the output whenever JIT is
38        actually used to run the match.
39    
40    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
41        or /S+[+] with a digit between 1 and 7.
42    
43    8.  OP_NOT now supports any UTF character not just single-byte ones.
44    
45    9.  (*MARK) control verb is now supported by the JIT compiler.
46    
47    10. The command "./RunTest list" lists the available tests without actually
48        running any of them. (Because I keep forgetting what they all are.)
49    
50    11. Add PCRE_INFO_MAXLOOKBEHIND.
51    
52    12. Applied a (slightly modified) user-supplied patch that improves performance
53        when the heap is used for recursion (compiled with --disable-stack-for-
54        recursion). Instead of malloc and free for each heap frame each time a
55        logical recursion happens, frames are retained on a chain and re-used where
56        possible. This sometimes gives as much as 30% improvement.
57    
58    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
59        call.
60    
61    14. As documented, (*COMMIT) is now confined to within a positive assertion.
62    
63    15. It is now possible to link pcretest with libedit as an alternative to
64        libreadline.
65    
66    16. (*COMMIT) control verb is now supported by the JIT compiler.
67    
68    17. The Unicode data tables have been updated to Unicode 6.1.0.
69    
70    18. Added --file-list option to pcregrep.
71    
72    19. Added binary file support to pcregrep, including the -a, --binary-files,
73        -I, and --text options.
74    
75    20. The madvise function is renamed for posix_madvise for QNX compatibility
76        reasons. Fixed by Giuseppe D'Angelo.
77    
78    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
79    
80    22. Changed the option for creating symbolic links for 16-bit man pages from
81        -s to -sf so that re-installing does not cause issues.
82    
83    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
84    
85    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
86        matches in certain environments (the workspace was not being correctly
87        retained). Also added to pcre_dfa_exec() a simple plausibility check on
88        some of the workspace data at the beginning of a restart.
89    
90    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
91        was not doing so when it should - probably a typo introduced by SVN 528
92        (change 8.10/14).
93    
94    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
95        \w+ when the character tables indicated that \x{c4} was a word character.
96        There were several related cases, all because the tests for doing a table
97        lookup were testing for characters less than 127 instead of 255.
98    
99    27. If a pattern contains capturing parentheses that are not used in a match,
100        their slots in the ovector are set to -1. For those that are higher than
101        any matched groups, this happens at the end of processing. In the case when
102        there were back references that the ovector was too small to contain
103        (causing temporary malloc'd memory to be used during matching), and the
104        highest capturing number was not used, memory off the end of the ovector
105        was incorrectly being set to -1. (It was using the size of the temporary
106        memory instead of the true size.)
107    
108    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
109        ovector size, it uses memory at the end of the block that it has got.
110    
111    
112    Version 8.30 04-February-2012
113    -----------------------------
114    
115    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
116        name is defined in ctype.h.
117    
118    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
119        only in quite long subpatterns.
120    
121    3.  Removed the function pcre_info(), which has been obsolete and deprecated
122        since it was replaced by pcre_fullinfo() in February 2000.
123    
124    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
125        match a (*MARK), and the match failed at the start of the subject, a
126        reference to memory before the start of the subject could occur. This bug
127        was introduced by fix 17 of release 8.21.
128    
129    5.  A reference to an unset group with zero minimum repetition was giving
130        totally wrong answers (in non-JavaScript-compatibility mode). For example,
131        /(another)?(\1?)test/ matched against "hello world test". This bug was
132        introduced in release 8.13.
133    
134    6.  Add support for 16-bit character strings (a large amount of work involving
135        many changes and refactorings).
136    
137    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
138        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
139        from a file.
140    
141    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
142        rounding is not applied in this particular case).
143    
144    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
145        if they appear, or are escaped, in patterns.
146    
147    10. Get rid of a number of -Wunused-but-set-variable warnings.
148    
149    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
150        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
151        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
152        also returns the mark "x". This bug applied to capturing parentheses,
153        non-capturing parentheses, and atomic parentheses. It also applied to some
154        assertions.
155    
156    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
157        information out of configure.ac instead of relying on pcre.h.generic, which
158        is not stored in the repository.
159    
160    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
161        -lreadline.
162    
163    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
164    
165    15. Applied Graycode's patch to put the top-level frame on the stack rather
166        than the heap when not using the stack for recursion. This gives a
167        performance improvement in many cases when recursion is not deep.
168    
169    16. Experimental code added to "pcretest -C" to output the stack frame size.
170    
171    
172    Version 8.21 12-Dec-2011
173    ------------------------
174    
175    1.  Updating the JIT compiler.
176    
177    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
178        are added as well.
179    
180    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
181        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
182        calling _pcre_jit_exec. Some extra comments are added.
183    
184    4.  (*MARK) settings inside atomic groups that do not contain any capturing
185        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
186        was introduced by change 18 for 8.20.
187    
188    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
189        ECMA-262 standard.
190    
191    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
192        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
193        This bug was probably introduced by change 9 of 8.13.
194    
195    7.  While fixing 6 above, I noticed that a number of other items were being
196        incorrectly rejected as "not fixed length". This arose partly because newer
197        opcodes had not been added to the fixed-length checking code. I have (a)
198        corrected the bug and added tests for these items, and (b) arranged for an
199        error to occur if an unknown opcode is encountered while checking for fixed
200        length instead of just assuming "not fixed length". The items that were
201        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
202        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
203        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
204    
205    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
206        being incorrectly compiled and would have given unpredicatble results.
207    
208    9.  A possessively repeated subpattern with minimum repeat count greater than
209        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
210        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
211        the first (A) could occur when it should not.
212    
213    10. Add a cast and remove a redundant test from the code.
214    
215    11. JIT should use pcre_malloc/pcre_free for allocation.
216    
217    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
218        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
219        is anything other than /usr, -L is still shown).
220    
221    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
222    
223    14. Perl does not support \N without a following name in a [] class; PCRE now
224        also gives an error.
225    
226    15. If a forward reference was repeated with an upper limit of around 2000,
227        it caused the error "internal error: overran compiling workspace". The
228        maximum number of forward references (including repeats) was limited by the
229        internal workspace, and dependent on the LINK_SIZE. The code has been
230        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
231        the default depends on LINK_SIZE. There is a new upper limit (for safety)
232        of around 200,000 forward references. While doing this, I also speeded up
233        the filling in of repeated forward references.
234    
235    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
236        incorrectly expecting the subject to contain another "a" after the start.
237    
238    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
239        in the match, the SKIP should be ignored. This was not happening; instead
240        the SKIP was being treated as NOMATCH. For patterns such as
241        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
242        tested.
243    
244    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
245        now much more compatible with Perl, in particular in cases where the result
246        is a non-match for a non-anchored pattern. For example, if
247        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
248        "m", where previously it did not return a name. A side effect of this
249        change is that for partial matches, the last encountered mark name is
250        returned, as for non matches. A number of tests that were previously not
251        Perl-compatible have been moved into the Perl-compatible test files. The
252        refactoring has had the pleasing side effect of removing one argument from
253        the match() function, thus reducing its stack requirements.
254    
255    19. If the /S+ option was used in pcretest to study a pattern using JIT,
256        subsequent uses of /S (without +) incorrectly behaved like /S+.
257    
258    21. Retrieve executable code size support for the JIT compiler and fixing
259        some warnings.
260    
261    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
262        not work when the shorter character appeared right at the end of the
263        subject string.
264    
265    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
266        systems.
267    
268    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
269        output it when the /M option is used in pcretest.
270    
271    25. The CheckMan script was not being included in the distribution. Also, added
272        an explicit "perl" to run Perl scripts from the PrepareRelease script
273        because this is reportedly needed in Windows.
274    
275    26. If study data was being save in a file and studying had not found a set of
276        "starts with" bytes for the pattern, the data written to the file (though
277        never used) was taken from uninitialized memory and so caused valgrind to
278        complain.
279    
280    27. Updated RunTest.bat as provided by Sheri Pierce.
281    
282    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
283    
284    29. Computation of memory usage for the table of capturing group names was
285        giving an unnecessarily large value.
286    
287    
288    Version 8.20 21-Oct-2011
289    ------------------------
290    
291    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
292        a POSIX class. After further experiments with Perl, which convinced me that
293        Perl has bugs and confusions, a closing square bracket is no longer allowed
294        in a POSIX name. This bug also affected patterns with classes that started
295        with full stops.
296    
297    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
298        captured substring, but while checking the failing first alternative,
299        substring 1 is temporarily captured. If the output vector supplied to
300        pcre_exec() was not big enough for this capture, the yield of the function
301        was still zero ("insufficient space for captured substrings"). This cannot
302        be totally fixed without adding another stack variable, which seems a lot
303        of expense for a edge case. However, I have improved the situation in cases
304        such as /(a)(b)x|abc/ matched against "abc", where the return code
305        indicates that fewer than the maximum number of slots in the ovector have
306        been set.
307    
308    3.  Related to (2) above: when there are more back references in a pattern than
309        slots in the output vector, pcre_exec() uses temporary memory during
310        matching, and copies in the captures as far as possible afterwards. It was
311        using the entire output vector, but this conflicts with the specification
312        that only 2/3 is used for passing back captured substrings. Now it uses
313        only the first 2/3, for compatibility. This is, of course, another edge
314        case.
315    
316    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
317        main code base, and can be used by building with --enable-jit. When this is
318        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
319        runtime --no-jit option is given.
320    
321    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
322        ovector, the return from the function was zero, implying that there were
323        other matches that did not fit. The correct "exactly full" value is now
324        returned.
325    
326    6.  If a subpattern that was called recursively or as a subroutine contained
327        (*PRUNE) or any other control that caused it to give a non-standard return,
328        invalid errors such as "Error -26 (nested recursion at the same subject
329        position)" or even infinite loops could occur.
330    
331    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
332        computing the minimum length on reaching *ACCEPT, and so ended up with the
333        wrong value of 1 rather than 0. Further investigation indicates that
334        computing a minimum subject length in the presence of *ACCEPT is difficult
335        (think back references, subroutine calls), and so I have changed the code
336        so that no minimum is registered for a pattern that contains *ACCEPT.
337    
338    8.  If (*THEN) was present in the first (true) branch of a conditional group,
339        it was not handled as intended. [But see 16 below.]
340    
341    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
342        Sheri Pierce.
343    
344    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
345        the first byte in a match must be "a".
346    
347    11. Change 17 for 8.13 increased the recursion depth for patterns like
348        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
349        pattern contains any instances of (*THEN). If it does not, the old
350        optimizations are restored. It would be nice to do this on a per-group
351        basis, but at the moment that is not feasible.
352    
353    12. In some environments, the output of pcretest -C is CRLF terminated. This
354        broke RunTest's code that checks for the link size. A single white space
355        character after the value is now allowed for.
356    
357    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
358        For "fr", it uses the Windows-specific input and output files.
359    
360    14. If (*THEN) appeared in a group that was called recursively or as a
361        subroutine, it did not work as intended. [But see next item.]
362    
363    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
364        pattern fragments (but not containing any | characters). If A and B are
365        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
366        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
367        D. In other words, Perl considers parentheses that do not contain any |
368        characters to be part of a surrounding alternative, whereas PCRE was
369        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
370        differently. PCRE now behaves in the same way as Perl, except in the case
371        of subroutine/recursion calls such as (?1) which have in any case always
372        been different (but PCRE had them first :-).
373    
374    16. Related to 15 above: Perl does not treat the | in a conditional group as
375        creating alternatives. Such a group is treated in the same way as an
376        ordinary group without any | characters when processing (*THEN). PCRE has
377        been changed to match Perl's behaviour.
378    
379    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
380        RunGrepTest script failed.
381    
382    18. Change 22 for version 13 caused atomic groups to use more stack. This is
383        inevitable for groups that contain captures, but it can lead to a lot of
384        stack use in large patterns. The old behaviour has been restored for atomic
385        groups that do not contain any capturing parentheses.
386    
387    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
388        suppress the check for a minimum subject length at run time. (If it was
389        given to pcre_exec() or pcre_dfa_exec() it did work.)
390    
391    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
392        fail to work when decoding hex characters in data strings in EBCDIC
393        environments.
394    
395    21. It appears that in at least one Mac OS environment, the isxdigit() function
396        is implemented as a macro that evaluates to its argument more than once,
397        contravening the C 90 Standard (I haven't checked a later standard). There
398        was an instance in pcretest which caused it to go wrong when processing
399        \x{...} escapes in subject strings. The has been rewritten to avoid using
400        things like p++ in the argument of isxdigit().
401    
402    
403    Version 8.13 16-Aug-2011
404    ------------------------
405    
406    1.  The Unicode data tables have been updated to Unicode 6.0.0.
407    
408    2.  Two minor typos in pcre_internal.h have been fixed.
409    
410    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
411        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
412        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
413    
414    4.  There were a number of related bugs in the code for matching backrefences
415        caselessly in UTF-8 mode when codes for the characters concerned were
416        different numbers of bytes. For example, U+023A and U+2C65 are an upper
417        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
418        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
419        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
420        2-byte code at the end of the subject (it thought there wasn't enough data
421        left).
422    
423    5.  Comprehensive information about what went wrong is now returned by
424        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
425        as the output vector has at least 2 elements. The offset of the start of
426        the failing character and a reason code are placed in the vector.
427    
428    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
429        now returned is for the first byte of the failing character, instead of the
430        last byte inspected. This is an incompatible change, but I hope it is small
431        enough not to be a problem. It makes the returned offset consistent with
432        pcre_exec() and pcre_dfa_exec().
433    
434    7.  pcretest now gives a text phrase as well as the error number when
435        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
436        failure, the offset and reason code are output.
437    
438    8.  When \R was used with a maximizing quantifier it failed to skip backwards
439        over a \r\n pair if the subsequent match failed. Instead, it just skipped
440        back over a single character (\n). This seems wrong (because it treated the
441        two characters as a single entity when going forwards), conflicts with the
442        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
443        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
444        has been changed.
445    
446    9.  Some internal refactoring has changed the processing so that the handling
447        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
448        time (the PCRE_DOTALL option was changed this way some time ago: version
449        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
450        which was always a bit of a fudge. It also means that there is one less
451        argument for the match() function, which reduces its stack requirements
452        slightly. This change also fixes an incompatibility with Perl: the pattern
453        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
454    
455    10. More internal refactoring has drastically reduced the number of recursive
456        calls to match() for possessively repeated groups such as (abc)++ when
457        using pcre_exec().
458    
459    11. While implementing 10, a number of bugs in the handling of groups were
460        discovered and fixed:
461    
462        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
463        (a|)*(?1) gave a compile-time internal error.
464        ((a|)+)+  did not notice that the outer group could match an empty string.
465        (^a|^)+   was not marked as anchored.
466        (.*a|.*)+ was not marked as matching at start or after a newline.
467    
468    12. Yet more internal refactoring has removed another argument from the match()
469        function. Special calls to this function are now indicated by setting a
470        value in a variable in the "match data" data block.
471    
472    13. Be more explicit in pcre_study() instead of relying on "default" for
473        opcodes that mean there is no starting character; this means that when new
474        ones are added and accidentally left out of pcre_study(), testing should
475        pick them up.
476    
477    14. The -s option of pcretest has been documented for ages as being an old
478        synonym of -m (show memory usage). I have changed it to mean "force study
479        for every regex", that is, assume /S for every regex. This is similar to -i
480        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
481        using it. It makes it easier to run collections of tests with and without
482        study enabled, and thereby test pcre_study() more easily. All the standard
483        tests are now run with and without -s (but some patterns can be marked as
484        "never study" - see 20 below).
485    
486    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
487        restoration of the capturing data to the outer values was not happening
488        correctly.
489    
490    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
491        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
492        pattern had matched an empty string, and so incorrectly returned a no
493        match.
494    
495    17. There was optimizing code for the last branch of non-capturing parentheses,
496        and also for the obeyed branch of a conditional subexpression, which used
497        tail recursion to cut down on stack usage. Unfortunately, now that there is
498        the possibility of (*THEN) occurring in these branches, tail recursion is
499        no longer possible because the return has to be checked for (*THEN). These
500        two optimizations have therefore been removed. [But see 8.20/11 above.]
501    
502    18. If a pattern containing \R was studied, it was assumed that \R always
503        matched two bytes, thus causing the minimum subject length to be
504        incorrectly computed because \R can also match just one byte.
505    
506    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
507        was incorrectly computed.
508    
509    20. If /S is present twice on a test pattern in pcretest input, it now
510        *disables* studying, thereby overriding the use of -s on the command line
511        (see 14 above). This is necessary for one or two tests to keep the output
512        identical in both cases.
513    
514    21. When (*ACCEPT) was used in an assertion that matched an empty string and
515        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
516    
517    22. When an atomic group that contained a capturing parenthesis was
518        successfully matched, but the branch in which it appeared failed, the
519        capturing was not being forgotten if a higher numbered group was later
520        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
521        group 1 to "a", when in fact it should be unset. This applied to multi-
522        branched capturing and non-capturing groups, repeated or not, and also to
523        positive assertions (capturing in negative assertions does not happen
524        in PCRE) and also to nested atomic groups.
525    
526    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
527        subject after a captured substring, to make it easier to tell which of a
528        number of identical substrings has been captured.
529    
530    24. The way atomic groups are processed by pcre_exec() has been changed so that
531        if they are repeated, backtracking one repetition now resets captured
532        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
533        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
534        "aaa". Previously, it would have been "a". As part of this code
535        refactoring, the way recursive calls are handled has also been changed.
536    
537    25. If an assertion condition captured any substrings, they were not passed
538        back unless some other capturing happened later. For example, if
539        (?(?=(a))a) was matched against "a", no capturing was returned.
540    
541    26. When studying a pattern that contained subroutine calls or assertions,
542        the code for finding the minimum length of a possible match was handling
543        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
544        group 1 called group 2 while simultaneously a separate group 2 called group
545        1). A stack overflow occurred in this case. I have fixed this by limiting
546        the recursion depth to 10.
547    
548    27. Updated RunTest.bat in the distribution to the version supplied by Tom
549        Fortmann. This supports explicit test numbers on the command line, and has
550        argument validation and error reporting.
551    
552    28. An instance of \X with an unlimited repeat could fail if at any point the
553        first character it looked at was a mark character.
554    
555    29. Some minor code refactoring concerning Unicode properties and scripts
556        should reduce the stack requirement of match() slightly.
557    
558    30. Added the '=' option to pcretest to check the setting of unused capturing
559        slots at the end of the pattern, which are documented as being -1, but are
560        not included in the return count.
561    
562    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
563        compiled something random. Now it gives a compile-time error (as does
564        Perl).
565    
566    32. A *MARK encountered during the processing of a positive assertion is now
567        recorded and passed back (compatible with Perl).
568    
569    33. If --only-matching or --colour was set on a pcregrep call whose pattern
570        had alternative anchored branches, the search for a second match in a line
571        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
572        matched the line "0102" twice. The same bug affected patterns that started
573        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
574        twice.
575    
576    34. Previously, PCRE did not allow quantification of assertions. However, Perl
577        does, and because of capturing effects, quantifying parenthesized
578        assertions may at times be useful. Quantifiers are now allowed for
579        parenthesized assertions.
580    
581    35. A minor code tidy in pcre_compile() when checking options for \R usage.
582    
583    36. \g was being checked for fancy things in a character class, when it should
584        just be a literal "g".
585    
586    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
587        appearance of a nested POSIX class supersedes an apparent external class.
588        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
589        unescaped square brackets may also appear as part of class names. For
590        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
591        more like Perl. (But see 8.20/1 above.)
592    
593    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
594        was because it thought it was \N{name}, which is not supported).
595    
596    39. Add minix to OS list not supporting the -S option in pcretest.
597    
598    40. PCRE tries to detect cases of infinite recursion at compile time, but it
599        cannot analyze patterns in sufficient detail to catch mutual recursions
600        such as ((?1))((?2)). There is now a runtime test that gives an error if a
601        subgroup is called recursively as a subpattern for a second time at the
602        same position in the subject string. In previous releases this might have
603        been caught by the recursion limit, or it might have run out of stack.
604    
605    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
606        happen only once. PCRE was, however incorrectly giving a compile time error
607        "recursive call could loop indefinitely" because it cannot analyze the
608        pattern in sufficient detail. The compile time test no longer happens when
609        PCRE is compiling a conditional subpattern, but actual runaway loops are
610        now caught at runtime (see 40 above).
611    
612    42. It seems that Perl allows any characters other than a closing parenthesis
613        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
614        has been changed to be the same.
615    
616    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
617        as not to get warnings when autogen.sh is called. Also changed
618        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
619    
620    44. To help people who use pcregrep to scan files containing exceedingly long
621        lines, the following changes have been made:
622    
623        (a) The default value of the buffer size parameter has been increased from
624            8K to 20K. (The actual buffer used is three times this size.)
625    
626        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
627            PCRE is built.
628    
629        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
630            to be set at run time.
631    
632        (d) Numerical values in pcregrep options can be followed by K or M, for
633            example --buffer-size=50K.
634    
635        (e) If a line being scanned overflows pcregrep's buffer, an error is now
636            given and the return code is set to 2.
637    
638    45. Add a pointer to the latest mark to the callout data block.
639    
640    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
641        partial match of an empty string instead of no match. This was specific to
642        the use of ".".
643    
644    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
645        complete match instead of a partial match. This bug was dependent on both
646        the PCRE_UTF8 and PCRE_DOTALL options being set.
647    
648    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
649        starting byte set, because \b was not being ignored.
650    
651    
652    Version 8.12 15-Jan-2011
653    ------------------------
654    
655    1.  Fixed some typos in the markup of the man pages, and wrote a script that
656        checks for such things as part of the documentation building process.
657    
658    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
659        --match-limit and --recursion-limit options (added for 8.11). In
660        particular, this made one of the standard tests fail. (The integer value
661        went into the wrong half of a long int.)
662    
663    3.  If the --colour option was given to pcregrep with -v (invert match), it
664        did strange things, either producing crazy output, or crashing. It should,
665        of course, ignore a request for colour when reporting lines that do not
666        match.
667    
668    4.  Another pcregrep bug caused similar problems if --colour was specified with
669        -M (multiline) and the pattern match finished with a line ending.
670    
671    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
672        matched in multiline mode, the following line was shown as part of the
673        match. This seems wrong, so I have changed it.
674    
675    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
676        the check for further matches in the same line (so they could be coloured)
677        to overrun the end of the current line. If another match was found, it was
678        incorrectly shown (and then shown again when found in the next line).
679    
680    7.  If pcregrep was compiled under Windows, there was a reference to the
681        function pcregrep_exit() before it was defined. I am assuming this was
682        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
683        reported by a user. I've moved the definition above the reference.
684    
685    
686    Version 8.11 10-Dec-2010
687    ------------------------
688    
689    1.  (*THEN) was not working properly if there were untried alternatives prior
690        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
691        backtracked to try for "b" instead of moving to the next alternative branch
692        at the same level (in this case, to look for "c"). The Perl documentation
693        is clear that when (*THEN) is backtracked onto, it goes to the "next
694        alternative in the innermost enclosing group".
695    
696    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
697        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
698        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
699        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
700        (*THEN).
701    
702    3.  If \s appeared in a character class, it removed the VT character from
703        the class, even if it had been included by some previous item, for example
704        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
705        of \s, but is part of the POSIX "space" class.)
706    
707    4.  A partial match never returns an empty string (because you can always
708        match an empty string at the end of the subject); however the checking for
709        an empty string was starting at the "start of match" point. This has been
710        changed to the "earliest inspected character" point, because the returned
711        data for a partial match starts at this character. This means that, for
712        example, /(?<=abc)def/ gives a partial match for the subject "abc"
713        (previously it gave "no match").
714    
715    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
716        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
717        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
718        has an implication that the given string is incomplete (because a partial
719        match is preferred over a full match). For this reason, these items now
720        give a partial match in this situation. [Aside: previously, the one case
721        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
722        match rather than a full match, which was wrong by the old rules, but is
723        now correct.]
724    
725    6.  There was a bug in the handling of #-introduced comments, recognized when
726        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
727        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
728        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
729        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
730        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
731        places in pcre_compile().
732    
733    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
734        comments when looking ahead for named forward references to subpatterns,
735        the only newline sequence it recognized was NL. It now handles newlines
736        according to the set newline convention.
737    
738    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
739        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
740        cater for a lack of strerror(). These oversights have been fixed.
741    
742    9.  Added --match-limit and --recursion-limit to pcregrep.
743    
744    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
745    
746    11. When the -o option was used, pcregrep was setting a return code of 1, even
747        when matches were found, and --line-buffered was not being honoured.
748    
749    12. Added an optional parentheses number to the -o and --only-matching options
750        of pcregrep.
751    
752    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
753        can match an empty string. The code to do it in pcretest and pcredemo
754        needed fixing:
755    
756        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
757            only one byte after an empty string match just before CRLF (this case
758            just got forgotten; "any" and "anycrlf" were OK).
759    
760        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
761            mode when an empty string match preceded an ASCII character followed by
762            a non-ASCII character. (The code for advancing by one character rather
763            than one byte was nonsense.)
764    
765        (c) The pcredemo.c sample program did not have any code at all to handle
766            the cases when CRLF is a valid newline sequence.
767    
768    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
769        as a starting offset was within the subject string. There is now a new
770        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
771        negative or greater than the length of the string. In order to test this,
772        pcretest is extended to allow the setting of negative starting offsets.
773    
774    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
775        starting offset points to the beginning of a UTF-8 character was
776        unnecessarily clumsy. I tidied it up.
777    
778    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
779        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
780    
781    17. Nobody had reported that the --include_dir option, which was added in
782        release 7.7 should have been called --include-dir (hyphen, not underscore)
783        for compatibility with GNU grep. I have changed it to --include-dir, but
784        left --include_dir as an undocumented synonym, and the same for
785        --exclude-dir, though that is not available in GNU grep, at least as of
786        release 2.5.4.
787    
788    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
789        characters from a string of bytes) have been redefined so as not to use
790        loops, in order to improve performance in some environments. At the same
791        time, I abstracted some of the common code into auxiliary macros to save
792        repetition (this should not affect the compiled code).
793    
794    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
795        compile-time error is now given if \c is not followed by an ASCII
796        character, that is, a byte less than 128. (In EBCDIC mode, the code is
797        different, and any byte value is allowed.)
798    
799    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
800        START_OPTIMIZE option, which is now allowed at compile time - but just
801        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
802        to pcregrep and other applications that have no direct access to PCRE
803        options. The new /Y option in pcretest sets this option when calling
804        pcre_compile().
805    
806    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
807        back references. Groups containing recursive back references were forced to
808        be atomic by that change, but in the case of named groups, the amount of
809        memory required was incorrectly computed, leading to "Failed: internal
810        error: code overflow". This has been fixed.
811    
812    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
813        pcretest.c, to avoid build problems in some Borland environments.
814    
815    
816    Version 8.10 25-Jun-2010
817    ------------------------
818    
819    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
820        THEN.
821    
822    2.  (*ACCEPT) was not working when inside an atomic group.
823    
824    3.  Inside a character class, \B is treated as a literal by default, but
825        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
826        causes the error). The code is unchanged, but I tidied the documentation.
827    
828    4.  Inside a character class, PCRE always treated \R and \X as literals,
829        whereas Perl faults them if its -w option is set. I have changed PCRE so
830        that it faults them when PCRE_EXTRA is set.
831    
832    5.  Added support for \N, which always matches any character other than
833        newline. (It is the same as "." when PCRE_DOTALL is not set.)
834    
835    6.  When compiling pcregrep with newer versions of gcc which may have
836        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
837        declared with attribute warn_unused_result" were given. Just casting the
838        result to (void) does not stop the warnings; a more elaborate fudge is
839        needed. I've used a macro to implement this.
840    
841    7.  Minor change to pcretest.c to avoid a compiler warning.
842    
843    8.  Added four artifical Unicode properties to help with an option to make
844        \s etc use properties (see next item). The new properties are: Xan
845        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
846    
847    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
848        use Unicode properties. (*UCP) at the start of a pattern can be used to set
849        this option. Modified pcretest to add /W to test this facility. Added
850        REG_UCP to make it available via the POSIX interface.
851    
852    10. Added --line-buffered to pcregrep.
853    
854    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
855        studied, and the match started with a letter with a code point greater than
856        127 whose first byte was different to the first byte of the other case of
857        the letter, the other case of this starting letter was not recognized
858        (#976).
859    
860    12. If a pattern that was studied started with a repeated Unicode property
861        test, for example, \p{Nd}+, there was the theoretical possibility of
862        setting up an incorrect bitmap of starting bytes, but fortunately it could
863        not have actually happened in practice until change 8 above was made (it
864        added property types that matched character-matching opcodes).
865    
866    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
867        possible starting bytes for non-anchored patterns.
868    
869    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
870        \R, and also a number of cases that involve Unicode properties, both
871        explicit and implicit when PCRE_UCP is set.
872    
873    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
874        input, it could crash or give wrong results if characters with values
875        greater than 0xc0 were present in the subject string. (Detail: it assumed
876        UTF-8 input when processing these items.)
877    
878    16. Added a lot of (int) casts to avoid compiler warnings in systems where
879        size_t is 64-bit (#991).
880    
881    17. Added a check for running out of memory when PCRE is compiled with
882        --disable-stack-for-recursion (#990).
883    
884    18. If the last data line in a file for pcretest does not have a newline on
885        the end, a newline was missing in the output.
886    
887    19. The default pcre_chartables.c file recognizes only ASCII characters (values
888        less than 128) in its various bitmaps. However, there is a facility for
889        generating tables according to the current locale when PCRE is compiled. It
890        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
891        characters, are recognized by isspace() and therefore were getting set in
892        these tables, and indeed these tables seem to approximate to ISO 8859. This
893        caused a problem in UTF-8 mode when pcre_study() was used to create a list
894        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
895        which of course cannot start UTF-8 characters. I have changed the code so
896        that only real ASCII characters (less than 128) and the correct starting
897        bytes for UTF-8 encodings are set for characters greater than 127 when in
898        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
899        altogether.)
900    
901    20. Added the /T option to pcretest so as to be able to run tests with non-
902        standard character tables, thus making it possible to include the tests
903        used for 19 above in the standard set of tests.
904    
905    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
906        reference to a subpattern the other side of a comment that contains an
907        opening parenthesis caused either an internal compiling error, or a
908        reference to the wrong subpattern.
909    
910    
911    Version 8.02 19-Mar-2010
912    ------------------------
913    
914    1.  The Unicode data tables have been updated to Unicode 5.2.0.
915    
916    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
917        configured.
918    
919    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
920        original author of that file, following a query about its status.
921    
922    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
923        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
924    
925    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
926        quantifier applied to a forward-referencing subroutine call, could compile
927        incorrect code or give the error "internal error: previously-checked
928        referenced subpattern not found".
929    
930    6.  Both MS Visual Studio and Symbian OS have problems with initializing
931        variables to point to external functions. For these systems, therefore,
932        pcre_malloc etc. are now initialized to local functions that call the
933        relevant global functions.
934    
935    7.  There were two entries missing in the vectors called coptable and poptable
936        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
937        I've fixed the data, and added a kludgy way of testing at compile time that
938        the lengths are correct (equal to the number of opcodes).
939    
940    8.  Following on from 7, I added a similar kludge to check the length of the
941        eint vector in pcreposix.c.
942    
943    9.  Error texts for pcre_compile() are held as one long string to avoid too
944        much relocation at load time. To find a text, the string is searched,
945        counting zeros. There was no check for running off the end of the string,
946        which could happen if a new error number was added without updating the
947        string.
948    
949    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
950    
951    11. \K was not working if it appeared in an atomic group or in a group that
952        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
953        \K is "not well defined" if used in an assertion. PCRE now accepts it if
954        the assertion is positive, but not if it is negative.
955    
956    12. Change 11 fortuitously reduced the size of the stack frame used in the
957        "match()" function of pcre_exec.c by one pointer. Forthcoming
958        implementation of support for (*MARK) will need an extra pointer on the
959        stack; I have reserved it now, so that the stack frame size does not
960        decrease.
961    
962    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
963        item in branch that calls a recursion is a subroutine call - as in the
964        second branch in the above example - was incorrectly given the compile-
965        time error "recursive call could loop indefinitely" because pcre_compile()
966        was not correctly checking the subroutine for matching a non-empty string.
967    
968    14. The checks for overrunning compiling workspace could trigger after an
969        overrun had occurred. This is a "should never occur" error, but it can be
970        triggered by pathological patterns such as hundreds of nested parentheses.
971        The checks now trigger 100 bytes before the end of the workspace.
972    
973    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
974    
975    
976    Version 8.01 19-Jan-2010
977    ------------------------
978    
979    1.  If a pattern contained a conditional subpattern with only one branch (in
980        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
981        computed the wrong minimum data length (which is of course zero for such
982        subpatterns). This could cause incorrect "no match" results.
983    
984    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
985        the pattern is reset in the first branch, pcre_compile() failed with
986        "internal error: code overflow at offset...". This happened only when
987        the reset was to the original external option setting. (An optimization
988        abstracts leading options settings into an external setting, which was the
989        cause of this.)
990    
991    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
992        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
993        assertion pattern did not match (meaning that the assertion was true), it
994        was incorrectly treated as false if the SKIP had been reached during the
995        matching. This also applied to assertions used as conditions.
996    
997    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
998        assertion subpattern, including such a pattern used as a condition,
999        unpredictable results occurred, instead of the error return
1000        PCRE_ERROR_DFA_UITEM.
1001    
1002    5.  The C++ GlobalReplace function was not working like Perl for the special
1003        situation when an empty string is matched. It now does the fancy magic
1004        stuff that is necessary.
1005    
1006    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1007        removed. (These were left over from very, very early versions of PCRE.)
1008    
1009    7.  Some cosmetic changes to the code to make life easier when compiling it
1010        as part of something else:
1011    
1012        (a) Change DEBUG to PCRE_DEBUG.
1013    
1014        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1015            called "current" as "current_branch", to prevent a collision with the
1016            Linux macro when compiled as a kernel module.
1017    
1018        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1019            prevent a collision with the Linux macro when compiled as a kernel
1020            module.
1021    
1022    8.  In pcre_compile() there are some checks for integer overflows that used to
1023        cast potentially large values to (double). This has been changed to that
1024        when building, a check for int64_t is made, and if it is found, it is used
1025        instead, thus avoiding the use of floating point arithmetic. (There is no
1026        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1027        double.
1028    
1029    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1030        2005 (difference between two addresses compared to an unsigned value).
1031    
1032    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1033        custom one, because of the following reported problem in Windows:
1034    
1035          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1036              under Win32.
1037          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1038              therefore missing the function definition.
1039          - The compiler thus generates a "C" signature for the test function.
1040          - The linker fails to find the "C" function.
1041          - PCRE fails to configure if asked to do so against libbz2.
1042    
1043    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1044        messages were output:
1045    
1046          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1047          rerunning libtoolize, to keep the correct libtool macros in-tree.
1048          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1049    
1050        I have done both of these things.
1051    
1052    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1053        most of the time, it *can* run out if it is given a pattern that contains a
1054        runaway infinite recursion. I updated the discussion in the pcrestack man
1055        page.
1056    
1057    13. Now that we have gone to the x.xx style of version numbers, the minor
1058        version may start with zero. Using 08 or 09 is a bad idea because users
1059        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1060        interpreted as invalid octal numbers. I've updated the previous comment in
1061        configure.ac, and also added a check that gives an error if 08 or 09 are
1062        used.
1063    
1064    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1065        causing partial matching to fail when the end of the subject matched \W
1066        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1067    
1068    15. There were some discrepancies between the declarations in pcre_internal.h
1069        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1070        their definitions. The declarations used "const uschar *" and the
1071        definitions used USPTR. Even though USPTR is normally defined as "const
1072        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1073        reported that: "This difference in casting confuses some C++ compilers, for
1074        example, SunCC recognizes above declarations as different functions and
1075        generates broken code for hbpcre." I have changed the declarations to use
1076        USPTR.
1077    
1078    16. GNU libtool is named differently on some systems. The autogen.sh script now
1079        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1080        (FreeBSD).
1081    
1082    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1083        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1084        comment: "Figure out how to create a longlong from a string: strtoll and
1085        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1086        instance, but it only takes 2 args instead of 3!"
1087    
1088    18. A subtle bug concerned with back references has been fixed by a change of
1089        specification, with a corresponding code fix. A pattern such as
1090        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1091        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1092        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1093        same bug. Such groups have to be quantified to be useful, or contained
1094        inside another quantified group. (If there's no repetition, the reference
1095        can never match.) The problem arises because, having left the group and
1096        moved on to the rest of the pattern, a later failure that backtracks into
1097        the group uses the captured value from the final iteration of the group
1098        rather than the correct earlier one. I have fixed this in PCRE by forcing
1099        any group that contains a reference to itself to be an atomic group; that
1100        is, there cannot be any backtracking into it once it has completed. This is
1101        similar to recursive and subroutine calls.
1102    
1103    
1104    Version 8.00 19-Oct-09
1105    ----------------------
1106    
1107    1.  The table for translating pcre_compile() error codes into POSIX error codes
1108        was out-of-date, and there was no check on the pcre_compile() error code
1109        being within the table. This could lead to an OK return being given in
1110        error.
1111    
1112    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1113        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1114        in a Windows environment.
1115    
1116    3.  The pcregrep --count option prints the count for each file even when it is
1117        zero, as does GNU grep. However, pcregrep was also printing all files when
1118        --files-with-matches was added. Now, when both options are given, it prints
1119        counts only for those files that have at least one match. (GNU grep just
1120        prints the file name in this circumstance, but including the count seems
1121        more useful - otherwise, why use --count?) Also ensured that the
1122        combination -clh just lists non-zero counts, with no names.
1123    
1124    4.  The long form of the pcregrep -F option was incorrectly implemented as
1125        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1126        but it seems right to fix it, and I didn't think it was worth preserving
1127        the old behaviour.
1128    
1129    5.  The command line items --regex=pattern and --regexp=pattern were not
1130        recognized by pcregrep, which required --regex pattern or --regexp pattern
1131        (with a space rather than an '='). The man page documented the '=' forms,
1132        which are compatible with GNU grep; these now work.
1133    
1134    6.  No libpcreposix.pc file was created for pkg-config; there was just
1135        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1136    
1137    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1138        when UCP support is not needed, by modifying the Python script that
1139        generates it from Unicode data files. This should not matter if the module
1140        is correctly used as a library, but I received one complaint about 50K of
1141        unwanted data. My guess is that the person linked everything into his
1142        program rather than using a library. Anyway, it does no harm.
1143    
1144    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1145        was a minimum greater than 1 for a wide character in a possessive
1146        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1147        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1148        character. Chaos in the form of incorrect output or a compiling loop could
1149        result.
1150    
1151    9.  The restrictions on what a pattern can contain when partial matching is
1152        requested for pcre_exec() have been removed. All patterns can now be
1153        partially matched by this function. In addition, if there are at least two
1154        slots in the offset vector, the offset of the earliest inspected character
1155        for the match and the offset of the end of the subject are set in them when
1156        PCRE_ERROR_PARTIAL is returned.
1157    
1158    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1159        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1160        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1161        and may be more useful for multi-segment matching.
1162    
1163    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1164        used to be given if ever the end of the subject was reached; now it is
1165        given only if matching could not proceed because another character was
1166        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1167        string "Z", which now yields "no match" instead of "partial match". In the
1168        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1169        final character ended with (*FAIL).
1170    
1171    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1172        if the pattern had a "must contain" character that was already found in the
1173        earlier partial match, unless partial matching was again requested. For
1174        example, with the pattern /dog.(body)?/, the "must contain" character is
1175        "g". If the first part-match was for the string "dog", restarting with
1176        "sbody" failed. This bug has been fixed.
1177    
1178    13. The string returned by pcre_dfa_exec() after a partial match has been
1179        changed so that it starts at the first inspected character rather than the
1180        first character of the match. This makes a difference only if the pattern
1181        starts with a lookbehind assertion or \b or \B (\K is not supported by
1182        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1183        matching functions compatible, and I think it's the right thing to do.
1184    
1185    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1186        so that the demonstration program is easily available in environments where
1187        PCRE has not been installed from source.
1188    
1189    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1190        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1191        library.
1192    
1193    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1194        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1195        is not the first non-POSIX option to be added. Clearly some people find
1196        these options useful.
1197    
1198    17. If a caller to the POSIX matching function regexec() passes a non-zero
1199        value for nmatch with a NULL value for pmatch, the value of
1200        nmatch is forced to zero.
1201    
1202    18. RunGrepTest did not have a test for the availability of the -u option of
1203        the diff command, as RunTest does. It now checks in the same way as
1204        RunTest, and also checks for the -b option.
1205    
1206    19. If an odd number of negated classes containing just a single character
1207        interposed, within parentheses, between a forward reference to a named
1208        subpattern and the definition of the subpattern, compilation crashed with
1209        an internal error, complaining that it could not find the referenced
1210        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1211        [The bug was that it was starting one character too far in when skipping
1212        over the character class, thus treating the ] as data rather than
1213        terminating the class. This meant it could skip too much.]
1214    
1215    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1216        /g option in pcretest when the pattern contains \K, which makes it possible
1217        to have an empty string match not at the start, even when the pattern is
1218        anchored. Updated pcretest and pcredemo to use this option.
1219    
1220    21. If the maximum number of capturing subpatterns in a recursion was greater
1221        than the maximum at the outer level, the higher number was returned, but
1222        with unset values at the outer level. The correct (outer level) value is
1223        now given.
1224    
1225    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1226        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1227        make it do so. The string so far is captured, making this feature
1228        compatible with Perl.
1229    
1230    23. The tests have been re-organized, adding tests 11 and 12, to make it
1231        possible to check the Perl 5.10 features against Perl 5.10.
1232    
1233    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1234        pattern matches a fixed length string. PCRE did not allow this; now it
1235        does. Neither allows recursion.
1236    
1237    25. I finally figured out how to implement a request to provide the minimum
1238        length of subject string that was needed in order to match a given pattern.
1239        (It was back references and recursion that I had previously got hung up
1240        on.) This code has now been added to pcre_study(); it finds a lower bound
1241        to the length of subject needed. It is not necessarily the greatest lower
1242        bound, but using it to avoid searching strings that are too short does give
1243        some useful speed-ups. The value is available to calling programs via
1244        pcre_fullinfo().
1245    
1246    26. While implementing 25, I discovered to my embarrassment that pcretest had
1247        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1248        study optimizations had never been tested with that matching function.
1249        Oops. What is worse, even when it was passed study data, there was a bug in
1250        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1251        were also very few tests of studied patterns with pcre_dfa_exec().
1252    
1253    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1254        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1255        on the other side of the coin, they are no longer allowed to have different
1256        names, because these cannot be distinguished in PCRE, and this has caused
1257        confusion. (This is a difference from Perl.)
1258    
1259    28. When duplicate subpattern names are present (necessarily with different
1260        numbers, as required by 27 above), and a test is made by name in a
1261        conditional pattern, either for a subpattern having been matched, or for
1262        recursion in such a pattern, all the associated numbered subpatterns are
1263        tested, and the overall condition is true if the condition is true for any
1264        one of them. This is the way Perl works, and is also more like the way
1265        testing by number works.
1266    
1267    
1268    Version 7.9 11-Apr-09
1269    ---------------------
1270    
1271    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1272        (pcretest), all targets were linked against these libraries. This included
1273        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1274        libraries. This caused unwanted dependencies to be created. This problem
1275        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1276        pcretest is linked with readline.
1277    
1278    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1279        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1280        moved outside it again, because FALSE and TRUE are already defined in AIX,
1281        but BOOL is not.
1282    
1283    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1284        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1285    
1286    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1287        hyphens) following file names and line numbers when outputting matching
1288        lines. This is not true; no spaces are inserted. I have also clarified the
1289        wording for the --colour (or --color) option.
1290    
1291    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1292        was not coloured; this is different to GNU grep, so I have changed it to be
1293        the same.
1294    
1295    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1296        each matching line was coloured. Now it goes on to look for further matches
1297        of any of the test patterns, which is the same behaviour as GNU grep.
1298    
1299    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1300        doesn't make sense to accept an empty string match in pcregrep, so I have
1301        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1302        seems to be how GNU grep behaves.
1303    
1304    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1305        start or after a newline", because the conditional assertion was not being
1306        correctly handled. The rule now is that both the assertion and what follows
1307        in the first alternative must satisfy the test.
1308    
1309    9.  If auto-callout was enabled in a pattern with a conditional group whose
1310        condition was an assertion, PCRE could crash during matching, both with
1311        pcre_exec() and pcre_dfa_exec().
1312    
1313    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1314        used for matching.
1315    
1316    11. Unicode property support in character classes was not working for
1317        characters (bytes) greater than 127 when not in UTF-8 mode.
1318    
1319    12. Added the -M command line option to pcretest.
1320    
1321    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1322    
1323    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1324    
1325    16. Added comments and documentation about mis-use of no_arg in the C++
1326        wrapper.
1327    
1328    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1329        from Martin Jerabek that uses macro names for all relevant character and
1330        string constants.
1331    
1332    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1333        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1334        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1335        these, but not everybody uses configure.
1336    
1337    19. A conditional group that had only one branch was not being correctly
1338        recognized as an item that could match an empty string. This meant that an
1339        enclosing group might also not be so recognized, causing infinite looping
1340        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1341        with the subject "ab", where knowledge that the repeated group can match
1342        nothing is needed in order to break the loop.
1343    
1344    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1345        exec(), but without supplying a callout function, matching went wrong.
1346    
1347    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1348        leak if the size of the offset vector was greater than 30. When the vector
1349        is smaller, the saved offsets during recursion go onto a local stack
1350        vector, but for larger vectors malloc() is used. It was failing to free
1351        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1352        error, in fact).
1353    
1354    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1355        heapframe that is used only when UTF-8 support is enabled. This caused no
1356        problem, but was untidy.
1357    
1358    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1359        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1360        included within another project.
1361    
1362    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1363        slightly modified by me:
1364    
1365          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1366              not building pcregrep.
1367    
1368          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1369              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1370    
1371    25. Forward references, both numeric and by name, in patterns that made use of
1372        duplicate group numbers, could behave incorrectly or give incorrect errors,
1373        because when scanning forward to find the reference group, PCRE was not
1374        taking into account the duplicate group numbers. A pattern such as
1375        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1376    
1377    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1378        the feature of a custom pointer more persuasive (as requested by a user).
1379    
1380    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1381        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1382        pre-defined.
1383    
1384    28. Added support for (*UTF8) at the start of a pattern.
1385    
1386    29. Arrange for flags added by the "release type" setting in CMake to be shown
1387        in the configuration summary.
1388    
1389    
1390    Version 7.8 05-Sep-08
1391    ---------------------
1392    
1393    1.  Replaced UCP searching code with optimized version as implemented for Ad
1394        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1395        stage table and inline lookup instead of a function, giving speed ups of 2
1396        to 5 times on some simple patterns that I tested. Permission was given to
1397        distribute the MultiStage2.py script that generates the tables (it's not in
1398        the tarball, but is in the Subversion repository).
1399    
1400    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1401        scripts.
1402    
1403    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1404        a group with a zero qualifier. The result of the study could be incorrect,
1405        or the function might crash, depending on the pattern.
1406    
1407    4.  Caseless matching was not working for non-ASCII characters in back
1408        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1409        It now works when Unicode Property Support is available.
1410    
1411    5.  In pcretest, an escape such as \x{de} in the data was always generating
1412        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1413        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1414        truncation.
1415    
1416    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1417    
1418    7.  Added two (int) casts to pcregrep when printing the difference of two
1419        pointers, in case they are 64-bit values.
1420    
1421    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1422        test 2 if it fails.
1423    
1424    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1425        and a #define of that name to empty if it is not externally set. This is to
1426        allow users of MSVC to set it if necessary.
1427    
1428    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1429        the convenience functions in the pcre_get.c source file.
1430    
1431    11. An option change at the start of a pattern that had top-level alternatives
1432        could cause overwriting and/or a crash. This command provoked a crash in
1433        some environments:
1434    
1435          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1436    
1437        This potential security problem was recorded as CVE-2008-2371.
1438    
1439    12. For a pattern where the match had to start at the beginning or immediately
1440        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1441        pcre_dfa_exec() could read past the end of the passed subject if there was
1442        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1443        pcretest so that it places the subject at the end of its malloc-ed buffer.
1444    
1445    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1446        exec() might read past the end of the data buffer in UTF-8 mode.
1447    
1448    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1449        the data contained the byte 0x85 as part of a UTF-8 character within its
1450        first line. This applied both to normal and DFA matching.
1451    
1452    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1453        /^[^d]*?$/8 failed to match "abc".
1454    
1455    16. Added a missing copyright notice to pcrecpp_internal.h.
1456    
1457    17. Make it more clear in the documentation that values returned from
1458        pcre_exec() in ovector are byte offsets, not character counts.
1459    
1460    18. Tidied a few places to stop certain compilers from issuing warnings.
1461    
1462    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1463        supplied by Stefan Weber. I made a further small update for 7.8 because
1464        there is a change of source arrangements: the pcre_searchfuncs.c module is
1465        replaced by pcre_ucd.c.
1466    
1467    
1468    Version 7.7 07-May-08
1469    ---------------------
1470    
1471    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1472        a string to a long long, pretend we don't even have a long long." This is
1473        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1474    
1475    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1476        pre-7.6 versions, which defined a global no_arg variable instead of putting
1477        it in the RE class. (See also #8 below.)
1478    
1479    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1480        Lopes.
1481    
1482    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1483    
1484        (1) The include/exclude patterns were being applied to the whole pathnames
1485            of files, instead of just to the final components.
1486    
1487        (2) If there was more than one level of directory, the subdirectories were
1488            skipped unless they satisfied the include/exclude conditions. This is
1489            inconsistent with GNU grep (and could even be seen as contrary to the
1490            pcregrep specification - which I improved to make it absolutely clear).
1491            The action now is always to scan all levels of directory, and just
1492            apply the include/exclude patterns to regular files.
1493    
1494    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1495        --exclude_dir in the tests to avoid scanning .svn directories.
1496    
1497    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1498        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1499        doesn't support NULs in patterns.
1500    
1501    7.  Added some missing "const"s to declarations of static tables in
1502        pcre_compile.c and pcre_dfa_exec.c.
1503    
1504    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1505        caused by fix #2  above. (Subsequently also a second patch to fix the
1506        first patch. And a third patch - this was a messy problem.)
1507    
1508    9.  Applied Craig's patch to remove the use of push_back().
1509    
1510    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1511        matching function regexec().
1512    
1513    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1514        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1515        references. PCRE supports relative numbers with this syntax (I don't think
1516        Oniguruma does).
1517    
1518    12. Previously, a group with a zero repeat such as (...){0} was completely
1519        omitted from the compiled regex. However, this means that if the group
1520        was called as a subroutine from elsewhere in the pattern, things went wrong
1521        (an internal error was given). Such groups are now left in the compiled
1522        pattern, with a new opcode that causes them to be skipped at execution
1523        time.
1524    
1525    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1526        to the way PCRE behaves:
1527    
1528        (a) A lone ] character is dis-allowed (Perl treats it as data).
1529    
1530        (b) A back reference to an unmatched subpattern matches an empty string
1531            (Perl fails the current match path).
1532    
1533        (c) A data ] in a character class must be notated as \] because if the
1534            first data character in a class is ], it defines an empty class. (In
1535            Perl it is not possible to have an empty class.) The empty class []
1536            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1537            The negative empty class [^] matches any one character, independently
1538            of the DOTALL setting.
1539    
1540    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1541        non-existent subpattern following a character class starting with ']' and
1542        containing () gave an internal compiling error instead of "reference to
1543        non-existent subpattern". Fortunately, when the pattern did exist, the
1544        compiled code was correct. (When scanning forwards to check for the
1545        existencd of the subpattern, it was treating the data ']' as terminating
1546        the class, so got the count wrong. When actually compiling, the reference
1547        was subsequently set up correctly.)
1548    
1549    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1550        it was being rejected as not supported by pcre_dfa_exec(), even though
1551        other assertions are supported. I have made pcre_dfa_exec() support
1552        (*FAIL).
1553    
1554    16. The implementation of 13c above involved the invention of a new opcode,
1555        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1556        cannot be changed at match time, I realized I could make a small
1557        improvement to matching performance by compiling OP_ALLANY instead of
1558        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1559        on the OP_ANY path.
1560    
1561    17. Compiling pcretest on Windows with readline support failed without the
1562        following two fixes: (1) Make the unistd.h include conditional on
1563        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1564    
1565    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1566        ncurses library to be included for pcretest when ReadLine support is
1567        requested, but also to allow for it to be overridden. This patch came from
1568        Daniel Bergström.
1569    
1570    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1571        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1572        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1573        spotting this.
1574    
1575    
1576    Version 7.6 28-Jan-08
1577    ---------------------
1578    
1579    1.  A character class containing a very large number of characters with
1580        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1581        overflow.
1582    
1583    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1584        HAVE_LONG_LONG is not defined.
1585    
1586    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1587        bring them up to date and include new features. This patch includes:
1588    
1589        - Fixed PH's badly added libz and libbz2 support.
1590        - Fixed a problem with static linking.
1591        - Added pcredemo. [But later removed - see 7 below.]
1592        - Fixed dftables problem and added an option.
1593        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1594            HAVE_LONG_LONG.
1595        - Added readline support for pcretest.
1596        - Added an listing of the option settings after cmake has run.
1597    
1598    4.  A user submitted a patch to Makefile that makes it easy to create
1599        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1600        Makefile.am that cause it to include this special target, without
1601        affecting anything else. Note that the same mingw target plus all
1602        the other distribution libraries and programs are now supported
1603        when configuring with CMake (see 6 below) instead of with
1604        Configure/Make.
1605    
1606    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1607        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1608        exported in the Windows port". It has not yet been confirmed that the patch
1609        solves the problem, but it does no harm.
1610    
1611    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1612        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1613        with CMake, and also correct the comment about stack recursion.
1614    
1615    7.  Remove the automatic building of pcredemo from the ./configure system and
1616        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1617        of a program that users should build themselves after PCRE is installed, so
1618        building it automatically is not really right. What is more, it gave
1619        trouble in some build environments.
1620    
1621    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1622    
1623    
1624    Version 7.5 10-Jan-08
1625    ---------------------
1626    
1627    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1628        values in parens when parsing an RE using the C++ wrapper."
1629    
1630    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1631        Characters greater than 255 were excluded from the class instead of being
1632        included.
1633    
1634    3.  The same bug as (2) above applied to negated POSIX classes such as
1635        [:^space:].
1636    
1637    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1638        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1639        I have changed it.
1640    
1641    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1642        first named subpattern) and a construct such as (?&a) would reference the
1643        first named subpattern whose name started with "a" (in other words, the
1644        length check was missing). Both these problems are fixed. "Subpattern name
1645        expected" is now given for (?&) (a zero-length name), and this patch also
1646        makes it give the same error for \k'' (previously it complained that that
1647        was a reference to a non-existent subpattern).
1648    
1649    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1650        this is right because (?- can be followed by option settings as well as by
1651        digits. I have, however, made the messages clearer.
1652    
1653    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1654        than the number used in the conditional) now cause a compile-time error.
1655        This is actually not compatible with Perl, which accepts such patterns, but
1656        treats the conditional as always being FALSE (as PCRE used to), but it
1657        seems to me that giving a diagnostic is better.
1658    
1659    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1660        and messages.
1661    
1662    9.  Fix two occurrences of "backslash" in comments that should have been
1663        "backspace".
1664    
1665    10. Remove two redundant lines of code that can never be obeyed (their function
1666        was moved elsewhere).
1667    
1668    11. The program that makes PCRE's Unicode character property table had a bug
1669        which caused it to generate incorrect table entries for sequences of
1670        characters that have the same character type, but are in different scripts.
1671        It amalgamated them into a single range, with the script of the first of
1672        them. In other words, some characters were in the wrong script. There were
1673        thirteen such cases, affecting characters in the following ranges:
1674    
1675          U+002b0 - U+002c1
1676          U+0060c - U+0060d
1677          U+0061e - U+00612
1678          U+0064b - U+0065e
1679          U+0074d - U+0076d
1680          U+01800 - U+01805
1681          U+01d00 - U+01d77
1682          U+01d9b - U+01dbf
1683          U+0200b - U+0200f
1684          U+030fc - U+030fe
1685          U+03260 - U+0327f
1686          U+0fb46 - U+0fbb1
1687          U+10450 - U+1049d
1688    
1689    12. The -o option (show only the matching part of a line) for pcregrep was not
1690        compatible with GNU grep in that, if there was more than one match in a
1691        line, it showed only the first of them. It now behaves in the same way as
1692        GNU grep.
1693    
1694    13. If the -o and -v options were combined for pcregrep, it printed a blank
1695        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1696        does the same. The return code can be used to tell if there were any
1697        non-matching lines.
1698    
1699    14. Added --file-offsets and --line-offsets to pcregrep.
1700    
1701    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1702        infinitely looping recursion. The bug was that positive lookaheads were not
1703        being skipped when checking for a possible empty match (negative lookaheads
1704        and both kinds of lookbehind were skipped).
1705    
1706    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1707        inclusion of <windows.h> to before rather than after the definition of
1708        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1709    
1710    17. Specifying a possessive quantifier with a specific limit for a Unicode
1711        character property caused pcre_compile() to compile bad code, which led at
1712        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1713        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1714        caused the error; without that there was no problem.
1715    
1716    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1717    
1718    19. Added --enable-pcretest-libreadline.
1719    
1720    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1721        RE::GlobalReplace(). As a result, the number of replacements returned was
1722        double what it should be. I removed one of the increments, but Craig sent a
1723        later patch that removed the other one (the right fix) and added unit tests
1724        that check the return values (which was not done before).
1725    
1726    21. Several CMake things:
1727    
1728        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1729            the names libpcre and libpcreposix, not just pcre and pcreposix.
1730    
1731        (2) The above change means that pcretest and pcregrep are now correctly
1732            linked with the newly-built libraries, not previously installed ones.
1733    
1734        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1735    
1736    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1737        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1738        UTF-8 newline character). The key issue is that the pattern starts .*;
1739        this means that the match must be either at the beginning, or after a
1740        newline. The bug was in the code for advancing after a failed match and
1741        checking that the new position followed a newline. It was not taking
1742        account of UTF-8 characters correctly.
1743    
1744    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1745        character classes. PCRE was not treating the sequence [:...:] as a
1746        character class unless the ... were all letters. Perl, however, seems to
1747        allow any characters between [: and :], though of course it rejects as
1748        unknown any "names" that contain non-letters, because all the known class
1749        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1750        for example, whereas PCRE did not - it did not recognize a POSIX character
1751        class. This seemed a bit dangerous, so the code has been changed to be
1752        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1753        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1754        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1755        Perl does, and where it didn't before.
1756    
1757    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1758        Windows environments %n is disabled by default.
1759    
1760    
1761    Version 7.4 21-Sep-07
1762    ---------------------
1763    
1764    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1765        means that a class such as [\s] counted as "explicit reference to CR or
1766        LF". That isn't really right - the whole point of the change was to try to
1767        help when there was an actual mention of one of the two characters. So now
1768        the change happens only if \r or \n (or a literal CR or LF) character is
1769        encountered.
1770    
1771    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1772        of both had grown to the point where there were only 3 bits left.
1773        Fortunately, there was spare space in the data structure, and so I have
1774        moved the internal flags into a new 16-bit field to free up more option
1775        bits.
1776    
1777    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1778        but did not set the internal JCHANGED flag - either of these is enough to
1779        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1780        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1781        start sets both bits.
1782    
1783    4.  Added options (at build time, compile time, exec time) to change \R from
1784        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1785    
1786    5.  doc/pcresyntax.html was missing from the distribution.
1787    
1788    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1789        compatibility, even though it is no longer used.
1790    
1791    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1792        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1793        windows.h file is present (where different names are used). [This was
1794        reversed later after testing - see 16 below.]
1795    
1796    8.  Changed all #include <config.h> to #include "config.h". There were also
1797        some further <pcre.h> cases that I changed to "pcre.h".
1798    
1799    9.  When pcregrep was used with the --colour option, it missed the line ending
1800        sequence off the lines that it output.
1801    
1802    10. It was pointed out to me that arrays of string pointers cause lots of
1803        relocations when a shared library is dynamically loaded. A technique of
1804        using a single long string with a table of offsets can drastically reduce
1805        these. I have refactored PCRE in four places to do this. The result is
1806        dramatic:
1807    
1808          Originally:                          290
1809          After changing UCP table:            187
1810          After changing error message table:   43
1811          After changing table of "verbs"       36
1812          After changing table of Posix names   22
1813    
1814        Thanks to the folks working on Gregex for glib for this insight.
1815    
1816    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1817        unicode-properties was also set.
1818    
1819    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1820    
1821    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1822        checked only for CRLF.
1823    
1824    14. Added casts to pcretest.c to avoid compiler warnings.
1825    
1826    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1827    
1828    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1829        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1830        entirely. This removes changes made in 7 above.
1831    
1832    17. The CMake files have been updated, and there is now more information about
1833        building with CMake in the NON-UNIX-USE document.
1834    
1835    
1836    Version 7.3 28-Aug-07
1837    ---------------------
1838    
1839     1. In the rejigging of the build system that eventually resulted in 7.1, the
1840        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1841        brackets there is not right, since it causes compilers to look for an
1842        installed pcre.h, not the version that is in the source that is being
1843        compiled (which of course may be different). I have changed it back to:
1844    
1845          #include "pcre.h"
1846    
1847        I have a vague recollection that the change was concerned with compiling in
1848        different directories, but in the new build system, that is taken care of
1849        by the VPATH setting the Makefile.
1850    
1851     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1852        when the subject happened to end in the byte 0x85 (e.g. if the last
1853        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1854        characters but of course it shouldn't be taken as a newline when it is part
1855        of another character. The bug was that, for an unlimited repeat of . in
1856        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1857        characters when looking for a newline.
1858    
1859     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1860    
1861     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1862        in debug output.
1863    
1864     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1865        long printing in the pcrecpp unittest when running under MinGW.
1866    
1867     6. ESC_K was left out of the EBCDIC table.
1868    
1869     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1870        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1871        limit also applies to "virtual nesting" when a pattern is recursive, and in
1872        this case 1000 isn't so big. I have been able to remove this limit at the
1873        expense of backing off one optimization in certain circumstances. Normally,
1874        when pcre_exec() would call its internal match() function recursively and
1875        immediately return the result unconditionally, it uses a "tail recursion"
1876        feature to save stack. However, when a subpattern that can match an empty
1877        string has an unlimited repetition quantifier, it no longer makes this
1878        optimization. That gives it a stack frame in which to save the data for
1879        checking that an empty string has been matched. Previously this was taken
1880        from the 1000-entry workspace that had been reserved. So now there is no
1881        explicit limit, but more stack is used.
1882    
1883     8. Applied Daniel's patches to solve problems with the import/export magic
1884        syntax that is required for Windows, and which was going wrong for the
1885        pcreposix and pcrecpp parts of the library. These were overlooked when this
1886        problem was solved for the main library.
1887    
1888     9. There were some crude static tests to avoid integer overflow when computing
1889        the size of patterns that contain repeated groups with explicit upper
1890        limits. As the maximum quantifier is 65535, the maximum group length was
1891        set at 30,000 so that the product of these two numbers did not overflow a
1892        32-bit integer. However, it turns out that people want to use groups that
1893        are longer than 30,000 bytes (though not repeat them that many times).
1894        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1895        made it possible to implement the integer overflow checks in a much more
1896        dynamic way, which I have now done. The artificial limitation on group
1897        length has been removed - we now have only the limit on the total length of
1898        the compiled pattern, which depends on the LINK_SIZE setting.
1899    
1900    10. Fixed a bug in the documentation for get/copy named substring when
1901        duplicate names are permitted. If none of the named substrings are set, the
1902        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1903        empty string.
1904    
1905    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1906        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1907        because the ] is interpreted as the first data character and the
1908        terminating ] is not found. PCRE has been made compatible with Perl in this
1909        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1910        cause memory overwriting.
1911    
1912    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1913        string has been matched (to stop an infinite loop). It was not recognizing
1914        a conditional subpattern that could match an empty string if that
1915        subpattern was within another subpattern. For example, it looped when
1916        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1917        condition was not nested. This bug has been fixed.
1918    
1919    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1920        past the start of the subject in the presence of bytes with the top bit
1921        set, for example "\x8aBCD".
1922    
1923    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1924        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1925    
1926    14. Optimized (?!) to (*FAIL).
1927    
1928    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1929        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1930        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1931        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1932        does: it's just the validity check that is more restrictive.
1933    
1934    16. Inserted checks for integer overflows during escape sequence (backslash)
1935        processing, and also fixed erroneous offset values for syntax errors during
1936        backslash processing.
1937    
1938    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1939        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1940    
1941    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1942        caused an overrun.
1943    
1944    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1945        something other than just ASCII characters) inside a group that had an
1946        unlimited repeat caused a loop at compile time (while checking to see
1947        whether the group could match an empty string).
1948    
1949    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1950        [\P{Any}] did so. (Error in the code for printing property names.)
1951    
1952    21. An orphan \E inside a character class could cause a crash.
1953    
1954    22. A repeated capturing bracket such as (A)? could cause a wild memory
1955        reference during compilation.
1956    
1957    23. There are several functions in pcre_compile() that scan along a compiled
1958        expression for various reasons (e.g. to see if it's fixed length for look
1959        behind). There were bugs in these functions when a repeated \p or \P was
1960        present in the pattern. These operators have additional parameters compared
1961        with \d, etc, and these were not being taken into account when moving along
1962        the compiled data. Specifically:
1963    
1964        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1965            length.
1966    
1967        (b) An item such as \pL+ within a repeated group could cause crashes or
1968            loops.
1969    
1970        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1971            "reference to non-existent subpattern" error.
1972    
1973        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1974    
1975    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1976        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1977    
1978    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1979    
1980    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1981        character were causing crashes (broken optimization).
1982    
1983    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1984        \p or \P) caused a compile-time loop.
1985    
1986    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1987        break. For example, the unstudied pattern [\r\n]A does not match the string
1988        "\r\nA" because change 7.0/46 below moves the current point on by two
1989        characters after failing to match at the start. However, the pattern \nA
1990        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1991        the same is true. There doesn't seem any very clean way out of this, but
1992        what I have chosen to do makes the common cases work: PCRE now takes note
1993        of whether there can be an explicit match for \r or \n anywhere in the
1994        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1995        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1996        pattern has explicit CR or LF references.
1997    
1998    29. Added (*CR) etc for changing newline setting at start of pattern.
1999    
2000    
2001    Version 7.2 19-Jun-07
2002    ---------------------
2003    
2004     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2005        which is apparently normally available under Windows.
2006    
2007     2. Re-jig the pcregrep tests with different newline settings in an attempt
2008        to make them independent of the local environment's newline setting.
2009    
2010     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2011    
2012     4. Some of the "internals" tests were previously cut out when the link size
2013        was not 2, because the output contained actual offsets. The recent new
2014        "Z" feature of pcretest means that these can be cut out, making the tests
2015        usable with all link sizes.
2016    
2017     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2018        stack recursion. This gives a massive performance boost under BSD, but just
2019        a small improvement under Linux. However, it saves one field in the frame
2020        in all cases.
2021    
2022     6. Added more features from the forthcoming Perl 5.10:
2023    
2024        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2025            recursion call. It refers to the nth most recently opened parentheses.
2026    
2027        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2028            to be opened parentheses.
2029    
2030        (c) Conditions that refer to capturing parentheses can be specified
2031            relatively, for example, (?(-2)... or (?(+3)...
2032    
2033        (d) \K resets the start of the current match so that everything before
2034            is not part of it.
2035    
2036        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2037    
2038        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2039            reference syntax.
2040    
2041        (g) (?| introduces a group in which the numbering of parentheses in each
2042            alternative starts with the same number.
2043    
2044        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2045    
2046     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2047        PCRE_INFO_JCHANGED.
2048    
2049     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2050        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2051        for detecting groups that can match an empty string.
2052    
2053     9. A pattern with a very large number of alternatives (more than several
2054        hundred) was running out of internal workspace during the pre-compile
2055        phase, where pcre_compile() figures out how much memory will be needed. A
2056        bit of new cunning has reduced the workspace needed for groups with
2057        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2058        workspace instead of running out of the 4096 that are available.
2059    
2060    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2061    
2062    11. Applied patch from Google to remove an optimization that didn't quite work.
2063        The report of the bug said:
2064    
2065          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2066          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2067          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2068    
2069    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2070        it matched the wrong number of bytes.
2071    
2072    
2073    Version 7.1 24-Apr-07
2074    ---------------------
2075    
2076     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2077        that is more "standard", making use of automake and other Autotools. There
2078        is some re-arrangement of the files and adjustment of comments consequent
2079        on this.
2080    
2081     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2082        for recursive directory scanning broke on some systems because the files
2083        are not scanned in any specific order and on different systems the order
2084        was different. A call to "sort" has been inserted into RunGrepTest for the
2085        approprate test as a short-term fix. In the longer term there may be an
2086        alternative.
2087    
2088     3. I had an email from Eric Raymond about problems translating some of PCRE's
2089        man pages to HTML (despite the fact that I distribute HTML pages, some
2090        people do their own conversions for various reasons). The problems
2091        concerned the use of low-level troff macros .br and .in. I have therefore
2092        removed all such uses from the man pages (some were redundant, some could
2093        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2094        HTML has been updated to handle .nf/.fi and to complain if it encounters
2095        .br or .in.
2096    
2097     4. Updated comments in configure.ac that get placed in config.h.in and also
2098        arranged for config.h to be included in the distribution, with the name
2099        config.h.generic, for the benefit of those who have to compile without
2100        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2101    
2102     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2103        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2104        makevp.bat for the current PCRE, using the additional files
2105        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2106    
2107     6. A Windows user reported a minor discrepancy with test 2, which turned out
2108        to be caused by a trailing space on an input line that had got lost in his
2109        copy. The trailing space was an accident, so I've just removed it.
2110    
2111     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2112        that is needed.
2113    
2114     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2115        as "const" (a) because they are and (b) because it helps the PHP
2116        maintainers who have recently made a script to detect big data structures
2117        in the php code that should be moved to the .rodata section. I remembered
2118        to update Builducptable as well, so it won't revert if ucptable.h is ever
2119        re-created.
2120    
2121     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2122        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2123        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2124        support is not required. This saves 1.5-2K of code, which is important in
2125        some applications.
2126    
2127        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2128        so as not to refer to the tables, even though these functions will never be
2129        called when UTF-8 support is disabled. Otherwise there are problems with a
2130        shared library.
2131    
2132    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2133    
2134        (a) It was defining its arguments as char * instead of void *.
2135    
2136        (b) It was assuming that all moves were upwards in memory; this was true
2137            a long time ago when I wrote it, but is no longer the case.
2138    
2139        The emulated memove() is provided for those environments that have neither
2140        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2141        is clearly not the case, as these two bugs were recently reported.
2142    
2143    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2144        and Detrail to create the HTML documentation, the .txt form of the man
2145        pages, and it removes trailing spaces from listed files. It also creates
2146        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2147        case, it wraps all the #defines with #ifndefs. This script should be run
2148        before "make dist".
2149    
2150    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2151        with Unicode property support.
2152    
2153        (a) For a maximizing quantifier, if the two different cases of the
2154            character were of different lengths in their UTF-8 codings (there are
2155            some cases like this - I found 11), and the matching function had to
2156            back up over a mixture of the two cases, it incorrectly assumed they
2157            were both the same length.
2158    
2159        (b) When PCRE was configured to use the heap rather than the stack for
2160            recursion during matching, it was not correctly preserving the data for
2161            the other case of a UTF-8 character when checking ahead for a match
2162            while processing a minimizing repeat. If the check also involved
2163            matching a wide character, but failed, corruption could cause an
2164            erroneous result when trying to check for a repeat of the original
2165            character.
2166    
2167    13. Some tidying changes to the testing mechanism:
2168    
2169        (a) The RunTest script now detects the internal link size and whether there
2170            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2171            values substituted by "configure". (The RunGrepTest script already did
2172            this for UTF-8.) The configure.ac script no longer substitutes the
2173            relevant variables.
2174    
2175        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2176            with length and offset values. This means that the output is different
2177            for different internal link sizes. Test 2 is skipped for link sizes
2178            other than 2 because of this, bypassing the problem. Unfortunately,
2179            there was also a test in test 3 (the locale tests) that used /B and
2180            failed for link sizes other than 2. Rather than cut the whole test out,
2181            I have added a new /Z option to pcretest that replaces the length and
2182            offset values with spaces. This is now used to make test 3 independent
2183            of link size. (Test 2 will be tidied up later.)
2184    
2185    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2186        segmentation fault instead of returning the appropriate error message.
2187    
2188    15. In multiline mode when the newline sequence was set to "any", the pattern
2189        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2190        This doesn't seem right; it now treats the CRLF combination as the line
2191        ending, and so does not match in that case. It's only a pattern such as ^$
2192        that would hit this one: something like ^ABC$ would have failed after \r
2193        and then tried again after \r\n.
2194    
2195    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2196        in an attempt to make files that differ only in their line terminators
2197        compare equal. This works on Linux.
2198    
2199    17. Under certain error circumstances pcregrep might try to free random memory
2200        as it exited. This is now fixed, thanks to valgrind.
2201    
2202    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2203        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2204        was because its rules for how to advance for /g after matching an empty
2205        string at the end of a line did not allow for this case. They now check for
2206        it specially.
2207    
2208    20. pcretest is supposed to handle patterns and data of any length, by
2209        extending its buffers when necessary. It was getting this wrong when the
2210        buffer for a data line had to be extended.
2211    
2212    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2213        CRLF as a newline sequence.
2214    
2215    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2216        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2217        I have nevertheless tidied it up.
2218    
2219    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2220    
2221    24. Added a man page for pcre-config.
2222    
2223    
2224    Version 7.0 19-Dec-06
2225    ---------------------
2226    
2227     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2228        moving to gcc 4.1.1.
2229    
2230     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2231        sys/time.h, which is documented as needed for this function. It doesn't
2232        seem to matter on Linux, but it showed up on some releases of OS X.
2233    
2234     3. It seems that there are systems where bytes whose values are greater than
2235        127 match isprint() in the "C" locale. The "C" locale should be the
2236        default when a C program starts up. In most systems, only ASCII printing
2237        characters match isprint(). This difference caused the output from pcretest
2238        to vary, making some of the tests fail. I have changed pcretest so that:
2239    
2240        (a) When it is outputting text in the compiled version of a pattern, bytes
2241            other than 32-126 are always shown as hex escapes.
2242    
2243        (b) When it is outputting text that is a matched part of a subject string,
2244            it does the same, unless a different locale has been set for the match
2245            (using the /L modifier). In this case, it uses isprint() to decide.
2246    
2247     4. Fixed a major bug that caused incorrect computation of the amount of memory
2248        required for a compiled pattern when options that changed within the
2249        pattern affected the logic of the preliminary scan that determines the
2250        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2251        that the computed length was too small. The symptoms of this bug were
2252        either the PCRE error "internal error: code overflow" from pcre_compile(),
2253        or a glibc crash with a message such as "pcretest: free(): invalid next
2254        size (fast)". Examples of patterns that provoked this bug (shown in
2255        pcretest format) are:
2256    
2257          /(?-x: )/x
2258          /(?x)(?-x: \s*#\s*)/
2259          /((?i)[\x{c0}])/8
2260          /(?i:[\x{c0}])/8
2261    
2262        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2263        is now done differently.
2264    
2265     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2266        wrapper classes; (b) implement a new function in the C++ scanner that is
2267        more efficient than the old way of doing things because it avoids levels of
2268        recursion in the regex matching; (c) add a paragraph to the documentation
2269        for the FullMatch() function.
2270    
2271     6. The escape sequence \n was being treated as whatever was defined as
2272        "newline". Not only was this contrary to the documentation, which states
2273        that \n is character 10 (hex 0A), but it also went horribly wrong when
2274        "newline" was defined as CRLF. This has been fixed.
2275    
2276     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2277        was being set to -1 for the "end of line" case (supposedly a value that no
2278        character can have). Though this value is never used (the check for end of
2279        line is "zero bytes in current character"), it caused compiler complaints.
2280        I've changed it to 0xffffffff.
2281    
2282     8. In pcre_version.c, the version string was being built by a sequence of
2283        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2284        string (as it is for production releases) called a macro with an empty
2285        argument. The C standard says the result of this is undefined. The gcc
2286        compiler treats it as an empty string (which was what was wanted) but it is
2287        reported that Visual C gives an error. The source has been hacked around to
2288        avoid this problem.
2289    
2290     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2291        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2292        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2293        of them did).
2294    
2295    10. Originally, pcretest opened its input and output without "b"; then I was
2296        told that "b" was needed in some environments, so it was added for release
2297        5.0 to both the input and output. (It makes no difference on Unix-like
2298        systems.) Later I was told that it is wrong for the input on Windows. I've
2299        now abstracted the modes into two macros, to make it easier to fiddle with
2300        them, and removed "b" from the input mode under Windows.
2301    
2302    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2303    
2304    12. Added -help and --help to pcretest as an official way of being reminded
2305        of the options.
2306    
2307    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2308        and pcrecpp.cc because they annoy compilers at high warning levels.
2309    
2310    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2311    
2312    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2313        scripts are not C programs :-) and which was not noticed because it works
2314        on Linux.
2315    
2316    16. pcretest is supposed to handle any length of pattern and data line (as one
2317        line or as a continued sequence of lines) by extending its input buffer if
2318        necessary. This feature was broken for very long pattern lines, leading to
2319        a string of junk being passed to pcre_compile() if the pattern was longer
2320        than about 50K.
2321    
2322    17. I have done a major re-factoring of the way pcre_compile() computes the
2323        amount of memory needed for a compiled pattern. Previously, there was code
2324        that made a preliminary scan of the pattern in order to do this. That was
2325        OK when PCRE was new, but as the facilities have expanded, it has become
2326        harder and harder to keep it in step with the real compile phase, and there
2327        have been a number of bugs (see for example, 4 above). I have now found a
2328        cunning way of running the real compile function in a "fake" mode that
2329        enables it to compute how much memory it would need, while actually only
2330        ever using a few hundred bytes of working memory and without too many
2331        tests of the mode. This should make future maintenance and development
2332        easier. A side effect of this work is that the limit of 200 on the nesting
2333        depth of parentheses has been removed (though this was never a serious
2334        limitation, I suspect). However, there is a downside: pcre_compile() now
2335        runs more slowly than before (30% or more, depending on the pattern). I
2336        hope this isn't a big issue. There is no effect on runtime performance.
2337    
2338    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2339        newline (only possible for the last line of a file) and it was a
2340        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2341    
2342    19. Added additional timing features to pcretest. (1) The -tm option now times
2343        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2344        separate command line item, by a number that specifies the number of
2345        repeats to use when timing. The default is 50000; this gives better
2346        precision, but takes uncomfortably long for very large patterns.
2347    
2348    20. Extended pcre_study() to be more clever in cases where a branch of a
2349        subpattern has no definite first character. For example, (a*|b*)[cd] would
2350        previously give no result from pcre_study(). Now it recognizes that the
2351        first character must be a, b, c, or d.
2352    
2353    21. There was an incorrect error "recursive call could loop indefinitely" if
2354        a subpattern (or the entire pattern) that was being tested for matching an
2355        empty string contained only one non-empty item after a nested subpattern.
2356        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2357        incorrectly, because the \d was being skipped in the check.
2358    
2359    22. The pcretest program now has a new pattern option /B and a command line
2360        option -b, which is equivalent to adding /B to every pattern. This causes
2361        it to show the compiled bytecode, without the additional information that
2362        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2363        is the same as /B/I).
2364    
2365    23. A new optimization is now able automatically to treat some sequences such
2366        as a*b as a*+b. More specifically, if something simple (such as a character
2367        or a simple class like \d) has an unlimited quantifier, and is followed by
2368        something that cannot possibly match the quantified thing, the quantifier
2369        is automatically "possessified".
2370    
2371    24. A recursive reference to a subpattern whose number was greater than 39
2372        went wrong under certain circumstances in UTF-8 mode. This bug could also
2373        have affected the operation of pcre_study().
2374    
2375    25. Realized that a little bit of performance could be had by replacing
2376        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2377    
2378    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2379    
2380    27. Possessive quantifiers such as a++ were previously implemented by turning
2381        them into atomic groups such as ($>a+). Now they have their own opcodes,
2382        which improves performance. This includes the automatically created ones
2383        from 23 above.
2384    
2385    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2386        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2387        the first matched character to be a colon. This applied both to named and
2388        numbered groups.
2389    
2390    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2391    
2392    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2393        building PCRE on Windows easier, so I have included it in the distribution.
2394    
2395    31. There is now a check in pcretest against a ridiculously large number being
2396        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2397        loop, the loop is abandoned.
2398    
2399    32. Forward references to subpatterns in conditions such as (?(2)...) where
2400        subpattern 2 is defined later cause pcre_compile() to search forwards in
2401        the pattern for the relevant set of parentheses. This search went wrong
2402        when there were unescaped parentheses in a character class, parentheses
2403        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2404    
2405    33. "Subroutine" calls and backreferences were previously restricted to
2406        referencing subpatterns earlier in the regex. This restriction has now
2407        been removed.
2408    
2409    34. Added a number of extra features that are going to be in Perl 5.10. On the
2410        whole, these are just syntactic alternatives for features that PCRE had
2411        previously implemented using the Python syntax or my own invention. The
2412        other formats are all retained for compatibility.
2413    
2414        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2415            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2416            also .NET compatible.
2417    
2418        (b) A recursion or subroutine call to a named group can now be defined as
2419            (?&name) as well as (?P>name).
2420    
2421        (c) A backreference to a named group can now be defined as \k<name> or
2422            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2423            5.10, are also .NET compatible.
2424    
2425        (d) A conditional reference to a named group can now use the syntax
2426            (?(<name>) or (?('name') as well as (?(name).
2427    
2428        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2429            groups (named and numbered) that are never evaluated inline, but can be
2430            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2431            is always false. There may be only one alternative in such a group.
2432    
2433        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2434            as the simple (?(R). The condition is true only if the most recent
2435            recursion is that of the given number or name. It does not search out
2436            through the entire recursion stack.
2437    
2438        (g) The escape \gN or \g{N} has been added, where N is a positive or
2439            negative number, specifying an absolute or relative reference.
2440    
2441    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2442        some "unreachable code" warnings.
2443    
2444    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2445        things, this adds five new scripts.
2446    
2447    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2448        There were also incompatibilities regarding the handling of \Q..\E inside
2449        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2450        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2451    
2452    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2453        matches an empty string, and forcibly breaks the loop. There were bugs in
2454        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2455        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2456        separate and independent bugs (that affected different cases) have been
2457        fixed.
2458    
2459    39. Refactored the code to abolish the use of different opcodes for small
2460        capturing bracket numbers. This is a tidy that I avoided doing when I
2461        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2462        The new approach is not only tidier, it makes it possible to reduce the
2463        memory needed to fix the previous bug (38).
2464    
2465    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2466        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2467        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2468        mode.
2469    
2470    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2471        report.
2472    
2473    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2474        copy construction and assignment in the C++ wrapper.
2475    
2476    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2477        couple of bugs that could have given wrong results in the "--newline=crlf"
2478        case.
2479    
2480    44. Added a number of casts and did some reorganization of signed/unsigned int
2481        variables following suggestions from Dair Grant. Also renamed the variable
2482        "this" as "item" because it is a C++ keyword.
2483    
2484    45. Arranged for dftables to add
2485    
2486          #include "pcre_internal.h"
2487    
2488        to pcre_chartables.c because without it, gcc 4.x may remove the array
2489        definition from the final binary if PCRE is built into a static library and
2490        dead code stripping is activated.
2491    
2492    46. For an unanchored pattern, if a match attempt fails at the start of a
2493        newline sequence, and the newline setting is CRLF or ANY, and the next two
2494        characters are CRLF, advance by two characters instead of one.
2495    
2496    
2497    Version 6.7 04-Jul-06
2498    ---------------------
2499    
2500     1. In order to handle tests when input lines are enormously long, pcretest has
2501        been re-factored so that it automatically extends its buffers when
2502        necessary. The code is crude, but this _is_ just a test program. The
2503        default size has been increased from 32K to 50K.
2504    
2505     2. The code in pcre_study() was using the value of the re argument before
2506        testing it for NULL. (Of course, in any sensible call of the function, it
2507        won't be NULL.)
2508    
2509     3. The memmove() emulation function in pcre_internal.h, which is used on
2510        systems that lack both memmove() and bcopy() - that is, hardly ever -
2511        was missing a "static" storage class specifier.
2512    
2513     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2514        containing an extended class (one that cannot be represented by a bitmap
2515        because it contains high-valued characters or Unicode property items, e.g.
2516        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2517        pattern, but PCRE should not loop if you do not (it no longer does).
2518        [Detail: two cases were found: (a) a repeated subpattern containing an
2519        extended class; (b) a recursive reference to a subpattern that followed a
2520        previous extended class. It wasn't skipping over the extended class
2521        correctly when UTF-8 mode was not set.]
2522    
2523     5. A negated single-character class was not being recognized as fixed-length
2524        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2525        compile error "lookbehind assertion is not fixed length".
2526    
2527     6. The RunPerlTest auxiliary script was showing an unexpected difference
2528        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2529        write a Perl script that can interpret lines of an input file either as
2530        byte characters or as UTF-8, which is what "perltest" was being required to
2531        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2532        can't do is switch easily at run time between having the "use utf8;" pragma
2533        or not. In the end, I fudged it by using the RunPerlTest script to insert
2534        "use utf8;" explicitly for the UTF-8 tests.
2535    
2536     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2537        the end of the subject string, contrary to the documentation and to what
2538        Perl does. This was true of both matching functions. Now it matches only at
2539        the start of the subject and immediately after *internal* newlines.
2540    
2541     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2542        a pointer to an int instead of a pointer to an unsigned long int. This
2543        caused problems on 64-bit systems.
2544    
2545     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2546        instance of the 'standard' template library not being so standard".
2547    
2548    10. There was no check on the number of named subpatterns nor the maximum
2549        length of a subpattern name. The product of these values is used to compute
2550        the size of the memory block for a compiled pattern. By supplying a very
2551        long subpattern name and a large number of named subpatterns, the size
2552        computation could be caused to overflow. This is now prevented by limiting
2553        the length of names to 32 characters, and the number of named subpatterns
2554        to 10,000.
2555    
2556    11. Subpatterns that are repeated with specific counts have to be replicated in
2557        the compiled pattern. The size of memory for this was computed from the
2558        length of the subpattern and the repeat count. The latter is limited to
2559        65535, but there was no limit on the former, meaning that integer overflow
2560        could in principle occur. The compiled length of a repeated subpattern is
2561        now limited to 30,000 bytes in order to prevent this.
2562    
2563    12. Added the optional facility to have named substrings with the same name.
2564    
2565    13. Added the ability to use a named substring as a condition, using the
2566        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2567        are numbers (not recommended). Forward references are permitted.
2568    
2569    14. Added forward references in named backreferences (if you see what I mean).
2570    
2571    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2572        pattern could run off the end of the subject. For example, the pattern
2573        "(?s)(.{1,5})"8 did this with the subject "ab".
2574    
2575    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2576        PCRE_CASELESS was set when matching characters that were quantified with ?
2577        or *.
2578    
2579    17. A character class other than a single negated character that had a minimum
2580        but no maximum quantifier - for example [ab]{6,} - was not handled
2581        correctly by pce_dfa_exec(). It would match only one character.
2582    
2583    18. A valid (though odd) pattern that looked like a POSIX character
2584        class but used an invalid character after [ (for example [[,abc,]]) caused
2585        pcre_compile() to give the error "Failed: internal error: code overflow" or
2586        in some cases to crash with a glibc free() error. This could even happen if
2587        the pattern terminated after [[ but there just happened to be a sequence of
2588        letters, a binary zero, and a closing ] in the memory that followed.
2589    
2590    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2591        over the years. Originally (before any Unicode support), just the bottom 8
2592        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2593        output from "man perlunicode" includes this:
2594    
2595          The regular expression compiler produces polymorphic opcodes.  That
2596          is, the pattern adapts to the data and automatically switches to
2597          the Unicode character scheme when presented with Unicode data--or
2598          instead uses a traditional byte scheme when presented with byte
2599          data.
2600    
2601        Sadly, a wide octal escape does not cause a switch, and in a string with
2602        no other multibyte characters, these octal escapes are treated as before.
2603        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2604        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2605        Unicode string.
2606    
2607        I have not perpetrated such confusion in PCRE. Up till now, it took just
2608        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2609        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2610        translate to the appropriate multibyte character.
2611    
2612    29. Applied some refactoring to reduce the number of warnings from Microsoft
2613        and Borland compilers. This has included removing the fudge introduced
2614        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2615        a warning about an unused variable.
2616    
2617    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2618        characters since release 4.0, because Perl (from release 5.004) does not.
2619        [Or at least, is documented not to: some releases seem to be in conflict
2620        with the documentation.] However, when a pattern was studied with
2621        pcre_study() and all its branches started with \s, PCRE still included VT
2622        as a possible starting character. Of course, this did no harm; it just
2623        caused an unnecessary match attempt.
2624    
2625    22. Removed a now-redundant internal flag bit that recorded the fact that case
2626        dependency changed within the pattern. This was once needed for "required
2627        byte" processing, but is no longer used. This recovers a now-scarce options
2628        bit. Also moved the least significant internal flag bit to the most-
2629        significant bit of the word, which was not previously used (hangover from
2630        the days when it was an int rather than a uint) to free up another bit for
2631        the future.
2632    
2633    23. Added support for CRLF line endings as well as CR and LF. As well as the
2634        default being selectable at build time, it can now be changed at runtime
2635        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2636        specify that it is scanning data with non-default line endings.
2637    
2638    24. Changed the definition of CXXLINK to make it agree with the definition of
2639        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2640    
2641    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2642        recursions. This makes a big different to stack usage for some patterns.
2643    
2644    26. If a subpattern containing a named recursion or subroutine reference such
2645        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2646        the space required for the compiled pattern went wrong and gave too small a
2647        value. Depending on the environment, this could lead to "Failed: internal
2648        error: code overflow at offset 49" or "glibc detected double free or
2649        corruption" errors.
2650    
2651    27. Applied patches from Google (a) to support the new newline modes and (b) to
2652        advance over multibyte UTF-8 characters in GlobalReplace.
2653    
2654    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2655        difference for some implementation of PCRE in some Windows version.
2656    
2657    29. Added some extra testing facilities to pcretest:
2658    
2659        \q<number>   in a data line sets the "match limit" value
2660        \Q<number>   in a data line sets the "match recursion limt" value
2661        -S <number>  sets the stack size, where <number> is in megabytes
2662    
2663        The -S option isn't available for Windows.
2664    
2665    
2666    Version 6.6 06-Feb-06
2667    ---------------------
2668    
2669     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2670        in pcreposix.h. I have copied the definition from pcre.h.
2671    
2672     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2673        because pcre.h is no longer a built file.
2674    
2675     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2676        not normally included in the compiled code.
2677    
2678    
2679    Version 6.5 01-Feb-06
2680    ---------------------
2681    
2682     1. When using the partial match feature with pcre_dfa_exec(), it was not
2683        anchoring the second and subsequent partial matches at the new starting
2684        point. This could lead to incorrect results. For example, with the pattern
2685        /1234/, partially matching against "123" and then "a4" gave a match.
2686    
2687     2. Changes to pcregrep:
2688    
2689        (a) All non-match returns from pcre_exec() were being treated as failures
2690            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2691            error message is output. Some extra information is given for the
2692            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2693            probably the only errors that are likely to be caused by users (by
2694            specifying a regex that has nested indefinite repeats, for instance).
2695            If there are more than 20 of these errors, pcregrep is abandoned.
2696    
2697        (b) A binary zero was treated as data while matching, but terminated the
2698            output line if it was written out. This has been fixed: binary zeroes
2699            are now no different to any other data bytes.
2700    
2701        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2702            used to set a locale for matching. The --locale=xxxx long option has
2703            been added (no short equivalent) to specify a locale explicitly on the
2704            pcregrep command, overriding the environment variables.
2705    
2706        (d) When -B was used with -n, some line numbers in the output were one less
2707            than they should have been.
2708    
2709        (e) Added the -o (--only-matching) option.
2710    
2711        (f) If -A or -C was used with -c (count only), some lines of context were
2712            accidentally printed for the final match.
2713    
2714        (g) Added the -H (--with-filename) option.
2715    
2716        (h) The combination of options -rh failed to suppress file names for files
2717            that were found from directory arguments.
2718    
2719        (i) Added the -D (--devices) and -d (--directories) options.
2720    
2721        (j) Added the -F (--fixed-strings) option.
2722    
2723        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2724    
2725        (l) Added the --colo(u)r option.
2726    
2727        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2728            is not present by default.
2729    
2730     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2731        items such as (?R) or (?1), when the recursion could match a number of
2732        alternatives. If it matched one of the alternatives, but subsequently,
2733        outside the recursion, there was a failure, the code tried to back up into
2734        the recursion. However, because of the way PCRE is implemented, this is not
2735        possible, and the result was an incorrect result from the match.
2736    
2737        In order to prevent this happening, the specification of recursion has
2738        been changed so that all such subpatterns are automatically treated as
2739        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2740    
2741     4. I had overlooked the fact that, in some locales, there are characters for
2742        which isalpha() is true but neither isupper() nor islower() are true. In
2743        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2744        and ordfeminine) are like this. This affected the treatment of \w and \W
2745        when they appeared in character classes, but not when they appeared outside
2746        a character class. The bit map for "word" characters is now created
2747        separately from the results of isalnum() instead of just taking it from the
2748        upper, lower, and digit maps. (Plus the underscore character, of course.)
2749    
2750     5. The above bug also affected the handling of POSIX character classes such as
2751        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2752        permanent tables. Instead, the bit maps for such a class were previously
2753        created as the appropriate unions of the upper, lower, and digit bitmaps.
2754        Now they are created by subtraction from the [[:word:]] class, which has
2755        its own bitmap.
2756    
2757     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2758        It is created by subtracting the vertical space characters (\x09, \x0a,
2759        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2760        subtraction was done in the overall bitmap for a character class, meaning
2761        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2762        be recognized. This bug has been fixed.
2763    
2764     7. Patches from the folks at Google:
2765    
2766          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2767          real life, but is still worth protecting against".
2768    
2769          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2770          regular expressions".
2771    
2772          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2773          have it.
2774    
2775          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2776          "configure" and the latter not, in order to fix a problem somebody had
2777          with compiling the Arg class on HP-UX.
2778    
2779          (e) Improve the error-handling of the C++ wrapper a little bit.
2780    
2781          (f) New tests for checking recursion limiting.
2782    
2783     8. The pcre_memmove() function, which is used only if the environment does not
2784        have a standard memmove() function (and is therefore rarely compiled),
2785        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2786        returning a result (though PCRE never actually uses the result).
2787    
2788     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2789        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2790        returned instead of calling malloc() with an overflowing number that would
2791        most likely cause subsequent chaos.
2792    
2793    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2794    
2795    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2796        with this option is matched, the nmatch and pmatch options of regexec() are
2797        ignored.
2798    
2799    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2800        provided in case anyone wants to the the POSIX interface with UTF-8
2801        strings.
2802    
2803    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2804        C++ linking (needed for some HP-UX environments).
2805    
2806    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2807        (unused parameter) and in the pcre_printint() function (omitted "default"
2808        switch label when the default is to do nothing).
2809    
2810    15. Added some code to make it possible, when PCRE is compiled as a C++
2811        library, to replace subject pointers for pcre_exec() with a smart pointer
2812        class, thus making it possible to process discontinuous strings.
2813    
2814    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2815        much the same function. They were added by different people who were trying
2816        to make PCRE easy to compile on non-Unix systems. It has been suggested
2817        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2818        for compiling on Windows systems. I have therefore replaced it with
2819        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2820        defaults to "extern" for C or "extern C" for C++, which works fine on
2821        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2822        SCOPE with something explicit in config.h. In addition:
2823    
2824        (a) pcreposix.h still had just "extern" instead of either of these macros;
2825            I have replaced it with PCRE_DATA_SCOPE.
2826    
2827        (b) Functions such as _pcre_xclass(), which are internal to the library,
2828            but external in the C sense, all had PCRE_EXPORT in their definitions.
2829            This is apparently wrong for the Windows case, so I have removed it.
2830            (It makes no difference on Unix-like systems.)
2831    
2832    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2833        of recursive calls to match(). This is different to MATCH_LIMIT because
2834        that limits the total number of calls to match(), not all of which increase
2835        the depth of recursion. Limiting the recursion depth limits the amount of
2836        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2837        when PCRE is compiled, and changed at run time. A patch from Google adds
2838        this functionality to the C++ interface.
2839    
2840    18. Changes to the handling of Unicode character properties:
2841    
2842        (a) Updated the table to Unicode 4.1.0.
2843    
2844        (b) Recognize characters that are not in the table as "Cn" (undefined).
2845    
2846        (c) I revised the way the table is implemented to a much improved format
2847            which includes recognition of ranges. It now supports the ranges that
2848            are defined in UnicodeData.txt, and it also amalgamates other
2849            characters into ranges. This has reduced the number of entries in the
2850            table from around 16,000 to around 3,000, thus reducing its size
2851            considerably. I realized I did not need to use a tree structure after
2852            all - a binary chop search is just as efficient. Having reduced the
2853            number of entries, I extended their size from 6 bytes to 8 bytes to
2854            allow for more data.
2855    
2856        (d) Added support for Unicode script names via properties such as \p{Han}.
2857    
2858    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2859        matching that character.
2860    
2861    20. When matching a repeated Unicode property with a minimum greater than zero,
2862        (for example \pL{2,}), PCRE could look past the end of the subject if it
2863        reached it while seeking the minimum number of characters. This could
2864        happen only if some of the characters were more than one byte long, because
2865        there is a check for at least the minimum number of bytes.
2866    
2867    21. Refactored the implementation of \p and \P so as to be more general, to
2868        allow for more different types of property in future. This has changed the
2869        compiled form incompatibly. Anybody with saved compiled patterns that use
2870        \p or \P will have to recompile them.
2871    
2872    22. Added "Any" and "L&" to the supported property types.
2873    
2874    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2875        but give a compile time error if the value is greater than 0xff.
2876    
2877    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2878        accidentally not being installed or uninstalled.
2879    
2880    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2881        made were to insert the current release number. This seemed silly, because
2882        it made things harder for people building PCRE on systems that don't run
2883        "configure". I have turned pcre.h into a distributed file, no longer built
2884        by "configure", with the version identification directly included. There is
2885        no longer a pcre.h.in file.
2886    
2887        However, this change necessitated a change to the pcre-config script as
2888        well. It is built from pcre-config.in, and one of the substitutions was the
2889        release number. I have updated configure.ac so that ./configure now finds
2890        the release number by grepping pcre.h.
2891    
2892    26. Added the ability to run the tests under valgrind.
2893    
2894    
2895    Version 6.4 05-Sep-05
2896    ---------------------
2897    
2898     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2899        "--" to be printed when multiple files were scanned, even when none of the
2900        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2901        consider it to be a bug, and have restored the previous behaviour.
2902    
2903     2. A couple of code tidies to get rid of compiler warnings.
2904    
2905     3. The pcretest program used to cheat by referring to symbols in the library
2906        whose names begin with _pcre_. These are internal symbols that are not
2907        really supposed to be visible externally, and in some environments it is
2908        possible to suppress them. The cheating is now confined to including
2909        certain files from the library's source, which is a bit cleaner.
2910    
2911     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2912        file's purpose clearer.
2913    
2914     5. Reorganized pcre_ucp_findchar().
2915    
2916    
2917    Version 6.3 15-Aug-05
2918    ---------------------
2919    
2920     1. The file libpcre.pc.in did not have general read permission in the tarball.
2921    
2922     2. There were some problems when building without C++ support:
2923    
2924        (a) If C++ support was not built, "make install" and "make test" still
2925            tried to test it.
2926    
2927        (b) There were problems when the value of CXX was explicitly set. Some
2928            changes have been made to try to fix these, and ...
2929    
2930        (c) --disable-cpp can now be used to explicitly disable C++ support.
2931    
2932        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2933            backslash in a target when C++ was disabled. This confuses some
2934            versions of "make", apparently. Using an intermediate variable solves
2935            this. (Same for CPP_LOBJ.)
2936    
2937     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2938        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2939        necessary on certain architectures.
2940    
2941     4. Added a setting of -export-symbols-regex to the link command to remove
2942        those symbols that are exported in the C sense, but actually are local
2943        within the library, and not documented. Their names all begin with
2944        "_pcre_". This is not a perfect job, because (a) we have to except some
2945        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2946        available (and never for static libraries). I have made a note to try to
2947        find a way round (a) in the future.
2948    
2949    
2950    Version 6.2 01-Aug-05
2951    ---------------------
2952    
2953     1. There was no test for integer overflow of quantifier values. A construction
2954        such as {1111111111111111} would give undefined results. What is worse, if
2955        a minimum quantifier for a parenthesized subpattern overflowed and became
2956        negative, the calculation of the memory size went wrong. This could have
2957        led to memory overwriting.
2958    
2959     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2960    
2961     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2962        operating environments where this matters.
2963    
2964     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2965        PCRE options from within the C++ wrapper.
2966    
2967     5. Named capturing subpatterns were not being correctly counted when a pattern
2968        was compiled. This caused two problems: (a) If there were more than 100
2969        such subpatterns, the calculation of the memory needed for the whole
2970        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2971        back references of the form \12, where the number was greater than 9, were
2972        not recognized as back references, even though there were sufficient
2973        previous subpatterns.
2974    
2975     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2976        versions of gcc, e.g. 2.95.4.
2977    
2978    
2979    Version 6.1 21-Jun-05
2980    ---------------------
2981    
2982     1. There was one reference to the variable "posix" in pcretest.c that was not
2983        surrounded by "#if !defined NOPOSIX".
2984    
2985     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2986        the cross-check on the old pcre_info() function, for the benefit of the
2987        cut-down version of PCRE that is currently imported into Exim.
2988    
2989     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2990        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2991        patterns that start (?i) but I don't think that matters. The use of (?i) is
2992        just an example; this all applies to the other options as well.
2993    
2994     4. Since libtool seems to echo the compile commands it is issuing, the output
2995        from "make" can be reduced a bit by putting "@" in front of each libtool
2996        compile command.
2997    
2998     5. Patch from the folks at Google for configure.in to be a bit more thorough
2999        in checking for a suitable C++ installation before trying to compile the
3000        C++ stuff. This should fix a reported problem when a compiler was present,
3001        but no suitable headers.
3002    
3003     6. The man pages all had just "PCRE" as their title. I have changed them to
3004        be the relevant file name. I have also arranged that these names are
3005        retained in the file doc/pcre.txt, which is a concatenation in text format
3006        of all the man pages except the little individual ones for each function.
3007    
3008     7. The NON-UNIX-USE file had not been updated for the different set of source
3009        files that come with release 6. I also added a few comments about the C++
3010        wrapper.
3011    
3012    
3013    Version 6.0 07-Jun-05
3014    ---------------------
3015    
3016     1. Some minor internal re-organization to help with my DFA experiments.
3017    
3018     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3019        didn't matter for the library itself when fully configured, but did matter
3020        when compiling without UCP support, or within Exim, where the ucp files are
3021        not imported.
3022    
3023     3. Refactoring of the library code to split up the various functions into
3024        different source modules. The addition of the new DFA matching code (see
3025        below) to a single monolithic source would have made it really too
3026        unwieldy, quite apart from causing all the code to be include in a
3027        statically linked application, when only some functions are used. This is
3028        relevant even without the DFA addition now that patterns can be compiled in
3029        one application and matched in another.
3030    
3031        The downside of splitting up is that there have to be some external
3032        functions and data tables that are used internally in different modules of
3033        the library but which are not part of the API. These have all had their
3034        names changed to start with "_pcre_" so that they are unlikely to clash
3035        with other external names.
3036    
3037     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3038        a different (DFA) algorithm. Although it is slower than the original
3039        function, it does have some advantages for certain types of matching
3040        problem.
3041    
3042     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3043        including restarting after a partial match.
3044    
3045     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3046        defined when compiling for Windows was sent to me. I have put it into the
3047        code, though I have no means of testing or verifying it.
3048    
3049     7. Added the pcre_refcount() auxiliary function.
3050    
3051     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3052        match before or at the first newline in the subject string. In pcretest,
3053        the /f option on a pattern can be used to set this.
3054    
3055     9. A repeated \w when used in UTF-8 mode with characters greater than 256
3056        would behave wrongly. This has been present in PCRE since release 4.0.
3057    
3058    10. A number of changes to the pcregrep command:
3059    
3060        (a) Refactored how -x works; insert ^(...)$ instead of setting
3061            PCRE_ANCHORED and checking the length, in preparation for adding
3062            something similar for -w.
3063    
3064        (b) Added the -w (match as a word) option.
3065    
3066        (c) Refactored the way lines are read and buffered so as to have more
3067            than one at a time available.
3068    
3069        (d) Implemented a pcregrep test script.
3070    
3071        (e) Added the -M (multiline match) option. This allows patterns to match
3072            over several lines of the subject. The buffering ensures that at least
3073            8K, or the rest of the document (whichever is the shorter) is available
3074            for matching (and similarly the previous 8K for lookbehind assertions).
3075    
3076        (f) Changed the --help output so that it now says
3077    
3078              -w, --word-regex(p)
3079    
3080            instead of two lines, one with "regex" and the other with "regexp"
3081            because that confused at least one person since the short forms are the
3082            same. (This required a bit of code, as the output is generated
3083            automatically from a table. It wasn't just a text change.)
3084    
3085        (g) -- can be used to terminate pcregrep options if the next thing isn't an
3086            option but starts with a hyphen. Could be a pattern or a path name
3087            starting with a hyphen, for instance.
3088    
3089        (h) "-" can be given as a file name to represent stdin.
3090    
3091        (i) When file names are being printed, "(standard input)" is used for
3092            the standard input, for compatibility with GNU grep. Previously
3093            "<stdin>" was used.
3094    
3095        (j) The option --label=xxx can be used to supply a name to be used for
3096            stdin when file names are being printed. There is no short form.
3097    
3098        (k) Re-factored the options decoding logic because we are going to add
3099            two more options that take data. Such options can now be given in four
3100            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3101    
3102        (l) Added the -A, -B, and -C options for requesting that lines of context
3103            around matches be printed.
3104    
3105        (m) Added the -L option to print the names of files that do not contain
3106            any matching lines, that is, the complement of -l.
3107    
3108        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3109            continue to scan other files.
3110    
3111        (o) The -s option was incorrectly implemented. For compatibility with other
3112            greps, it now suppresses the error message for a non-existent or non-
3113            accessible file (but not the return code). There is a new option called
3114            -q that suppresses the output of matching lines, which was what -s was
3115            previously doing.
3116    
3117        (p) Added --include and --exclude options to specify files for inclusion
3118            and exclusion when recursing.
3119    
3120    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3121        Hopefully, it now does.
3122    
3123    12. Missing cast in pcre_study().
3124    
3125    13. Added an "uninstall" target to the makefile.
3126    
3127    14. Replaced "extern" in the function prototypes in Makefile.in with
3128        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3129        world, but is set differently for Windows.
3130    
3131    15. Added a second compiling function called pcre_compile2(). The only
3132        difference is that it has an extra argument, which is a pointer to an
3133        integer error code. When there is a compile-time failure, this is set
3134        non-zero, in addition to the error test pointer being set to point to an
3135        error message. The new argument may be NULL if no error number is required
3136        (but then you may as well call pcre_compile(), which is now just a
3137        wrapper). This facility is provided because some applications need a
3138        numeric error indication, but it has also enabled me to tidy up the way
3139        compile-time errors are handled in the POSIX wrapper.
3140    
3141    16. Added VPATH=.libs to the makefile; this should help when building with one
3142        prefix path and installing with another. (Or so I'm told by someone who
3143        knows more about this stuff than I do.)
3144    
3145    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3146        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3147        match everything, including newlines. This is not POSIX-compatible, but
3148        somebody wanted the feature. From pcretest it can be activated by using
3149        both the P and the s flags.
3150    
3151    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3152    
3153    19. libpcre.pc was being incorrectly installed as executable.
3154    
3155    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3156        it now also looks for '\r' so that it will work unmodified on Windows.
3157    
3158    21. Added Google's contributed C++ wrapper to the distribution.
3159    
3160    22. Added some untidy missing memory free() calls in pcretest, to keep
3161        Electric Fence happy when testing.
3162    
3163    
3164    
3165    Version 5.0 13-Sep-04
3166    ---------------------
3167    
3168     1. Internal change: literal characters are no longer packed up into items
3169        containing multiple characters in a single byte-string. Each character
3170        is now matched using a separate opcode. However, there may be more than one
3171        byte in the character in UTF-8 mode.
3172    
3173     2. The pcre_callout_block structure has two new fields: pattern_position and
3174        next_item_length. These contain the offset in the pattern to the next match
3175        item, and its length, respectively.
3176    
3177     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3178        insertion of callouts before each pattern item. Added the /C option to
3179        pcretest to make use of this.
3180    
3181     4. On the advice of a Windows user, the lines
3182    
3183          #if defined(_WIN32) || defined(WIN32)
3184          _setmode( _fileno( stdout ), 0x8000 );
3185          #endif  /* defined(_WIN32) || defined(WIN32) */
3186    
3187        have been added to the source of pcretest. This apparently does useful
3188        magic in relation to line terminators.
3189    
3190     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3191        for the benefit of those environments where the "b" makes a difference.
3192    
3193     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3194        to know about it. I have put a hack into configure.in that adds in code
3195        to set GCC=yes if CC=icc. This seems to end up at a point in the
3196        generated configure script that is early enough to affect the setting of
3197        compiler options, which is what is needed, but I have no means of testing
3198        whether it really works. (The user who reported this had patched the
3199        generated configure script, which of course I cannot do.)
3200    
3201        LATER: After change 22 below (new libtool files), the configure script
3202        seems to know about icc (and also ecc). Therefore, I have commented out
3203        this hack in configure.in.
3204    
3205     7. Added support for pkg-config (2 patches were sent in).
3206    
3207     8. Negated POSIX character classes that used a combination of internal tables
3208        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3209        [[:^ascii]]. Typically, they would match almost any characters. The other
3210        POSIX classes were not broken in this way.
3211    
3212     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3213        to find the match, as PCRE was deluded into thinking that the match had to
3214        start at the start point or following a newline. The same bug applied to
3215        patterns with negative forward assertions or any backward assertions
3216        preceding ".*" at the start, unless the pattern required a fixed first
3217        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3218    
3219    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3220        starting at the last subject character, bytes beyond the end of the subject
3221        string were read.
3222    
3223    11. Renamed the variable "class" as "classbits" to make life easier for C++
3224        users. (Previously there was a macro definition, but it apparently wasn't
3225        enough.)
3226    
3227    12. Added the new field "tables" to the extra data so that tables can be passed
3228        in at exec time, or the internal tables can be re-selected. This allows
3229        a compiled regex to be saved and re-used at a later time by a different
3230        program that might have everything at different addresses.
3231    
3232    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3233        -R library as well as a -L library.
3234    
3235    14. The debugging options of pcretest (-d on the command line or D on a
3236        pattern) showed incorrect output for anything following an extended class
3237        that contained multibyte characters and which was followed by a quantifier.
3238    
3239    15. Added optional support for general category Unicode character properties
3240        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3241        support. It adds about 90K to the size of the library. The meanings of the
3242        inbuilt class escapes such as \d and \s have NOT been changed.
3243    
3244    16. Updated pcredemo.c to include calls to free() to release the memory for the
3245        compiled pattern.
3246    
3247    17. The generated file chartables.c was being created in the source directory
3248        instead of in the building directory. This caused the build to fail if the
3249        source directory was different from the building directory, and was
3250        read-only.
3251    
3252    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3253        file. No doubt somebody will tell me if they don't make sense... Also added
3254        Dan Mooney's comments about building on OpenVMS.
3255    
3256    19. Added support for partial matching via the PCRE_PARTIAL option for
3257        pcre_exec() and the \P data escape in pcretest.
3258    
3259    20. Extended pcretest with 3 new pattern features:
3260    
3261        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3262              write the compiled pattern to the file whose name is "rest-of-line".
3263              This is a straight binary dump of the data, with the saved pointer to
3264              the character tables forced to be NULL. The study data, if any, is
3265              written too. After writing, pcretest reads a new pattern.
3266    
3267        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3268              compiled pattern from the given file. There must not be any
3269              occurrences of "<" in the file name (pretty unlikely); if there are,
3270              pcretest will instead treat the initial "<" as a pattern delimiter.
3271              After reading in the pattern, pcretest goes on to read data lines as
3272              usual.
3273    
3274        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3275              and 16-bit fields in a compiled pattern, to simulate a pattern that
3276              was compiled on a host of opposite endianness.
3277    
3278    21. The pcre-exec() function can now cope with patterns that were compiled on
3279        hosts of opposite endianness, with this restriction:
3280    
3281          As for any compiled expression that is saved and used later, the tables
3282          pointer field cannot be preserved; the extra_data field in the arguments
3283          to pcre_exec() should be used to pass in a tables address if a value
3284          other than the default internal tables were used at compile time.
3285    
3286    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3287        now diagnosed as an error. Previously, most of the time, a negative number
3288        would have been treated as zero, but if in addition "ovector" was passed as
3289        NULL, a crash could occur.
3290    
3291    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3292        new versions from the libtool 1.5 distribution (the last one is a copy of
3293        a file called libtool.m4). This seems to have fixed the need to patch
3294        "configure" to support Darwin 1.3 (which I used to do). However, I still
3295        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3296        workstation).
3297    
3298    24. Changed the PCRE licence to be the more standard "BSD" licence.
3299    
3300    
3301    Version 4.5 01-Dec-03
3302    ---------------------
3303    
3304     1. There has been some re-arrangement of the code for the match() function so
3305        that it can be compiled in a version that does not call itself recursively.
3306        Instead, it keeps those local variables that need separate instances for
3307        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3308        needs to "recurse". Keeping track of where control must go is done by means
3309        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3310        hide most of the details from the main code, and operates only if
3311        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3312        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3313        operating.
3314    
3315        To make it easier for callers to provide specially tailored get/free
3316        functions for this usage, two new functions, pcre_stack_malloc, and
3317        pcre_stack_free, are used. They are always called in strict stacking order,
3318        and the size of block requested is always the same.
3319    
3320        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3321        PCRE has been compiled to use the stack or the heap for recursion. The
3322        -C option of pcretest uses this to show which version is compiled.
3323    
3324        A new data escape \S, is added to pcretest; it causes the amounts of store
3325        obtained and freed by both kinds of malloc/free at match time to be added
3326        to the output.
3327    
3328     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3329        what's available on my current Linux desktop machine.
3330    
3331     3. When matching a UTF-8 string, the test for a valid string at the start has
3332        been extended. If start_offset is not zero, PCRE now checks that it points
3333        to a byte that is the start of a UTF-8 character. If not, it returns
3334        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3335        this is necessary because there may be backward assertions in the pattern.
3336        When matching the same subject several times, it may save resources to use
3337        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3338    
3339     4. The code for checking the validity of UTF-8 strings has been tightened so
3340        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3341        containing "overlong sequences".
3342    
3343     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3344        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3345        should have been "&", but it just so happened that all the cases this let
3346        through by mistake were picked up later in the function.
3347    
3348     6. I had used a variable called "isblank" - this is a C99 function, causing
3349        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3350    
3351     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3352        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3353        so the version is shown; (c) stop "make test" echoing "./RunTest".
3354    
3355     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3356    
3357     9. The prototype for memmove() for systems that don't have it was using
3358        size_t, but the inclusion of the header that defines size_t was later. I've
3359        moved the #includes for the C headers earlier to avoid this.
3360    
3361    10. Added some adjustments to the code to make it easier to compiler on certain
3362        special systems:
3363    
3364          (a) Some "const" qualifiers were missing.
3365          (b) Added the macro EXPORT before all exported functions; by default this
3366              is defined to be empty.
3367          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3368              that it reads its output file name as an argument instead of writing
3369              to the standard output and assuming this can be redirected.
3370    
3371    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3372        class containing characters with values greater than 255, PCRE compilation
3373        went into a loop.
3374    
3375    12. A recursive reference to a subpattern that was within another subpattern
3376        that had a minimum quantifier of zero caused PCRE to crash. For example,
3377        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3378        recursion. If the recursively-called subpattern itself had a zero repeat,
3379        that was OK.
3380    
3381    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3382        buffer into which it was copied (for escape processing) was still set at
3383        1024, so long lines caused crashes.
3384    
3385    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3386        "internal error: code overflow...". This applied to any character class
3387        that was followed by a possessive quantifier.
3388    
3389    15. Modified the Makefile to add libpcre.la as a prerequisite for
3390        libpcreposix.la because I was told this is needed for a parallel build to
3391        work.
3392    
3393    16. If a pattern that contained .* following optional items at the start was
3394        studied, the wrong optimizing data was generated, leading to matching
3395        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3396        matching string must start with a or b or c. The correct conclusion for
3397        this pattern is that a match can start with any character.
3398    
3399    
3400    Version 4.4 13-Aug-03
3401    ---------------------
3402    
3403     1. In UTF-8 mode, a character class containing characters with values between
3404        127 and 255 was not handled correctly if the compiled pattern was studied.
3405        In fixing this, I have also improved the studying algorithm for such
3406        classes (slightly).
3407    
3408     2. Three internal functions had redundant arguments passed to them. Removal
3409        might give a very teeny performance improvement.
3410    
3411     3. Documentation bug: the value of the capture_top field in a callout is *one
3412        more than* the number of the hightest numbered captured substring.
3413    
3414     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3415        in incorrectly linking with a previously installed version. They now link
3416        explicitly with libpcre.la.
3417    
3418     5. configure.in no longer needs to recognize Cygwin specially.
3419    
3420     6. A problem in pcre.in for Windows platforms is fixed.
3421    
3422     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3423        pcretest, it used to include the size of the study block as part of its
3424        output. Unfortunately, the structure contains a field that has a different
3425        size on different hardware architectures. This meant that the tests that
3426        showed this size failed. As the block is currently always of a fixed size,
3427        this information isn't actually particularly useful in pcretest output, so
3428        I have just removed it.
3429    
3430     8. Three pre-processor statements accidentally did not start in column 1.
3431        Sadly, there are *still* compilers around that complain, even though
3432        standard C has not required this for well over a decade. Sigh.
3433    
3434     9. In pcretest, the code for checking callouts passed small integers in the
3435        callout_data field, which is a void * field. However, some picky compilers
3436        complained about the casts involved for this on 64-bit systems. Now
3437        pcretest passes the address of the small integer instead, which should get
3438        rid of the warnings.
3439    
3440    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3441        both compile and run time, and gives an error if an invalid UTF-8 sequence
3442        is found. There is a option for disabling this check in cases where the
3443        string is known to be correct and/or the maximum performance is wanted.
3444    
3445    11. In response to a bug report, I changed one line in Makefile.in from
3446    
3447            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3448        to
3449            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3450    
3451        to look similar to other lines, but I have no way of telling whether this
3452        is the right thing to do, as I do not use Windows. No doubt I'll get told
3453        if it's wrong...
3454    
3455    
3456    Version 4.3 21-May-03
3457    ---------------------
3458    
3459    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3460       Makefile.
3461    
3462    2. Some refactoring to improve the quality of the code:
3463    
3464       (i)   The utf8_table... variables are now declared "const".
3465    
3466       (ii)  The code for \cx, which used the "case flipping" table to upper case
3467             lower case letters, now just substracts 32. This is ASCII-specific,
3468             but the whole concept of \cx is ASCII-specific, so it seems
3469             reasonable.
3470    
3471       (iii) PCRE was using its character types table to recognize decimal and
3472             hexadecimal digits in the pattern. This is silly, because it handles
3473             only 0-9, a-f, and A-F, but the character types table is locale-
3474             specific, which means strange things might happen. A private
3475             table is now used for this - though it costs 256 bytes, a table is
3476             much faster than multiple explicit tests. Of course, the standard
3477             character types table is still used for matching digits in subject
3478             strings against \d.
3479    
3480       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3481             ending in _t are). So I've renamed it as ESC_tee.
3482    
3483    3. The first argument for regexec() in the POSIX wrapper should have been
3484       defined as "const".
3485    
3486    4. Changed pcretest to use malloc() for its buffers so that they can be
3487       Electric Fenced for debugging.
3488    
3489    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3490       to read one or more bytes before the start of the subject string. Often this
3491       had no effect on PCRE's behaviour, but in some circumstances it could
3492       provoke a segmentation fault.
3493    
3494    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3495       to try to read one or more bytes before the start of the subject string.
3496    
3497    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3498       UTF-8 support could misbehave in various ways if the subject string
3499       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3500       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3501       back over UTF-8 characters.)
3502    
3503    
3504    Version 4.2 14-Apr-03
3505    ---------------------
3506    
3507    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3508    
3509    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3510         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3511         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3512         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3513         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3514           and BUILD_EXEEXT
3515         Note: automatic setting of the BUILD variables is not yet working
3516         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3517           compile-time but not at link-time
3518         [LINK]: use for linking executables only
3519         make different versions for Windows and non-Windows
3520         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3521           libraries
3522         [LINK_FOR_BUILD]: new variable
3523         [OBJEXT]: use throughout
3524         [EXEEXT]: use throughout
3525         <winshared>: new target
3526         <wininstall>: new target
3527         <dftables.o>: use native compiler
3528         <dftables>: use native linker
3529         <install>: handle Windows platform correctly
3530         <clean>: ditto
3531         <check>: ditto
3532         copy DLL to top builddir before testing
3533    
3534       As part of these changes, -no-undefined was removed again. This was reported
3535       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3536       in any case.
3537    
3538    3. Some tidies to get rid of compiler warnings:
3539