ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 5 by nigel, Sat Feb 24 21:38:05 2007 UTC revision 913 by ph10, Sun Feb 12 17:06:59 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.31
5    -----------------------------
7    1.  Fixing a wrong JIT test case and some compiler warnings.
9    2.  Removed a bashism from the RunTest script.
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)".
16    Version 8.30 04-February-2012
17    -----------------------------
19    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
20        name is defined in ctype.h.
22    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
23        only in quite long subpatterns.
25    3.  Removed the function pcre_info(), which has been obsolete and deprecated
26        since it was replaced by pcre_fullinfo() in February 2000.
28    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
29        match a (*MARK), and the match failed at the start of the subject, a
30        reference to memory before the start of the subject could occur. This bug
31        was introduced by fix 17 of release 8.21.
33    5.  A reference to an unset group with zero minimum repetition was giving
34        totally wrong answers (in non-JavaScript-compatibility mode). For example,
35        /(another)?(\1?)test/ matched against "hello world test". This bug was
36        introduced in release 8.13.
38    6.  Add support for 16-bit character strings (a large amount of work involving
39        many changes and refactorings).
41    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
42        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
43        from a file.
45    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
46        rounding is not applied in this particular case).
48    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
49        if they appear, or are escaped, in patterns.
51    10. Get rid of a number of -Wunused-but-set-variable warnings.
53    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
54        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
55        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
56        also returns the mark "x". This bug applied to capturing parentheses,
57        non-capturing parentheses, and atomic parentheses. It also applied to some
58        assertions.
60    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
61        information out of configure.ac instead of relying on pcre.h.generic, which
62        is not stored in the repository.
64    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
65        -lreadline.
67    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
69    15. Applied Graycode's patch to put the top-level frame on the stack rather
70        than the heap when not using the stack for recursion. This gives a
71        performance improvement in many cases when recursion is not deep.
73    16. Experimental code added to "pcretest -C" to output the stack frame size.
76    Version 8.21 12-Dec-2011
77    ------------------------
79    1.  Updating the JIT compiler.
81    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
82        are added as well.
84    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
85        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
86        calling _pcre_jit_exec. Some extra comments are added.
88    4.  (*MARK) settings inside atomic groups that do not contain any capturing
89        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
90        was introduced by change 18 for 8.20.
92    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
93        ECMA-262 standard.
95    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
96        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
97        This bug was probably introduced by change 9 of 8.13.
99    7.  While fixing 6 above, I noticed that a number of other items were being
100        incorrectly rejected as "not fixed length". This arose partly because newer
101        opcodes had not been added to the fixed-length checking code. I have (a)
102        corrected the bug and added tests for these items, and (b) arranged for an
103        error to occur if an unknown opcode is encountered while checking for fixed
104        length instead of just assuming "not fixed length". The items that were
105        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
106        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
107        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
109    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
110        being incorrectly compiled and would have given unpredicatble results.
112    9.  A possessively repeated subpattern with minimum repeat count greater than
113        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
114        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
115        the first (A) could occur when it should not.
117    10. Add a cast and remove a redundant test from the code.
119    11. JIT should use pcre_malloc/pcre_free for allocation.
121    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
122        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
123        is anything other than /usr, -L is still shown).
125    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
127    14. Perl does not support \N without a following name in a [] class; PCRE now
128        also gives an error.
130    15. If a forward reference was repeated with an upper limit of around 2000,
131        it caused the error "internal error: overran compiling workspace". The
132        maximum number of forward references (including repeats) was limited by the
133        internal workspace, and dependent on the LINK_SIZE. The code has been
134        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
135        the default depends on LINK_SIZE. There is a new upper limit (for safety)
136        of around 200,000 forward references. While doing this, I also speeded up
137        the filling in of repeated forward references.
139    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
140        incorrectly expecting the subject to contain another "a" after the start.
142    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
143        in the match, the SKIP should be ignored. This was not happening; instead
144        the SKIP was being treated as NOMATCH. For patterns such as
145        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
146        tested.
148    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
149        now much more compatible with Perl, in particular in cases where the result
150        is a non-match for a non-anchored pattern. For example, if
151        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
152        "m", where previously it did not return a name. A side effect of this
153        change is that for partial matches, the last encountered mark name is
154        returned, as for non matches. A number of tests that were previously not
155        Perl-compatible have been moved into the Perl-compatible test files. The
156        refactoring has had the pleasing side effect of removing one argument from
157        the match() function, thus reducing its stack requirements.
159    19. If the /S+ option was used in pcretest to study a pattern using JIT,
160        subsequent uses of /S (without +) incorrectly behaved like /S+.
162    21. Retrieve executable code size support for the JIT compiler and fixing
163        some warnings.
165    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
166        not work when the shorter character appeared right at the end of the
167        subject string.
169    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
170        systems.
172    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
173        output it when the /M option is used in pcretest.
175    25. The CheckMan script was not being included in the distribution. Also, added
176        an explicit "perl" to run Perl scripts from the PrepareRelease script
177        because this is reportedly needed in Windows.
179    26. If study data was being save in a file and studying had not found a set of
180        "starts with" bytes for the pattern, the data written to the file (though
181        never used) was taken from uninitialized memory and so caused valgrind to
182        complain.
184    27. Updated RunTest.bat as provided by Sheri Pierce.
186    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
188    29. Computation of memory usage for the table of capturing group names was
189        giving an unnecessarily large value.
192    Version 8.20 21-Oct-2011
193    ------------------------
195    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
196        a POSIX class. After further experiments with Perl, which convinced me that
197        Perl has bugs and confusions, a closing square bracket is no longer allowed
198        in a POSIX name. This bug also affected patterns with classes that started
199        with full stops.
201    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
202        captured substring, but while checking the failing first alternative,
203        substring 1 is temporarily captured. If the output vector supplied to
204        pcre_exec() was not big enough for this capture, the yield of the function
205        was still zero ("insufficient space for captured substrings"). This cannot
206        be totally fixed without adding another stack variable, which seems a lot
207        of expense for a edge case. However, I have improved the situation in cases
208        such as /(a)(b)x|abc/ matched against "abc", where the return code
209        indicates that fewer than the maximum number of slots in the ovector have
210        been set.
212    3.  Related to (2) above: when there are more back references in a pattern than
213        slots in the output vector, pcre_exec() uses temporary memory during
214        matching, and copies in the captures as far as possible afterwards. It was
215        using the entire output vector, but this conflicts with the specification
216        that only 2/3 is used for passing back captured substrings. Now it uses
217        only the first 2/3, for compatibility. This is, of course, another edge
218        case.
220    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
221        main code base, and can be used by building with --enable-jit. When this is
222        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
223        runtime --no-jit option is given.
225    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
226        ovector, the return from the function was zero, implying that there were
227        other matches that did not fit. The correct "exactly full" value is now
228        returned.
230    6.  If a subpattern that was called recursively or as a subroutine contained
231        (*PRUNE) or any other control that caused it to give a non-standard return,
232        invalid errors such as "Error -26 (nested recursion at the same subject
233        position)" or even infinite loops could occur.
235    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
236        computing the minimum length on reaching *ACCEPT, and so ended up with the
237        wrong value of 1 rather than 0. Further investigation indicates that
238        computing a minimum subject length in the presence of *ACCEPT is difficult
239        (think back references, subroutine calls), and so I have changed the code
240        so that no minimum is registered for a pattern that contains *ACCEPT.
242    8.  If (*THEN) was present in the first (true) branch of a conditional group,
243        it was not handled as intended. [But see 16 below.]
245    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
246        Sheri Pierce.
248    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
249        the first byte in a match must be "a".
251    11. Change 17 for 8.13 increased the recursion depth for patterns like
252        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
253        pattern contains any instances of (*THEN). If it does not, the old
254        optimizations are restored. It would be nice to do this on a per-group
255        basis, but at the moment that is not feasible.
257    12. In some environments, the output of pcretest -C is CRLF terminated. This
258        broke RunTest's code that checks for the link size. A single white space
259        character after the value is now allowed for.
261    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
262        For "fr", it uses the Windows-specific input and output files.
264    14. If (*THEN) appeared in a group that was called recursively or as a
265        subroutine, it did not work as intended. [But see next item.]
267    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
268        pattern fragments (but not containing any | characters). If A and B are
269        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
270        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
271        D. In other words, Perl considers parentheses that do not contain any |
272        characters to be part of a surrounding alternative, whereas PCRE was
273        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
274        differently. PCRE now behaves in the same way as Perl, except in the case
275        of subroutine/recursion calls such as (?1) which have in any case always
276        been different (but PCRE had them first :-).
278    16. Related to 15 above: Perl does not treat the | in a conditional group as
279        creating alternatives. Such a group is treated in the same way as an
280        ordinary group without any | characters when processing (*THEN). PCRE has
281        been changed to match Perl's behaviour.
283    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
284        RunGrepTest script failed.
286    18. Change 22 for version 13 caused atomic groups to use more stack. This is
287        inevitable for groups that contain captures, but it can lead to a lot of
288        stack use in large patterns. The old behaviour has been restored for atomic
289        groups that do not contain any capturing parentheses.
291    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
292        suppress the check for a minimum subject length at run time. (If it was
293        given to pcre_exec() or pcre_dfa_exec() it did work.)
295    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
296        fail to work when decoding hex characters in data strings in EBCDIC
297        environments.
299    21. It appears that in at least one Mac OS environment, the isxdigit() function
300        is implemented as a macro that evaluates to its argument more than once,
301        contravening the C 90 Standard (I haven't checked a later standard). There
302        was an instance in pcretest which caused it to go wrong when processing
303        \x{...} escapes in subject strings. The has been rewritten to avoid using
304        things like p++ in the argument of isxdigit().
307    Version 8.13 16-Aug-2011
308    ------------------------
310    1.  The Unicode data tables have been updated to Unicode 6.0.0.
312    2.  Two minor typos in pcre_internal.h have been fixed.
314    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
315        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
316        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
318    4.  There were a number of related bugs in the code for matching backrefences
319        caselessly in UTF-8 mode when codes for the characters concerned were
320        different numbers of bytes. For example, U+023A and U+2C65 are an upper
321        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
322        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
323        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
324        2-byte code at the end of the subject (it thought there wasn't enough data
325        left).
327    5.  Comprehensive information about what went wrong is now returned by
328        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
329        as the output vector has at least 2 elements. The offset of the start of
330        the failing character and a reason code are placed in the vector.
332    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
333        now returned is for the first byte of the failing character, instead of the
334        last byte inspected. This is an incompatible change, but I hope it is small
335        enough not to be a problem. It makes the returned offset consistent with
336        pcre_exec() and pcre_dfa_exec().
338    7.  pcretest now gives a text phrase as well as the error number when
339        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
340        failure, the offset and reason code are output.
342    8.  When \R was used with a maximizing quantifier it failed to skip backwards
343        over a \r\n pair if the subsequent match failed. Instead, it just skipped
344        back over a single character (\n). This seems wrong (because it treated the
345        two characters as a single entity when going forwards), conflicts with the
346        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
347        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
348        has been changed.
350    9.  Some internal refactoring has changed the processing so that the handling
351        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
352        time (the PCRE_DOTALL option was changed this way some time ago: version
353        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
354        which was always a bit of a fudge. It also means that there is one less
355        argument for the match() function, which reduces its stack requirements
356        slightly. This change also fixes an incompatibility with Perl: the pattern
357        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
359    10. More internal refactoring has drastically reduced the number of recursive
360        calls to match() for possessively repeated groups such as (abc)++ when
361        using pcre_exec().
363    11. While implementing 10, a number of bugs in the handling of groups were
364        discovered and fixed:
366        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
367        (a|)*(?1) gave a compile-time internal error.
368        ((a|)+)+  did not notice that the outer group could match an empty string.
369        (^a|^)+   was not marked as anchored.
370        (.*a|.*)+ was not marked as matching at start or after a newline.
372    12. Yet more internal refactoring has removed another argument from the match()
373        function. Special calls to this function are now indicated by setting a
374        value in a variable in the "match data" data block.
376    13. Be more explicit in pcre_study() instead of relying on "default" for
377        opcodes that mean there is no starting character; this means that when new
378        ones are added and accidentally left out of pcre_study(), testing should
379        pick them up.
381    14. The -s option of pcretest has been documented for ages as being an old
382        synonym of -m (show memory usage). I have changed it to mean "force study
383        for every regex", that is, assume /S for every regex. This is similar to -i
384        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
385        using it. It makes it easier to run collections of tests with and without
386        study enabled, and thereby test pcre_study() more easily. All the standard
387        tests are now run with and without -s (but some patterns can be marked as
388        "never study" - see 20 below).
390    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
391        restoration of the capturing data to the outer values was not happening
392        correctly.
394    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
395        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
396        pattern had matched an empty string, and so incorrectly returned a no
397        match.
399    17. There was optimizing code for the last branch of non-capturing parentheses,
400        and also for the obeyed branch of a conditional subexpression, which used
401        tail recursion to cut down on stack usage. Unfortunately, now that there is
402        the possibility of (*THEN) occurring in these branches, tail recursion is
403        no longer possible because the return has to be checked for (*THEN). These
404        two optimizations have therefore been removed. [But see 8.20/11 above.]
406    18. If a pattern containing \R was studied, it was assumed that \R always
407        matched two bytes, thus causing the minimum subject length to be
408        incorrectly computed because \R can also match just one byte.
410    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
411        was incorrectly computed.
413    20. If /S is present twice on a test pattern in pcretest input, it now
414        *disables* studying, thereby overriding the use of -s on the command line
415        (see 14 above). This is necessary for one or two tests to keep the output
416        identical in both cases.
418    21. When (*ACCEPT) was used in an assertion that matched an empty string and
419        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
421    22. When an atomic group that contained a capturing parenthesis was
422        successfully matched, but the branch in which it appeared failed, the
423        capturing was not being forgotten if a higher numbered group was later
424        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
425        group 1 to "a", when in fact it should be unset. This applied to multi-
426        branched capturing and non-capturing groups, repeated or not, and also to
427        positive assertions (capturing in negative assertions does not happen
428        in PCRE) and also to nested atomic groups.
430    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
431        subject after a captured substring, to make it easier to tell which of a
432        number of identical substrings has been captured.
434    24. The way atomic groups are processed by pcre_exec() has been changed so that
435        if they are repeated, backtracking one repetition now resets captured
436        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
437        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
438        "aaa". Previously, it would have been "a". As part of this code
439        refactoring, the way recursive calls are handled has also been changed.
441    25. If an assertion condition captured any substrings, they were not passed
442        back unless some other capturing happened later. For example, if
443        (?(?=(a))a) was matched against "a", no capturing was returned.
445    26. When studying a pattern that contained subroutine calls or assertions,
446        the code for finding the minimum length of a possible match was handling
447        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
448        group 1 called group 2 while simultaneously a separate group 2 called group
449        1). A stack overflow occurred in this case. I have fixed this by limiting
450        the recursion depth to 10.
452    27. Updated RunTest.bat in the distribution to the version supplied by Tom
453        Fortmann. This supports explicit test numbers on the command line, and has
454        argument validation and error reporting.
456    28. An instance of \X with an unlimited repeat could fail if at any point the
457        first character it looked at was a mark character.
459    29. Some minor code refactoring concerning Unicode properties and scripts
460        should reduce the stack requirement of match() slightly.
462    30. Added the '=' option to pcretest to check the setting of unused capturing
463        slots at the end of the pattern, which are documented as being -1, but are
464        not included in the return count.
466    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
467        compiled something random. Now it gives a compile-time error (as does
468        Perl).
470    32. A *MARK encountered during the processing of a positive assertion is now
471        recorded and passed back (compatible with Perl).
473    33. If --only-matching or --colour was set on a pcregrep call whose pattern
474        had alternative anchored branches, the search for a second match in a line
475        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
476        matched the line "0102" twice. The same bug affected patterns that started
477        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
478        twice.
480    34. Previously, PCRE did not allow quantification of assertions. However, Perl
481        does, and because of capturing effects, quantifying parenthesized
482        assertions may at times be useful. Quantifiers are now allowed for
483        parenthesized assertions.
485    35. A minor code tidy in pcre_compile() when checking options for \R usage.
487    36. \g was being checked for fancy things in a character class, when it should
488        just be a literal "g".
490    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
491        appearance of a nested POSIX class supersedes an apparent external class.
492        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
493        unescaped square brackets may also appear as part of class names. For
494        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
495        more like Perl. (But see 8.20/1 above.)
497    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
498        was because it thought it was \N{name}, which is not supported).
500    39. Add minix to OS list not supporting the -S option in pcretest.
502    40. PCRE tries to detect cases of infinite recursion at compile time, but it
503        cannot analyze patterns in sufficient detail to catch mutual recursions
504        such as ((?1))((?2)). There is now a runtime test that gives an error if a
505        subgroup is called recursively as a subpattern for a second time at the
506        same position in the subject string. In previous releases this might have
507        been caught by the recursion limit, or it might have run out of stack.
509    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
510        happen only once. PCRE was, however incorrectly giving a compile time error
511        "recursive call could loop indefinitely" because it cannot analyze the
512        pattern in sufficient detail. The compile time test no longer happens when
513        PCRE is compiling a conditional subpattern, but actual runaway loops are
514        now caught at runtime (see 40 above).
516    42. It seems that Perl allows any characters other than a closing parenthesis
517        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
518        has been changed to be the same.
520    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
521        as not to get warnings when autogen.sh is called. Also changed
522        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
524    44. To help people who use pcregrep to scan files containing exceedingly long
525        lines, the following changes have been made:
527        (a) The default value of the buffer size parameter has been increased from
528            8K to 20K. (The actual buffer used is three times this size.)
530        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
531            PCRE is built.
533        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
534            to be set at run time.
536        (d) Numerical values in pcregrep options can be followed by K or M, for
537            example --buffer-size=50K.
539        (e) If a line being scanned overflows pcregrep's buffer, an error is now
540            given and the return code is set to 2.
542    45. Add a pointer to the latest mark to the callout data block.
544    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
545        partial match of an empty string instead of no match. This was specific to
546        the use of ".".
548    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
549        complete match instead of a partial match. This bug was dependent on both
550        the PCRE_UTF8 and PCRE_DOTALL options being set.
552    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
553        starting byte set, because \b was not being ignored.
556    Version 8.12 15-Jan-2011
557    ------------------------
559    1.  Fixed some typos in the markup of the man pages, and wrote a script that
560        checks for such things as part of the documentation building process.
562    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
563        --match-limit and --recursion-limit options (added for 8.11). In
564        particular, this made one of the standard tests fail. (The integer value
565        went into the wrong half of a long int.)
567    3.  If the --colour option was given to pcregrep with -v (invert match), it
568        did strange things, either producing crazy output, or crashing. It should,
569        of course, ignore a request for colour when reporting lines that do not
570        match.
572    4.  Another pcregrep bug caused similar problems if --colour was specified with
573        -M (multiline) and the pattern match finished with a line ending.
575    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
576        matched in multiline mode, the following line was shown as part of the
577        match. This seems wrong, so I have changed it.
579    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
580        the check for further matches in the same line (so they could be coloured)
581        to overrun the end of the current line. If another match was found, it was
582        incorrectly shown (and then shown again when found in the next line).
584    7.  If pcregrep was compiled under Windows, there was a reference to the
585        function pcregrep_exit() before it was defined. I am assuming this was
586        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
587        reported by a user. I've moved the definition above the reference.
590    Version 8.11 10-Dec-2010
591    ------------------------
593    1.  (*THEN) was not working properly if there were untried alternatives prior
594        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
595        backtracked to try for "b" instead of moving to the next alternative branch
596        at the same level (in this case, to look for "c"). The Perl documentation
597        is clear that when (*THEN) is backtracked onto, it goes to the "next
598        alternative in the innermost enclosing group".
600    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
601        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
602        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
603        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
604        (*THEN).
606    3.  If \s appeared in a character class, it removed the VT character from
607        the class, even if it had been included by some previous item, for example
608        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
609        of \s, but is part of the POSIX "space" class.)
611    4.  A partial match never returns an empty string (because you can always
612        match an empty string at the end of the subject); however the checking for
613        an empty string was starting at the "start of match" point. This has been
614        changed to the "earliest inspected character" point, because the returned
615        data for a partial match starts at this character. This means that, for
616        example, /(?<=abc)def/ gives a partial match for the subject "abc"
617        (previously it gave "no match").
619    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
620        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
621        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
622        has an implication that the given string is incomplete (because a partial
623        match is preferred over a full match). For this reason, these items now
624        give a partial match in this situation. [Aside: previously, the one case
625        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
626        match rather than a full match, which was wrong by the old rules, but is
627        now correct.]
629    6.  There was a bug in the handling of #-introduced comments, recognized when
630        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
631        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
632        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
633        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
634        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
635        places in pcre_compile().
637    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
638        comments when looking ahead for named forward references to subpatterns,
639        the only newline sequence it recognized was NL. It now handles newlines
640        according to the set newline convention.
642    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
643        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
644        cater for a lack of strerror(). These oversights have been fixed.
646    9.  Added --match-limit and --recursion-limit to pcregrep.
648    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
650    11. When the -o option was used, pcregrep was setting a return code of 1, even
651        when matches were found, and --line-buffered was not being honoured.
653    12. Added an optional parentheses number to the -o and --only-matching options
654        of pcregrep.
656    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
657        can match an empty string. The code to do it in pcretest and pcredemo
658        needed fixing:
660        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
661            only one byte after an empty string match just before CRLF (this case
662            just got forgotten; "any" and "anycrlf" were OK).
664        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
665            mode when an empty string match preceded an ASCII character followed by
666            a non-ASCII character. (The code for advancing by one character rather
667            than one byte was nonsense.)
669        (c) The pcredemo.c sample program did not have any code at all to handle
670            the cases when CRLF is a valid newline sequence.
672    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
673        as a starting offset was within the subject string. There is now a new
674        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
675        negative or greater than the length of the string. In order to test this,
676        pcretest is extended to allow the setting of negative starting offsets.
678    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
679        starting offset points to the beginning of a UTF-8 character was
680        unnecessarily clumsy. I tidied it up.
682    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
683        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
685    17. Nobody had reported that the --include_dir option, which was added in
686        release 7.7 should have been called --include-dir (hyphen, not underscore)
687        for compatibility with GNU grep. I have changed it to --include-dir, but
688        left --include_dir as an undocumented synonym, and the same for
689        --exclude-dir, though that is not available in GNU grep, at least as of
690        release 2.5.4.
692    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
693        characters from a string of bytes) have been redefined so as not to use
694        loops, in order to improve performance in some environments. At the same
695        time, I abstracted some of the common code into auxiliary macros to save
696        repetition (this should not affect the compiled code).
698    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
699        compile-time error is now given if \c is not followed by an ASCII
700        character, that is, a byte less than 128. (In EBCDIC mode, the code is
701        different, and any byte value is allowed.)
703    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
704        START_OPTIMIZE option, which is now allowed at compile time - but just
705        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
706        to pcregrep and other applications that have no direct access to PCRE
707        options. The new /Y option in pcretest sets this option when calling
708        pcre_compile().
710    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
711        back references. Groups containing recursive back references were forced to
712        be atomic by that change, but in the case of named groups, the amount of
713        memory required was incorrectly computed, leading to "Failed: internal
714        error: code overflow". This has been fixed.
716    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
717        pcretest.c, to avoid build problems in some Borland environments.
720    Version 8.10 25-Jun-2010
721    ------------------------
723    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
724        THEN.
726    2.  (*ACCEPT) was not working when inside an atomic group.
728    3.  Inside a character class, \B is treated as a literal by default, but
729        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
730        causes the error). The code is unchanged, but I tidied the documentation.
732    4.  Inside a character class, PCRE always treated \R and \X as literals,
733        whereas Perl faults them if its -w option is set. I have changed PCRE so
734        that it faults them when PCRE_EXTRA is set.
736    5.  Added support for \N, which always matches any character other than
737        newline. (It is the same as "." when PCRE_DOTALL is not set.)
739    6.  When compiling pcregrep with newer versions of gcc which may have
740        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
741        declared with attribute warn_unused_result" were given. Just casting the
742        result to (void) does not stop the warnings; a more elaborate fudge is
743        needed. I've used a macro to implement this.
745    7.  Minor change to pcretest.c to avoid a compiler warning.
747    8.  Added four artifical Unicode properties to help with an option to make
748        \s etc use properties (see next item). The new properties are: Xan
749        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
751    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
752        use Unicode properties. (*UCP) at the start of a pattern can be used to set
753        this option. Modified pcretest to add /W to test this facility. Added
754        REG_UCP to make it available via the POSIX interface.
756    10. Added --line-buffered to pcregrep.
758    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
759        studied, and the match started with a letter with a code point greater than
760        127 whose first byte was different to the first byte of the other case of
761        the letter, the other case of this starting letter was not recognized
762        (#976).
764    12. If a pattern that was studied started with a repeated Unicode property
765        test, for example, \p{Nd}+, there was the theoretical possibility of
766        setting up an incorrect bitmap of starting bytes, but fortunately it could
767        not have actually happened in practice until change 8 above was made (it
768        added property types that matched character-matching opcodes).
770    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
771        possible starting bytes for non-anchored patterns.
773    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
774        \R, and also a number of cases that involve Unicode properties, both
775        explicit and implicit when PCRE_UCP is set.
777    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
778        input, it could crash or give wrong results if characters with values
779        greater than 0xc0 were present in the subject string. (Detail: it assumed
780        UTF-8 input when processing these items.)
782    16. Added a lot of (int) casts to avoid compiler warnings in systems where
783        size_t is 64-bit (#991).
785    17. Added a check for running out of memory when PCRE is compiled with
786        --disable-stack-for-recursion (#990).
788    18. If the last data line in a file for pcretest does not have a newline on
789        the end, a newline was missing in the output.
791    19. The default pcre_chartables.c file recognizes only ASCII characters (values
792        less than 128) in its various bitmaps. However, there is a facility for
793        generating tables according to the current locale when PCRE is compiled. It
794        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
795        characters, are recognized by isspace() and therefore were getting set in
796        these tables, and indeed these tables seem to approximate to ISO 8859. This
797        caused a problem in UTF-8 mode when pcre_study() was used to create a list
798        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
799        which of course cannot start UTF-8 characters. I have changed the code so
800        that only real ASCII characters (less than 128) and the correct starting
801        bytes for UTF-8 encodings are set for characters greater than 127 when in
802        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
803        altogether.)
805    20. Added the /T option to pcretest so as to be able to run tests with non-
806        standard character tables, thus making it possible to include the tests
807        used for 19 above in the standard set of tests.
809    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
810        reference to a subpattern the other side of a comment that contains an
811        opening parenthesis caused either an internal compiling error, or a
812        reference to the wrong subpattern.
815    Version 8.02 19-Mar-2010
816    ------------------------
818    1.  The Unicode data tables have been updated to Unicode 5.2.0.
820    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
821        configured.
823    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
824        original author of that file, following a query about its status.
826    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
827        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
829    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
830        quantifier applied to a forward-referencing subroutine call, could compile
831        incorrect code or give the error "internal error: previously-checked
832        referenced subpattern not found".
834    6.  Both MS Visual Studio and Symbian OS have problems with initializing
835        variables to point to external functions. For these systems, therefore,
836        pcre_malloc etc. are now initialized to local functions that call the
837        relevant global functions.
839    7.  There were two entries missing in the vectors called coptable and poptable
840        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
841        I've fixed the data, and added a kludgy way of testing at compile time that
842        the lengths are correct (equal to the number of opcodes).
844    8.  Following on from 7, I added a similar kludge to check the length of the
845        eint vector in pcreposix.c.
847    9.  Error texts for pcre_compile() are held as one long string to avoid too
848        much relocation at load time. To find a text, the string is searched,
849        counting zeros. There was no check for running off the end of the string,
850        which could happen if a new error number was added without updating the
851        string.
853    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
855    11. \K was not working if it appeared in an atomic group or in a group that
856        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
857        \K is "not well defined" if used in an assertion. PCRE now accepts it if
858        the assertion is positive, but not if it is negative.
860    12. Change 11 fortuitously reduced the size of the stack frame used in the
861        "match()" function of pcre_exec.c by one pointer. Forthcoming
862        implementation of support for (*MARK) will need an extra pointer on the
863        stack; I have reserved it now, so that the stack frame size does not
864        decrease.
866    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
867        item in branch that calls a recursion is a subroutine call - as in the
868        second branch in the above example - was incorrectly given the compile-
869        time error "recursive call could loop indefinitely" because pcre_compile()
870        was not correctly checking the subroutine for matching a non-empty string.
872    14. The checks for overrunning compiling workspace could trigger after an
873        overrun had occurred. This is a "should never occur" error, but it can be
874        triggered by pathological patterns such as hundreds of nested parentheses.
875        The checks now trigger 100 bytes before the end of the workspace.
877    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
880    Version 8.01 19-Jan-2010
881    ------------------------
883    1.  If a pattern contained a conditional subpattern with only one branch (in
884        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
885        computed the wrong minimum data length (which is of course zero for such
886        subpatterns). This could cause incorrect "no match" results.
888    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
889        the pattern is reset in the first branch, pcre_compile() failed with
890        "internal error: code overflow at offset...". This happened only when
891        the reset was to the original external option setting. (An optimization
892        abstracts leading options settings into an external setting, which was the
893        cause of this.)
895    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
896        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
897        assertion pattern did not match (meaning that the assertion was true), it
898        was incorrectly treated as false if the SKIP had been reached during the
899        matching. This also applied to assertions used as conditions.
901    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
902        assertion subpattern, including such a pattern used as a condition,
903        unpredictable results occurred, instead of the error return
906    5.  The C++ GlobalReplace function was not working like Perl for the special
907        situation when an empty string is matched. It now does the fancy magic
908        stuff that is necessary.
910    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
911        removed. (These were left over from very, very early versions of PCRE.)
913    7.  Some cosmetic changes to the code to make life easier when compiling it
914        as part of something else:
916        (a) Change DEBUG to PCRE_DEBUG.
918        (b) In pcre_compile(), rename the member of the "branch_chain" structure
919            called "current" as "current_branch", to prevent a collision with the
920            Linux macro when compiled as a kernel module.
922        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
923            prevent a collision with the Linux macro when compiled as a kernel
924            module.
926    8.  In pcre_compile() there are some checks for integer overflows that used to
927        cast potentially large values to (double). This has been changed to that
928        when building, a check for int64_t is made, and if it is found, it is used
929        instead, thus avoiding the use of floating point arithmetic. (There is no
930        other use of FP in PCRE.) If int64_t is not found, the fallback is to
931        double.
933    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
934        2005 (difference between two addresses compared to an unsigned value).
936    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
937        custom one, because of the following reported problem in Windows:
939          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
940              under Win32.
941          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
942              therefore missing the function definition.
943          - The compiler thus generates a "C" signature for the test function.
944          - The linker fails to find the "C" function.
945          - PCRE fails to configure if asked to do so against libbz2.
947    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
948        messages were output:
950          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
951          rerunning libtoolize, to keep the correct libtool macros in-tree.
952          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
954        I have done both of these things.
956    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
957        most of the time, it *can* run out if it is given a pattern that contains a
958        runaway infinite recursion. I updated the discussion in the pcrestack man
959        page.
961    13. Now that we have gone to the x.xx style of version numbers, the minor
962        version may start with zero. Using 08 or 09 is a bad idea because users
963        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
964        interpreted as invalid octal numbers. I've updated the previous comment in
965        configure.ac, and also added a check that gives an error if 08 or 09 are
966        used.
968    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
969        causing partial matching to fail when the end of the subject matched \W
970        in a UTF-8 pattern where \W was quantified with a minimum of 3.
972    15. There were some discrepancies between the declarations in pcre_internal.h
973        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
974        their definitions. The declarations used "const uschar *" and the
975        definitions used USPTR. Even though USPTR is normally defined as "const
976        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
977        reported that: "This difference in casting confuses some C++ compilers, for
978        example, SunCC recognizes above declarations as different functions and
979        generates broken code for hbpcre." I have changed the declarations to use
980        USPTR.
982    16. GNU libtool is named differently on some systems. The autogen.sh script now
983        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
984        (FreeBSD).
986    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
987        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
988        comment: "Figure out how to create a longlong from a string: strtoll and
989        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
990        instance, but it only takes 2 args instead of 3!"
992    18. A subtle bug concerned with back references has been fixed by a change of
993        specification, with a corresponding code fix. A pattern such as
994        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
995        refers, was giving matches when it shouldn't. For example, xa=xaaa would
996        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
997        same bug. Such groups have to be quantified to be useful, or contained
998        inside another quantified group. (If there's no repetition, the reference
999        can never match.) The problem arises because, having left the group and
1000        moved on to the rest of the pattern, a later failure that backtracks into
1001        the group uses the captured value from the final iteration of the group
1002        rather than the correct earlier one. I have fixed this in PCRE by forcing
1003        any group that contains a reference to itself to be an atomic group; that
1004        is, there cannot be any backtracking into it once it has completed. This is
1005        similar to recursive and subroutine calls.
1008    Version 8.00 19-Oct-09
1009    ----------------------
1011    1.  The table for translating pcre_compile() error codes into POSIX error codes
1012        was out-of-date, and there was no check on the pcre_compile() error code
1013        being within the table. This could lead to an OK return being given in
1014        error.
1016    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1017        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1018        in a Windows environment.
1020    3.  The pcregrep --count option prints the count for each file even when it is
1021        zero, as does GNU grep. However, pcregrep was also printing all files when
1022        --files-with-matches was added. Now, when both options are given, it prints
1023        counts only for those files that have at least one match. (GNU grep just
1024        prints the file name in this circumstance, but including the count seems
1025        more useful - otherwise, why use --count?) Also ensured that the
1026        combination -clh just lists non-zero counts, with no names.
1028    4.  The long form of the pcregrep -F option was incorrectly implemented as
1029        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1030        but it seems right to fix it, and I didn't think it was worth preserving
1031        the old behaviour.
1033    5.  The command line items --regex=pattern and --regexp=pattern were not
1034        recognized by pcregrep, which required --regex pattern or --regexp pattern
1035        (with a space rather than an '='). The man page documented the '=' forms,
1036        which are compatible with GNU grep; these now work.
1038    6.  No libpcreposix.pc file was created for pkg-config; there was just
1039        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1041    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1042        when UCP support is not needed, by modifying the Python script that
1043        generates it from Unicode data files. This should not matter if the module
1044        is correctly used as a library, but I received one complaint about 50K of
1045        unwanted data. My guess is that the person linked everything into his
1046        program rather than using a library. Anyway, it does no harm.
1048    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1049        was a minimum greater than 1 for a wide character in a possessive
1050        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1051        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1052        character. Chaos in the form of incorrect output or a compiling loop could
1053        result.
1055    9.  The restrictions on what a pattern can contain when partial matching is
1056        requested for pcre_exec() have been removed. All patterns can now be
1057        partially matched by this function. In addition, if there are at least two
1058        slots in the offset vector, the offset of the earliest inspected character
1059        for the match and the offset of the end of the subject are set in them when
1060        PCRE_ERROR_PARTIAL is returned.
1062    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1063        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1064        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1065        and may be more useful for multi-segment matching.
1067    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1068        used to be given if ever the end of the subject was reached; now it is
1069        given only if matching could not proceed because another character was
1070        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1071        string "Z", which now yields "no match" instead of "partial match". In the
1072        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1073        final character ended with (*FAIL).
1075    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1076        if the pattern had a "must contain" character that was already found in the
1077        earlier partial match, unless partial matching was again requested. For
1078        example, with the pattern /dog.(body)?/, the "must contain" character is
1079        "g". If the first part-match was for the string "dog", restarting with
1080        "sbody" failed. This bug has been fixed.
1082    13. The string returned by pcre_dfa_exec() after a partial match has been
1083        changed so that it starts at the first inspected character rather than the
1084        first character of the match. This makes a difference only if the pattern
1085        starts with a lookbehind assertion or \b or \B (\K is not supported by
1086        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1087        matching functions compatible, and I think it's the right thing to do.
1089    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1090        so that the demonstration program is easily available in environments where
1091        PCRE has not been installed from source.
1093    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1094        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1095        library.
1097    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1098        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1099        is not the first non-POSIX option to be added. Clearly some people find
1100        these options useful.
1102    17. If a caller to the POSIX matching function regexec() passes a non-zero
1103        value for nmatch with a NULL value for pmatch, the value of
1104        nmatch is forced to zero.
1106    18. RunGrepTest did not have a test for the availability of the -u option of
1107        the diff command, as RunTest does. It now checks in the same way as
1108        RunTest, and also checks for the -b option.
1110    19. If an odd number of negated classes containing just a single character
1111        interposed, within parentheses, between a forward reference to a named
1112        subpattern and the definition of the subpattern, compilation crashed with
1113        an internal error, complaining that it could not find the referenced
1114        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1115        [The bug was that it was starting one character too far in when skipping
1116        over the character class, thus treating the ] as data rather than
1117        terminating the class. This meant it could skip too much.]
1119    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1120        /g option in pcretest when the pattern contains \K, which makes it possible
1121        to have an empty string match not at the start, even when the pattern is
1122        anchored. Updated pcretest and pcredemo to use this option.
1124    21. If the maximum number of capturing subpatterns in a recursion was greater
1125        than the maximum at the outer level, the higher number was returned, but
1126        with unset values at the outer level. The correct (outer level) value is
1127        now given.
1129    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1130        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1131        make it do so. The string so far is captured, making this feature
1132        compatible with Perl.
1134    23. The tests have been re-organized, adding tests 11 and 12, to make it
1135        possible to check the Perl 5.10 features against Perl 5.10.
1137    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1138        pattern matches a fixed length string. PCRE did not allow this; now it
1139        does. Neither allows recursion.
1141    25. I finally figured out how to implement a request to provide the minimum
1142        length of subject string that was needed in order to match a given pattern.
1143        (It was back references and recursion that I had previously got hung up
1144        on.) This code has now been added to pcre_study(); it finds a lower bound
1145        to the length of subject needed. It is not necessarily the greatest lower
1146        bound, but using it to avoid searching strings that are too short does give
1147        some useful speed-ups. The value is available to calling programs via
1148        pcre_fullinfo().
1150    26. While implementing 25, I discovered to my embarrassment that pcretest had
1151        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1152        study optimizations had never been tested with that matching function.
1153        Oops. What is worse, even when it was passed study data, there was a bug in
1154        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1155        were also very few tests of studied patterns with pcre_dfa_exec().
1157    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1158        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1159        on the other side of the coin, they are no longer allowed to have different
1160        names, because these cannot be distinguished in PCRE, and this has caused
1161        confusion. (This is a difference from Perl.)
1163    28. When duplicate subpattern names are present (necessarily with different
1164        numbers, as required by 27 above), and a test is made by name in a
1165        conditional pattern, either for a subpattern having been matched, or for
1166        recursion in such a pattern, all the associated numbered subpatterns are
1167        tested, and the overall condition is true if the condition is true for any
1168        one of them. This is the way Perl works, and is also more like the way
1169        testing by number works.
1172    Version 7.9 11-Apr-09
1173    ---------------------
1175    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1176        (pcretest), all targets were linked against these libraries. This included
1177        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1178        libraries. This caused unwanted dependencies to be created. This problem
1179        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1180        pcretest is linked with readline.
1182    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1183        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1184        moved outside it again, because FALSE and TRUE are already defined in AIX,
1185        but BOOL is not.
1187    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1188        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1190    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1191        hyphens) following file names and line numbers when outputting matching
1192        lines. This is not true; no spaces are inserted. I have also clarified the
1193        wording for the --colour (or --color) option.
1195    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1196        was not coloured; this is different to GNU grep, so I have changed it to be
1197        the same.
1199    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1200        each matching line was coloured. Now it goes on to look for further matches
1201        of any of the test patterns, which is the same behaviour as GNU grep.
1203    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1204        doesn't make sense to accept an empty string match in pcregrep, so I have
1205        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1206        seems to be how GNU grep behaves.
1208    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1209        start or after a newline", because the conditional assertion was not being
1210        correctly handled. The rule now is that both the assertion and what follows
1211        in the first alternative must satisfy the test.
1213    9.  If auto-callout was enabled in a pattern with a conditional group whose
1214        condition was an assertion, PCRE could crash during matching, both with
1215        pcre_exec() and pcre_dfa_exec().
1217    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1218        used for matching.
1220    11. Unicode property support in character classes was not working for
1221        characters (bytes) greater than 127 when not in UTF-8 mode.
1223    12. Added the -M command line option to pcretest.
1225    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1227    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1229    16. Added comments and documentation about mis-use of no_arg in the C++
1230        wrapper.
1232    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1233        from Martin Jerabek that uses macro names for all relevant character and
1234        string constants.
1236    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1237        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1238        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1239        these, but not everybody uses configure.
1241    19. A conditional group that had only one branch was not being correctly
1242        recognized as an item that could match an empty string. This meant that an
1243        enclosing group might also not be so recognized, causing infinite looping
1244        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1245        with the subject "ab", where knowledge that the repeated group can match
1246        nothing is needed in order to break the loop.
1248    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1249        exec(), but without supplying a callout function, matching went wrong.
1251    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1252        leak if the size of the offset vector was greater than 30. When the vector
1253        is smaller, the saved offsets during recursion go onto a local stack
1254        vector, but for larger vectors malloc() is used. It was failing to free
1255        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1256        error, in fact).
1258    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1259        heapframe that is used only when UTF-8 support is enabled. This caused no
1260        problem, but was untidy.
1262    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1263        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1264        included within another project.
1266    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1267        slightly modified by me:
1269          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1270              not building pcregrep.
1272          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1273              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1275    25. Forward references, both numeric and by name, in patterns that made use of
1276        duplicate group numbers, could behave incorrectly or give incorrect errors,
1277        because when scanning forward to find the reference group, PCRE was not
1278        taking into account the duplicate group numbers. A pattern such as
1279        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1281    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1282        the feature of a custom pointer more persuasive (as requested by a user).
1284    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1285        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1286        pre-defined.
1288    28. Added support for (*UTF8) at the start of a pattern.
1290    29. Arrange for flags added by the "release type" setting in CMake to be shown
1291        in the configuration summary.
1294    Version 7.8 05-Sep-08
1295    ---------------------
1297    1.  Replaced UCP searching code with optimized version as implemented for Ad
1298        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1299        stage table and inline lookup instead of a function, giving speed ups of 2
1300        to 5 times on some simple patterns that I tested. Permission was given to
1301        distribute the MultiStage2.py script that generates the tables (it's not in
1302        the tarball, but is in the Subversion repository).
1304    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1305        scripts.
1307    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1308        a group with a zero qualifier. The result of the study could be incorrect,
1309        or the function might crash, depending on the pattern.
1311    4.  Caseless matching was not working for non-ASCII characters in back
1312        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1313        It now works when Unicode Property Support is available.
1315    5.  In pcretest, an escape such as \x{de} in the data was always generating
1316        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1317        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1318        truncation.
1320    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1322    7.  Added two (int) casts to pcregrep when printing the difference of two
1323        pointers, in case they are 64-bit values.
1325    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1326        test 2 if it fails.
1328    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1329        and a #define of that name to empty if it is not externally set. This is to
1330        allow users of MSVC to set it if necessary.
1332    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1333        the convenience functions in the pcre_get.c source file.
1335    11. An option change at the start of a pattern that had top-level alternatives
1336        could cause overwriting and/or a crash. This command provoked a crash in
1337        some environments:
1339          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1341        This potential security problem was recorded as CVE-2008-2371.
1343    12. For a pattern where the match had to start at the beginning or immediately
1344        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1345        pcre_dfa_exec() could read past the end of the passed subject if there was
1346        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1347        pcretest so that it places the subject at the end of its malloc-ed buffer.
1349    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1350        exec() might read past the end of the data buffer in UTF-8 mode.
1352    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1353        the data contained the byte 0x85 as part of a UTF-8 character within its
1354        first line. This applied both to normal and DFA matching.
1356    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1357        /^[^d]*?$/8 failed to match "abc".
1359    16. Added a missing copyright notice to pcrecpp_internal.h.
1361    17. Make it more clear in the documentation that values returned from
1362        pcre_exec() in ovector are byte offsets, not character counts.
1364    18. Tidied a few places to stop certain compilers from issuing warnings.
1366    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1367        supplied by Stefan Weber. I made a further small update for 7.8 because
1368        there is a change of source arrangements: the pcre_searchfuncs.c module is
1369        replaced by pcre_ucd.c.
1372    Version 7.7 07-May-08
1373    ---------------------
1375    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1376        a string to a long long, pretend we don't even have a long long." This is
1377        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1379    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1380        pre-7.6 versions, which defined a global no_arg variable instead of putting
1381        it in the RE class. (See also #8 below.)
1383    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1384        Lopes.
1386    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1388        (1) The include/exclude patterns were being applied to the whole pathnames
1389            of files, instead of just to the final components.
1391        (2) If there was more than one level of directory, the subdirectories were
1392            skipped unless they satisfied the include/exclude conditions. This is
1393            inconsistent with GNU grep (and could even be seen as contrary to the
1394            pcregrep specification - which I improved to make it absolutely clear).
1395            The action now is always to scan all levels of directory, and just
1396            apply the include/exclude patterns to regular files.
1398    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1399        --exclude_dir in the tests to avoid scanning .svn directories.
1401    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1402        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1403        doesn't support NULs in patterns.
1405    7.  Added some missing "const"s to declarations of static tables in
1406        pcre_compile.c and pcre_dfa_exec.c.
1408    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1409        caused by fix #2  above. (Subsequently also a second patch to fix the
1410        first patch. And a third patch - this was a messy problem.)
1412    9.  Applied Craig's patch to remove the use of push_back().
1414    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1415        matching function regexec().
1417    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1418        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1419        references. PCRE supports relative numbers with this syntax (I don't think
1420        Oniguruma does).
1422    12. Previously, a group with a zero repeat such as (...){0} was completely
1423        omitted from the compiled regex. However, this means that if the group
1424        was called as a subroutine from elsewhere in the pattern, things went wrong
1425        (an internal error was given). Such groups are now left in the compiled
1426        pattern, with a new opcode that causes them to be skipped at execution
1427        time.
1429    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1430        to the way PCRE behaves:
1432        (a) A lone ] character is dis-allowed (Perl treats it as data).
1434        (b) A back reference to an unmatched subpattern matches an empty string
1435            (Perl fails the current match path).
1437        (c) A data ] in a character class must be notated as \] because if the
1438            first data character in a class is ], it defines an empty class. (In
1439            Perl it is not possible to have an empty class.) The empty class []
1440            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1441            The negative empty class [^] matches any one character, independently
1442            of the DOTALL setting.
1444    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1445        non-existent subpattern following a character class starting with ']' and
1446        containing () gave an internal compiling error instead of "reference to
1447        non-existent subpattern". Fortunately, when the pattern did exist, the
1448        compiled code was correct. (When scanning forwards to check for the
1449        existencd of the subpattern, it was treating the data ']' as terminating
1450        the class, so got the count wrong. When actually compiling, the reference
1451        was subsequently set up correctly.)
1453    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1454        it was being rejected as not supported by pcre_dfa_exec(), even though
1455        other assertions are supported. I have made pcre_dfa_exec() support
1456        (*FAIL).
1458    16. The implementation of 13c above involved the invention of a new opcode,
1459        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1460        cannot be changed at match time, I realized I could make a small
1461        improvement to matching performance by compiling OP_ALLANY instead of
1462        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1463        on the OP_ANY path.
1465    17. Compiling pcretest on Windows with readline support failed without the
1466        following two fixes: (1) Make the unistd.h include conditional on
1467        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1469    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1470        ncurses library to be included for pcretest when ReadLine support is
1471        requested, but also to allow for it to be overridden. This patch came from
1472        Daniel Bergström.
1474    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1475        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1476        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1477        spotting this.
1480    Version 7.6 28-Jan-08
1481    ---------------------
1483    1.  A character class containing a very large number of characters with
1484        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1485        overflow.
1487    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1488        HAVE_LONG_LONG is not defined.
1490    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1491        bring them up to date and include new features. This patch includes:
1493        - Fixed PH's badly added libz and libbz2 support.
1494        - Fixed a problem with static linking.
1495        - Added pcredemo. [But later removed - see 7 below.]
1496        - Fixed dftables problem and added an option.
1497        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1498            HAVE_LONG_LONG.
1499        - Added readline support for pcretest.
1500        - Added an listing of the option settings after cmake has run.
1502    4.  A user submitted a patch to Makefile that makes it easy to create
1503        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1504        Makefile.am that cause it to include this special target, without
1505        affecting anything else. Note that the same mingw target plus all
1506        the other distribution libraries and programs are now supported
1507        when configuring with CMake (see 6 below) instead of with
1508        Configure/Make.
1510    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1511        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1512        exported in the Windows port". It has not yet been confirmed that the patch
1513        solves the problem, but it does no harm.
1515    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1516        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1517        with CMake, and also correct the comment about stack recursion.
1519    7.  Remove the automatic building of pcredemo from the ./configure system and
1520        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1521        of a program that users should build themselves after PCRE is installed, so
1522        building it automatically is not really right. What is more, it gave
1523        trouble in some build environments.
1525    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1528    Version 7.5 10-Jan-08
1529    ---------------------
1531    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1532        values in parens when parsing an RE using the C++ wrapper."
1534    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1535        Characters greater than 255 were excluded from the class instead of being
1536        included.
1538    3.  The same bug as (2) above applied to negated POSIX classes such as
1539        [:^space:].
1541    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1542        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1543        I have changed it.
1545    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1546        first named subpattern) and a construct such as (?&a) would reference the
1547        first named subpattern whose name started with "a" (in other words, the
1548        length check was missing). Both these problems are fixed. "Subpattern name
1549        expected" is now given for (?&) (a zero-length name), and this patch also
1550        makes it give the same error for \k'' (previously it complained that that
1551        was a reference to a non-existent subpattern).
1553    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1554        this is right because (?- can be followed by option settings as well as by
1555        digits. I have, however, made the messages clearer.
1557    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1558        than the number used in the conditional) now cause a compile-time error.
1559        This is actually not compatible with Perl, which accepts such patterns, but
1560        treats the conditional as always being FALSE (as PCRE used to), but it
1561        seems to me that giving a diagnostic is better.
1563    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1564        and messages.
1566    9.  Fix two occurrences of "backslash" in comments that should have been
1567        "backspace".
1569    10. Remove two redundant lines of code that can never be obeyed (their function
1570        was moved elsewhere).
1572    11. The program that makes PCRE's Unicode character property table had a bug
1573        which caused it to generate incorrect table entries for sequences of
1574        characters that have the same character type, but are in different scripts.
1575        It amalgamated them into a single range, with the script of the first of
1576        them. In other words, some characters were in the wrong script. There were
1577        thirteen such cases, affecting characters in the following ranges:
1579          U+002b0 - U+002c1
1580          U+0060c - U+0060d
1581          U+0061e - U+00612
1582          U+0064b - U+0065e
1583          U+0074d - U+0076d
1584          U+01800 - U+01805
1585          U+01d00 - U+01d77
1586          U+01d9b - U+01dbf
1587          U+0200b - U+0200f
1588          U+030fc - U+030fe
1589          U+03260 - U+0327f
1590          U+0fb46 - U+0fbb1
1591          U+10450 - U+1049d
1593    12. The -o option (show only the matching part of a line) for pcregrep was not
1594        compatible with GNU grep in that, if there was more than one match in a
1595        line, it showed only the first of them. It now behaves in the same way as
1596        GNU grep.
1598    13. If the -o and -v options were combined for pcregrep, it printed a blank
1599        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1600        does the same. The return code can be used to tell if there were any
1601        non-matching lines.
1603    14. Added --file-offsets and --line-offsets to pcregrep.
1605    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1606        infinitely looping recursion. The bug was that positive lookaheads were not
1607        being skipped when checking for a possible empty match (negative lookaheads
1608        and both kinds of lookbehind were skipped).
1610    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1611        inclusion of <windows.h> to before rather than after the definition of
1612        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1614    17. Specifying a possessive quantifier with a specific limit for a Unicode
1615        character property caused pcre_compile() to compile bad code, which led at
1616        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1617        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1618        caused the error; without that there was no problem.
1620    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1622    19. Added --enable-pcretest-libreadline.
1624    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1625        RE::GlobalReplace(). As a result, the number of replacements returned was
1626        double what it should be. I removed one of the increments, but Craig sent a
1627        later patch that removed the other one (the right fix) and added unit tests
1628        that check the return values (which was not done before).
1630    21. Several CMake things:
1632        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1633            the names libpcre and libpcreposix, not just pcre and pcreposix.
1635        (2) The above change means that pcretest and pcregrep are now correctly
1636            linked with the newly-built libraries, not previously installed ones.
1640    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1641        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1642        UTF-8 newline character). The key issue is that the pattern starts .*;
1643        this means that the match must be either at the beginning, or after a
1644        newline. The bug was in the code for advancing after a failed match and
1645        checking that the new position followed a newline. It was not taking
1646        account of UTF-8 characters correctly.
1648    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1649        character classes. PCRE was not treating the sequence [:...:] as a
1650        character class unless the ... were all letters. Perl, however, seems to
1651        allow any characters between [: and :], though of course it rejects as
1652        unknown any "names" that contain non-letters, because all the known class
1653        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1654        for example, whereas PCRE did not - it did not recognize a POSIX character
1655        class. This seemed a bit dangerous, so the code has been changed to be
1656        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1657        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1658        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1659        Perl does, and where it didn't before.
1661    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1662        Windows environments %n is disabled by default.
1665    Version 7.4 21-Sep-07
1666    ---------------------
1668    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1669        means that a class such as [\s] counted as "explicit reference to CR or
1670        LF". That isn't really right - the whole point of the change was to try to
1671        help when there was an actual mention of one of the two characters. So now
1672        the change happens only if \r or \n (or a literal CR or LF) character is
1673        encountered.
1675    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1676        of both had grown to the point where there were only 3 bits left.
1677        Fortunately, there was spare space in the data structure, and so I have
1678        moved the internal flags into a new 16-bit field to free up more option
1679        bits.
1681    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1682        but did not set the internal JCHANGED flag - either of these is enough to
1683        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1684        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1685        start sets both bits.
1687    4.  Added options (at build time, compile time, exec time) to change \R from
1688        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1690    5.  doc/pcresyntax.html was missing from the distribution.
1692    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1693        compatibility, even though it is no longer used.
1695    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1696        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1697        windows.h file is present (where different names are used). [This was
1698        reversed later after testing - see 16 below.]
1700    8.  Changed all #include <config.h> to #include "config.h". There were also
1701        some further <pcre.h> cases that I changed to "pcre.h".
1703    9.  When pcregrep was used with the --colour option, it missed the line ending
1704        sequence off the lines that it output.
1706    10. It was pointed out to me that arrays of string pointers cause lots of
1707        relocations when a shared library is dynamically loaded. A technique of
1708        using a single long string with a table of offsets can drastically reduce
1709        these. I have refactored PCRE in four places to do this. The result is
1710        dramatic:
1712          Originally:                          290
1713          After changing UCP table:            187
1714          After changing error message table:   43
1715          After changing table of "verbs"       36
1716          After changing table of Posix names   22
1718        Thanks to the folks working on Gregex for glib for this insight.
1720    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1721        unicode-properties was also set.
1723    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1725    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1726        checked only for CRLF.
1728    14. Added casts to pcretest.c to avoid compiler warnings.
1730    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1732    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1733        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1734        entirely. This removes changes made in 7 above.
1736    17. The CMake files have been updated, and there is now more information about
1737        building with CMake in the NON-UNIX-USE document.
1740    Version 7.3 28-Aug-07
1741    ---------------------
1743     1. In the rejigging of the build system that eventually resulted in 7.1, the
1744        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1745        brackets there is not right, since it causes compilers to look for an
1746        installed pcre.h, not the version that is in the source that is being
1747        compiled (which of course may be different). I have changed it back to:
1749          #include "pcre.h"
1751        I have a vague recollection that the change was concerned with compiling in
1752        different directories, but in the new build system, that is taken care of
1753        by the VPATH setting the Makefile.
1755     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1756        when the subject happened to end in the byte 0x85 (e.g. if the last
1757        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1758        characters but of course it shouldn't be taken as a newline when it is part
1759        of another character. The bug was that, for an unlimited repeat of . in
1760        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1761        characters when looking for a newline.
1763     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1765     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1766        in debug output.
1768     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1769        long printing in the pcrecpp unittest when running under MinGW.
1771     6. ESC_K was left out of the EBCDIC table.
1773     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1774        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1775        limit also applies to "virtual nesting" when a pattern is recursive, and in
1776        this case 1000 isn't so big. I have been able to remove this limit at the
1777        expense of backing off one optimization in certain circumstances. Normally,
1778        when pcre_exec() would call its internal match() function recursively and
1779        immediately return the result unconditionally, it uses a "tail recursion"
1780        feature to save stack. However, when a subpattern that can match an empty
1781        string has an unlimited repetition quantifier, it no longer makes this
1782        optimization. That gives it a stack frame in which to save the data for
1783        checking that an empty string has been matched. Previously this was taken
1784        from the 1000-entry workspace that had been reserved. So now there is no
1785        explicit limit, but more stack is used.
1787     8. Applied Daniel's patches to solve problems with the import/export magic
1788        syntax that is required for Windows, and which was going wrong for the
1789        pcreposix and pcrecpp parts of the library. These were overlooked when this
1790        problem was solved for the main library.
1792     9. There were some crude static tests to avoid integer overflow when computing
1793        the size of patterns that contain repeated groups with explicit upper
1794        limits. As the maximum quantifier is 65535, the maximum group length was
1795        set at 30,000 so that the product of these two numbers did not overflow a
1796        32-bit integer. However, it turns out that people want to use groups that
1797        are longer than 30,000 bytes (though not repeat them that many times).
1798        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1799        made it possible to implement the integer overflow checks in a much more
1800        dynamic way, which I have now done. The artificial limitation on group
1801        length has been removed - we now have only the limit on the total length of
1802        the compiled pattern, which depends on the LINK_SIZE setting.
1804    10. Fixed a bug in the documentation for get/copy named substring when
1805        duplicate names are permitted. If none of the named substrings are set, the
1806        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1807        empty string.
1809    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1810        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1811        because the ] is interpreted as the first data character and the
1812        terminating ] is not found. PCRE has been made compatible with Perl in this
1813        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1814        cause memory overwriting.
1816    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1817        string has been matched (to stop an infinite loop). It was not recognizing
1818        a conditional subpattern that could match an empty string if that
1819        subpattern was within another subpattern. For example, it looped when
1820        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1821        condition was not nested. This bug has been fixed.
1823    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1824        past the start of the subject in the presence of bytes with the top bit
1825        set, for example "\x8aBCD".
1827    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1828        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1830    14. Optimized (?!) to (*FAIL).
1832    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1833        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1834        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1835        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1836        does: it's just the validity check that is more restrictive.
1838    16. Inserted checks for integer overflows during escape sequence (backslash)
1839        processing, and also fixed erroneous offset values for syntax errors during
1840        backslash processing.
1842    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1843        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1845    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1846        caused an overrun.
1848    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1849        something other than just ASCII characters) inside a group that had an
1850        unlimited repeat caused a loop at compile time (while checking to see
1851        whether the group could match an empty string).
1853    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1854        [\P{Any}] did so. (Error in the code for printing property names.)
1856    21. An orphan \E inside a character class could cause a crash.
1858    22. A repeated capturing bracket such as (A)? could cause a wild memory
1859        reference during compilation.
1861    23. There are several functions in pcre_compile() that scan along a compiled
1862        expression for various reasons (e.g. to see if it's fixed length for look
1863        behind). There were bugs in these functions when a repeated \p or \P was
1864        present in the pattern. These operators have additional parameters compared
1865        with \d, etc, and these were not being taken into account when moving along
1866        the compiled data. Specifically:
1868        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1869            length.
1871        (b) An item such as \pL+ within a repeated group could cause crashes or
1872            loops.
1874        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1875            "reference to non-existent subpattern" error.
1877        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1879    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1880        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1882    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1884    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1885        character were causing crashes (broken optimization).
1887    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1888        \p or \P) caused a compile-time loop.
1890    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1891        break. For example, the unstudied pattern [\r\n]A does not match the string
1892        "\r\nA" because change 7.0/46 below moves the current point on by two
1893        characters after failing to match at the start. However, the pattern \nA
1894        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1895        the same is true. There doesn't seem any very clean way out of this, but
1896        what I have chosen to do makes the common cases work: PCRE now takes note
1897        of whether there can be an explicit match for \r or \n anywhere in the
1898        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1899        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1900        pattern has explicit CR or LF references.
1902    29. Added (*CR) etc for changing newline setting at start of pattern.
1905    Version 7.2 19-Jun-07
1906    ---------------------
1908     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1909        which is apparently normally available under Windows.
1911     2. Re-jig the pcregrep tests with different newline settings in an attempt
1912        to make them independent of the local environment's newline setting.
1914     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1916     4. Some of the "internals" tests were previously cut out when the link size
1917        was not 2, because the output contained actual offsets. The recent new
1918        "Z" feature of pcretest means that these can be cut out, making the tests
1919        usable with all link sizes.
1921     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1922        stack recursion. This gives a massive performance boost under BSD, but just
1923        a small improvement under Linux. However, it saves one field in the frame
1924        in all cases.
1926     6. Added more features from the forthcoming Perl 5.10:
1928        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1929            recursion call. It refers to the nth most recently opened parentheses.
1931        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1932            to be opened parentheses.
1934        (c) Conditions that refer to capturing parentheses can be specified
1935            relatively, for example, (?(-2)... or (?(+3)...
1937        (d) \K resets the start of the current match so that everything before
1938            is not part of it.
1940        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1942        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1943            reference syntax.
1945        (g) (?| introduces a group in which the numbering of parentheses in each
1946            alternative starts with the same number.
1948        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1950     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1953     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1954        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1955        for detecting groups that can match an empty string.
1957     9. A pattern with a very large number of alternatives (more than several
1958        hundred) was running out of internal workspace during the pre-compile
1959        phase, where pcre_compile() figures out how much memory will be needed. A
1960        bit of new cunning has reduced the workspace needed for groups with
1961        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1962        workspace instead of running out of the 4096 that are available.
1964    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1966    11. Applied patch from Google to remove an optimization that didn't quite work.
1967        The report of the bug said:
1969          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1970          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1971          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1973    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1974        it matched the wrong number of bytes.
1977    Version 7.1 24-Apr-07
1978    ---------------------
1980     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1981        that is more "standard", making use of automake and other Autotools. There
1982        is some re-arrangement of the files and adjustment of comments consequent
1983        on this.
1985     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1986        for recursive directory scanning broke on some systems because the files
1987        are not scanned in any specific order and on different systems the order
1988        was different. A call to "sort" has been inserted into RunGrepTest for the
1989        approprate test as a short-term fix. In the longer term there may be an
1990        alternative.
1992     3. I had an email from Eric Raymond about problems translating some of PCRE's
1993        man pages to HTML (despite the fact that I distribute HTML pages, some
1994        people do their own conversions for various reasons). The problems
1995        concerned the use of low-level troff macros .br and .in. I have therefore
1996        removed all such uses from the man pages (some were redundant, some could
1997        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1998        HTML has been updated to handle .nf/.fi and to complain if it encounters
1999        .br or .in.
2001     4. Updated comments in configure.ac that get placed in config.h.in and also
2002        arranged for config.h to be included in the distribution, with the name
2003        config.h.generic, for the benefit of those who have to compile without
2004        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2006     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2007        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2008        makevp.bat for the current PCRE, using the additional files
2009        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2011     6. A Windows user reported a minor discrepancy with test 2, which turned out
2012        to be caused by a trailing space on an input line that had got lost in his
2013        copy. The trailing space was an accident, so I've just removed it.
2015     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2016        that is needed.
2018     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2019        as "const" (a) because they are and (b) because it helps the PHP
2020        maintainers who have recently made a script to detect big data structures
2021        in the php code that should be moved to the .rodata section. I remembered
2022        to update Builducptable as well, so it won't revert if ucptable.h is ever
2023        re-created.
2025     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2026        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2027        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2028        support is not required. This saves 1.5-2K of code, which is important in
2029        some applications.
2031        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2032        so as not to refer to the tables, even though these functions will never be
2033        called when UTF-8 support is disabled. Otherwise there are problems with a
2034        shared library.
2036    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2038        (a) It was defining its arguments as char * instead of void *.
2040        (b) It was assuming that all moves were upwards in memory; this was true
2041            a long time ago when I wrote it, but is no longer the case.
2043        The emulated memove() is provided for those environments that have neither
2044        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2045        is clearly not the case, as these two bugs were recently reported.
2047    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2048        and Detrail to create the HTML documentation, the .txt form of the man
2049        pages, and it removes trailing spaces from listed files. It also creates
2050        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2051        case, it wraps all the #defines with #ifndefs. This script should be run
2052        before "make dist".
2054    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2055        with Unicode property support.
2057        (a) For a maximizing quantifier, if the two different cases of the
2058            character were of different lengths in their UTF-8 codings (there are
2059            some cases like this - I found 11), and the matching function had to
2060            back up over a mixture of the two cases, it incorrectly assumed they
2061            were both the same length.
2063        (b) When PCRE was configured to use the heap rather than the stack for
2064            recursion during matching, it was not correctly preserving the data for
2065            the other case of a UTF-8 character when checking ahead for a match
2066            while processing a minimizing repeat. If the check also involved
2067            matching a wide character, but failed, corruption could cause an
2068            erroneous result when trying to check for a repeat of the original
2069            character.
2071    13. Some tidying changes to the testing mechanism:
2073        (a) The RunTest script now detects the internal link size and whether there
2074            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2075            values substituted by "configure". (The RunGrepTest script already did
2076            this for UTF-8.) The configure.ac script no longer substitutes the
2077            relevant variables.
2079        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2080            with length and offset values. This means that the output is different
2081            for different internal link sizes. Test 2 is skipped for link sizes
2082            other than 2 because of this, bypassing the problem. Unfortunately,
2083            there was also a test in test 3 (the locale tests) that used /B and
2084            failed for link sizes other than 2. Rather than cut the whole test out,
2085            I have added a new /Z option to pcretest that replaces the length and
2086            offset values with spaces. This is now used to make test 3 independent
2087            of link size. (Test 2 will be tidied up later.)
2089    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2090        segmentation fault instead of returning the appropriate error message.
2092    15. In multiline mode when the newline sequence was set to "any", the pattern
2093        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2094        This doesn't seem right; it now treats the CRLF combination as the line
2095        ending, and so does not match in that case. It's only a pattern such as ^$
2096        that would hit this one: something like ^ABC$ would have failed after \r
2097        and then tried again after \r\n.
2099    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2100        in an attempt to make files that differ only in their line terminators
2101        compare equal. This works on Linux.
2103    17. Under certain error circumstances pcregrep might try to free random memory
2104        as it exited. This is now fixed, thanks to valgrind.
2106    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2107        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2108        was because its rules for how to advance for /g after matching an empty
2109        string at the end of a line did not allow for this case. They now check for
2110        it specially.
2112    20. pcretest is supposed to handle patterns and data of any length, by
2113        extending its buffers when necessary. It was getting this wrong when the
2114        buffer for a data line had to be extended.
2116    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2117        CRLF as a newline sequence.
2119    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2120        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2121        I have nevertheless tidied it up.
2123    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2125    24. Added a man page for pcre-config.
2128    Version 7.0 19-Dec-06
2129    ---------------------
2131     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2132        moving to gcc 4.1.1.
2134     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2135        sys/time.h, which is documented as needed for this function. It doesn't
2136        seem to matter on Linux, but it showed up on some releases of OS X.
2138     3. It seems that there are systems where bytes whose values are greater than
2139        127 match isprint() in the "C" locale. The "C" locale should be the
2140        default when a C program starts up. In most systems, only ASCII printing
2141        characters match isprint(). This difference caused the output from pcretest
2142        to vary, making some of the tests fail. I have changed pcretest so that:
2144        (a) When it is outputting text in the compiled version of a pattern, bytes
2145            other than 32-126 are always shown as hex escapes.
2147        (b) When it is outputting text that is a matched part of a subject string,
2148            it does the same, unless a different locale has been set for the match
2149            (using the /L modifier). In this case, it uses isprint() to decide.
2151     4. Fixed a major bug that caused incorrect computation of the amount of memory
2152        required for a compiled pattern when options that changed within the
2153        pattern affected the logic of the preliminary scan that determines the
2154        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2155        that the computed length was too small. The symptoms of this bug were
2156        either the PCRE error "internal error: code overflow" from pcre_compile(),
2157        or a glibc crash with a message such as "pcretest: free(): invalid next
2158        size (fast)". Examples of patterns that provoked this bug (shown in
2159        pcretest format) are:
2161          /(?-x: )/x
2162          /(?x)(?-x: \s*#\s*)/
2163          /((?i)[\x{c0}])/8
2164          /(?i:[\x{c0}])/8
2166        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2167        is now done differently.
2169     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2170        wrapper classes; (b) implement a new function in the C++ scanner that is
2171        more efficient than the old way of doing things because it avoids levels of
2172        recursion in the regex matching; (c) add a paragraph to the documentation
2173        for the FullMatch() function.
2175     6. The escape sequence \n was being treated as whatever was defined as
2176        "newline". Not only was this contrary to the documentation, which states
2177        that \n is character 10 (hex 0A), but it also went horribly wrong when
2178        "newline" was defined as CRLF. This has been fixed.
2180     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2181        was being set to -1 for the "end of line" case (supposedly a value that no
2182        character can have). Though this value is never used (the check for end of
2183        line is "zero bytes in current character"), it caused compiler complaints.
2184        I've changed it to 0xffffffff.
2186     8. In pcre_version.c, the version string was being built by a sequence of
2187        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2188        string (as it is for production releases) called a macro with an empty
2189        argument. The C standard says the result of this is undefined. The gcc
2190        compiler treats it as an empty string (which was what was wanted) but it is
2191        reported that Visual C gives an error. The source has been hacked around to
2192        avoid this problem.
2194     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2195        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2196        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2197        of them did).
2199    10. Originally, pcretest opened its input and output without "b"; then I was
2200        told that "b" was needed in some environments, so it was added for release
2201        5.0 to both the input and output. (It makes no difference on Unix-like
2202        systems.) Later I was told that it is wrong for the input on Windows. I've
2203        now abstracted the modes into two macros, to make it easier to fiddle with
2204        them, and removed "b" from the input mode under Windows.
2206    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2208    12. Added -help and --help to pcretest as an official way of being reminded
2209        of the options.
2211    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2212        and pcrecpp.cc because they annoy compilers at high warning levels.
2214    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2216    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2217        scripts are not C programs :-) and which was not noticed because it works
2218        on Linux.
2220    16. pcretest is supposed to handle any length of pattern and data line (as one
2221        line or as a continued sequence of lines) by extending its input buffer if
2222        necessary. This feature was broken for very long pattern lines, leading to
2223        a string of junk being passed to pcre_compile() if the pattern was longer
2224        than about 50K.
2226    17. I have done a major re-factoring of the way pcre_compile() computes the
2227        amount of memory needed for a compiled pattern. Previously, there was code
2228        that made a preliminary scan of the pattern in order to do this. That was
2229        OK when PCRE was new, but as the facilities have expanded, it has become
2230        harder and harder to keep it in step with the real compile phase, and there
2231        have been a number of bugs (see for example, 4 above). I have now found a
2232        cunning way of running the real compile function in a "fake" mode that
2233        enables it to compute how much memory it would need, while actually only
2234        ever using a few hundred bytes of working memory and without too many
2235        tests of the mode. This should make future maintenance and development
2236        easier. A side effect of this work is that the limit of 200 on the nesting
2237        depth of parentheses has been removed (though this was never a serious
2238        limitation, I suspect). However, there is a downside: pcre_compile() now
2239        runs more slowly than before (30% or more, depending on the pattern). I
2240        hope this isn't a big issue. There is no effect on runtime performance.
2242    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2243        newline (only possible for the last line of a file) and it was a
2244        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2246    19. Added additional timing features to pcretest. (1) The -tm option now times
2247        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2248        separate command line item, by a number that specifies the number of
2249        repeats to use when timing. The default is 50000; this gives better
2250        precision, but takes uncomfortably long for very large patterns.
2252    20. Extended pcre_study() to be more clever in cases where a branch of a
2253        subpattern has no definite first character. For example, (a*|b*)[cd] would
2254        previously give no result from pcre_study(). Now it recognizes that the
2255        first character must be a, b, c, or d.
2257    21. There was an incorrect error "recursive call could loop indefinitely" if
2258        a subpattern (or the entire pattern) that was being tested for matching an
2259        empty string contained only one non-empty item after a nested subpattern.
2260        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2261        incorrectly, because the \d was being skipped in the check.
2263    22. The pcretest program now has a new pattern option /B and a command line
2264        option -b, which is equivalent to adding /B to every pattern. This causes
2265        it to show the compiled bytecode, without the additional information that
2266        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2267        is the same as /B/I).
2269    23. A new optimization is now able automatically to treat some sequences such
2270        as a*b as a*+b. More specifically, if something simple (such as a character
2271        or a simple class like \d) has an unlimited quantifier, and is followed by
2272        something that cannot possibly match the quantified thing, the quantifier
2273        is automatically "possessified".
2275    24. A recursive reference to a subpattern whose number was greater than 39
2276        went wrong under certain circumstances in UTF-8 mode. This bug could also
2277        have affected the operation of pcre_study().
2279    25. Realized that a little bit of performance could be had by replacing
2280        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2282    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2284    27. Possessive quantifiers such as a++ were previously implemented by turning
2285        them into atomic groups such as ($>a+). Now they have their own opcodes,
2286        which improves performance. This includes the automatically created ones
2287        from 23 above.
2289    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2290        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2291        the first matched character to be a colon. This applied both to named and
2292        numbered groups.
2294    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2296    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2297        building PCRE on Windows easier, so I have included it in the distribution.
2299    31. There is now a check in pcretest against a ridiculously large number being
2300        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2301        loop, the loop is abandoned.
2303    32. Forward references to subpatterns in conditions such as (?(2)...) where
2304        subpattern 2 is defined later cause pcre_compile() to search forwards in
2305        the pattern for the relevant set of parentheses. This search went wrong
2306        when there were unescaped parentheses in a character class, parentheses
2307        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2309    33. "Subroutine" calls and backreferences were previously restricted to
2310        referencing subpatterns earlier in the regex. This restriction has now
2311        been removed.
2313    34. Added a number of extra features that are going to be in Perl 5.10. On the
2314        whole, these are just syntactic alternatives for features that PCRE had
2315        previously implemented using the Python syntax or my own invention. The
2316        other formats are all retained for compatibility.
2318        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2319            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2320            also .NET compatible.
2322        (b) A recursion or subroutine call to a named group can now be defined as
2323            (?&name) as well as (?P>name).
2325        (c) A backreference to a named group can now be defined as \k<name> or
2326            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2327            5.10, are also .NET compatible.
2329        (d) A conditional reference to a named group can now use the syntax
2330            (?(<name>) or (?('name') as well as (?(name).
2332        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2333            groups (named and numbered) that are never evaluated inline, but can be
2334            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2335            is always false. There may be only one alternative in such a group.
2337        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2338            as the simple (?(R). The condition is true only if the most recent
2339            recursion is that of the given number or name. It does not search out
2340            through the entire recursion stack.
2342        (g) The escape \gN or \g{N} has been added, where N is a positive or
2343            negative number, specifying an absolute or relative reference.
2345    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2346        some "unreachable code" warnings.
2348    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2349        things, this adds five new scripts.
2351    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2352        There were also incompatibilities regarding the handling of \Q..\E inside
2353        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2354        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2356    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2357        matches an empty string, and forcibly breaks the loop. There were bugs in
2358        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2359        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2360        separate and independent bugs (that affected different cases) have been
2361        fixed.
2363    39. Refactored the code to abolish the use of different opcodes for small
2364        capturing bracket numbers. This is a tidy that I avoided doing when I
2365        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2366        The new approach is not only tidier, it makes it possible to reduce the
2367        memory needed to fix the previous bug (38).
2369    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2370        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2371        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2372        mode.
2374    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2375        report.
2377    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2378        copy construction and assignment in the C++ wrapper.
2380    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2381        couple of bugs that could have given wrong results in the "--newline=crlf"
2382        case.
2384    44. Added a number of casts and did some reorganization of signed/unsigned int
2385        variables following suggestions from Dair Grant. Also renamed the variable
2386        "this" as "item" because it is a C++ keyword.
2388    45. Arranged for dftables to add
2390          #include "pcre_internal.h"
2392        to pcre_chartables.c because without it, gcc 4.x may remove the array
2393        definition from the final binary if PCRE is built into a static library and
2394        dead code stripping is activated.
2396    46. For an unanchored pattern, if a match attempt fails at the start of a
2397        newline sequence, and the newline setting is CRLF or ANY, and the next two
2398        characters are CRLF, advance by two characters instead of one.
2401    Version 6.7 04-Jul-06
2402    ---------------------
2404     1. In order to handle tests when input lines are enormously long, pcretest has
2405        been re-factored so that it automatically extends its buffers when
2406        necessary. The code is crude, but this _is_ just a test program. The
2407        default size has been increased from 32K to 50K.
2409     2. The code in pcre_study() was using the value of the re argument before
2410        testing it for NULL. (Of course, in any sensible call of the function, it
2411        won't be NULL.)
2413     3. The memmove() emulation function in pcre_internal.h, which is used on
2414        systems that lack both memmove() and bcopy() - that is, hardly ever -
2415        was missing a "static" storage class specifier.
2417     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2418        containing an extended class (one that cannot be represented by a bitmap
2419        because it contains high-valued characters or Unicode property items, e.g.
2420        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2421        pattern, but PCRE should not loop if you do not (it no longer does).
2422        [Detail: two cases were found: (a) a repeated subpattern containing an
2423        extended class; (b) a recursive reference to a subpattern that followed a
2424        previous extended class. It wasn't skipping over the extended class
2425        correctly when UTF-8 mode was not set.]
2427     5. A negated single-character class was not being recognized as fixed-length
2428        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2429        compile error "lookbehind assertion is not fixed length".
2431     6. The RunPerlTest auxiliary script was showing an unexpected difference
2432        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2433        write a Perl script that can interpret lines of an input file either as
2434        byte characters or as UTF-8, which is what "perltest" was being required to
2435        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2436        can't do is switch easily at run time between having the "use utf8;" pragma
2437        or not. In the end, I fudged it by using the RunPerlTest script to insert
2438        "use utf8;" explicitly for the UTF-8 tests.
2440     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2441        the end of the subject string, contrary to the documentation and to what
2442        Perl does. This was true of both matching functions. Now it matches only at
2443        the start of the subject and immediately after *internal* newlines.
2445     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2446        a pointer to an int instead of a pointer to an unsigned long int. This
2447        caused problems on 64-bit systems.
2449     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2450        instance of the 'standard' template library not being so standard".
2452    10. There was no check on the number of named subpatterns nor the maximum
2453        length of a subpattern name. The product of these values is used to compute
2454        the size of the memory block for a compiled pattern. By supplying a very
2455        long subpattern name and a large number of named subpatterns, the size
2456        computation could be caused to overflow. This is now prevented by limiting
2457        the length of names to 32 characters, and the number of named subpatterns
2458        to 10,000.
2460    11. Subpatterns that are repeated with specific counts have to be replicated in
2461        the compiled pattern. The size of memory for this was computed from the
2462        length of the subpattern and the repeat count. The latter is limited to
2463        65535, but there was no limit on the former, meaning that integer overflow
2464        could in principle occur. The compiled length of a repeated subpattern is
2465        now limited to 30,000 bytes in order to prevent this.
2467    12. Added the optional facility to have named substrings with the same name.
2469    13. Added the ability to use a named substring as a condition, using the
2470        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2471        are numbers (not recommended). Forward references are permitted.
2473    14. Added forward references in named backreferences (if you see what I mean).
2475    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2476        pattern could run off the end of the subject. For example, the pattern
2477        "(?s)(.{1,5})"8 did this with the subject "ab".
2479    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2480        PCRE_CASELESS was set when matching characters that were quantified with ?
2481        or *.
2483    17. A character class other than a single negated character that had a minimum
2484        but no maximum quantifier - for example [ab]{6,} - was not handled
2485        correctly by pce_dfa_exec(). It would match only one character.
2487    18. A valid (though odd) pattern that looked like a POSIX character
2488        class but used an invalid character after [ (for example [[,abc,]]) caused
2489        pcre_compile() to give the error "Failed: internal error: code overflow" or
2490        in some cases to crash with a glibc free() error. This could even happen if
2491        the pattern terminated after [[ but there just happened to be a sequence of
2492        letters, a binary zero, and a closing ] in the memory that followed.
2494    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2495        over the years. Originally (before any Unicode support), just the bottom 8
2496        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2497        output from "man perlunicode" includes this:
2499          The regular expression compiler produces polymorphic opcodes.  That
2500          is, the pattern adapts to the data and automatically switches to
2501          the Unicode character scheme when presented with Unicode data--or
2502          instead uses a traditional byte scheme when presented with byte
2503          data.
2505        Sadly, a wide octal escape does not cause a switch, and in a string with
2506        no other multibyte characters, these octal escapes are treated as before.
2507        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2508        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2509        Unicode string.
2511        I have not perpetrated such confusion in PCRE. Up till now, it took just
2512        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2513        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2514        translate to the appropriate multibyte character.
2516    29. Applied some refactoring to reduce the number of warnings from Microsoft
2517        and Borland compilers. This has included removing the fudge introduced
2518        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2519        a warning about an unused variable.
2521    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2522        characters since release 4.0, because Perl (from release 5.004) does not.
2523        [Or at least, is documented not to: some releases seem to be in conflict
2524        with the documentation.] However, when a pattern was studied with
2525        pcre_study() and all its branches started with \s, PCRE still included VT
2526        as a possible starting character. Of course, this did no harm; it just
2527        caused an unnecessary match attempt.
2529    22. Removed a now-redundant internal flag bit that recorded the fact that case
2530        dependency changed within the pattern. This was once needed for "required
2531        byte" processing, but is no longer used. This recovers a now-scarce options
2532        bit. Also moved the least significant internal flag bit to the most-
2533        significant bit of the word, which was not previously used (hangover from
2534        the days when it was an int rather than a uint) to free up another bit for
2535        the future.
2537    23. Added support for CRLF line endings as well as CR and LF. As well as the
2538        default being selectable at build time, it can now be changed at runtime
2539        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2540        specify that it is scanning data with non-default line endings.
2542    24. Changed the definition of CXXLINK to make it agree with the definition of
2543        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2545    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2546        recursions. This makes a big different to stack usage for some patterns.
2548    26. If a subpattern containing a named recursion or subroutine reference such
2549        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2550        the space required for the compiled pattern went wrong and gave too small a
2551        value. Depending on the environment, this could lead to "Failed: internal
2552        error: code overflow at offset 49" or "glibc detected double free or
2553        corruption" errors.
2555    27. Applied patches from Google (a) to support the new newline modes and (b) to
2556        advance over multibyte UTF-8 characters in GlobalReplace.
2558    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2559        difference for some implementation of PCRE in some Windows version.
2561    29. Added some extra testing facilities to pcretest:
2563        \q<number>   in a data line sets the "match limit" value
2564        \Q<number>   in a data line sets the "match recursion limt" value
2565        -S <number>  sets the stack size, where <number> is in megabytes
2567        The -S option isn't available for Windows.
2570    Version 6.6 06-Feb-06
2571    ---------------------
2573     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2574        in pcreposix.h. I have copied the definition from pcre.h.
2576     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2577        because pcre.h is no longer a built file.
2579     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2580        not normally included in the compiled code.
2583    Version 6.5 01-Feb-06
2584    ---------------------
2586     1. When using the partial match feature with pcre_dfa_exec(), it was not
2587        anchoring the second and subsequent partial matches at the new starting
2588        point. This could lead to incorrect results. For example, with the pattern
2589        /1234/, partially matching against "123" and then "a4" gave a match.
2591     2. Changes to pcregrep:
2593        (a) All non-match returns from pcre_exec() were being treated as failures
2594            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2595            error message is output. Some extra information is given for the
2596            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2597            probably the only errors that are likely to be caused by users (by
2598            specifying a regex that has nested indefinite repeats, for instance).
2599            If there are more than 20 of these errors, pcregrep is abandoned.
2601        (b) A binary zero was treated as data while matching, but terminated the
2602            output line if it was written out. This has been fixed: binary zeroes
2603            are now no different to any other data bytes.
2605        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2606            used to set a locale for matching. The --locale=xxxx long option has
2607            been added (no short equivalent) to specify a locale explicitly on the
2608            pcregrep command, overriding the environment variables.
2610        (d) When -B was used with -n, some line numbers in the output were one less
2611            than they should have been.
2613        (e) Added the -o (--only-matching) option.
2615        (f) If -A or -C was used with -c (count only), some lines of context were
2616            accidentally printed for the final match.
2618        (g) Added the -H (--with-filename) option.
2620        (h) The combination of options -rh failed to suppress file names for files
2621            that were found from directory arguments.
2623        (i) Added the -D (--devices) and -d (--directories) options.
2625        (j) Added the -F (--fixed-strings) option.
2627        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2629        (l) Added the --colo(u)r option.
2631        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2632            is not present by default.
2634     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2635        items such as (?R) or (?1), when the recursion could match a number of
2636        alternatives. If it matched one of the alternatives, but subsequently,
2637        outside the recursion, there was a failure, the code tried to back up into
2638        the recursion. However, because of the way PCRE is implemented, this is not
2639        possible, and the result was an incorrect result from the match.
2641        In order to prevent this happening, the specification of recursion has
2642        been changed so that all such subpatterns are automatically treated as
2643        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2645     4. I had overlooked the fact that, in some locales, there are characters for
2646        which isalpha() is true but neither isupper() nor islower() are true. In
2647        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2648        and ordfeminine) are like this. This affected the treatment of \w and \W
2649        when they appeared in character classes, but not when they appeared outside
2650        a character class. The bit map for "word" characters is now created
2651        separately from the results of isalnum() instead of just taking it from the
2652        upper, lower, and digit maps. (Plus the underscore character, of course.)
2654     5. The above bug also affected the handling of POSIX character classes such as
2655        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2656        permanent tables. Instead, the bit maps for such a class were previously
2657        created as the appropriate unions of the upper, lower, and digit bitmaps.
2658        Now they are created by subtraction from the [[:word:]] class, which has
2659        its own bitmap.
2661     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2662        It is created by subtracting the vertical space characters (\x09, \x0a,
2663        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2664        subtraction was done in the overall bitmap for a character class, meaning
2665        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2666        be recognized. This bug has been fixed.
2668     7. Patches from the folks at Google:
2670          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2671          real life, but is still worth protecting against".
2673          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2674          regular expressions".
2676          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2677          have it.
2679          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2680          "configure" and the latter not, in order to fix a problem somebody had
2681          with compiling the Arg class on HP-UX.
2683          (e) Improve the error-handling of the C++ wrapper a little bit.
2685          (f) New tests for checking recursion limiting.
2687     8. The pcre_memmove() function, which is used only if the environment does not
2688        have a standard memmove() function (and is therefore rarely compiled),
2689        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2690        returning a result (though PCRE never actually uses the result).
2692     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2693        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2694        returned instead of calling malloc() with an overflowing number that would
2695        most likely cause subsequent chaos.
2697    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2699    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2700        with this option is matched, the nmatch and pmatch options of regexec() are
2701        ignored.
2703    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2704        provided in case anyone wants to the the POSIX interface with UTF-8
2705        strings.
2707    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2708        C++ linking (needed for some HP-UX environments).
2710    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2711        (unused parameter) and in the pcre_printint() function (omitted "default"
2712        switch label when the default is to do nothing).
2714    15. Added some code to make it possible, when PCRE is compiled as a C++
2715        library, to replace subject pointers for pcre_exec() with a smart pointer
2716        class, thus making it possible to process discontinuous strings.
2718    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2719        much the same function. They were added by different people who were trying
2720        to make PCRE easy to compile on non-Unix systems. It has been suggested
2721        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2722        for compiling on Windows systems. I have therefore replaced it with
2723        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2724        defaults to "extern" for C or "extern C" for C++, which works fine on
2725        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2726        SCOPE with something explicit in config.h. In addition:
2728        (a) pcreposix.h still had just "extern" instead of either of these macros;
2729            I have replaced it with PCRE_DATA_SCOPE.
2731        (b) Functions such as _pcre_xclass(), which are internal to the library,
2732            but external in the C sense, all had PCRE_EXPORT in their definitions.
2733            This is apparently wrong for the Windows case, so I have removed it.
2734            (It makes no difference on Unix-like systems.)
2736    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2737        of recursive calls to match(). This is different to MATCH_LIMIT because
2738        that limits the total number of calls to match(), not all of which increase
2739        the depth of recursion. Limiting the recursion depth limits the amount of
2740        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2741        when PCRE is compiled, and changed at run time. A patch from Google adds
2742        this functionality to the C++ interface.
2744    18. Changes to the handling of Unicode character properties:
2746        (a) Updated the table to Unicode 4.1.0.
2748        (b) Recognize characters that are not in the table as "Cn" (undefined).
2750        (c) I revised the way the table is implemented to a much improved format
2751            which includes recognition of ranges. It now supports the ranges that
2752            are defined in UnicodeData.txt, and it also amalgamates other
2753            characters into ranges. This has reduced the number of entries in the
2754            table from around 16,000 to around 3,000, thus reducing its size
2755            considerably. I realized I did not need to use a tree structure after
2756            all - a binary chop search is just as efficient. Having reduced the
2757            number of entries, I extended their size from 6 bytes to 8 bytes to
2758            allow for more data.
2760        (d) Added support for Unicode script names via properties such as \p{Han}.
2762    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2763        matching that character.
2765    20. When matching a repeated Unicode property with a minimum greater than zero,
2766        (for example \pL{2,}), PCRE could look past the end of the subject if it
2767        reached it while seeking the minimum number of characters. This could
2768        happen only if some of the characters were more than one byte long, because
2769        there is a check for at least the minimum number of bytes.
2771    21. Refactored the implementation of \p and \P so as to be more general, to
2772        allow for more different types of property in future. This has changed the
2773        compiled form incompatibly. Anybody with saved compiled patterns that use
2774        \p or \P will have to recompile them.
2776    22. Added "Any" and "L&" to the supported property types.
2778    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2779        but give a compile time error if the value is greater than 0xff.
2781    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2782        accidentally not being installed or uninstalled.
2784    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2785        made were to insert the current release number. This seemed silly, because
2786        it made things harder for people building PCRE on systems that don't run
2787        "configure". I have turned pcre.h into a distributed file, no longer built
2788        by "configure", with the version identification directly included. There is
2789        no longer a pcre.h.in file.
2791        However, this change necessitated a change to the pcre-config script as
2792        well. It is built from pcre-config.in, and one of the substitutions was the
2793        release number. I have updated configure.ac so that ./configure now finds
2794        the release number by grepping pcre.h.
2796    26. Added the ability to run the tests under valgrind.
2799    Version 6.4 05-Sep-05
2800    ---------------------
2802     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2803        "--" to be printed when multiple files were scanned, even when none of the
2804        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2805        consider it to be a bug, and have restored the previous behaviour.
2807     2. A couple of code tidies to get rid of compiler warnings.
2809     3. The pcretest program used to cheat by referring to symbols in the library
2810        whose names begin with _pcre_. These are internal symbols that are not
2811        really supposed to be visible externally, and in some environments it is
2812        possible to suppress them. The cheating is now confined to including
2813        certain files from the library's source, which is a bit cleaner.
2815     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2816        file's purpose clearer.
2818     5. Reorganized pcre_ucp_findchar().
2821    Version 6.3 15-Aug-05
2822    ---------------------
2824     1. The file libpcre.pc.in did not have general read permission in the tarball.
2826     2. There were some problems when building without C++ support:
2828        (a) If C++ support was not built, "make install" and "make test" still
2829            tried to test it.
2831        (b) There were problems when the value of CXX was explicitly set. Some
2832            changes have been made to try to fix these, and ...
2834        (c) --disable-cpp can now be used to explicitly disable C++ support.
2836        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2837            backslash in a target when C++ was disabled. This confuses some
2838            versions of "make", apparently. Using an intermediate variable solves
2839            this. (Same for CPP_LOBJ.)
2841     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2842        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2843        necessary on certain architectures.
2845     4. Added a setting of -export-symbols-regex to the link command to remove
2846        those symbols that are exported in the C sense, but actually are local
2847        within the library, and not documented. Their names all begin with
2848        "_pcre_". This is not a perfect job, because (a) we have to except some
2849        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2850        available (and never for static libraries). I have made a note to try to
2851        find a way round (a) in the future.
2854    Version 6.2 01-Aug-05
2855    ---------------------
2857     1. There was no test for integer overflow of quantifier values. A construction
2858        such as {1111111111111111} would give undefined results. What is worse, if
2859        a minimum quantifier for a parenthesized subpattern overflowed and became
2860        negative, the calculation of the memory size went wrong. This could have
2861        led to memory overwriting.
2863     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2865     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2866        operating environments where this matters.
2868     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2869        PCRE options from within the C++ wrapper.
2871     5. Named capturing subpatterns were not being correctly counted when a pattern
2872        was compiled. This caused two problems: (a) If there were more than 100
2873        such subpatterns, the calculation of the memory needed for the whole
2874        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2875        back references of the form \12, where the number was greater than 9, were
2876        not recognized as back references, even though there were sufficient
2877        previous subpatterns.
2879     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2880        versions of gcc, e.g. 2.95.4.
2883    Version 6.1 21-Jun-05
2884    ---------------------
2886     1. There was one reference to the variable "posix" in pcretest.c that was not
2887        surrounded by "#if !defined NOPOSIX".
2889     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2890        the cross-check on the old pcre_info() function, for the benefit of the
2891        cut-down version of PCRE that is currently imported into Exim.
2893     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2894        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2895        patterns that start (?i) but I don't think that matters. The use of (?i) is
2896        just an example; this all applies to the other options as well.
2898     4. Since libtool seems to echo the compile commands it is issuing, the output
2899        from "make" can be reduced a bit by putting "@" in front of each libtool
2900        compile command.
2902     5. Patch from the folks at Google for configure.in to be a bit more thorough
2903        in checking for a suitable C++ installation before trying to compile the
2904        C++ stuff. This should fix a reported problem when a compiler was present,
2905        but no suitable headers.
2907     6. The man pages all had just "PCRE" as their title. I have changed them to
2908        be the relevant file name. I have also arranged that these names are
2909        retained in the file doc/pcre.txt, which is a concatenation in text format
2910        of all the man pages except the little individual ones for each function.
2912     7. The NON-UNIX-USE file had not been updated for the different set of source
2913        files that come with release 6. I also added a few comments about the C++
2914        wrapper.
2917    Version 6.0 07-Jun-05
2918    ---------------------
2920     1. Some minor internal re-organization to help with my DFA experiments.
2922     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2923        didn't matter for the library itself when fully configured, but did matter
2924        when compiling without UCP support, or within Exim, where the ucp files are
2925        not imported.
2927     3. Refactoring of the library code to split up the various functions into
2928        different source modules. The addition of the new DFA matching code (see
2929        below) to a single monolithic source would have made it really too
2930        unwieldy, quite apart from causing all the code to be include in a
2931        statically linked application, when only some functions are used. This is
2932        relevant even without the DFA addition now that patterns can be compiled in
2933        one application and matched in another.
2935        The downside of splitting up is that there have to be some external
2936        functions and data tables that are used internally in different modules of
2937        the library but which are not part of the API. These have all had their
2938        names changed to start with "_pcre_" so that they are unlikely to clash
2939        with other external names.
2941     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2942        a different (DFA) algorithm. Although it is slower than the original
2943        function, it does have some advantages for certain types of matching
2944        problem.
2946     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2947        including restarting after a partial match.
2949     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2950        defined when compiling for Windows was sent to me. I have put it into the
2951        code, though I have no means of testing or verifying it.
2953     7. Added the pcre_refcount() auxiliary function.
2955     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2956        match before or at the first newline in the subject string. In pcretest,
2957        the /f option on a pattern can be used to set this.
2959     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2960        would behave wrongly. This has been present in PCRE since release 4.0.
2962    10. A number of changes to the pcregrep command:
2964        (a) Refactored how -x works; insert ^(...)$ instead of setting
2965            PCRE_ANCHORED and checking the length, in preparation for adding
2966            something similar for -w.
2968        (b) Added the -w (match as a word) option.
2970        (c) Refactored the way lines are read and buffered so as to have more
2971            than one at a time available.
2973        (d) Implemented a pcregrep test script.
2975        (e) Added the -M (multiline match) option. This allows patterns to match
2976            over several lines of the subject. The buffering ensures that at least
2977            8K, or the rest of the document (whichever is the shorter) is available
2978            for matching (and similarly the previous 8K for lookbehind assertions).
2980        (f) Changed the --help output so that it now says
2982              -w, --word-regex(p)
2984            instead of two lines, one with "regex" and the other with "regexp"
2985            because that confused at least one person since the short forms are the
2986            same. (This required a bit of code, as the output is generated
2987            automatically from a table. It wasn't just a text change.)
2989        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2990            option but starts with a hyphen. Could be a pattern or a path name
2991            starting with a hyphen, for instance.
2993        (h) "-" can be given as a file name to represent stdin.
2995        (i) When file names are being printed, "(standard input)" is used for
2996            the standard input, for compatibility with GNU grep. Previously
2997            "<stdin>" was used.
2999        (j) The option --label=xxx can be used to supply a name to be used for
3000            stdin when file names are being printed. There is no short form.
3002        (k) Re-factored the options decoding logic because we are going to add
3003            two more options that take data. Such options can now be given in four
3004            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3006        (l) Added the -A, -B, and -C options for requesting that lines of context
3007            around matches be printed.
3009        (m) Added the -L option to print the names of files that do not contain
3010            any matching lines, that is, the complement of -l.
3012        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3013            continue to scan other files.
3015        (o) The -s option was incorrectly implemented. For compatibility with other
3016            greps, it now suppresses the error message for a non-existent or non-
3017            accessible file (but not the return code). There is a new option called
3018            -q that suppresses the output of matching lines, which was what -s was
3019            previously doing.
3021        (p) Added --include and --exclude options to specify files for inclusion
3022            and exclusion when recursing.
3024    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3025        Hopefully, it now does.
3027    12. Missing cast in pcre_study().
3029    13. Added an "uninstall" target to the makefile.
3031    14. Replaced "extern" in the function prototypes in Makefile.in with
3032        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3033        world, but is set differently for Windows.
3035    15. Added a second compiling function called pcre_compile2(). The only
3036        difference is that it has an extra argument, which is a pointer to an
3037        integer error code. When there is a compile-time failure, this is set
3038        non-zero, in addition to the error test pointer being set to point to an
3039        error message. The new argument may be NULL if no error number is required
3040        (but then you may as well call pcre_compile(), which is now just a
3041        wrapper). This facility is provided because some applications need a
3042        numeric error indication, but it has also enabled me to tidy up the way
3043        compile-time errors are handled in the POSIX wrapper.
3045    16. Added VPATH=.libs to the makefile; this should help when building with one
3046        prefix path and installing with another. (Or so I'm told by someone who
3047        knows more about this stuff than I do.)
3049    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3050        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3051        match everything, including newlines. This is not POSIX-compatible, but
3052        somebody wanted the feature. From pcretest it can be activated by using
3053        both the P and the s flags.
3055    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3057    19. libpcre.pc was being incorrectly installed as executable.
3059    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3060        it now also looks for '\r' so that it will work unmodified on Windows.
3062    21. Added Google's contributed C++ wrapper to the distribution.
3064    22. Added some untidy missing memory free() calls in pcretest, to keep
3065        Electric Fence happy when testing.
3069    Version 5.0 13-Sep-04
3070    ---------------------
3072     1. Internal change: literal characters are no longer packed up into items
3073        containing multiple characters in a single byte-string. Each character
3074        is now matched using a separate opcode. However, there may be more than one
3075        byte in the character in UTF-8 mode.
3077     2. The pcre_callout_block structure has two new fields: pattern_position and
3078        next_item_length. These contain the offset in the pattern to the next match
3079        item, and its length, respectively.
3081     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3082        insertion of callouts before each pattern item. Added the /C option to
3083        pcretest to make use of this.
3085     4. On the advice of a Windows user, the lines
3087          #if defined(_WIN32) || defined(WIN32)
3088          _setmode( _fileno( stdout ), 0x8000 );
3089          #endif  /* defined(_WIN32) || defined(WIN32) */
3091        have been added to the source of pcretest. This apparently does useful
3092        magic in relation to line terminators.
3094     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3095        for the benefit of those environments where the "b" makes a difference.
3097     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3098        to know about it. I have put a hack into configure.in that adds in code
3099        to set GCC=yes if CC=icc. This seems to end up at a point in the
3100        generated configure script that is early enough to affect the setting of
3101        compiler options, which is what is needed, but I have no means of testing
3102        whether it really works. (The user who reported this had patched the
3103        generated configure script, which of course I cannot do.)
3105        LATER: After change 22 below (new libtool files), the configure script
3106        seems to know about icc (and also ecc). Therefore, I have commented out
3107        this hack in configure.in.
3109     7. Added support for pkg-config (2 patches were sent in).
3111     8. Negated POSIX character classes that used a combination of internal tables
3112        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3113        [[:^ascii]]. Typically, they would match almost any characters. The other
3114        POSIX classes were not broken in this way.
3116     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3117        to find the match, as PCRE was deluded into thinking that the match had to
3118        start at the start point or following a newline. The same bug applied to
3119        patterns with negative forward assertions or any backward assertions
3120        preceding ".*" at the start, unless the pattern required a fixed first
3121        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3123    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3124        starting at the last subject character, bytes beyond the end of the subject
3125        string were read.
3127    11. Renamed the variable "class" as "classbits" to make life easier for C++
3128        users. (Previously there was a macro definition, but it apparently wasn't
3129        enough.)
3131    12. Added the new field "tables" to the extra data so that tables can be passed
3132        in at exec time, or the internal tables can be re-selected. This allows
3133        a compiled regex to be saved and re-used at a later time by a different
3134        program that might have everything at different addresses.
3136    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3137        -R library as well as a -L library.
3139    14. The debugging options of pcretest (-d on the command line or D on a
3140        pattern) showed incorrect output for anything following an extended class
3141        that contained multibyte characters and which was followed by a quantifier.
3143    15. Added optional support for general category Unicode character properties
3144        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3145        support. It adds about 90K to the size of the library. The meanings of the
3146        inbuilt class escapes such as \d and \s have NOT been changed.
3148    16. Updated pcredemo.c to include calls to free() to release the memory for the
3149        compiled pattern.
3151    17. The generated file chartables.c was being created in the source directory
3152        instead of in the building directory. This caused the build to fail if the
3153        source directory was different from the building directory, and was
3154        read-only.
3156    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3157        file. No doubt somebody will tell me if they don't make sense... Also added
3158        Dan Mooney's comments about building on OpenVMS.
3160    19. Added support for partial matching via the PCRE_PARTIAL option for
3161        pcre_exec() and the \P data escape in pcretest.
3163    20. Extended pcretest with 3 new pattern features:
3165        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3166              write the compiled pattern to the file whose name is "rest-of-line".
3167              This is a straight binary dump of the data, with the saved pointer to
3168              the character tables forced to be NULL. The study data, if any, is
3169              written too. After writing, pcretest reads a new pattern.
3171        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3172              compiled pattern from the given file. There must not be any
3173              occurrences of "<" in the file name (pretty unlikely); if there are,
3174              pcretest will instead treat the initial "<" as a pattern delimiter.
3175              After reading in the pattern, pcretest goes on to read data lines as
3176              usual.
3178        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3179              and 16-bit fields in a compiled pattern, to simulate a pattern that
3180              was compiled on a host of opposite endianness.
3182    21. The pcre-exec() function can now cope with patterns that were compiled on
3183        hosts of opposite endianness, with this restriction:
3185          As for any compiled expression that is saved and used later, the tables
3186          pointer field cannot be preserved; the extra_data field in the arguments
3187          to pcre_exec() should be used to pass in a tables address if a value
3188          other than the default internal tables were used at compile time.
3190    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3191        now diagnosed as an error. Previously, most of the time, a negative number
3192        would have been treated as zero, but if in addition "ovector" was passed as
3193        NULL, a crash could occur.
3195    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3196        new versions from the libtool 1.5 distribution (the last one is a copy of
3197        a file called libtool.m4). This seems to have fixed the need to patch
3198        "configure" to support Darwin 1.3 (which I used to do). However, I still
3199        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3200        workstation).
3202    24. Changed the PCRE licence to be the more standard "BSD" licence.
3205    Version 4.5 01-Dec-03
3206    ---------------------
3208     1. There has been some re-arrangement of the code for the match() function so
3209        that it can be compiled in a version that does not call itself recursively.
3210        Instead, it keeps those local variables that need separate instances for
3211        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3212        needs to "recurse". Keeping track of where control must go is done by means
3213        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3214        hide most of the details from the main code, and operates only if
3215        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3216        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3217        operating.
3219        To make it easier for callers to provide specially tailored get/free
3220        functions for this usage, two new functions, pcre_stack_malloc, and
3221        pcre_stack_free, are used. They are always called in strict stacking order,
3222        and the size of block requested is always the same.
3224        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3225        PCRE has been compiled to use the stack or the heap for recursion. The
3226        -C option of pcretest uses this to show which version is compiled.
3228        A new data escape \S, is added to pcretest; it causes the amounts of store
3229        obtained and freed by both kinds of malloc/free at match time to be added
3230        to the output.
3232     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3233        what's available on my current Linux desktop machine.
3235     3. When matching a UTF-8 string, the test for a valid string at the start has
3236        been extended. If start_offset is not zero, PCRE now checks that it points
3237        to a byte that is the start of a UTF-8 character. If not, it returns
3238        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3239        this is necessary because there may be backward assertions in the pattern.
3240        When matching the same subject several times, it may save resources to use
3241        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3243     4. The code for checking the validity of UTF-8 strings has been tightened so
3244        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3245        containing "overlong sequences".
3247     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3248        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3249        should have been "&", but it just so happened that all the cases this let
3250        through by mistake were picked up later in the function.
3252     6. I had used a variable called "isblank" - this is a C99 function, causing
3253        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3255     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3256        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3257        so the version is shown; (c) stop "make test" echoing "./RunTest".
3259     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3261     9. The prototype for memmove() for systems that don't have it was using
3262        size_t, but the inclusion of the header that defines size_t was later. I've
3263        moved the #includes for the C headers earlier to avoid this.
3265    10. Added some adjustments to the code to make it easier to compiler on certain
3266        special systems:
3268          (a) Some "const" qualifiers were missing.
3269          (b) Added the macro EXPORT before all exported functions; by default this
3270              is defined to be empty.
3271          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3272              that it reads its output file name as an argument instead of writing
3273              to the standard output and assuming this can be redirected.
3275    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3276        class containing characters with values greater than 255, PCRE compilation
3277        went into a loop.
3279    12. A recursive reference to a subpattern that was within another subpattern
3280        that had a minimum quantifier of zero caused PCRE to crash. For example,
3281        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3282        recursion. If the recursively-called subpattern itself had a zero repeat,
3283        that was OK.
3285    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3286        buffer into which it was copied (for escape processing) was still set at
3287        1024, so long lines caused crashes.
3289    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3290        "internal error: code overflow...". This applied to any character class
3291        that was followed by a possessive quantifier.
3293    15. Modified the Makefile to add libpcre.la as a prerequisite for
3294        libpcreposix.la because I was told this is needed for a parallel build to
3295        work.
3297    16. If a pattern that contained .* following optional items at the start was
3298        studied, the wrong optimizing data was generated, leading to matching
3299        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3300        matching string must start with a or b or c. The correct conclusion for
3301        this pattern is that a match can start with any character.
3304    Version 4.4 13-Aug-03
3305    ---------------------
3307     1. In UTF-8 mode, a character class containing characters with values between
3308        127 and 255 was not handled correctly if the compiled pattern was studied.
3309        In fixing this, I have also improved the studying algorithm for such
3310        classes (slightly).
3312     2. Three internal functions had redundant arguments passed to them. Removal
3313        might give a very teeny performance improvement.
3315     3. Documentation bug: the value of the capture_top field in a callout is *one
3316        more than* the number of the hightest numbered captured substring.
3318     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3319        in incorrectly linking with a previously installed version. They now link
3320        explicitly with libpcre.la.
3322     5. configure.in no longer needs to recognize Cygwin specially.
3324     6. A problem in pcre.in for Windows platforms is fixed.
3326     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3327        pcretest, it used to include the size of the study block as part of its
3328        output. Unfortunately, the structure contains a field that has a different
3329        size on different hardware architectures. This meant that the tests that
3330        showed this size failed. As the block is currently always of a fixed size,
3331        this information isn't actually particularly useful in pcretest output, so
3332        I have just removed it.
3334     8. Three pre-processor statements accidentally did not start in column 1.
3335        Sadly, there are *still* compilers around that complain, even though
3336        standard C has not required this for well over a decade. Sigh.
3338     9. In pcretest, the code for checking callouts passed small integers in the
3339        callout_data field, which is a void * field. However, some picky compilers
3340        complained about the casts involved for this on 64-bit systems. Now
3341        pcretest passes the address of the small integer instead, which should get
3342        rid of the warnings.
3344    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3345        both compile and run time, and gives an error if an invalid UTF-8 sequence
3346        is found. There is a option for disabling this check in cases where the
3347        string is known to be correct and/or the maximum performance is wanted.
3349    11. In response to a bug report, I changed one line in Makefile.in from
3351            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3352        to
3353            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3355        to look similar to other lines, but I have no way of telling whether this
3356        is the right thing to do, as I do not use Windows. No doubt I'll get told
3357        if it's wrong...
3360    Version 4.3 21-May-03
3361    ---------------------
3363    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3364       Makefile.
3366    2. Some refactoring to improve the quality of the code:
3368       (i)   The utf8_table... variables are now declared "const".
3370       (ii)  The code for \cx, which used the "case flipping" table to upper case
3371             lower case letters, now just substracts 32. This is ASCII-specific,
3372             but the whole concept of \cx is ASCII-specific, so it seems
3373             reasonable.
3375       (iii) PCRE was using its character types table to recognize decimal and
3376             hexadecimal digits in the pattern. This is silly, because it handles
3377             only 0-9, a-f, and A-F, but the character types table is locale-
3378             specific, which means strange things might happen. A private
3379             table is now used for this - though it costs 256 bytes, a table is
3380             much faster than multiple explicit tests. Of course, the standard
3381             character types table is still used for matching digits in subject
3382             strings against \d.
3384       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3385             ending in _t are). So I've renamed it as ESC_tee.
3387    3. The first argument for regexec() in the POSIX wrapper should have been
3388       defined as "const".
3390    4. Changed pcretest to use malloc() for its buffers so that they can be
3391       Electric Fenced for debugging.
3393    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3394       to read one or more bytes before the start of the subject string. Often this
3395       had no effect on PCRE's behaviour, but in some circumstances it could
3396       provoke a segmentation fault.
3398    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3399       to try to read one or more bytes before the start of the subject string.
3401    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3402       UTF-8 support could misbehave in various ways if the subject string
3403       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3404       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3405       back over UTF-8 characters.)
3408    Version 4.2 14-Apr-03
3409    ---------------------
3411    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3413    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3414         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3415         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3416         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3417         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3418           and BUILD_EXEEXT
3419         Note: automatic setting of the BUILD variables is not yet working
3420         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3421           compile-time but not at link-time
3422         [LINK]: use for linking executables only
3423         make different versions for Windows and non-Windows
3424         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3425           libraries
3426         [LINK_FOR_BUILD]: new variable
3427         [OBJEXT]: use throughout
3428         [EXEEXT]: use throughout
3429         <winshared>: new target
3430         <wininstall>: new target
3431         <dftables.o>: use native compiler
3432         <dftables>: use native linker
3433         <install>: handle Windows platform correctly
3434         <clean>: ditto
3435         <check>: ditto
3436         copy DLL to top builddir before testing
3438       As part of these changes, -no-undefined was removed again. This was reported
3439       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3440       in any case.
3442    3. Some tidies to get rid of compiler warnings:
3444       . In the match_data structure, match_limit was an unsigned long int, whereas
3445         match_call_count was an int. I've made them both unsigned long ints.
3447       . In pcretest the fact that a const uschar * doesn't automatically cast to
3448         a void * provoked a warning.
3450       . Turning on some more compiler warnings threw up some "shadow" variables
3451         and a few more missing casts.
3453    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3454       option, a class that contained a single character with a value between 128
3455       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3457    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3458       option, a class that contained several characters, but with at least one
3459       whose value was between 128 and 255 caused PCRE to crash.
3462    Version 4.1 12-Mar-03
3463    ---------------------
3465    1. Compiling with gcc -pedantic found a couple of places where casts were
3466    needed, and a string in dftables.c that was longer than standard compilers are
3467    required to support.
3469    2. Compiling with Sun's compiler found a few more places where the code could
3470    be tidied up in order to avoid warnings.
3472    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3473    first of these names is deprecated in the latest Autoconf in favour of the name
3474    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3475    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3476    analogy I have changed it to CFLAGS_FOR_BUILD.
3478    4. Added -no-undefined to the linking command in the Makefile, because this is
3479    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3480    linking step for the pcreposix library.
3482    5. PCRE was failing to diagnose the case of two named groups with the same
3483    name.
3485    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3486    literal character that is needed in the subject for a match, and scans along to
3487    ensure that it is present before embarking on the full matching process. This
3488    saves time in cases of nested unlimited repeats that are never going to match.
3489    Problem: the scan can take a lot of time if the subject is very long (e.g.
3490    megabytes), thus penalizing straightforward matches. It is now done only if the
3491    amount of subject to be scanned is less than 1000 bytes.
3493    7. A lesser problem with the same optimization is that it was recording the
3494    first character of an anchored pattern as "needed", thus provoking a search
3495    right along the subject, even when the first match of the pattern was going to
3496    fail. The "needed" character is now not set for anchored patterns, unless it
3497    follows something in the pattern that is of non-fixed length. Thus, it still
3498    fulfils its original purpose of finding quick non-matches in cases of nested
3499    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3502    Version 4.0 17-Feb-03
3503    ---------------------
3505    1. If a comment in an extended regex that started immediately after a meta-item
3506    extended to the end of string, PCRE compiled incorrect data. This could lead to
3507    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3509    2. Moved to autoconf 2.53 and libtool 1.4.2.
3511    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3512    the special perltest8 script is no longer needed - all the tests can be run
3513    from a single perltest script.
3515    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3516    by \s. It has now been removed in PCRE. This means it isn't recognized as
3517    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3518    class [:space:] *does* include VT, thereby creating a mess.
3520    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3521    space and tab.
3523    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3524    its new features into the main test script, reducing the number of scripts.
3526    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3527    were backward compatible, and made the (?i) apply to the whole pattern, as if
3528    /i were given. Now it behaves more logically, and applies the option setting
3529    only to what follows. PCRE has been changed to follow suit. However, if it
3530    finds options settings right at the start of the pattern, it extracts them into
3531    the global options, as before. Thus, they show up in the info data.
3533    8. Added support for the \Q...\E escape sequence. Characters in between are
3534    treated as literals. This is slightly different from Perl in that $ and @ are
3535    also handled as literals inside the quotes. In Perl, they will cause variable
3536    interpolation. Note the following examples:
3538        Pattern            PCRE matches      Perl matches
3540        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3541        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3542        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3544    For compatibility with Perl, \Q...\E sequences are recognized inside character
3545    classes as well as outside them.
3547    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3548    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3549    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3550    signed/unsigned warnings.
3552    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3553    option for pcretest, so I've replaced it by a simple function that does just
3554    that job.
3556    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3557    "pcregrep -".
3559    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3560    Java package. This provides some syntactic sugar for simple cases of what my
3561    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3562    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3563    item, you can use this simplified notation. Note that only makes sense with
3564    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3565    greediness, whatever the setting of the PCRE_UNGREEDY option.
3567    13. A change of greediness default within a pattern was not taking effect at
3568    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3569    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3570    was abstracted outside.
3572    14. PCRE now supports the \G assertion. It is true when the current matching
3573    position is at the start point of the match. This differs from \A when the
3574    starting offset is non-zero. Used with the /g option of pcretest (or similar
3575    code), it works in the same way as it does for Perl's /g option. If all
3576    alternatives of a regex begin with \G, the expression is anchored to the start
3577    match position, and the "anchored" flag is set in the compiled expression.
3579    15. Some bugs concerning the handling of certain option changes within patterns
3580    have been fixed. These applied to options other than (?ims). For example,
3581    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3582    been the other way round. Some of this was related to change 7 above.
3584    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3585    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3586    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3587    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3589    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3590    mode. Unlike ".", it always matches newline, whatever the setting of
3591    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3592    assertions. Perl allows it, but it doesn't (in general) work because it can't
3593    calculate the length of the lookbehind. At least, that's the case for Perl
3594    5.8.0 - I've been told they are going to document that it doesn't work in
3595    future.
3597    18. Added an error diagnosis for escapes that PCRE does not support: these are
3598    \L, \l, \N, \P, \p, \U, \u, and \X.
3600    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3601    reading past the end of the pattern in cases such as /[abcd/.
3603    20. PCRE was getting more memory than necessary for patterns with classes that
3604    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3606    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3607    compiling PCRE for use with Virtual Pascal.
3609    22. Small fix to the Makefile to make it work properly if the build is done
3610    outside the source tree.
3612    23. Added a new extension: a condition to go with recursion. If a conditional
3613    subpattern starts with (?(R) the "true" branch is used if recursion has
3614    happened, whereas the "false" branch is used only at the top level.
3616    24. When there was a very long string of literal characters (over 255 bytes
3617    without UTF support, over 250 bytes with UTF support), the computation of how
3618    much memory was required could be incorrect, leading to segfaults or other
3619    strange effects.
3621    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3622    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3623    there was a subsequent back reference to those brackets. This meant that, for
3624    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3625    possible to check for precisely this case. All we can do is abandon the
3626    optimization if .* occurs inside capturing brackets when there are any back
3627    references whatsoever. (See below for a better fix that came later.)
3629    26. The handling of the optimization for finding the first character of a
3630    non-anchored pattern, and for finding a character that is required later in the
3631    match were failing in some cases. This didn't break the matching; it just
3632    failed to optimize when it could. The way this is done has been re-implemented.
3634    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3636    28. Added a new feature that provides some of the functionality that Perl
3637    provides with (?{...}). The facility is termed a "callout". The way it is done
3638    in PCRE is for the caller to provide an optional function, by setting
3639    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3640    global variable. By default it is unset, which disables all calling out. To get
3641    the function called, the regex must include (?C) at appropriate points. This
3642    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3643    This provides a means of identifying different callout points. When PCRE
3644    reaches such a point in the regex, if pcre_callout has been set, the external
3645    function is called. It is provided with data in a structure called
3646    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3647    matching continues; if it returns a non-zero value, the match at the current
3648    point fails. However, backtracking will occur if possible. [This was changed
3649    later and other features added - see item 49 below.]
3651    29. pcretest is upgraded to test the callout functionality. It provides a
3652    callout function that displays information. By default, it shows the start of
3653    the match and the current position in the text. There are some new data escapes
3654    to vary what happens:
3656        \C+         in addition, show current contents of captured substrings
3657        \C-         do not supply a callout function
3658        \C!n        return 1 when callout number n is reached
3659        \C!n!m      return 1 when callout number n is reached for the mth time
3661    30. If pcregrep was called with the -l option and just a single file name, it
3662    output "<stdin>" if a match was found, instead of the file name.
3664    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3665    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3666    pcre_exec(). This saves a malloc/free per call. The default value of
3667    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3668    when configuring.
3670    32. The default maximum size of a compiled pattern is 64K. There have been a
3671    few cases of people hitting this limit. The code now uses macros to handle the
3672    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3673    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3674    configuring. Tests 2 and 5 work only with 2-byte links because they output
3675    debugging information about compiled patterns.
3677    33. Internal code re-arrangements:
3679    (a) Moved the debugging function for printing out a compiled regex into
3680        its own source file (printint.c) and used #include to pull it into
3681        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3682        separate copies.
3684    (b) Defined the list of op-code names for debugging as a macro in
3685        internal.h so that it is next to the definition of the opcodes.
3687    (c) Defined a table of op-code lengths for simpler skipping along compiled
3688        code. This is again a macro in internal.h so that it is next to the
3689        definition of the opcodes.
3691    34. Added support for recursive calls to individual subpatterns, along the
3692    lines of Robin Houston's patch (but implemented somewhat differently).
3694    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3695    allow it to read and process whole directories in Win32. This code was
3696    contributed by Lionel Fourquaux; it has not been tested by me.
3698    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3699    used to name a group. Names consist of alphanumerics and underscores, and must
3700    be unique. Back references use the syntax (?P=name) and recursive calls use
3701    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3702    numbers. The function pcre_fullinfo() can be used after compilation to extract
3703    a name/number map. There are three relevant calls:
3705      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3706      PCRE_INFO_NAMECOUNT            yields the number of entries
3707      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3709    The map is a vector of fixed-size entries. The size of each entry depends on
3710    the length of the longest name used. The first two bytes of each entry are the
3711    group number, most significant byte first. There follows the corresponding
3712    name, zero terminated. The names are in alphabetical order.
3714    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3715    case instead of 255. Making it the same both with and without UTF-8 support
3716    means that the same test output works with both.
3718    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3719    calling malloc() with a zero argument.
3721    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3722    optimization. I've improved things by keeping a bitmap of backreferences with
3723    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3724    fact referenced, the optimization can be applied. It is unlikely that a
3725    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3726    the match to follow \n) will appear inside brackets with a number greater than
3727    31, but if it does, any back reference > 31 suppresses the optimization.
3729    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3730    of disabling numbered capturing parentheses. Any opening parenthesis that is
3731    not followed by ? behaves as if it were followed by ?: but named parentheses
3732    can still be used for capturing (and they will acquire numbers in the usual
3733    way).
3735    41. Redesigned the return codes from the match() function into yes/no/error so
3736    that errors can be passed back from deep inside the nested calls. A malloc
3737    failure while inside a recursive subpattern call now causes the
3738    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3740    42. It is now possible to set a limit on the number of times the match()
3741    function is called in a call to pcre_exec(). This facility makes it possible to
3742    limit the amount of recursion and backtracking, though not in a directly
3743    obvious way, because the match() function is used in a number of different
3744    circumstances. The count starts from zero for each position in the subject
3745    string (for non-anchored patterns). The default limit is, for compatibility, a
3746    large number, namely 10 000 000. You can change this in two ways:
3748    (a) When configuring PCRE before making, you can use --with-match-limit=n
3749        to set a default value for the compiled library.
3751    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3752        a different value is set. See 45 below.
3754    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3756    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3757    of things that can be changed at compile time. The first argument specifies
3758    what is wanted and the second points to where the information is to be placed.
3759    The current list of available information is:
3761      PCRE_CONFIG_UTF8
3763    The output is an integer that is set to one if UTF-8 support is available;
3764    otherwise it is set to zero.
3768    The output is an integer that it set to the value of the code that is used for
3769    newline. It is either LF (10) or CR (13).
3773    The output is an integer that contains the number of bytes used for internal
3774    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3778    The output is an integer that contains the threshold above which the POSIX
3779    interface uses malloc() for output vectors. See item 31 above.
3783    The output is an unsigned integer that contains the default limit of the number
3784    of match() calls in a pcre_exec() execution. See 42 above.
3786    44. pcretest has been upgraded by the addition of the -C option. This causes it
3787    to extract all the available output from the new pcre_config() function, and to
3788    output it. The program then exits immediately.
3790    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3791    order to support additional features. One way would have been to define
3792    pcre_exec2() (for example) with extra arguments, but this would not have been
3793    extensible, and would also have required all calls to the original function to
3794    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3795    is used for passing in "extra" data from pcre_study().
3797    The pcre_extra structure is now exposed and defined in pcre.h. It currently