ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 35 by nigel, Sat Feb 24 21:39:05 2007 UTC revision 738 by ph10, Fri Oct 21 09:04:01 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.20 21-Oct-2011
5    ------------------------
7    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
8        a POSIX class. After further experiments with Perl, which convinced me that
9        Perl has bugs and confusions, a closing square bracket is no longer allowed
10        in a POSIX name. This bug also affected patterns with classes that started
11        with full stops.
13    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
14        captured substring, but while checking the failing first alternative,
15        substring 1 is temporarily captured. If the output vector supplied to
16        pcre_exec() was not big enough for this capture, the yield of the function
17        was still zero ("insufficient space for captured substrings"). This cannot
18        be totally fixed without adding another stack variable, which seems a lot
19        of expense for a edge case. However, I have improved the situation in cases
20        such as /(a)(b)x|abc/ matched against "abc", where the return code
21        indicates that fewer than the maximum number of slots in the ovector have
22        been set.
24    3.  Related to (2) above: when there are more back references in a pattern than
25        slots in the output vector, pcre_exec() uses temporary memory during
26        matching, and copies in the captures as far as possible afterwards. It was
27        using the entire output vector, but this conflicts with the specification
28        that only 2/3 is used for passing back captured substrings. Now it uses
29        only the first 2/3, for compatibility. This is, of course, another edge
30        case.
32    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
33        main code base, and can be used by building with --enable-jit. When this is
34        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
35        runtime --no-jit option is given.
37    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
38        ovector, the return from the function was zero, implying that there were
39        other matches that did not fit. The correct "exactly full" value is now
40        returned.
42    6.  If a subpattern that was called recursively or as a subroutine contained
43        (*PRUNE) or any other control that caused it to give a non-standard return,
44        invalid errors such as "Error -26 (nested recursion at the same subject
45        position)" or even infinite loops could occur.
47    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
48        computing the minimum length on reaching *ACCEPT, and so ended up with the
49        wrong value of 1 rather than 0. Further investigation indicates that
50        computing a minimum subject length in the presence of *ACCEPT is difficult
51        (think back references, subroutine calls), and so I have changed the code
52        so that no minimum is registered for a pattern that contains *ACCEPT.
54    8.  If (*THEN) was present in the first (true) branch of a conditional group,
55        it was not handled as intended. [But see 16 below.]
57    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
58        Sheri Pierce.
60    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
61        the first byte in a match must be "a".
63    11. Change 17 for 8.13 increased the recursion depth for patterns like
64        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
65        pattern contains any instances of (*THEN). If it does not, the old
66        optimizations are restored. It would be nice to do this on a per-group
67        basis, but at the moment that is not feasible.
69    12. In some environments, the output of pcretest -C is CRLF terminated. This
70        broke RunTest's code that checks for the link size. A single white space
71        character after the value is now allowed for.
73    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
74        For "fr", it uses the Windows-specific input and output files.
76    14. If (*THEN) appeared in a group that was called recursively or as a
77        subroutine, it did not work as intended. [But see next item.]
79    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
80        pattern fragments (but not containing any | characters). If A and B are
81        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
82        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
83        D. In other words, Perl considers parentheses that do not contain any |
84        characters to be part of a surrounding alternative, whereas PCRE was
85        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
86        differently. PCRE now behaves in the same way as Perl, except in the case
87        of subroutine/recursion calls such as (?1) which have in any case always
88        been different (but PCRE had them first :-).
90    16. Related to 15 above: Perl does not treat the | in a conditional group as
91        creating alternatives. Such a group is treated in the same way as an
92        ordinary group without any | characters when processing (*THEN). PCRE has
93        been changed to match Perl's behaviour.
95    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
96        RunGrepTest script failed.
98    18. Change 22 for version 13 caused atomic groups to use more stack. This is
99        inevitable for groups that contain captures, but it can lead to a lot of
100        stack use in large patterns. The old behaviour has been restored for atomic
101        groups that do not contain any capturing parentheses.
103    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
104        suppress the check for a minimum subject length at run time. (If it was
105        given to pcre_exec() or pcre_dfa_exec() it did work.)
107    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
108        fail to work when decoding hex characters in data strings in EBCDIC
109        environments.
111    21. It appears that in at least one Mac OS environment, the isxdigit() function
112        is implemented as a macro that evaluates to its argument more than once,
113        contravening the C 90 Standard (I haven't checked a later standard). There
114        was an instance in pcretest which caused it to go wrong when processing
115        \x{...} escapes in subject strings. The has been rewritten to avoid using
116        things like p++ in the argument of isxdigit().
119    Version 8.13 16-Aug-2011
120    ------------------------
122    1.  The Unicode data tables have been updated to Unicode 6.0.0.
124    2.  Two minor typos in pcre_internal.h have been fixed.
126    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
127        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
128        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
130    4.  There were a number of related bugs in the code for matching backrefences
131        caselessly in UTF-8 mode when codes for the characters concerned were
132        different numbers of bytes. For example, U+023A and U+2C65 are an upper
133        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
134        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
135        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
136        2-byte code at the end of the subject (it thought there wasn't enough data
137        left).
139    5.  Comprehensive information about what went wrong is now returned by
140        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
141        as the output vector has at least 2 elements. The offset of the start of
142        the failing character and a reason code are placed in the vector.
144    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
145        now returned is for the first byte of the failing character, instead of the
146        last byte inspected. This is an incompatible change, but I hope it is small
147        enough not to be a problem. It makes the returned offset consistent with
148        pcre_exec() and pcre_dfa_exec().
150    7.  pcretest now gives a text phrase as well as the error number when
151        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
152        failure, the offset and reason code are output.
154    8.  When \R was used with a maximizing quantifier it failed to skip backwards
155        over a \r\n pair if the subsequent match failed. Instead, it just skipped
156        back over a single character (\n). This seems wrong (because it treated the
157        two characters as a single entity when going forwards), conflicts with the
158        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
159        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
160        has been changed.
162    9.  Some internal refactoring has changed the processing so that the handling
163        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
164        time (the PCRE_DOTALL option was changed this way some time ago: version
165        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
166        which was always a bit of a fudge. It also means that there is one less
167        argument for the match() function, which reduces its stack requirements
168        slightly. This change also fixes an incompatibility with Perl: the pattern
169        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
171    10. More internal refactoring has drastically reduced the number of recursive
172        calls to match() for possessively repeated groups such as (abc)++ when
173        using pcre_exec().
175    11. While implementing 10, a number of bugs in the handling of groups were
176        discovered and fixed:
178        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
179        (a|)*(?1) gave a compile-time internal error.
180        ((a|)+)+  did not notice that the outer group could match an empty string.
181        (^a|^)+   was not marked as anchored.
182        (.*a|.*)+ was not marked as matching at start or after a newline.
184    12. Yet more internal refactoring has removed another argument from the match()
185        function. Special calls to this function are now indicated by setting a
186        value in a variable in the "match data" data block.
188    13. Be more explicit in pcre_study() instead of relying on "default" for
189        opcodes that mean there is no starting character; this means that when new
190        ones are added and accidentally left out of pcre_study(), testing should
191        pick them up.
193    14. The -s option of pcretest has been documented for ages as being an old
194        synonym of -m (show memory usage). I have changed it to mean "force study
195        for every regex", that is, assume /S for every regex. This is similar to -i
196        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
197        using it. It makes it easier to run collections of tests with and without
198        study enabled, and thereby test pcre_study() more easily. All the standard
199        tests are now run with and without -s (but some patterns can be marked as
200        "never study" - see 20 below).
202    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
203        restoration of the capturing data to the outer values was not happening
204        correctly.
206    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
207        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
208        pattern had matched an empty string, and so incorrectly returned a no
209        match.
211    17. There was optimizing code for the last branch of non-capturing parentheses,
212        and also for the obeyed branch of a conditional subexpression, which used
213        tail recursion to cut down on stack usage. Unfortunately, now that there is
214        the possibility of (*THEN) occurring in these branches, tail recursion is
215        no longer possible because the return has to be checked for (*THEN). These
216        two optimizations have therefore been removed. [But see 8.20/11 above.]
218    18. If a pattern containing \R was studied, it was assumed that \R always
219        matched two bytes, thus causing the minimum subject length to be
220        incorrectly computed because \R can also match just one byte.
222    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
223        was incorrectly computed.
225    20. If /S is present twice on a test pattern in pcretest input, it now
226        *disables* studying, thereby overriding the use of -s on the command line
227        (see 14 above). This is necessary for one or two tests to keep the output
228        identical in both cases.
230    21. When (*ACCEPT) was used in an assertion that matched an empty string and
231        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
233    22. When an atomic group that contained a capturing parenthesis was
234        successfully matched, but the branch in which it appeared failed, the
235        capturing was not being forgotten if a higher numbered group was later
236        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
237        group 1 to "a", when in fact it should be unset. This applied to multi-
238        branched capturing and non-capturing groups, repeated or not, and also to
239        positive assertions (capturing in negative assertions does not happen
240        in PCRE) and also to nested atomic groups.
242    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
243        subject after a captured substring, to make it easier to tell which of a
244        number of identical substrings has been captured.
246    24. The way atomic groups are processed by pcre_exec() has been changed so that
247        if they are repeated, backtracking one repetition now resets captured
248        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
249        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
250        "aaa". Previously, it would have been "a". As part of this code
251        refactoring, the way recursive calls are handled has also been changed.
253    25. If an assertion condition captured any substrings, they were not passed
254        back unless some other capturing happened later. For example, if
255        (?(?=(a))a) was matched against "a", no capturing was returned.
257    26. When studying a pattern that contained subroutine calls or assertions,
258        the code for finding the minimum length of a possible match was handling
259        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
260        group 1 called group 2 while simultaneously a separate group 2 called group
261        1). A stack overflow occurred in this case. I have fixed this by limiting
262        the recursion depth to 10.
264    27. Updated RunTest.bat in the distribution to the version supplied by Tom
265        Fortmann. This supports explicit test numbers on the command line, and has
266        argument validation and error reporting.
268    28. An instance of \X with an unlimited repeat could fail if at any point the
269        first character it looked at was a mark character.
271    29. Some minor code refactoring concerning Unicode properties and scripts
272        should reduce the stack requirement of match() slightly.
274    30. Added the '=' option to pcretest to check the setting of unused capturing
275        slots at the end of the pattern, which are documented as being -1, but are
276        not included in the return count.
278    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
279        compiled something random. Now it gives a compile-time error (as does
280        Perl).
282    32. A *MARK encountered during the processing of a positive assertion is now
283        recorded and passed back (compatible with Perl).
285    33. If --only-matching or --colour was set on a pcregrep call whose pattern
286        had alternative anchored branches, the search for a second match in a line
287        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
288        matched the line "0102" twice. The same bug affected patterns that started
289        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
290        twice.
292    34. Previously, PCRE did not allow quantification of assertions. However, Perl
293        does, and because of capturing effects, quantifying parenthesized
294        assertions may at times be useful. Quantifiers are now allowed for
295        parenthesized assertions.
297    35. A minor code tidy in pcre_compile() when checking options for \R usage.
299    36. \g was being checked for fancy things in a character class, when it should
300        just be a literal "g".
302    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
303        appearance of a nested POSIX class supersedes an apparent external class.
304        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
305        unescaped square brackets may also appear as part of class names. For
306        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
307        more like Perl. (But see 8.20/1 above.)
309    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
310        was because it thought it was \N{name}, which is not supported).
312    39. Add minix to OS list not supporting the -S option in pcretest.
314    40. PCRE tries to detect cases of infinite recursion at compile time, but it
315        cannot analyze patterns in sufficient detail to catch mutual recursions
316        such as ((?1))((?2)). There is now a runtime test that gives an error if a
317        subgroup is called recursively as a subpattern for a second time at the
318        same position in the subject string. In previous releases this might have
319        been caught by the recursion limit, or it might have run out of stack.
321    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
322        happen only once. PCRE was, however incorrectly giving a compile time error
323        "recursive call could loop indefinitely" because it cannot analyze the
324        pattern in sufficient detail. The compile time test no longer happens when
325        PCRE is compiling a conditional subpattern, but actual runaway loops are
326        now caught at runtime (see 40 above).
328    42. It seems that Perl allows any characters other than a closing parenthesis
329        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
330        has been changed to be the same.
332    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
333        as not to get warnings when autogen.sh is called. Also changed
334        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
336    44. To help people who use pcregrep to scan files containing exceedingly long
337        lines, the following changes have been made:
339        (a) The default value of the buffer size parameter has been increased from
340            8K to 20K. (The actual buffer used is three times this size.)
342        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
343            PCRE is built.
345        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
346            to be set at run time.
348        (d) Numerical values in pcregrep options can be followed by K or M, for
349            example --buffer-size=50K.
351        (e) If a line being scanned overflows pcregrep's buffer, an error is now
352            given and the return code is set to 2.
354    45. Add a pointer to the latest mark to the callout data block.
356    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
357        partial match of an empty string instead of no match. This was specific to
358        the use of ".".
360    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
361        complete match instead of a partial match. This bug was dependent on both
362        the PCRE_UTF8 and PCRE_DOTALL options being set.
364    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
365        starting byte set, because \b was not being ignored.
368    Version 8.12 15-Jan-2011
369    ------------------------
371    1.  Fixed some typos in the markup of the man pages, and wrote a script that
372        checks for such things as part of the documentation building process.
374    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
375        --match-limit and --recursion-limit options (added for 8.11). In
376        particular, this made one of the standard tests fail. (The integer value
377        went into the wrong half of a long int.)
379    3.  If the --colour option was given to pcregrep with -v (invert match), it
380        did strange things, either producing crazy output, or crashing. It should,
381        of course, ignore a request for colour when reporting lines that do not
382        match.
384    4.  Another pcregrep bug caused similar problems if --colour was specified with
385        -M (multiline) and the pattern match finished with a line ending.
387    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
388        matched in multiline mode, the following line was shown as part of the
389        match. This seems wrong, so I have changed it.
391    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
392        the check for further matches in the same line (so they could be coloured)
393        to overrun the end of the current line. If another match was found, it was
394        incorrectly shown (and then shown again when found in the next line).
396    7.  If pcregrep was compiled under Windows, there was a reference to the
397        function pcregrep_exit() before it was defined. I am assuming this was
398        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
399        reported by a user. I've moved the definition above the reference.
402    Version 8.11 10-Dec-2010
403    ------------------------
405    1.  (*THEN) was not working properly if there were untried alternatives prior
406        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
407        backtracked to try for "b" instead of moving to the next alternative branch
408        at the same level (in this case, to look for "c"). The Perl documentation
409        is clear that when (*THEN) is backtracked onto, it goes to the "next
410        alternative in the innermost enclosing group".
412    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
413        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
414        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
415        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
416        (*THEN).
418    3.  If \s appeared in a character class, it removed the VT character from
419        the class, even if it had been included by some previous item, for example
420        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
421        of \s, but is part of the POSIX "space" class.)
423    4.  A partial match never returns an empty string (because you can always
424        match an empty string at the end of the subject); however the checking for
425        an empty string was starting at the "start of match" point. This has been
426        changed to the "earliest inspected character" point, because the returned
427        data for a partial match starts at this character. This means that, for
428        example, /(?<=abc)def/ gives a partial match for the subject "abc"
429        (previously it gave "no match").
431    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
432        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
433        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
434        has an implication that the given string is incomplete (because a partial
435        match is preferred over a full match). For this reason, these items now
436        give a partial match in this situation. [Aside: previously, the one case
437        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
438        match rather than a full match, which was wrong by the old rules, but is
439        now correct.]
441    6.  There was a bug in the handling of #-introduced comments, recognized when
442        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
443        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
444        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
445        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
446        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
447        places in pcre_compile().
449    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
450        comments when looking ahead for named forward references to subpatterns,
451        the only newline sequence it recognized was NL. It now handles newlines
452        according to the set newline convention.
454    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
455        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
456        cater for a lack of strerror(). These oversights have been fixed.
458    9.  Added --match-limit and --recursion-limit to pcregrep.
460    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
462    11. When the -o option was used, pcregrep was setting a return code of 1, even
463        when matches were found, and --line-buffered was not being honoured.
465    12. Added an optional parentheses number to the -o and --only-matching options
466        of pcregrep.
468    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
469        can match an empty string. The code to do it in pcretest and pcredemo
470        needed fixing:
472        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
473            only one byte after an empty string match just before CRLF (this case
474            just got forgotten; "any" and "anycrlf" were OK).
476        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
477            mode when an empty string match preceded an ASCII character followed by
478            a non-ASCII character. (The code for advancing by one character rather
479            than one byte was nonsense.)
481        (c) The pcredemo.c sample program did not have any code at all to handle
482            the cases when CRLF is a valid newline sequence.
484    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
485        as a starting offset was within the subject string. There is now a new
486        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
487        negative or greater than the length of the string. In order to test this,
488        pcretest is extended to allow the setting of negative starting offsets.
490    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
491        starting offset points to the beginning of a UTF-8 character was
492        unnecessarily clumsy. I tidied it up.
494    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
495        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
497    17. Nobody had reported that the --include_dir option, which was added in
498        release 7.7 should have been called --include-dir (hyphen, not underscore)
499        for compatibility with GNU grep. I have changed it to --include-dir, but
500        left --include_dir as an undocumented synonym, and the same for
501        --exclude-dir, though that is not available in GNU grep, at least as of
502        release 2.5.4.
504    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
505        characters from a string of bytes) have been redefined so as not to use
506        loops, in order to improve performance in some environments. At the same
507        time, I abstracted some of the common code into auxiliary macros to save
508        repetition (this should not affect the compiled code).
510    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
511        compile-time error is now given if \c is not followed by an ASCII
512        character, that is, a byte less than 128. (In EBCDIC mode, the code is
513        different, and any byte value is allowed.)
515    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
516        START_OPTIMIZE option, which is now allowed at compile time - but just
517        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
518        to pcregrep and other applications that have no direct access to PCRE
519        options. The new /Y option in pcretest sets this option when calling
520        pcre_compile().
522    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
523        back references. Groups containing recursive back references were forced to
524        be atomic by that change, but in the case of named groups, the amount of
525        memory required was incorrectly computed, leading to "Failed: internal
526        error: code overflow". This has been fixed.
528    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
529        pcretest.c, to avoid build problems in some Borland environments.
532    Version 8.10 25-Jun-2010
533    ------------------------
535    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
536        THEN.
538    2.  (*ACCEPT) was not working when inside an atomic group.
540    3.  Inside a character class, \B is treated as a literal by default, but
541        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
542        causes the error). The code is unchanged, but I tidied the documentation.
544    4.  Inside a character class, PCRE always treated \R and \X as literals,
545        whereas Perl faults them if its -w option is set. I have changed PCRE so
546        that it faults them when PCRE_EXTRA is set.
548    5.  Added support for \N, which always matches any character other than
549        newline. (It is the same as "." when PCRE_DOTALL is not set.)
551    6.  When compiling pcregrep with newer versions of gcc which may have
552        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
553        declared with attribute warn_unused_result" were given. Just casting the
554        result to (void) does not stop the warnings; a more elaborate fudge is
555        needed. I've used a macro to implement this.
557    7.  Minor change to pcretest.c to avoid a compiler warning.
559    8.  Added four artifical Unicode properties to help with an option to make
560        \s etc use properties (see next item). The new properties are: Xan
561        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
563    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
564        use Unicode properties. (*UCP) at the start of a pattern can be used to set
565        this option. Modified pcretest to add /W to test this facility. Added
566        REG_UCP to make it available via the POSIX interface.
568    10. Added --line-buffered to pcregrep.
570    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
571        studied, and the match started with a letter with a code point greater than
572        127 whose first byte was different to the first byte of the other case of
573        the letter, the other case of this starting letter was not recognized
574        (#976).
576    12. If a pattern that was studied started with a repeated Unicode property
577        test, for example, \p{Nd}+, there was the theoretical possibility of
578        setting up an incorrect bitmap of starting bytes, but fortunately it could
579        not have actually happened in practice until change 8 above was made (it
580        added property types that matched character-matching opcodes).
582    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
583        possible starting bytes for non-anchored patterns.
585    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
586        \R, and also a number of cases that involve Unicode properties, both
587        explicit and implicit when PCRE_UCP is set.
589    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
590        input, it could crash or give wrong results if characters with values
591        greater than 0xc0 were present in the subject string. (Detail: it assumed
592        UTF-8 input when processing these items.)
594    16. Added a lot of (int) casts to avoid compiler warnings in systems where
595        size_t is 64-bit (#991).
597    17. Added a check for running out of memory when PCRE is compiled with
598        --disable-stack-for-recursion (#990).
600    18. If the last data line in a file for pcretest does not have a newline on
601        the end, a newline was missing in the output.
603    19. The default pcre_chartables.c file recognizes only ASCII characters (values
604        less than 128) in its various bitmaps. However, there is a facility for
605        generating tables according to the current locale when PCRE is compiled. It
606        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
607        characters, are recognized by isspace() and therefore were getting set in
608        these tables, and indeed these tables seem to approximate to ISO 8859. This
609        caused a problem in UTF-8 mode when pcre_study() was used to create a list
610        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
611        which of course cannot start UTF-8 characters. I have changed the code so
612        that only real ASCII characters (less than 128) and the correct starting
613        bytes for UTF-8 encodings are set for characters greater than 127 when in
614        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
615        altogether.)
617    20. Added the /T option to pcretest so as to be able to run tests with non-
618        standard character tables, thus making it possible to include the tests
619        used for 19 above in the standard set of tests.
621    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
622        reference to a subpattern the other side of a comment that contains an
623        opening parenthesis caused either an internal compiling error, or a
624        reference to the wrong subpattern.
627    Version 8.02 19-Mar-2010
628    ------------------------
630    1.  The Unicode data tables have been updated to Unicode 5.2.0.
632    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
633        configured.
635    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
636        original author of that file, following a query about its status.
638    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
639        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
641    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
642        quantifier applied to a forward-referencing subroutine call, could compile
643        incorrect code or give the error "internal error: previously-checked
644        referenced subpattern not found".
646    6.  Both MS Visual Studio and Symbian OS have problems with initializing
647        variables to point to external functions. For these systems, therefore,
648        pcre_malloc etc. are now initialized to local functions that call the
649        relevant global functions.
651    7.  There were two entries missing in the vectors called coptable and poptable
652        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
653        I've fixed the data, and added a kludgy way of testing at compile time that
654        the lengths are correct (equal to the number of opcodes).
656    8.  Following on from 7, I added a similar kludge to check the length of the
657        eint vector in pcreposix.c.
659    9.  Error texts for pcre_compile() are held as one long string to avoid too
660        much relocation at load time. To find a text, the string is searched,
661        counting zeros. There was no check for running off the end of the string,
662        which could happen if a new error number was added without updating the
663        string.
665    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
667    11. \K was not working if it appeared in an atomic group or in a group that
668        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
669        \K is "not well defined" if used in an assertion. PCRE now accepts it if
670        the assertion is positive, but not if it is negative.
672    12. Change 11 fortuitously reduced the size of the stack frame used in the
673        "match()" function of pcre_exec.c by one pointer. Forthcoming
674        implementation of support for (*MARK) will need an extra pointer on the
675        stack; I have reserved it now, so that the stack frame size does not
676        decrease.
678    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
679        item in branch that calls a recursion is a subroutine call - as in the
680        second branch in the above example - was incorrectly given the compile-
681        time error "recursive call could loop indefinitely" because pcre_compile()
682        was not correctly checking the subroutine for matching a non-empty string.
684    14. The checks for overrunning compiling workspace could trigger after an
685        overrun had occurred. This is a "should never occur" error, but it can be
686        triggered by pathological patterns such as hundreds of nested parentheses.
687        The checks now trigger 100 bytes before the end of the workspace.
689    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
692    Version 8.01 19-Jan-2010
693    ------------------------
695    1.  If a pattern contained a conditional subpattern with only one branch (in
696        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
697        computed the wrong minimum data length (which is of course zero for such
698        subpatterns). This could cause incorrect "no match" results.
700    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
701        the pattern is reset in the first branch, pcre_compile() failed with
702        "internal error: code overflow at offset...". This happened only when
703        the reset was to the original external option setting. (An optimization
704        abstracts leading options settings into an external setting, which was the
705        cause of this.)
707    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
708        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
709        assertion pattern did not match (meaning that the assertion was true), it
710        was incorrectly treated as false if the SKIP had been reached during the
711        matching. This also applied to assertions used as conditions.
713    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
714        assertion subpattern, including such a pattern used as a condition,
715        unpredictable results occurred, instead of the error return
718    5.  The C++ GlobalReplace function was not working like Perl for the special
719        situation when an empty string is matched. It now does the fancy magic
720        stuff that is necessary.
722    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
723        removed. (These were left over from very, very early versions of PCRE.)
725    7.  Some cosmetic changes to the code to make life easier when compiling it
726        as part of something else:
728        (a) Change DEBUG to PCRE_DEBUG.
730        (b) In pcre_compile(), rename the member of the "branch_chain" structure
731            called "current" as "current_branch", to prevent a collision with the
732            Linux macro when compiled as a kernel module.
734        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
735            prevent a collision with the Linux macro when compiled as a kernel
736            module.
738    8.  In pcre_compile() there are some checks for integer overflows that used to
739        cast potentially large values to (double). This has been changed to that
740        when building, a check for int64_t is made, and if it is found, it is used
741        instead, thus avoiding the use of floating point arithmetic. (There is no
742        other use of FP in PCRE.) If int64_t is not found, the fallback is to
743        double.
745    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
746        2005 (difference between two addresses compared to an unsigned value).
748    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
749        custom one, because of the following reported problem in Windows:
751          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
752              under Win32.
753          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
754              therefore missing the function definition.
755          - The compiler thus generates a "C" signature for the test function.
756          - The linker fails to find the "C" function.
757          - PCRE fails to configure if asked to do so against libbz2.
759    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
760        messages were output:
762          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
763          rerunning libtoolize, to keep the correct libtool macros in-tree.
764          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
766        I have done both of these things.
768    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
769        most of the time, it *can* run out if it is given a pattern that contains a
770        runaway infinite recursion. I updated the discussion in the pcrestack man
771        page.
773    13. Now that we have gone to the x.xx style of version numbers, the minor
774        version may start with zero. Using 08 or 09 is a bad idea because users
775        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
776        interpreted as invalid octal numbers. I've updated the previous comment in
777        configure.ac, and also added a check that gives an error if 08 or 09 are
778        used.
780    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
781        causing partial matching to fail when the end of the subject matched \W
782        in a UTF-8 pattern where \W was quantified with a minimum of 3.
784    15. There were some discrepancies between the declarations in pcre_internal.h
785        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
786        their definitions. The declarations used "const uschar *" and the
787        definitions used USPTR. Even though USPTR is normally defined as "const
788        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
789        reported that: "This difference in casting confuses some C++ compilers, for
790        example, SunCC recognizes above declarations as different functions and
791        generates broken code for hbpcre." I have changed the declarations to use
792        USPTR.
794    16. GNU libtool is named differently on some systems. The autogen.sh script now
795        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
796        (FreeBSD).
798    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
799        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
800        comment: "Figure out how to create a longlong from a string: strtoll and
801        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
802        instance, but it only takes 2 args instead of 3!"
804    18. A subtle bug concerned with back references has been fixed by a change of
805        specification, with a corresponding code fix. A pattern such as
806        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
807        refers, was giving matches when it shouldn't. For example, xa=xaaa would
808        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
809        same bug. Such groups have to be quantified to be useful, or contained
810        inside another quantified group. (If there's no repetition, the reference
811        can never match.) The problem arises because, having left the group and
812        moved on to the rest of the pattern, a later failure that backtracks into
813        the group uses the captured value from the final iteration of the group
814        rather than the correct earlier one. I have fixed this in PCRE by forcing
815        any group that contains a reference to itself to be an atomic group; that
816        is, there cannot be any backtracking into it once it has completed. This is
817        similar to recursive and subroutine calls.
820    Version 8.00 19-Oct-09
821    ----------------------
823    1.  The table for translating pcre_compile() error codes into POSIX error codes
824        was out-of-date, and there was no check on the pcre_compile() error code
825        being within the table. This could lead to an OK return being given in
826        error.
828    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
829        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
830        in a Windows environment.
832    3.  The pcregrep --count option prints the count for each file even when it is
833        zero, as does GNU grep. However, pcregrep was also printing all files when
834        --files-with-matches was added. Now, when both options are given, it prints
835        counts only for those files that have at least one match. (GNU grep just
836        prints the file name in this circumstance, but including the count seems
837        more useful - otherwise, why use --count?) Also ensured that the
838        combination -clh just lists non-zero counts, with no names.
840    4.  The long form of the pcregrep -F option was incorrectly implemented as
841        --fixed_strings instead of --fixed-strings. This is an incompatible change,
842        but it seems right to fix it, and I didn't think it was worth preserving
843        the old behaviour.
845    5.  The command line items --regex=pattern and --regexp=pattern were not
846        recognized by pcregrep, which required --regex pattern or --regexp pattern
847        (with a space rather than an '='). The man page documented the '=' forms,
848        which are compatible with GNU grep; these now work.
850    6.  No libpcreposix.pc file was created for pkg-config; there was just
851        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
853    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
854        when UCP support is not needed, by modifying the Python script that
855        generates it from Unicode data files. This should not matter if the module
856        is correctly used as a library, but I received one complaint about 50K of
857        unwanted data. My guess is that the person linked everything into his
858        program rather than using a library. Anyway, it does no harm.
860    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
861        was a minimum greater than 1 for a wide character in a possessive
862        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
863        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
864        character. Chaos in the form of incorrect output or a compiling loop could
865        result.
867    9.  The restrictions on what a pattern can contain when partial matching is
868        requested for pcre_exec() have been removed. All patterns can now be
869        partially matched by this function. In addition, if there are at least two
870        slots in the offset vector, the offset of the earliest inspected character
871        for the match and the offset of the end of the subject are set in them when
872        PCRE_ERROR_PARTIAL is returned.
874    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
875        synonymous with PCRE_PARTIAL, for backwards compatibility, and
876        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
877        and may be more useful for multi-segment matching.
879    11. Partial matching with pcre_exec() is now more intuitive. A partial match
880        used to be given if ever the end of the subject was reached; now it is
881        given only if matching could not proceed because another character was
882        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
883        string "Z", which now yields "no match" instead of "partial match". In the
884        case of pcre_dfa_exec(), "no match" is given if every matching path for the
885        final character ended with (*FAIL).
887    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
888        if the pattern had a "must contain" character that was already found in the
889        earlier partial match, unless partial matching was again requested. For
890        example, with the pattern /dog.(body)?/, the "must contain" character is
891        "g". If the first part-match was for the string "dog", restarting with
892        "sbody" failed. This bug has been fixed.
894    13. The string returned by pcre_dfa_exec() after a partial match has been
895        changed so that it starts at the first inspected character rather than the
896        first character of the match. This makes a difference only if the pattern
897        starts with a lookbehind assertion or \b or \B (\K is not supported by
898        pcre_dfa_exec()). It's an incompatible change, but it makes the two
899        matching functions compatible, and I think it's the right thing to do.
901    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
902        so that the demonstration program is easily available in environments where
903        PCRE has not been installed from source.
905    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
906        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
907        library.
909    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
910        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
911        is not the first non-POSIX option to be added. Clearly some people find
912        these options useful.
914    17. If a caller to the POSIX matching function regexec() passes a non-zero
915        value for nmatch with a NULL value for pmatch, the value of
916        nmatch is forced to zero.
918    18. RunGrepTest did not have a test for the availability of the -u option of
919        the diff command, as RunTest does. It now checks in the same way as
920        RunTest, and also checks for the -b option.
922    19. If an odd number of negated classes containing just a single character
923        interposed, within parentheses, between a forward reference to a named
924        subpattern and the definition of the subpattern, compilation crashed with
925        an internal error, complaining that it could not find the referenced
926        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
927        [The bug was that it was starting one character too far in when skipping
928        over the character class, thus treating the ] as data rather than
929        terminating the class. This meant it could skip too much.]
931    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
932        /g option in pcretest when the pattern contains \K, which makes it possible
933        to have an empty string match not at the start, even when the pattern is
934        anchored. Updated pcretest and pcredemo to use this option.
936    21. If the maximum number of capturing subpatterns in a recursion was greater
937        than the maximum at the outer level, the higher number was returned, but
938        with unset values at the outer level. The correct (outer level) value is
939        now given.
941    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
942        PCRE did not set those parentheses (unlike Perl). I have now found a way to
943        make it do so. The string so far is captured, making this feature
944        compatible with Perl.
946    23. The tests have been re-organized, adding tests 11 and 12, to make it
947        possible to check the Perl 5.10 features against Perl 5.10.
949    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
950        pattern matches a fixed length string. PCRE did not allow this; now it
951        does. Neither allows recursion.
953    25. I finally figured out how to implement a request to provide the minimum
954        length of subject string that was needed in order to match a given pattern.
955        (It was back references and recursion that I had previously got hung up
956        on.) This code has now been added to pcre_study(); it finds a lower bound
957        to the length of subject needed. It is not necessarily the greatest lower
958        bound, but using it to avoid searching strings that are too short does give
959        some useful speed-ups. The value is available to calling programs via
960        pcre_fullinfo().
962    26. While implementing 25, I discovered to my embarrassment that pcretest had
963        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
964        study optimizations had never been tested with that matching function.
965        Oops. What is worse, even when it was passed study data, there was a bug in
966        pcre_dfa_exec() that meant it never actually used it. Double oops. There
967        were also very few tests of studied patterns with pcre_dfa_exec().
969    27. If (?| is used to create subpatterns with duplicate numbers, they are now
970        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
971        on the other side of the coin, they are no longer allowed to have different
972        names, because these cannot be distinguished in PCRE, and this has caused
973        confusion. (This is a difference from Perl.)
975    28. When duplicate subpattern names are present (necessarily with different
976        numbers, as required by 27 above), and a test is made by name in a
977        conditional pattern, either for a subpattern having been matched, or for
978        recursion in such a pattern, all the associated numbered subpatterns are
979        tested, and the overall condition is true if the condition is true for any
980        one of them. This is the way Perl works, and is also more like the way
981        testing by number works.
984    Version 7.9 11-Apr-09
985    ---------------------
987    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
988        (pcretest), all targets were linked against these libraries. This included
989        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
990        libraries. This caused unwanted dependencies to be created. This problem
991        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
992        pcretest is linked with readline.
994    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
995        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
996        moved outside it again, because FALSE and TRUE are already defined in AIX,
997        but BOOL is not.
999    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1000        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1002    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1003        hyphens) following file names and line numbers when outputting matching
1004        lines. This is not true; no spaces are inserted. I have also clarified the
1005        wording for the --colour (or --color) option.
1007    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1008        was not coloured; this is different to GNU grep, so I have changed it to be
1009        the same.
1011    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1012        each matching line was coloured. Now it goes on to look for further matches
1013        of any of the test patterns, which is the same behaviour as GNU grep.
1015    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1016        doesn't make sense to accept an empty string match in pcregrep, so I have
1017        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1018        seems to be how GNU grep behaves.
1020    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1021        start or after a newline", because the conditional assertion was not being
1022        correctly handled. The rule now is that both the assertion and what follows
1023        in the first alternative must satisfy the test.
1025    9.  If auto-callout was enabled in a pattern with a conditional group whose
1026        condition was an assertion, PCRE could crash during matching, both with
1027        pcre_exec() and pcre_dfa_exec().
1029    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1030        used for matching.
1032    11. Unicode property support in character classes was not working for
1033        characters (bytes) greater than 127 when not in UTF-8 mode.
1035    12. Added the -M command line option to pcretest.
1037    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1039    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1041    16. Added comments and documentation about mis-use of no_arg in the C++
1042        wrapper.
1044    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1045        from Martin Jerabek that uses macro names for all relevant character and
1046        string constants.
1048    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1049        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1050        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1051        these, but not everybody uses configure.
1053    19. A conditional group that had only one branch was not being correctly
1054        recognized as an item that could match an empty string. This meant that an
1055        enclosing group might also not be so recognized, causing infinite looping
1056        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1057        with the subject "ab", where knowledge that the repeated group can match
1058        nothing is needed in order to break the loop.
1060    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1061        exec(), but without supplying a callout function, matching went wrong.
1063    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1064        leak if the size of the offset vector was greater than 30. When the vector
1065        is smaller, the saved offsets during recursion go onto a local stack
1066        vector, but for larger vectors malloc() is used. It was failing to free
1067        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1068        error, in fact).
1070    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1071        heapframe that is used only when UTF-8 support is enabled. This caused no
1072        problem, but was untidy.
1074    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1075        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1076        included within another project.
1078    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1079        slightly modified by me:
1081          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1082              not building pcregrep.
1084          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1085              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1087    25. Forward references, both numeric and by name, in patterns that made use of
1088        duplicate group numbers, could behave incorrectly or give incorrect errors,
1089        because when scanning forward to find the reference group, PCRE was not
1090        taking into account the duplicate group numbers. A pattern such as
1091        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1093    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1094        the feature of a custom pointer more persuasive (as requested by a user).
1096    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1097        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1098        pre-defined.
1100    28. Added support for (*UTF8) at the start of a pattern.
1102    29. Arrange for flags added by the "release type" setting in CMake to be shown
1103        in the configuration summary.
1106    Version 7.8 05-Sep-08
1107    ---------------------
1109    1.  Replaced UCP searching code with optimized version as implemented for Ad
1110        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1111        stage table and inline lookup instead of a function, giving speed ups of 2
1112        to 5 times on some simple patterns that I tested. Permission was given to
1113        distribute the MultiStage2.py script that generates the tables (it's not in
1114        the tarball, but is in the Subversion repository).
1116    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1117        scripts.
1119    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1120        a group with a zero qualifier. The result of the study could be incorrect,
1121        or the function might crash, depending on the pattern.
1123    4.  Caseless matching was not working for non-ASCII characters in back
1124        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1125        It now works when Unicode Property Support is available.
1127    5.  In pcretest, an escape such as \x{de} in the data was always generating
1128        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1129        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1130        truncation.
1132    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1134    7.  Added two (int) casts to pcregrep when printing the difference of two
1135        pointers, in case they are 64-bit values.
1137    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1138        test 2 if it fails.
1140    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1141        and a #define of that name to empty if it is not externally set. This is to
1142        allow users of MSVC to set it if necessary.
1144    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1145        the convenience functions in the pcre_get.c source file.
1147    11. An option change at the start of a pattern that had top-level alternatives
1148        could cause overwriting and/or a crash. This command provoked a crash in
1149        some environments:
1151          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1153        This potential security problem was recorded as CVE-2008-2371.
1155    12. For a pattern where the match had to start at the beginning or immediately
1156        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1157        pcre_dfa_exec() could read past the end of the passed subject if there was
1158        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1159        pcretest so that it places the subject at the end of its malloc-ed buffer.
1161    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1162        exec() might read past the end of the data buffer in UTF-8 mode.
1164    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1165        the data contained the byte 0x85 as part of a UTF-8 character within its
1166        first line. This applied both to normal and DFA matching.
1168    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1169        /^[^d]*?$/8 failed to match "abc".
1171    16. Added a missing copyright notice to pcrecpp_internal.h.
1173    17. Make it more clear in the documentation that values returned from
1174        pcre_exec() in ovector are byte offsets, not character counts.
1176    18. Tidied a few places to stop certain compilers from issuing warnings.
1178    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1179        supplied by Stefan Weber. I made a further small update for 7.8 because
1180        there is a change of source arrangements: the pcre_searchfuncs.c module is
1181        replaced by pcre_ucd.c.
1184    Version 7.7 07-May-08
1185    ---------------------
1187    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1188        a string to a long long, pretend we don't even have a long long." This is
1189        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1191    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1192        pre-7.6 versions, which defined a global no_arg variable instead of putting
1193        it in the RE class. (See also #8 below.)
1195    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1196        Lopes.
1198    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1200        (1) The include/exclude patterns were being applied to the whole pathnames
1201            of files, instead of just to the final components.
1203        (2) If there was more than one level of directory, the subdirectories were
1204            skipped unless they satisfied the include/exclude conditions. This is
1205            inconsistent with GNU grep (and could even be seen as contrary to the
1206            pcregrep specification - which I improved to make it absolutely clear).
1207            The action now is always to scan all levels of directory, and just
1208            apply the include/exclude patterns to regular files.
1210    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1211        --exclude_dir in the tests to avoid scanning .svn directories.
1213    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1214        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1215        doesn't support NULs in patterns.
1217    7.  Added some missing "const"s to declarations of static tables in
1218        pcre_compile.c and pcre_dfa_exec.c.
1220    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1221        caused by fix #2  above. (Subsequently also a second patch to fix the
1222        first patch. And a third patch - this was a messy problem.)
1224    9.  Applied Craig's patch to remove the use of push_back().
1226    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1227        matching function regexec().
1229    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1230        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1231        references. PCRE supports relative numbers with this syntax (I don't think
1232        Oniguruma does).
1234    12. Previously, a group with a zero repeat such as (...){0} was completely
1235        omitted from the compiled regex. However, this means that if the group
1236        was called as a subroutine from elsewhere in the pattern, things went wrong
1237        (an internal error was given). Such groups are now left in the compiled
1238        pattern, with a new opcode that causes them to be skipped at execution
1239        time.
1241    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1242        to the way PCRE behaves:
1244        (a) A lone ] character is dis-allowed (Perl treats it as data).
1246        (b) A back reference to an unmatched subpattern matches an empty string
1247            (Perl fails the current match path).
1249        (c) A data ] in a character class must be notated as \] because if the
1250            first data character in a class is ], it defines an empty class. (In
1251            Perl it is not possible to have an empty class.) The empty class []
1252            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1253            The negative empty class [^] matches any one character, independently
1254            of the DOTALL setting.
1256    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1257        non-existent subpattern following a character class starting with ']' and
1258        containing () gave an internal compiling error instead of "reference to
1259        non-existent subpattern". Fortunately, when the pattern did exist, the
1260        compiled code was correct. (When scanning forwards to check for the
1261        existencd of the subpattern, it was treating the data ']' as terminating
1262        the class, so got the count wrong. When actually compiling, the reference
1263        was subsequently set up correctly.)
1265    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1266        it was being rejected as not supported by pcre_dfa_exec(), even though
1267        other assertions are supported. I have made pcre_dfa_exec() support
1268        (*FAIL).
1270    16. The implementation of 13c above involved the invention of a new opcode,
1271        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1272        cannot be changed at match time, I realized I could make a small
1273        improvement to matching performance by compiling OP_ALLANY instead of
1274        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1275        on the OP_ANY path.
1277    17. Compiling pcretest on Windows with readline support failed without the
1278        following two fixes: (1) Make the unistd.h include conditional on
1279        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1281    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1282        ncurses library to be included for pcretest when ReadLine support is
1283        requested, but also to allow for it to be overridden. This patch came from
1284        Daniel Bergström.
1286    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1287        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1288        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1289        spotting this.
1292    Version 7.6 28-Jan-08
1293    ---------------------
1295    1.  A character class containing a very large number of characters with
1296        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1297        overflow.
1299    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1300        HAVE_LONG_LONG is not defined.
1302    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1303        bring them up to date and include new features. This patch includes:
1305        - Fixed PH's badly added libz and libbz2 support.
1306        - Fixed a problem with static linking.
1307        - Added pcredemo. [But later removed - see 7 below.]
1308        - Fixed dftables problem and added an option.
1309        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1310            HAVE_LONG_LONG.
1311        - Added readline support for pcretest.
1312        - Added an listing of the option settings after cmake has run.
1314    4.  A user submitted a patch to Makefile that makes it easy to create
1315        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1316        Makefile.am that cause it to include this special target, without
1317        affecting anything else. Note that the same mingw target plus all
1318        the other distribution libraries and programs are now supported
1319        when configuring with CMake (see 6 below) instead of with
1320        Configure/Make.
1322    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1323        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1324        exported in the Windows port". It has not yet been confirmed that the patch
1325        solves the problem, but it does no harm.
1327    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1328        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1329        with CMake, and also correct the comment about stack recursion.
1331    7.  Remove the automatic building of pcredemo from the ./configure system and
1332        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1333        of a program that users should build themselves after PCRE is installed, so
1334        building it automatically is not really right. What is more, it gave
1335        trouble in some build environments.
1337    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1340    Version 7.5 10-Jan-08
1341    ---------------------
1343    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1344        values in parens when parsing an RE using the C++ wrapper."
1346    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1347        Characters greater than 255 were excluded from the class instead of being
1348        included.
1350    3.  The same bug as (2) above applied to negated POSIX classes such as
1351        [:^space:].
1353    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1354        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1355        I have changed it.
1357    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1358        first named subpattern) and a construct such as (?&a) would reference the
1359        first named subpattern whose name started with "a" (in other words, the
1360        length check was missing). Both these problems are fixed. "Subpattern name
1361        expected" is now given for (?&) (a zero-length name), and this patch also
1362        makes it give the same error for \k'' (previously it complained that that
1363        was a reference to a non-existent subpattern).
1365    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1366        this is right because (?- can be followed by option settings as well as by
1367        digits. I have, however, made the messages clearer.
1369    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1370        than the number used in the conditional) now cause a compile-time error.
1371        This is actually not compatible with Perl, which accepts such patterns, but
1372        treats the conditional as always being FALSE (as PCRE used to), but it
1373        seems to me that giving a diagnostic is better.
1375    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1376        and messages.
1378    9.  Fix two occurrences of "backslash" in comments that should have been
1379        "backspace".
1381    10. Remove two redundant lines of code that can never be obeyed (their function
1382        was moved elsewhere).
1384    11. The program that makes PCRE's Unicode character property table had a bug
1385        which caused it to generate incorrect table entries for sequences of
1386        characters that have the same character type, but are in different scripts.
1387        It amalgamated them into a single range, with the script of the first of
1388        them. In other words, some characters were in the wrong script. There were
1389        thirteen such cases, affecting characters in the following ranges:
1391          U+002b0 - U+002c1
1392          U+0060c - U+0060d
1393          U+0061e - U+00612
1394          U+0064b - U+0065e
1395          U+0074d - U+0076d
1396          U+01800 - U+01805
1397          U+01d00 - U+01d77
1398          U+01d9b - U+01dbf
1399          U+0200b - U+0200f
1400          U+030fc - U+030fe
1401          U+03260 - U+0327f
1402          U+0fb46 - U+0fbb1
1403          U+10450 - U+1049d
1405    12. The -o option (show only the matching part of a line) for pcregrep was not
1406        compatible with GNU grep in that, if there was more than one match in a
1407        line, it showed only the first of them. It now behaves in the same way as
1408        GNU grep.
1410    13. If the -o and -v options were combined for pcregrep, it printed a blank
1411        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1412        does the same. The return code can be used to tell if there were any
1413        non-matching lines.
1415    14. Added --file-offsets and --line-offsets to pcregrep.
1417    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1418        infinitely looping recursion. The bug was that positive lookaheads were not
1419        being skipped when checking for a possible empty match (negative lookaheads
1420        and both kinds of lookbehind were skipped).
1422    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1423        inclusion of <windows.h> to before rather than after the definition of
1424        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1426    17. Specifying a possessive quantifier with a specific limit for a Unicode
1427        character property caused pcre_compile() to compile bad code, which led at
1428        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1429        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1430        caused the error; without that there was no problem.
1432    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1434    19. Added --enable-pcretest-libreadline.
1436    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1437        RE::GlobalReplace(). As a result, the number of replacements returned was
1438        double what it should be. I removed one of the increments, but Craig sent a
1439        later patch that removed the other one (the right fix) and added unit tests
1440        that check the return values (which was not done before).
1442    21. Several CMake things:
1444        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1445            the names libpcre and libpcreposix, not just pcre and pcreposix.
1447        (2) The above change means that pcretest and pcregrep are now correctly
1448            linked with the newly-built libraries, not previously installed ones.
1452    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1453        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1454        UTF-8 newline character). The key issue is that the pattern starts .*;
1455        this means that the match must be either at the beginning, or after a
1456        newline. The bug was in the code for advancing after a failed match and
1457        checking that the new position followed a newline. It was not taking
1458        account of UTF-8 characters correctly.
1460    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1461        character classes. PCRE was not treating the sequence [:...:] as a
1462        character class unless the ... were all letters. Perl, however, seems to
1463        allow any characters between [: and :], though of course it rejects as
1464        unknown any "names" that contain non-letters, because all the known class
1465        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1466        for example, whereas PCRE did not - it did not recognize a POSIX character
1467        class. This seemed a bit dangerous, so the code has been changed to be
1468        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1469        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1470        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1471        Perl does, and where it didn't before.
1473    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1474        Windows environments %n is disabled by default.
1477    Version 7.4 21-Sep-07
1478    ---------------------
1480    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1481        means that a class such as [\s] counted as "explicit reference to CR or
1482        LF". That isn't really right - the whole point of the change was to try to
1483        help when there was an actual mention of one of the two characters. So now
1484        the change happens only if \r or \n (or a literal CR or LF) character is
1485        encountered.
1487    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1488        of both had grown to the point where there were only 3 bits left.
1489        Fortunately, there was spare space in the data structure, and so I have
1490        moved the internal flags into a new 16-bit field to free up more option
1491        bits.
1493    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1494        but did not set the internal JCHANGED flag - either of these is enough to
1495        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1496        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1497        start sets both bits.
1499    4.  Added options (at build time, compile time, exec time) to change \R from
1500        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1502    5.  doc/pcresyntax.html was missing from the distribution.
1504    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1505        compatibility, even though it is no longer used.
1507    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1508        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1509        windows.h file is present (where different names are used). [This was
1510        reversed later after testing - see 16 below.]
1512    8.  Changed all #include <config.h> to #include "config.h". There were also
1513        some further <pcre.h> cases that I changed to "pcre.h".
1515    9.  When pcregrep was used with the --colour option, it missed the line ending
1516        sequence off the lines that it output.
1518    10. It was pointed out to me that arrays of string pointers cause lots of
1519        relocations when a shared library is dynamically loaded. A technique of
1520        using a single long string with a table of offsets can drastically reduce
1521        these. I have refactored PCRE in four places to do this. The result is
1522        dramatic:
1524          Originally:                          290
1525          After changing UCP table:            187
1526          After changing error message table:   43
1527          After changing table of "verbs"       36
1528          After changing table of Posix names   22
1530        Thanks to the folks working on Gregex for glib for this insight.
1532    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1533        unicode-properties was also set.
1535    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1537    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1538        checked only for CRLF.
1540    14. Added casts to pcretest.c to avoid compiler warnings.
1542    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1544    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1545        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1546        entirely. This removes changes made in 7 above.
1548    17. The CMake files have been updated, and there is now more information about
1549        building with CMake in the NON-UNIX-USE document.
1552    Version 7.3 28-Aug-07
1553    ---------------------
1555     1. In the rejigging of the build system that eventually resulted in 7.1, the
1556        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1557        brackets there is not right, since it causes compilers to look for an
1558        installed pcre.h, not the version that is in the source that is being
1559        compiled (which of course may be different). I have changed it back to:
1561          #include "pcre.h"
1563        I have a vague recollection that the change was concerned with compiling in
1564        different directories, but in the new build system, that is taken care of
1565        by the VPATH setting the Makefile.
1567     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1568        when the subject happened to end in the byte 0x85 (e.g. if the last
1569        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1570        characters but of course it shouldn't be taken as a newline when it is part
1571        of another character. The bug was that, for an unlimited repeat of . in
1572        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1573        characters when looking for a newline.
1575     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1577     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1578        in debug output.
1580     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1581        long printing in the pcrecpp unittest when running under MinGW.
1583     6. ESC_K was left out of the EBCDIC table.
1585     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1586        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1587        limit also applies to "virtual nesting" when a pattern is recursive, and in
1588        this case 1000 isn't so big. I have been able to remove this limit at the
1589        expense of backing off one optimization in certain circumstances. Normally,
1590        when pcre_exec() would call its internal match() function recursively and
1591        immediately return the result unconditionally, it uses a "tail recursion"
1592        feature to save stack. However, when a subpattern that can match an empty
1593        string has an unlimited repetition quantifier, it no longer makes this
1594        optimization. That gives it a stack frame in which to save the data for
1595        checking that an empty string has been matched. Previously this was taken
1596        from the 1000-entry workspace that had been reserved. So now there is no
1597        explicit limit, but more stack is used.
1599     8. Applied Daniel's patches to solve problems with the import/export magic
1600        syntax that is required for Windows, and which was going wrong for the
1601        pcreposix and pcrecpp parts of the library. These were overlooked when this
1602        problem was solved for the main library.
1604     9. There were some crude static tests to avoid integer overflow when computing
1605        the size of patterns that contain repeated groups with explicit upper
1606        limits. As the maximum quantifier is 65535, the maximum group length was
1607        set at 30,000 so that the product of these two numbers did not overflow a
1608        32-bit integer. However, it turns out that people want to use groups that
1609        are longer than 30,000 bytes (though not repeat them that many times).
1610        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1611        made it possible to implement the integer overflow checks in a much more
1612        dynamic way, which I have now done. The artificial limitation on group
1613        length has been removed - we now have only the limit on the total length of
1614        the compiled pattern, which depends on the LINK_SIZE setting.
1616    10. Fixed a bug in the documentation for get/copy named substring when
1617        duplicate names are permitted. If none of the named substrings are set, the
1618        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1619        empty string.
1621    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1622        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1623        because the ] is interpreted as the first data character and the
1624        terminating ] is not found. PCRE has been made compatible with Perl in this
1625        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1626        cause memory overwriting.
1628    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1629        string has been matched (to stop an infinite loop). It was not recognizing
1630        a conditional subpattern that could match an empty string if that
1631        subpattern was within another subpattern. For example, it looped when
1632        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1633        condition was not nested. This bug has been fixed.
1635    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1636        past the start of the subject in the presence of bytes with the top bit
1637        set, for example "\x8aBCD".
1639    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1640        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1642    14. Optimized (?!) to (*FAIL).
1644    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1645        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1646        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1647        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1648        does: it's just the validity check that is more restrictive.
1650    16. Inserted checks for integer overflows during escape sequence (backslash)
1651        processing, and also fixed erroneous offset values for syntax errors during
1652        backslash processing.
1654    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1655        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1657    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1658        caused an overrun.
1660    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1661        something other than just ASCII characters) inside a group that had an
1662        unlimited repeat caused a loop at compile time (while checking to see
1663        whether the group could match an empty string).
1665    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1666        [\P{Any}] did so. (Error in the code for printing property names.)
1668    21. An orphan \E inside a character class could cause a crash.
1670    22. A repeated capturing bracket such as (A)? could cause a wild memory
1671        reference during compilation.
1673    23. There are several functions in pcre_compile() that scan along a compiled
1674        expression for various reasons (e.g. to see if it's fixed length for look
1675        behind). There were bugs in these functions when a repeated \p or \P was
1676        present in the pattern. These operators have additional parameters compared
1677        with \d, etc, and these were not being taken into account when moving along
1678        the compiled data. Specifically:
1680        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1681            length.
1683        (b) An item such as \pL+ within a repeated group could cause crashes or
1684            loops.
1686        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1687            "reference to non-existent subpattern" error.
1689        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1691    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1692        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1694    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1696    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1697        character were causing crashes (broken optimization).
1699    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1700        \p or \P) caused a compile-time loop.
1702    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1703        break. For example, the unstudied pattern [\r\n]A does not match the string
1704        "\r\nA" because change 7.0/46 below moves the current point on by two
1705        characters after failing to match at the start. However, the pattern \nA
1706        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1707        the same is true. There doesn't seem any very clean way out of this, but
1708        what I have chosen to do makes the common cases work: PCRE now takes note
1709        of whether there can be an explicit match for \r or \n anywhere in the
1710        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1711        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1712        pattern has explicit CR or LF references.
1714    29. Added (*CR) etc for changing newline setting at start of pattern.
1717    Version 7.2 19-Jun-07
1718    ---------------------
1720     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1721        which is apparently normally available under Windows.
1723     2. Re-jig the pcregrep tests with different newline settings in an attempt
1724        to make them independent of the local environment's newline setting.
1726     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1728     4. Some of the "internals" tests were previously cut out when the link size
1729        was not 2, because the output contained actual offsets. The recent new
1730        "Z" feature of pcretest means that these can be cut out, making the tests
1731        usable with all link sizes.
1733     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1734        stack recursion. This gives a massive performance boost under BSD, but just
1735        a small improvement under Linux. However, it saves one field in the frame
1736        in all cases.
1738     6. Added more features from the forthcoming Perl 5.10:
1740        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1741            recursion call. It refers to the nth most recently opened parentheses.
1743        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1744            to be opened parentheses.
1746        (c) Conditions that refer to capturing parentheses can be specified
1747            relatively, for example, (?(-2)... or (?(+3)...
1749        (d) \K resets the start of the current match so that everything before
1750            is not part of it.
1752        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1754        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1755            reference syntax.
1757        (g) (?| introduces a group in which the numbering of parentheses in each
1758            alternative starts with the same number.
1760        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1762     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1765     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1766        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1767        for detecting groups that can match an empty string.
1769     9. A pattern with a very large number of alternatives (more than several
1770        hundred) was running out of internal workspace during the pre-compile
1771        phase, where pcre_compile() figures out how much memory will be needed. A
1772        bit of new cunning has reduced the workspace needed for groups with
1773        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1774        workspace instead of running out of the 4096 that are available.
1776    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1778    11. Applied patch from Google to remove an optimization that didn't quite work.
1779        The report of the bug said:
1781          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1782          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1783          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1785    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1786        it matched the wrong number of bytes.
1789    Version 7.1 24-Apr-07
1790    ---------------------
1792     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1793        that is more "standard", making use of automake and other Autotools. There
1794        is some re-arrangement of the files and adjustment of comments consequent
1795        on this.
1797     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1798        for recursive directory scanning broke on some systems because the files
1799        are not scanned in any specific order and on different systems the order
1800        was different. A call to "sort" has been inserted into RunGrepTest for the
1801        approprate test as a short-term fix. In the longer term there may be an
1802        alternative.
1804     3. I had an email from Eric Raymond about problems translating some of PCRE's
1805        man pages to HTML (despite the fact that I distribute HTML pages, some
1806        people do their own conversions for various reasons). The problems
1807        concerned the use of low-level troff macros .br and .in. I have therefore
1808        removed all such uses from the man pages (some were redundant, some could
1809        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1810        HTML has been updated to handle .nf/.fi and to complain if it encounters
1811        .br or .in.
1813     4. Updated comments in configure.ac that get placed in config.h.in and also
1814        arranged for config.h to be included in the distribution, with the name
1815        config.h.generic, for the benefit of those who have to compile without
1816        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1818     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1819        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1820        makevp.bat for the current PCRE, using the additional files
1821        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1823     6. A Windows user reported a minor discrepancy with test 2, which turned out
1824        to be caused by a trailing space on an input line that had got lost in his
1825        copy. The trailing space was an accident, so I've just removed it.
1827     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1828        that is needed.
1830     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1831        as "const" (a) because they are and (b) because it helps the PHP
1832        maintainers who have recently made a script to detect big data structures
1833        in the php code that should be moved to the .rodata section. I remembered
1834        to update Builducptable as well, so it won't revert if ucptable.h is ever
1835        re-created.
1837     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1838        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1839        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1840        support is not required. This saves 1.5-2K of code, which is important in
1841        some applications.
1843        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1844        so as not to refer to the tables, even though these functions will never be
1845        called when UTF-8 support is disabled. Otherwise there are problems with a
1846        shared library.
1848    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1850        (a) It was defining its arguments as char * instead of void *.
1852        (b) It was assuming that all moves were upwards in memory; this was true
1853            a long time ago when I wrote it, but is no longer the case.
1855        The emulated memove() is provided for those environments that have neither
1856        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1857        is clearly not the case, as these two bugs were recently reported.
1859    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1860        and Detrail to create the HTML documentation, the .txt form of the man
1861        pages, and it removes trailing spaces from listed files. It also creates
1862        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1863        case, it wraps all the #defines with #ifndefs. This script should be run
1864        before "make dist".
1866    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1867        with Unicode property support.
1869        (a) For a maximizing quantifier, if the two different cases of the
1870            character were of different lengths in their UTF-8 codings (there are
1871            some cases like this - I found 11), and the matching function had to
1872            back up over a mixture of the two cases, it incorrectly assumed they
1873            were both the same length.
1875        (b) When PCRE was configured to use the heap rather than the stack for
1876            recursion during matching, it was not correctly preserving the data for
1877            the other case of a UTF-8 character when checking ahead for a match
1878            while processing a minimizing repeat. If the check also involved
1879            matching a wide character, but failed, corruption could cause an
1880            erroneous result when trying to check for a repeat of the original
1881            character.
1883    13. Some tidying changes to the testing mechanism:
1885        (a) The RunTest script now detects the internal link size and whether there
1886            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1887            values substituted by "configure". (The RunGrepTest script already did
1888            this for UTF-8.) The configure.ac script no longer substitutes the
1889            relevant variables.
1891        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1892            with length and offset values. This means that the output is different
1893            for different internal link sizes. Test 2 is skipped for link sizes
1894            other than 2 because of this, bypassing the problem. Unfortunately,
1895            there was also a test in test 3 (the locale tests) that used /B and
1896            failed for link sizes other than 2. Rather than cut the whole test out,
1897            I have added a new /Z option to pcretest that replaces the length and
1898            offset values with spaces. This is now used to make test 3 independent
1899            of link size. (Test 2 will be tidied up later.)
1901    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1902        segmentation fault instead of returning the appropriate error message.
1904    15. In multiline mode when the newline sequence was set to "any", the pattern
1905        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1906        This doesn't seem right; it now treats the CRLF combination as the line
1907        ending, and so does not match in that case. It's only a pattern such as ^$
1908        that would hit this one: something like ^ABC$ would have failed after \r
1909        and then tried again after \r\n.
1911    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1912        in an attempt to make files that differ only in their line terminators
1913        compare equal. This works on Linux.
1915    17. Under certain error circumstances pcregrep might try to free random memory
1916        as it exited. This is now fixed, thanks to valgrind.
1918    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1919        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1920        was because its rules for how to advance for /g after matching an empty
1921        string at the end of a line did not allow for this case. They now check for
1922        it specially.
1924    20. pcretest is supposed to handle patterns and data of any length, by
1925        extending its buffers when necessary. It was getting this wrong when the
1926        buffer for a data line had to be extended.
1928    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1929        CRLF as a newline sequence.
1931    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1932        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1933        I have nevertheless tidied it up.
1935    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1937    24. Added a man page for pcre-config.
1940    Version 7.0 19-Dec-06
1941    ---------------------
1943     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1944        moving to gcc 4.1.1.
1946     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1947        sys/time.h, which is documented as needed for this function. It doesn't
1948        seem to matter on Linux, but it showed up on some releases of OS X.
1950     3. It seems that there are systems where bytes whose values are greater than
1951        127 match isprint() in the "C" locale. The "C" locale should be the
1952        default when a C program starts up. In most systems, only ASCII printing
1953        characters match isprint(). This difference caused the output from pcretest
1954        to vary, making some of the tests fail. I have changed pcretest so that:
1956        (a) When it is outputting text in the compiled version of a pattern, bytes
1957            other than 32-126 are always shown as hex escapes.
1959        (b) When it is outputting text that is a matched part of a subject string,
1960            it does the same, unless a different locale has been set for the match
1961            (using the /L modifier). In this case, it uses isprint() to decide.
1963     4. Fixed a major bug that caused incorrect computation of the amount of memory
1964        required for a compiled pattern when options that changed within the
1965        pattern affected the logic of the preliminary scan that determines the
1966        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1967        that the computed length was too small. The symptoms of this bug were
1968        either the PCRE error "internal error: code overflow" from pcre_compile(),
1969        or a glibc crash with a message such as "pcretest: free(): invalid next
1970        size (fast)". Examples of patterns that provoked this bug (shown in
1971        pcretest format) are:
1973          /(?-x: )/x
1974          /(?x)(?-x: \s*#\s*)/
1975          /((?i)[\x{c0}])/8
1976          /(?i:[\x{c0}])/8
1978        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1979        is now done differently.
1981     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1982        wrapper classes; (b) implement a new function in the C++ scanner that is
1983        more efficient than the old way of doing things because it avoids levels of
1984        recursion in the regex matching; (c) add a paragraph to the documentation
1985        for the FullMatch() function.
1987     6. The escape sequence \n was being treated as whatever was defined as
1988        "newline". Not only was this contrary to the documentation, which states
1989        that \n is character 10 (hex 0A), but it also went horribly wrong when
1990        "newline" was defined as CRLF. This has been fixed.
1992     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1993        was being set to -1 for the "end of line" case (supposedly a value that no
1994        character can have). Though this value is never used (the check for end of
1995        line is "zero bytes in current character"), it caused compiler complaints.
1996        I've changed it to 0xffffffff.
1998     8. In pcre_version.c, the version string was being built by a sequence of
1999        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2000        string (as it is for production releases) called a macro with an empty
2001        argument. The C standard says the result of this is undefined. The gcc
2002        compiler treats it as an empty string (which was what was wanted) but it is
2003        reported that Visual C gives an error. The source has been hacked around to
2004        avoid this problem.
2006     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2007        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2008        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2009        of them did).
2011    10. Originally, pcretest opened its input and output without "b"; then I was
2012        told that "b" was needed in some environments, so it was added for release
2013        5.0 to both the input and output. (It makes no difference on Unix-like
2014        systems.) Later I was told that it is wrong for the input on Windows. I've
2015        now abstracted the modes into two macros, to make it easier to fiddle with
2016        them, and removed "b" from the input mode under Windows.
2018    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2020    12. Added -help and --help to pcretest as an official way of being reminded
2021        of the options.
2023    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2024        and pcrecpp.cc because they annoy compilers at high warning levels.
2026    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2028    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2029        scripts are not C programs :-) and which was not noticed because it works
2030        on Linux.
2032    16. pcretest is supposed to handle any length of pattern and data line (as one
2033        line or as a continued sequence of lines) by extending its input buffer if
2034        necessary. This feature was broken for very long pattern lines, leading to
2035        a string of junk being passed to pcre_compile() if the pattern was longer
2036        than about 50K.
2038    17. I have done a major re-factoring of the way pcre_compile() computes the
2039        amount of memory needed for a compiled pattern. Previously, there was code
2040        that made a preliminary scan of the pattern in order to do this. That was
2041        OK when PCRE was new, but as the facilities have expanded, it has become
2042        harder and harder to keep it in step with the real compile phase, and there
2043        have been a number of bugs (see for example, 4 above). I have now found a
2044        cunning way of running the real compile function in a "fake" mode that
2045        enables it to compute how much memory it would need, while actually only
2046        ever using a few hundred bytes of working memory and without too many
2047        tests of the mode. This should make future maintenance and development
2048        easier. A side effect of this work is that the limit of 200 on the nesting
2049        depth of parentheses has been removed (though this was never a serious
2050        limitation, I suspect). However, there is a downside: pcre_compile() now
2051        runs more slowly than before (30% or more, depending on the pattern). I
2052        hope this isn't a big issue. There is no effect on runtime performance.
2054    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2055        newline (only possible for the last line of a file) and it was a
2056        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2058    19. Added additional timing features to pcretest. (1) The -tm option now times
2059        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2060        separate command line item, by a number that specifies the number of
2061        repeats to use when timing. The default is 50000; this gives better
2062        precision, but takes uncomfortably long for very large patterns.
2064    20. Extended pcre_study() to be more clever in cases where a branch of a
2065        subpattern has no definite first character. For example, (a*|b*)[cd] would
2066        previously give no result from pcre_study(). Now it recognizes that the
2067        first character must be a, b, c, or d.
2069    21. There was an incorrect error "recursive call could loop indefinitely" if
2070        a subpattern (or the entire pattern) that was being tested for matching an
2071        empty string contained only one non-empty item after a nested subpattern.
2072        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2073        incorrectly, because the \d was being skipped in the check.
2075    22. The pcretest program now has a new pattern option /B and a command line
2076        option -b, which is equivalent to adding /B to every pattern. This causes
2077        it to show the compiled bytecode, without the additional information that
2078        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2079        is the same as /B/I).
2081    23. A new optimization is now able automatically to treat some sequences such
2082        as a*b as a*+b. More specifically, if something simple (such as a character
2083        or a simple class like \d) has an unlimited quantifier, and is followed by
2084        something that cannot possibly match the quantified thing, the quantifier
2085        is automatically "possessified".
2087    24. A recursive reference to a subpattern whose number was greater than 39
2088        went wrong under certain circumstances in UTF-8 mode. This bug could also
2089        have affected the operation of pcre_study().
2091    25. Realized that a little bit of performance could be had by replacing
2092        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2094    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2096    27. Possessive quantifiers such as a++ were previously implemented by turning
2097        them into atomic groups such as ($>a+). Now they have their own opcodes,
2098        which improves performance. This includes the automatically created ones
2099        from 23 above.
2101    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2102        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2103        the first matched character to be a colon. This applied both to named and
2104        numbered groups.
2106    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2108    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2109        building PCRE on Windows easier, so I have included it in the distribution.
2111    31. There is now a check in pcretest against a ridiculously large number being
2112        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2113        loop, the loop is abandoned.
2115    32. Forward references to subpatterns in conditions such as (?(2)...) where
2116        subpattern 2 is defined later cause pcre_compile() to search forwards in
2117        the pattern for the relevant set of parentheses. This search went wrong
2118        when there were unescaped parentheses in a character class, parentheses
2119        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2121    33. "Subroutine" calls and backreferences were previously restricted to
2122        referencing subpatterns earlier in the regex. This restriction has now
2123        been removed.
2125    34. Added a number of extra features that are going to be in Perl 5.10. On the
2126        whole, these are just syntactic alternatives for features that PCRE had
2127        previously implemented using the Python syntax or my own invention. The
2128        other formats are all retained for compatibility.
2130        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2131            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2132            also .NET compatible.
2134        (b) A recursion or subroutine call to a named group can now be defined as
2135            (?&name) as well as (?P>name).
2137        (c) A backreference to a named group can now be defined as \k<name> or
2138            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2139            5.10, are also .NET compatible.
2141        (d) A conditional reference to a named group can now use the syntax
2142            (?(<name>) or (?('name') as well as (?(name).
2144        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2145            groups (named and numbered) that are never evaluated inline, but can be
2146            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2147            is always false. There may be only one alternative in such a group.
2149        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2150            as the simple (?(R). The condition is true only if the most recent
2151            recursion is that of the given number or name. It does not search out
2152            through the entire recursion stack.
2154        (g) The escape \gN or \g{N} has been added, where N is a positive or
2155            negative number, specifying an absolute or relative reference.
2157    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2158        some "unreachable code" warnings.
2160    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2161        things, this adds five new scripts.
2163    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2164        There were also incompatibilities regarding the handling of \Q..\E inside
2165        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2166        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2168    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2169        matches an empty string, and forcibly breaks the loop. There were bugs in
2170        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2171        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2172        separate and independent bugs (that affected different cases) have been
2173        fixed.
2175    39. Refactored the code to abolish the use of different opcodes for small
2176        capturing bracket numbers. This is a tidy that I avoided doing when I
2177        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2178        The new approach is not only tidier, it makes it possible to reduce the
2179        memory needed to fix the previous bug (38).
2181    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2182        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2183        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2184        mode.
2186    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2187        report.
2189    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2190        copy construction and assignment in the C++ wrapper.
2192    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2193        couple of bugs that could have given wrong results in the "--newline=crlf"
2194        case.
2196    44. Added a number of casts and did some reorganization of signed/unsigned int
2197        variables following suggestions from Dair Grant. Also renamed the variable
2198        "this" as "item" because it is a C++ keyword.
2200    45. Arranged for dftables to add
2202          #include "pcre_internal.h"
2204        to pcre_chartables.c because without it, gcc 4.x may remove the array
2205        definition from the final binary if PCRE is built into a static library and
2206        dead code stripping is activated.
2208    46. For an unanchored pattern, if a match attempt fails at the start of a
2209        newline sequence, and the newline setting is CRLF or ANY, and the next two
2210        characters are CRLF, advance by two characters instead of one.
2213    Version 6.7 04-Jul-06
2214    ---------------------
2216     1. In order to handle tests when input lines are enormously long, pcretest has
2217        been re-factored so that it automatically extends its buffers when
2218        necessary. The code is crude, but this _is_ just a test program. The
2219        default size has been increased from 32K to 50K.
2221     2. The code in pcre_study() was using the value of the re argument before
2222        testing it for NULL. (Of course, in any sensible call of the function, it
2223        won't be NULL.)
2225     3. The memmove() emulation function in pcre_internal.h, which is used on
2226        systems that lack both memmove() and bcopy() - that is, hardly ever -
2227        was missing a "static" storage class specifier.
2229     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2230        containing an extended class (one that cannot be represented by a bitmap
2231        because it contains high-valued characters or Unicode property items, e.g.
2232        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2233        pattern, but PCRE should not loop if you do not (it no longer does).
2234        [Detail: two cases were found: (a) a repeated subpattern containing an
2235        extended class; (b) a recursive reference to a subpattern that followed a
2236        previous extended class. It wasn't skipping over the extended class
2237        correctly when UTF-8 mode was not set.]
2239     5. A negated single-character class was not being recognized as fixed-length
2240        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2241        compile error "lookbehind assertion is not fixed length".
2243     6. The RunPerlTest auxiliary script was showing an unexpected difference
2244        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2245        write a Perl script that can interpret lines of an input file either as
2246        byte characters or as UTF-8, which is what "perltest" was being required to
2247        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2248        can't do is switch easily at run time between having the "use utf8;" pragma
2249        or not. In the end, I fudged it by using the RunPerlTest script to insert
2250        "use utf8;" explicitly for the UTF-8 tests.
2252     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2253        the end of the subject string, contrary to the documentation and to what
2254        Perl does. This was true of both matching functions. Now it matches only at
2255        the start of the subject and immediately after *internal* newlines.
2257     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2258        a pointer to an int instead of a pointer to an unsigned long int. This
2259        caused problems on 64-bit systems.
2261     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2262        instance of the 'standard' template library not being so standard".
2264    10. There was no check on the number of named subpatterns nor the maximum
2265        length of a subpattern name. The product of these values is used to compute
2266        the size of the memory block for a compiled pattern. By supplying a very
2267        long subpattern name and a large number of named subpatterns, the size
2268        computation could be caused to overflow. This is now prevented by limiting
2269        the length of names to 32 characters, and the number of named subpatterns
2270        to 10,000.
2272    11. Subpatterns that are repeated with specific counts have to be replicated in
2273        the compiled pattern. The size of memory for this was computed from the
2274        length of the subpattern and the repeat count. The latter is limited to
2275        65535, but there was no limit on the former, meaning that integer overflow
2276        could in principle occur. The compiled length of a repeated subpattern is
2277        now limited to 30,000 bytes in order to prevent this.
2279    12. Added the optional facility to have named substrings with the same name.
2281    13. Added the ability to use a named substring as a condition, using the
2282        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2283        are numbers (not recommended). Forward references are permitted.
2285    14. Added forward references in named backreferences (if you see what I mean).
2287    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2288        pattern could run off the end of the subject. For example, the pattern
2289        "(?s)(.{1,5})"8 did this with the subject "ab".
2291    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2292        PCRE_CASELESS was set when matching characters that were quantified with ?
2293        or *.
2295    17. A character class other than a single negated character that had a minimum
2296        but no maximum quantifier - for example [ab]{6,} - was not handled
2297        correctly by pce_dfa_exec(). It would match only one character.
2299    18. A valid (though odd) pattern that looked like a POSIX character
2300        class but used an invalid character after [ (for example [[,abc,]]) caused
2301        pcre_compile() to give the error "Failed: internal error: code overflow" or
2302        in some cases to crash with a glibc free() error. This could even happen if
2303        the pattern terminated after [[ but there just happened to be a sequence of
2304        letters, a binary zero, and a closing ] in the memory that followed.
2306    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2307        over the years. Originally (before any Unicode support), just the bottom 8
2308        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2309        output from "man perlunicode" includes this:
2311          The regular expression compiler produces polymorphic opcodes.  That
2312          is, the pattern adapts to the data and automatically switches to
2313          the Unicode character scheme when presented with Unicode data--or
2314          instead uses a traditional byte scheme when presented with byte
2315          data.
2317        Sadly, a wide octal escape does not cause a switch, and in a string with
2318        no other multibyte characters, these octal escapes are treated as before.
2319        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2320        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2321        Unicode string.
2323        I have not perpetrated such confusion in PCRE. Up till now, it took just
2324        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2325        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2326        translate to the appropriate multibyte character.
2328    29. Applied some refactoring to reduce the number of warnings from Microsoft
2329        and Borland compilers. This has included removing the fudge introduced
2330        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2331        a warning about an unused variable.
2333    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2334        characters since release 4.0, because Perl (from release 5.004) does not.
2335        [Or at least, is documented not to: some releases seem to be in conflict
2336        with the documentation.] However, when a pattern was studied with
2337        pcre_study() and all its branches started with \s, PCRE still included VT
2338        as a possible starting character. Of course, this did no harm; it just
2339        caused an unnecessary match attempt.
2341    22. Removed a now-redundant internal flag bit that recorded the fact that case
2342        dependency changed within the pattern. This was once needed for "required
2343        byte" processing, but is no longer used. This recovers a now-scarce options
2344        bit. Also moved the least significant internal flag bit to the most-
2345        significant bit of the word, which was not previously used (hangover from
2346        the days when it was an int rather than a uint) to free up another bit for
2347        the future.
2349    23. Added support for CRLF line endings as well as CR and LF. As well as the
2350        default being selectable at build time, it can now be changed at runtime
2351        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2352        specify that it is scanning data with non-default line endings.
2354    24. Changed the definition of CXXLINK to make it agree with the definition of
2355        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2357    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2358        recursions. This makes a big different to stack usage for some patterns.
2360    26. If a subpattern containing a named recursion or subroutine reference such
2361        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2362        the space required for the compiled pattern went wrong and gave too small a
2363        value. Depending on the environment, this could lead to "Failed: internal
2364        error: code overflow at offset 49" or "glibc detected double free or
2365        corruption" errors.
2367    27. Applied patches from Google (a) to support the new newline modes and (b) to
2368        advance over multibyte UTF-8 characters in GlobalReplace.
2370    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2371        difference for some implementation of PCRE in some Windows version.
2373    29. Added some extra testing facilities to pcretest:
2375        \q<number>   in a data line sets the "match limit" value
2376        \Q<number>   in a data line sets the "match recursion limt" value
2377        -S <number>  sets the stack size, where <number> is in megabytes
2379        The -S option isn't available for Windows.
2382    Version 6.6 06-Feb-06
2383    ---------------------
2385     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2386        in pcreposix.h. I have copied the definition from pcre.h.
2388     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2389        because pcre.h is no longer a built file.
2391     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2392        not normally included in the compiled code.
2395    Version 6.5 01-Feb-06
2396    ---------------------
2398     1. When using the partial match feature with pcre_dfa_exec(), it was not
2399        anchoring the second and subsequent partial matches at the new starting
2400        point. This could lead to incorrect results. For example, with the pattern
2401        /1234/, partially matching against "123" and then "a4" gave a match.
2403     2. Changes to pcregrep:
2405        (a) All non-match returns from pcre_exec() were being treated as failures
2406            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2407            error message is output. Some extra information is given for the
2408            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2409            probably the only errors that are likely to be caused by users (by
2410            specifying a regex that has nested indefinite repeats, for instance).
2411            If there are more than 20 of these errors, pcregrep is abandoned.
2413        (b) A binary zero was treated as data while matching, but terminated the
2414            output line if it was written out. This has been fixed: binary zeroes
2415            are now no different to any other data bytes.
2417        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2418            used to set a locale for matching. The --locale=xxxx long option has
2419            been added (no short equivalent) to specify a locale explicitly on the
2420            pcregrep command, overriding the environment variables.
2422        (d) When -B was used with -n, some line numbers in the output were one less
2423            than they should have been.
2425        (e) Added the -o (--only-matching) option.
2427        (f) If -A or -C was used with -c (count only), some lines of context were
2428            accidentally printed for the final match.
2430        (g) Added the -H (--with-filename) option.
2432        (h) The combination of options -rh failed to suppress file names for files
2433            that were found from directory arguments.
2435        (i) Added the -D (--devices) and -d (--directories) options.
2437        (j) Added the -F (--fixed-strings) option.
2439        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2441        (l) Added the --colo(u)r option.
2443        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2444            is not present by default.
2446     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2447        items such as (?R) or (?1), when the recursion could match a number of
2448        alternatives. If it matched one of the alternatives, but subsequently,
2449        outside the recursion, there was a failure, the code tried to back up into
2450        the recursion. However, because of the way PCRE is implemented, this is not
2451        possible, and the result was an incorrect result from the match.
2453        In order to prevent this happening, the specification of recursion has
2454        been changed so that all such subpatterns are automatically treated as
2455        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2457     4. I had overlooked the fact that, in some locales, there are characters for
2458        which isalpha() is true but neither isupper() nor islower() are true. In
2459        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2460        and ordfeminine) are like this. This affected the treatment of \w and \W
2461        when they appeared in character classes, but not when they appeared outside
2462        a character class. The bit map for "word" characters is now created
2463        separately from the results of isalnum() instead of just taking it from the
2464        upper, lower, and digit maps. (Plus the underscore character, of course.)
2466     5. The above bug also affected the handling of POSIX character classes such as
2467        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2468        permanent tables. Instead, the bit maps for such a class were previously
2469        created as the appropriate unions of the upper, lower, and digit bitmaps.
2470        Now they are created by subtraction from the [[:word:]] class, which has
2471        its own bitmap.
2473     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2474        It is created by subtracting the vertical space characters (\x09, \x0a,
2475        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2476        subtraction was done in the overall bitmap for a character class, meaning
2477        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2478        be recognized. This bug has been fixed.
2480     7. Patches from the folks at Google:
2482          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2483          real life, but is still worth protecting against".
2485          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2486          regular expressions".
2488          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2489          have it.
2491          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2492          "configure" and the latter not, in order to fix a problem somebody had
2493          with compiling the Arg class on HP-UX.
2495          (e) Improve the error-handling of the C++ wrapper a little bit.
2497          (f) New tests for checking recursion limiting.
2499     8. The pcre_memmove() function, which is used only if the environment does not
2500        have a standard memmove() function (and is therefore rarely compiled),
2501        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2502        returning a result (though PCRE never actually uses the result).
2504     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2505        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2506        returned instead of calling malloc() with an overflowing number that would
2507        most likely cause subsequent chaos.
2509    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2511    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2512        with this option is matched, the nmatch and pmatch options of regexec() are
2513        ignored.
2515    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2516        provided in case anyone wants to the the POSIX interface with UTF-8
2517        strings.
2519    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2520        C++ linking (needed for some HP-UX environments).
2522    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2523        (unused parameter) and in the pcre_printint() function (omitted "default"
2524        switch label when the default is to do nothing).
2526    15. Added some code to make it possible, when PCRE is compiled as a C++
2527        library, to replace subject pointers for pcre_exec() with a smart pointer
2528        class, thus making it possible to process discontinuous strings.
2530    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2531        much the same function. They were added by different people who were trying
2532        to make PCRE easy to compile on non-Unix systems. It has been suggested
2533        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2534        for compiling on Windows systems. I have therefore replaced it with
2535        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2536        defaults to "extern" for C or "extern C" for C++, which works fine on
2537        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2538        SCOPE with something explicit in config.h. In addition:
2540        (a) pcreposix.h still had just "extern" instead of either of these macros;
2541            I have replaced it with PCRE_DATA_SCOPE.
2543        (b) Functions such as _pcre_xclass(), which are internal to the library,
2544            but external in the C sense, all had PCRE_EXPORT in their definitions.
2545            This is apparently wrong for the Windows case, so I have removed it.
2546            (It makes no difference on Unix-like systems.)
2548    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2549        of recursive calls to match(). This is different to MATCH_LIMIT because
2550        that limits the total number of calls to match(), not all of which increase
2551        the depth of recursion. Limiting the recursion depth limits the amount of
2552        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2553        when PCRE is compiled, and changed at run time. A patch from Google adds
2554        this functionality to the C++ interface.
2556    18. Changes to the handling of Unicode character properties:
2558        (a) Updated the table to Unicode 4.1.0.
2560        (b) Recognize characters that are not in the table as "Cn" (undefined).
2562        (c) I revised the way the table is implemented to a much improved format
2563            which includes recognition of ranges. It now supports the ranges that
2564            are defined in UnicodeData.txt, and it also amalgamates other
2565            characters into ranges. This has reduced the number of entries in the
2566            table from around 16,000 to around 3,000, thus reducing its size
2567            considerably. I realized I did not need to use a tree structure after
2568            all - a binary chop search is just as efficient. Having reduced the
2569            number of entries, I extended their size from 6 bytes to 8 bytes to
2570            allow for more data.
2572        (d) Added support for Unicode script names via properties such as \p{Han}.
2574    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2575        matching that character.
2577    20. When matching a repeated Unicode property with a minimum greater than zero,
2578        (for example \pL{2,}), PCRE could look past the end of the subject if it
2579        reached it while seeking the minimum number of characters. This could
2580        happen only if some of the characters were more than one byte long, because
2581        there is a check for at least the minimum number of bytes.
2583    21. Refactored the implementation of \p and \P so as to be more general, to
2584        allow for more different types of property in future. This has changed the
2585        compiled form incompatibly. Anybody with saved compiled patterns that use
2586        \p or \P will have to recompile them.
2588    22. Added "Any" and "L&" to the supported property types.
2590    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2591        but give a compile time error if the value is greater than 0xff.
2593    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2594        accidentally not being installed or uninstalled.
2596    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2597        made were to insert the current release number. This seemed silly, because
2598        it made things harder for people building PCRE on systems that don't run
2599        "configure". I have turned pcre.h into a distributed file, no longer built
2600        by "configure", with the version identification directly included. There is
2601        no longer a pcre.h.in file.
2603        However, this change necessitated a change to the pcre-config script as
2604        well. It is built from pcre-config.in, and one of the substitutions was the
2605        release number. I have updated configure.ac so that ./configure now finds
2606        the release number by grepping pcre.h.
2608    26. Added the ability to run the tests under valgrind.
2611    Version 6.4 05-Sep-05
2612    ---------------------
2614     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2615        "--" to be printed when multiple files were scanned, even when none of the
2616        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2617        consider it to be a bug, and have restored the previous behaviour.
2619     2. A couple of code tidies to get rid of compiler warnings.
2621     3. The pcretest program used to cheat by referring to symbols in the library
2622        whose names begin with _pcre_. These are internal symbols that are not
2623        really supposed to be visible externally, and in some environments it is
2624        possible to suppress them. The cheating is now confined to including
2625        certain files from the library's source, which is a bit cleaner.
2627     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2628        file's purpose clearer.
2630     5. Reorganized pcre_ucp_findchar().
2633    Version 6.3 15-Aug-05
2634    ---------------------
2636     1. The file libpcre.pc.in did not have general read permission in the tarball.
2638     2. There were some problems when building without C++ support:
2640        (a) If C++ support was not built, "make install" and "make test" still
2641            tried to test it.
2643        (b) There were problems when the value of CXX was explicitly set. Some
2644            changes have been made to try to fix these, and ...
2646        (c) --disable-cpp can now be used to explicitly disable C++ support.
2648        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2649            backslash in a target when C++ was disabled. This confuses some
2650            versions of "make", apparently. Using an intermediate variable solves
2651            this. (Same for CPP_LOBJ.)
2653     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2654        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2655        necessary on certain architectures.
2657     4. Added a setting of -export-symbols-regex to the link command to remove
2658        those symbols that are exported in the C sense, but actually are local
2659        within the library, and not documented. Their names all begin with
2660        "_pcre_". This is not a perfect job, because (a) we have to except some
2661        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2662        available (and never for static libraries). I have made a note to try to
2663        find a way round (a) in the future.
2666    Version 6.2 01-Aug-05
2667    ---------------------
2669     1. There was no test for integer overflow of quantifier values. A construction
2670        such as {1111111111111111} would give undefined results. What is worse, if
2671        a minimum quantifier for a parenthesized subpattern overflowed and became
2672        negative, the calculation of the memory size went wrong. This could have
2673        led to memory overwriting.
2675     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2677     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2678        operating environments where this matters.
2680     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2681        PCRE options from within the C++ wrapper.
2683     5. Named capturing subpatterns were not being correctly counted when a pattern
2684        was compiled. This caused two problems: (a) If there were more than 100
2685        such subpatterns, the calculation of the memory needed for the whole
2686        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2687        back references of the form \12, where the number was greater than 9, were
2688        not recognized as back references, even though there were sufficient
2689        previous subpatterns.
2691     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2692        versions of gcc, e.g. 2.95.4.
2695    Version 6.1 21-Jun-05
2696    ---------------------
2698     1. There was one reference to the variable "posix" in pcretest.c that was not
2699        surrounded by "#if !defined NOPOSIX".
2701     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2702        the cross-check on the old pcre_info() function, for the benefit of the
2703        cut-down version of PCRE that is currently imported into Exim.
2705     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2706        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2707        patterns that start (?i) but I don't think that matters. The use of (?i) is
2708        just an example; this all applies to the other options as well.
2710     4. Since libtool seems to echo the compile commands it is issuing, the output
2711        from "make" can be reduced a bit by putting "@" in front of each libtool
2712        compile command.
2714     5. Patch from the folks at Google for configure.in to be a bit more thorough
2715        in checking for a suitable C++ installation before trying to compile the
2716        C++ stuff. This should fix a reported problem when a compiler was present,
2717        but no suitable headers.
2719     6. The man pages all had just "PCRE" as their title. I have changed them to
2720        be the relevant file name. I have also arranged that these names are
2721        retained in the file doc/pcre.txt, which is a concatenation in text format
2722        of all the man pages except the little individual ones for each function.
2724     7. The NON-UNIX-USE file had not been updated for the different set of source
2725        files that come with release 6. I also added a few comments about the C++
2726        wrapper.
2729    Version 6.0 07-Jun-05
2730    ---------------------
2732     1. Some minor internal re-organization to help with my DFA experiments.
2734     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2735        didn't matter for the library itself when fully configured, but did matter
2736        when compiling without UCP support, or within Exim, where the ucp files are
2737        not imported.
2739     3. Refactoring of the library code to split up the various functions into
2740        different source modules. The addition of the new DFA matching code (see
2741        below) to a single monolithic source would have made it really too
2742        unwieldy, quite apart from causing all the code to be include in a
2743        statically linked application, when only some functions are used. This is
2744        relevant even without the DFA addition now that patterns can be compiled in
2745        one application and matched in another.
2747        The downside of splitting up is that there have to be some external
2748        functions and data tables that are used internally in different modules of
2749        the library but which are not part of the API. These have all had their
2750        names changed to start with "_pcre_" so that they are unlikely to clash
2751        with other external names.
2753     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2754        a different (DFA) algorithm. Although it is slower than the original
2755        function, it does have some advantages for certain types of matching
2756        problem.
2758     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2759        including restarting after a partial match.
2761     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2762        defined when compiling for Windows was sent to me. I have put it into the
2763        code, though I have no means of testing or verifying it.
2765     7. Added the pcre_refcount() auxiliary function.
2767     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2768        match before or at the first newline in the subject string. In pcretest,
2769        the /f option on a pattern can be used to set this.
2771     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2772        would behave wrongly. This has been present in PCRE since release 4.0.
2774    10. A number of changes to the pcregrep command:
2776        (a) Refactored how -x works; insert ^(...)$ instead of setting
2777            PCRE_ANCHORED and checking the length, in preparation for adding
2778            something similar for -w.
2780        (b) Added the -w (match as a word) option.
2782        (c) Refactored the way lines are read and buffered so as to have more
2783            than one at a time available.
2785        (d) Implemented a pcregrep test script.
2787        (e) Added the -M (multiline match) option. This allows patterns to match
2788            over several lines of the subject. The buffering ensures that at least
2789            8K, or the rest of the document (whichever is the shorter) is available
2790            for matching (and similarly the previous 8K for lookbehind assertions).
2792        (f) Changed the --help output so that it now says
2794              -w, --word-regex(p)
2796            instead of two lines, one with "regex" and the other with "regexp"
2797            because that confused at least one person since the short forms are the
2798            same. (This required a bit of code, as the output is generated
2799            automatically from a table. It wasn't just a text change.)
2801        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2802            option but starts with a hyphen. Could be a pattern or a path name
2803            starting with a hyphen, for instance.
2805        (h) "-" can be given as a file name to represent stdin.
2807        (i) When file names are being printed, "(standard input)" is used for
2808            the standard input, for compatibility with GNU grep. Previously
2809            "<stdin>" was used.
2811        (j) The option --label=xxx can be used to supply a name to be used for
2812            stdin when file names are being printed. There is no short form.
2814        (k) Re-factored the options decoding logic because we are going to add
2815            two more options that take data. Such options can now be given in four
2816            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2818        (l) Added the -A, -B, and -C options for requesting that lines of context
2819            around matches be printed.
2821        (m) Added the -L option to print the names of files that do not contain
2822            any matching lines, that is, the complement of -l.
2824        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2825            continue to scan other files.
2827        (o) The -s option was incorrectly implemented. For compatibility with other
2828            greps, it now suppresses the error message for a non-existent or non-
2829            accessible file (but not the return code). There is a new option called
2830            -q that suppresses the output of matching lines, which was what -s was
2831            previously doing.
2833        (p) Added --include and --exclude options to specify files for inclusion
2834            and exclusion when recursing.
2836    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2837        Hopefully, it now does.
2839    12. Missing cast in pcre_study().
2841    13. Added an "uninstall" target to the makefile.
2843    14. Replaced "extern" in the function prototypes in Makefile.in with
2844        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2845        world, but is set differently for Windows.
2847    15. Added a second compiling function called pcre_compile2(). The only
2848        difference is that it has an extra argument, which is a pointer to an
2849        integer error code. When there is a compile-time failure, this is set
2850        non-zero, in addition to the error test pointer being set to point to an
2851        error message. The new argument may be NULL if no error number is required
2852        (but then you may as well call pcre_compile(), which is now just a
2853        wrapper). This facility is provided because some applications need a
2854        numeric error indication, but it has also enabled me to tidy up the way
2855        compile-time errors are handled in the POSIX wrapper.
2857    16. Added VPATH=.libs to the makefile; this should help when building with one
2858        prefix path and installing with another. (Or so I'm told by someone who
2859        knows more about this stuff than I do.)
2861    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2862        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2863        match everything, including newlines. This is not POSIX-compatible, but
2864        somebody wanted the feature. From pcretest it can be activated by using
2865        both the P and the s flags.
2867    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2869    19. libpcre.pc was being incorrectly installed as executable.
2871    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2872        it now also looks for '\r' so that it will work unmodified on Windows.
2874    21. Added Google's contributed C++ wrapper to the distribution.
2876    22. Added some untidy missing memory free() calls in pcretest, to keep
2877        Electric Fence happy when testing.
2881    Version 5.0 13-Sep-04
2882    ---------------------
2884     1. Internal change: literal characters are no longer packed up into items
2885        containing multiple characters in a single byte-string. Each character
2886        is now matched using a separate opcode. However, there may be more than one
2887        byte in the character in UTF-8 mode.
2889     2. The pcre_callout_block structure has two new fields: pattern_position and
2890        next_item_length. These contain the offset in the pattern to the next match
2891        item, and its length, respectively.
2893     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2894        insertion of callouts before each pattern item. Added the /C option to
2895        pcretest to make use of this.
2897     4. On the advice of a Windows user, the lines
2899          #if defined(_WIN32) || defined(WIN32)
2900          _setmode( _fileno( stdout ), 0x8000 );
2901          #endif  /* defined(_WIN32) || defined(WIN32) */
2903        have been added to the source of pcretest. This apparently does useful
2904        magic in relation to line terminators.
2906     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2907        for the benefit of those environments where the "b" makes a difference.
2909     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2910        to know about it. I have put a hack into configure.in that adds in code
2911        to set GCC=yes if CC=icc. This seems to end up at a point in the
2912        generated configure script that is early enough to affect the setting of
2913        compiler options, which is what is needed, but I have no means of testing
2914        whether it really works. (The user who reported this had patched the
2915        generated configure script, which of course I cannot do.)
2917        LATER: After change 22 below (new libtool files), the configure script
2918        seems to know about icc (and also ecc). Therefore, I have commented out
2919        this hack in configure.in.
2921     7. Added support for pkg-config (2 patches were sent in).
2923     8. Negated POSIX character classes that used a combination of internal tables
2924        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2925        [[:^ascii]]. Typically, they would match almost any characters. The other
2926        POSIX classes were not broken in this way.
2928     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2929        to find the match, as PCRE was deluded into thinking that the match had to
2930        start at the start point or following a newline. The same bug applied to
2931        patterns with negative forward assertions or any backward assertions
2932        preceding ".*" at the start, unless the pattern required a fixed first
2933        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2935    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2936        starting at the last subject character, bytes beyond the end of the subject
2937        string were read.
2939    11. Renamed the variable "class" as "classbits" to make life easier for C++
2940        users. (Previously there was a macro definition, but it apparently wasn't
2941        enough.)
2943    12. Added the new field "tables" to the extra data so that tables can be passed
2944        in at exec time, or the internal tables can be re-selected. This allows
2945        a compiled regex to be saved and re-used at a later time by a different
2946        program that might have everything at different addresses.
2948    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2949        -R library as well as a -L library.
2951    14. The debugging options of pcretest (-d on the command line or D on a
2952        pattern) showed incorrect output for anything following an extended class
2953        that contained multibyte characters and which was followed by a quantifier.
2955    15. Added optional support for general category Unicode character properties
2956        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2957        support. It adds about 90K to the size of the library. The meanings of the
2958        inbuilt class escapes such as \d and \s have NOT been changed.
2960    16. Updated pcredemo.c to include calls to free() to release the memory for the
2961        compiled pattern.
2963    17. The generated file chartables.c was being created in the source directory
2964        instead of in the building directory. This caused the build to fail if the
2965        source directory was different from the building directory, and was
2966        read-only.
2968    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2969        file. No doubt somebody will tell me if they don't make sense... Also added
2970        Dan Mooney's comments about building on OpenVMS.
2972    19. Added support for partial matching via the PCRE_PARTIAL option for
2973        pcre_exec() and the \P data escape in pcretest.
2975    20. Extended pcretest with 3 new pattern features:
2977        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2978              write the compiled pattern to the file whose name is "rest-of-line".
2979              This is a straight binary dump of the data, with the saved pointer to
2980              the character tables forced to be NULL. The study data, if any, is
2981              written too. After writing, pcretest reads a new pattern.
2983        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2984              compiled pattern from the given file. There must not be any
2985              occurrences of "<" in the file name (pretty unlikely); if there are,
2986              pcretest will instead treat the initial "<" as a pattern delimiter.
2987              After reading in the pattern, pcretest goes on to read data lines as
2988              usual.
2990        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2991              and 16-bit fields in a compiled pattern, to simulate a pattern that
2992              was compiled on a host of opposite endianness.
2994    21. The pcre-exec() function can now cope with patterns that were compiled on
2995        hosts of opposite endianness, with this restriction:
2997          As for any compiled expression that is saved and used later, the tables
2998          pointer field cannot be preserved; the extra_data field in the arguments
2999          to pcre_exec() should be used to pass in a tables address if a value
3000          other than the default internal tables were used at compile time.
3002    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3003        now diagnosed as an error. Previously, most of the time, a negative number
3004        would have been treated as zero, but if in addition "ovector" was passed as
3005        NULL, a crash could occur.
3007    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3008        new versions from the libtool 1.5 distribution (the last one is a copy of
3009        a file called libtool.m4). This seems to have fixed the need to patch
3010        "configure" to support Darwin 1.3 (which I used to do). However, I still
3011        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3012        workstation).
3014    24. Changed the PCRE licence to be the more standard "BSD" licence.
3017    Version 4.5 01-Dec-03
3018    ---------------------
3020     1. There has been some re-arrangement of the code for the match() function so
3021        that it can be compiled in a version that does not call itself recursively.
3022        Instead, it keeps those local variables that need separate instances for
3023        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3024        needs to "recurse". Keeping track of where control must go is done by means
3025        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3026        hide most of the details from the main code, and operates only if
3027        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3028        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3029        operating.
3031        To make it easier for callers to provide specially tailored get/free
3032        functions for this usage, two new functions, pcre_stack_malloc, and
3033        pcre_stack_free, are used. They are always called in strict stacking order,
3034        and the size of block requested is always the same.
3036        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3037        PCRE has been compiled to use the stack or the heap for recursion. The
3038        -C option of pcretest uses this to show which version is compiled.
3040        A new data escape \S, is added to pcretest; it causes the amounts of store
3041        obtained and freed by both kinds of malloc/free at match time to be added
3042        to the output.
3044     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3045        what's available on my current Linux desktop machine.
3047     3. When matching a UTF-8 string, the test for a valid string at the start has
3048        been extended. If start_offset is not zero, PCRE now checks that it points
3049        to a byte that is the start of a UTF-8 character. If not, it returns
3050        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3051        this is necessary because there may be backward assertions in the pattern.
3052        When matching the same subject several times, it may save resources to use
3053        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3055     4. The code for checking the validity of UTF-8 strings has been tightened so
3056        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3057        containing "overlong sequences".
3059     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3060        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3061        should have been "&", but it just so happened that all the cases this let
3062        through by mistake were picked up later in the function.
3064     6. I had used a variable called "isblank" - this is a C99 function, causing
3065        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3067     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3068        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3069        so the version is shown; (c) stop "make test" echoing "./RunTest".
3071     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3073     9. The prototype for memmove() for systems that don't have it was using
3074        size_t, but the inclusion of the header that defines size_t was later. I've
3075        moved the #includes for the C headers earlier to avoid this.
3077    10. Added some adjustments to the code to make it easier to compiler on certain
3078        special systems:
3080          (a) Some "const" qualifiers were missing.
3081          (b) Added the macro EXPORT before all exported functions; by default this
3082              is defined to be empty.
3083          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3084              that it reads its output file name as an argument instead of writing
3085              to the standard output and assuming this can be redirected.
3087    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3088        class containing characters with values greater than 255, PCRE compilation
3089        went into a loop.
3091    12. A recursive reference to a subpattern that was within another subpattern
3092        that had a minimum quantifier of zero caused PCRE to crash. For example,
3093        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3094        recursion. If the recursively-called subpattern itself had a zero repeat,
3095        that was OK.
3097    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3098        buffer into which it was copied (for escape processing) was still set at
3099        1024, so long lines caused crashes.
3101    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3102        "internal error: code overflow...". This applied to any character class
3103        that was followed by a possessive quantifier.
3105    15. Modified the Makefile to add libpcre.la as a prerequisite for
3106        libpcreposix.la because I was told this is needed for a parallel build to
3107        work.
3109    16. If a pattern that contained .* following optional items at the start was
3110        studied, the wrong optimizing data was generated, leading to matching
3111        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3112        matching string must start with a or b or c. The correct conclusion for
3113        this pattern is that a match can start with any character.
3116    Version 4.4 13-Aug-03
3117    ---------------------
3119     1. In UTF-8 mode, a character class containing characters with values between
3120        127 and 255 was not handled correctly if the compiled pattern was studied.
3121        In fixing this, I have also improved the studying algorithm for such
3122        classes (slightly).
3124     2. Three internal functions had redundant arguments passed to them. Removal
3125        might give a very teeny performance improvement.
3127     3. Documentation bug: the value of the capture_top field in a callout is *one
3128        more than* the number of the hightest numbered captured substring.
3130     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3131        in incorrectly linking with a previously installed version. They now link
3132        explicitly with libpcre.la.
3134     5. configure.in no longer needs to recognize Cygwin specially.
3136     6. A problem in pcre.in for Windows platforms is fixed.
3138     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3139        pcretest, it used to include the size of the study block as part of its
3140        output. Unfortunately, the structure contains a field that has a different
3141        size on different hardware architectures. This meant that the tests that
3142        showed this size failed. As the block is currently always of a fixed size,
3143        this information isn't actually particularly useful in pcretest output, so
3144        I have just removed it.
3146     8. Three pre-processor statements accidentally did not start in column 1.
3147        Sadly, there are *still* compilers around that complain, even though
3148        standard C has not required this for well over a decade. Sigh.
3150     9. In pcretest, the code for checking callouts passed small integers in the
3151        callout_data field, which is a void * field. However, some picky compilers
3152        complained about the casts involved for this on 64-bit systems. Now
3153        pcretest passes the address of the small integer instead, which should get
3154        rid of the warnings.
3156    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3157        both compile and run time, and gives an error if an invalid UTF-8 sequence
3158        is found. There is a option for disabling this check in cases where the
3159        string is known to be correct and/or the maximum performance is wanted.
3161    11. In response to a bug report, I changed one line in Makefile.in from
3163            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3164        to
3165            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3167        to look similar to other lines, but I have no way of telling whether this
3168        is the right thing to do, as I do not use Windows. No doubt I'll get told
3169        if it's wrong...
3172    Version 4.3 21-May-03
3173    ---------------------
3175    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3176       Makefile.
3178    2. Some refactoring to improve the quality of the code:
3180       (i)   The utf8_table... variables are now declared "const".
3182       (ii)  The code for \cx, which used the "case flipping" table to upper case
3183             lower case letters, now just substracts 32. This is ASCII-specific,
3184             but the whole concept of \cx is ASCII-specific, so it seems
3185             reasonable.
3187       (iii) PCRE was using its character types table to recognize decimal and
3188             hexadecimal digits in the pattern. This is silly, because it handles
3189             only 0-9, a-f, and A-F, but the character types table is locale-
3190             specific, which means strange things might happen. A private
3191             table is now used for this - though it costs 256 bytes, a table is
3192             much faster than multiple explicit tests. Of course, the standard
3193             character types table is still used for matching digits in subject
3194             strings against \d.
3196       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3197             ending in _t are). So I've renamed it as ESC_tee.
3199    3. The first argument for regexec() in the POSIX wrapper should have been
3200       defined as "const".
3202    4. Changed pcretest to use malloc() for its buffers so that they can be
3203       Electric Fenced for debugging.
3205    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3206       to read one or more bytes before the start of the subject string. Often this
3207       had no effect on PCRE's behaviour, but in some circumstances it could
3208       provoke a segmentation fault.
3210    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3211       to try to read one or more bytes before the start of the subject string.
3213    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3214       UTF-8 support could misbehave in various ways if the subject string
3215       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3216       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3217       back over UTF-8 characters.)
3220    Version 4.2 14-Apr-03
3221    ---------------------
3223    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3225    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3226         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3227         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3228         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3229         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3230           and BUILD_EXEEXT
3231         Note: automatic setting of the BUILD variables is not yet working
3232         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3233           compile-time but not at link-time
3234         [LINK]: use for linking executables only
3235         make different versions for Windows and non-Windows
3236         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3237           libraries
3238         [LINK_FOR_BUILD]: new variable
3239         [OBJEXT]: use throughout
3240         [EXEEXT]: use throughout
3241         <winshared>: new target
3242         <wininstall>: new target
3243         <dftables.o>: use native compiler
3244         <dftables>: use native linker
3245         <install>: handle Windows platform correctly
3246         <clean>: ditto
3247         <check>: ditto
3248         copy DLL to top builddir before testing
3250       As part of these changes, -no-undefined was removed again. This was reported
3251       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3252       in any case.
3254    3. Some tidies to get rid of compiler warnings:
3256       . In the match_data structure, match_limit was an unsigned long int, whereas
3257         match_call_count was an int. I've made them both unsigned long ints.
3259       . In pcretest the fact that a const uschar * doesn't automatically cast to
3260         a void * provoked a warning.
3262       . Turning on some more compiler warnings threw up some "shadow" variables
3263         and a few more missing casts.
3265    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3266       option, a class that contained a single character with a value between 128
3267       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3269    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3270       option, a class that contained several characters, but with at least one
3271       whose value was between 128 and 255 caused PCRE to crash.
3274    Version 4.1 12-Mar-03
3275    ---------------------
3277    1. Compiling with gcc -pedantic found a couple of places where casts were
3278    needed, and a string in dftables.c that was longer than standard compilers are
3279    required to support.
3281    2. Compiling with Sun's compiler found a few more places where the code could
3282    be tidied up in order to avoid warnings.
3284    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3285    first of these names is deprecated in the latest Autoconf in favour of the name
3286    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3287    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3288    analogy I have changed it to CFLAGS_FOR_BUILD.
3290    4. Added -no-undefined to the linking command in the Makefile, because this is
3291    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3292    linking step for the pcreposix library.
3294    5. PCRE was failing to diagnose the case of two named groups with the same
3295    name.
3297    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3298    literal character that is needed in the subject for a match, and scans along to
3299    ensure that it is present before embarking on the full matching process. This
3300    saves time in cases of nested unlimited repeats that are never going to match.
3301    Problem: the scan can take a lot of time if the subject is very long (e.g.
3302    megabytes), thus penalizing straightforward matches. It is now done only if the
3303    amount of subject to be scanned is less than 1000 bytes.
3305    7. A lesser problem with the same optimization is that it was recording the
3306    first character of an anchored pattern as "needed", thus provoking a search
3307    right along the subject, even when the first match of the pattern was going to
3308    fail. The "needed" character is now not set for anchored patterns, unless it
3309    follows something in the pattern that is of non-fixed length. Thus, it still
3310    fulfils its original purpose of finding quick non-matches in cases of nested
3311    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3314    Version 4.0 17-Feb-03
3315    ---------------------
3317    1. If a comment in an extended regex that started immediately after a meta-item
3318    extended to the end of string, PCRE compiled incorrect data. This could lead to
3319    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3321    2. Moved to autoconf 2.53 and libtool 1.4.2.
3323    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3324    the special perltest8 script is no longer needed - all the tests can be run
3325    from a single perltest script.
3327    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3328    by \s. It has now been removed in PCRE. This means it isn't recognized as
3329    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3330    class [:space:] *does* include VT, thereby creating a mess.
3332    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3333    space and tab.
3335    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3336    its new features into the main test script, reducing the number of scripts.
3338    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3339    were backward compatible, and made the (?i) apply to the whole pattern, as if
3340    /i were given. Now it behaves more logically, and applies the option setting
3341    only to what follows. PCRE has been changed to follow suit. However, if it
3342    finds options settings right at the start of the pattern, it extracts them into
3343    the global options, as before. Thus, they show up in the info data.
3345    8. Added support for the \Q...\E escape sequence. Characters in between are
3346    treated as literals. This is slightly different from Perl in that $ and @ are
3347    also handled as literals inside the quotes. In Perl, they will cause variable
3348    interpolation. Note the following examples:
3350        Pattern            PCRE matches      Perl matches
3352        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3353        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3354        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3356    For compatibility with Perl, \Q...\E sequences are recognized inside character
3357    classes as well as outside them.
3359    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3360    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3361    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3362    signed/unsigned warnings.
3364    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3365    option for pcretest, so I've replaced it by a simple function that does just
3366    that job.
3368    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3369    "pcregrep -".
3371    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3372    Java package. This provides some syntactic sugar for simple cases of what my
3373    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3374    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3375    item, you can use this simplified notation. Note that only makes sense with
3376    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3377    greediness, whatever the setting of the PCRE_UNGREEDY option.
3379    13. A change of greediness default within a pattern was not taking effect at
3380    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3381    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3382    was abstracted outside.
3384    14. PCRE now supports the \G assertion. It is true when the current matching
3385    position is at the start point of the match. This differs from \A when the
3386    starting offset is non-zero. Used with the /g option of pcretest (or similar
3387    code), it works in the same way as it does for Perl's /g option. If all
3388    alternatives of a regex begin with \G, the expression is anchored to the start
3389    match position, and the "anchored" flag is set in the compiled expression.
3391    15. Some bugs concerning the handling of certain option changes within patterns
3392    have been fixed. These applied to options other than (?ims). For example,
3393    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3394    been the other way round. Some of this was related to change 7 above.
3396    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3397    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3398    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3399    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3401    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3402    mode. Unlike ".", it always matches newline, whatever the setting of
3403    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3404    assertions. Perl allows it, but it doesn't (in general) work because it can't
3405    calculate the length of the lookbehind. At least, that's the case for Perl
3406    5.8.0 - I've been told they are going to document that it doesn't work in
3407    future.
3409    18. Added an error diagnosis for escapes that PCRE does not support: these are
3410    \L, \l, \N, \P, \p, \U, \u, and \X.
3412    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3413    reading past the end of the pattern in cases such as /[abcd/.
3415    20. PCRE was getting more memory than necessary for patterns with classes that
3416    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3418    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3419    compiling PCRE for use with Virtual Pascal.
3421    22. Small fix to the Makefile to make it work properly if the build is done
3422    outside the source tree.
3424    23. Added a new extension: a condition to go with recursion. If a conditional
3425    subpattern starts with (?(R) the "true" branch is used if recursion has
3426    happened, whereas the "false" branch is used only at the top level.
3428    24. When there was a very long string of literal characters (over 255 bytes
3429    without UTF support, over 250 bytes with UTF support), the computation of how
3430    much memory was required could be incorrect, leading to segfaults or other
3431    strange effects.
3433    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3434    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3435    there was a subsequent back reference to those brackets. This meant that, for
3436    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3437    possible to check for precisely this case. All we can do is abandon the
3438    optimization if .* occurs inside capturing brackets when there are any back
3439    references whatsoever. (See below for a better fix that came later.)
3441    26. The handling of the optimization for finding the first character of a
3442    non-anchored pattern, and for finding a character that is required later in the
3443    match were failing in some cases. This didn't break the matching; it just
3444    failed to optimize when it could. The way this is done has been re-implemented.
3446    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3448    28. Added a new feature that provides some of the functionality that Perl
3449    provides with (?{...}). The facility is termed a "callout". The way it is done
3450    in PCRE is for the caller to provide an optional function, by setting
3451    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3452    global variable. By default it is unset, which disables all calling out. To get
3453    the function called, the regex must include (?C) at appropriate points. This
3454    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3455    This provides a means of identifying different callout points. When PCRE
3456    reaches such a point in the regex, if pcre_callout has been set, the external
3457    function is called. It is provided with data in a structure called
3458    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3459    matching continues; if it returns a non-zero value, the match at the current
3460    point fails. However, backtracking will occur if possible. [This was changed
3461    later and other features added - see item 49 below.]
3463    29. pcretest is upgraded to test the callout functionality. It provides a
3464    callout function that displays information. By default, it shows the start of
3465    the match and the current position in the text. There are some new data escapes
3466    to vary what happens:
3468        \C+         in addition, show current contents of captured substrings
3469        \C-         do not supply a callout function
3470        \C!n        return 1 when callout number n is reached
3471        \C!n!m      return 1 when callout number n is reached for the mth time
3473    30. If pcregrep was called with the -l option and just a single file name, it
3474    output "<stdin>" if a match was found, instead of the file name.
3476    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3477    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3478    pcre_exec(). This saves a malloc/free per call. The default value of
3479    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3480    when configuring.
3482    32. The default maximum size of a compiled pattern is 64K. There have been a
3483    few cases of people hitting this limit. The code now uses macros to handle the
3484    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3485    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3486    configuring. Tests 2 and 5 work only with 2-byte links because they output
3487    debugging information about compiled patterns.
3489    33. Internal code re-arrangements:
3491    (a) Moved the debugging function for printing out a compiled regex into
3492        its own source file (printint.c) and used #include to pull it into
3493        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3494        separate copies.
3496    (b) Defined the list of op-code names for debugging as a macro in
3497        internal.h so that it is next to the definition of the opcodes.
3499    (c) Defined a table of op-code lengths for simpler skipping along compiled
3500        code. This is again a macro in internal.h so that it is next to the
3501        definition of the opcodes.
3503    34. Added support for recursive calls to individual subpatterns, along the
3504    lines of Robin Houston's patch (but implemented somewhat differently).
3506    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3507    allow it to read and process whole directories in Win32. This code was
3508    contributed by Lionel Fourquaux; it has not been tested by me.
3510    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3511    used to name a group. Names consist of alphanumerics and underscores, and must
3512    be unique. Back references use the syntax (?P=name) and recursive calls use
3513    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3514    numbers. The function pcre_fullinfo() can be used after compilation to extract
3515    a name/number map. There are three relevant calls:
3517      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3518      PCRE_INFO_NAMECOUNT            yields the number of entries
3519      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3521    The map is a vector of fixed-size entries. The size of each entry depends on
3522    the length of the longest name used. The first two bytes of each entry are the
3523    group number, most significant byte first. There follows the corresponding
3524    name, zero terminated. The names are in alphabetical order.
3526    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3527    case instead of 255. Making it the same both with and without UTF-8 support
3528    means that the same test output works with both.
3530    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3531    calling malloc() with a zero argument.
3533    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3534    optimization. I've improved things by keeping a bitmap of backreferences with
3535    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3536    fact referenced, the optimization can be applied. It is unlikely that a
3537    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3538    the match to follow \n) will appear inside brackets with a number greater than
3539    31, but if it does, any back reference > 31 suppresses the optimization.
3541    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3542    of disabling numbered capturing parentheses. Any opening parenthesis that is
3543    not followed by ? behaves as if it were followed by ?: but named parentheses
3544    can still be used for capturing (and they will acquire numbers in the usual
3545    way).
3547    41. Redesigned the return codes from the match() function into yes/no/error so
3548    that errors can be passed back from deep inside the nested calls. A malloc
3549    failure while inside a recursive subpattern call now causes the
3550    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3552    42. It is now possible to set a limit on the number of times the match()
3553    function is called in a call to pcre_exec(). This facility makes it possible to
3554    limit the amount of recursion and backtracking, though not in a directly
3555    obvious way, because the match() function is used in a number of different
3556    circumstances. The count starts from zero for each position in the subject
3557    string (for non-anchored patterns). The default limit is, for compatibility, a
3558    large number, namely 10 000 000. You can change this in two ways:
3560    (a) When configuring PCRE before making, you can use --with-match-limit=n
3561        to set a default value for the compiled library.
3563    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3564        a different value is set. See 45 below.
3566    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3568    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3569    of things that can be changed at compile time. The first argument specifies
3570    what is wanted and the second points to where the information is to be placed.
3571    The current list of available information is:
3573      PCRE_CONFIG_UTF8
3575    The output is an integer that is set to one if UTF-8 support is available;
3576    otherwise it is set to zero.
3580    The output is an integer that it set to the value of the code that is used for
3581    newline. It is either LF (10) or CR (13).
3585    The output is an integer that contains the number of bytes used for internal
3586    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3590    The output is an integer that contains the threshold above which the POSIX
3591    interface uses malloc() for output vectors. See item 31 above.
3595    The output is an unsigned integer that contains the default limit of the number
3596    of match() calls in a pcre_exec() execution. See 42 above.
3598    44. pcretest has been upgraded by the addition of the -C option. This causes it
3599    to extract all the available output from the new pcre_config() function, and to
3600    output it. The program then exits immediately.
3602    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3603    order to support additional features. One way would have been to define
3604    pcre_exec2() (for example) with extra arguments, but this would not have been
3605    extensible, and would also have required all calls to the original function to
3606    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3607    is used for passing in "extra" data from pcre_study().
3609    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3610    contains the following fields: