ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 27 by nigel, Sat Feb 24 21:38:49 2007 UTC revision 716 by ph10, Tue Oct 4 16:38:05 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.20 23-Sep-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 with the much improved version provided by Sheri
58        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        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.
96    Version 8.13 16-Aug-2011
97    ------------------------
99    1.  The Unicode data tables have been updated to Unicode 6.0.0.
101    2.  Two minor typos in pcre_internal.h have been fixed.
103    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
104        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
105        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
107    4.  There were a number of related bugs in the code for matching backrefences
108        caselessly in UTF-8 mode when codes for the characters concerned were
109        different numbers of bytes. For example, U+023A and U+2C65 are an upper
110        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
111        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
112        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
113        2-byte code at the end of the subject (it thought there wasn't enough data
114        left).
116    5.  Comprehensive information about what went wrong is now returned by
117        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
118        as the output vector has at least 2 elements. The offset of the start of
119        the failing character and a reason code are placed in the vector.
121    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
122        now returned is for the first byte of the failing character, instead of the
123        last byte inspected. This is an incompatible change, but I hope it is small
124        enough not to be a problem. It makes the returned offset consistent with
125        pcre_exec() and pcre_dfa_exec().
127    7.  pcretest now gives a text phrase as well as the error number when
128        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
129        failure, the offset and reason code are output.
131    8.  When \R was used with a maximizing quantifier it failed to skip backwards
132        over a \r\n pair if the subsequent match failed. Instead, it just skipped
133        back over a single character (\n). This seems wrong (because it treated the
134        two characters as a single entity when going forwards), conflicts with the
135        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
136        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
137        has been changed.
139    9.  Some internal refactoring has changed the processing so that the handling
140        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
141        time (the PCRE_DOTALL option was changed this way some time ago: version
142        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
143        which was always a bit of a fudge. It also means that there is one less
144        argument for the match() function, which reduces its stack requirements
145        slightly. This change also fixes an incompatibility with Perl: the pattern
146        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
148    10. More internal refactoring has drastically reduced the number of recursive
149        calls to match() for possessively repeated groups such as (abc)++ when
150        using pcre_exec().
152    11. While implementing 10, a number of bugs in the handling of groups were
153        discovered and fixed:
155        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
156        (a|)*(?1) gave a compile-time internal error.
157        ((a|)+)+  did not notice that the outer group could match an empty string.
158        (^a|^)+   was not marked as anchored.
159        (.*a|.*)+ was not marked as matching at start or after a newline.
161    12. Yet more internal refactoring has removed another argument from the match()
162        function. Special calls to this function are now indicated by setting a
163        value in a variable in the "match data" data block.
165    13. Be more explicit in pcre_study() instead of relying on "default" for
166        opcodes that mean there is no starting character; this means that when new
167        ones are added and accidentally left out of pcre_study(), testing should
168        pick them up.
170    14. The -s option of pcretest has been documented for ages as being an old
171        synonym of -m (show memory usage). I have changed it to mean "force study
172        for every regex", that is, assume /S for every regex. This is similar to -i
173        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
174        using it. It makes it easier to run collections of tests with and without
175        study enabled, and thereby test pcre_study() more easily. All the standard
176        tests are now run with and without -s (but some patterns can be marked as
177        "never study" - see 20 below).
179    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
180        restoration of the capturing data to the outer values was not happening
181        correctly.
183    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
184        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
185        pattern had matched an empty string, and so incorrectly returned a no
186        match.
188    17. There was optimizing code for the last branch of non-capturing parentheses,
189        and also for the obeyed branch of a conditional subexpression, which used
190        tail recursion to cut down on stack usage. Unfortunately, now that there is
191        the possibility of (*THEN) occurring in these branches, tail recursion is
192        no longer possible because the return has to be checked for (*THEN). These
193        two optimizations have therefore been removed. [But see 8.20/11 above.]
195    18. If a pattern containing \R was studied, it was assumed that \R always
196        matched two bytes, thus causing the minimum subject length to be
197        incorrectly computed because \R can also match just one byte.
199    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
200        was incorrectly computed.
202    20. If /S is present twice on a test pattern in pcretest input, it now
203        *disables* studying, thereby overriding the use of -s on the command line
204        (see 14 above). This is necessary for one or two tests to keep the output
205        identical in both cases.
207    21. When (*ACCEPT) was used in an assertion that matched an empty string and
208        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
210    22. When an atomic group that contained a capturing parenthesis was
211        successfully matched, but the branch in which it appeared failed, the
212        capturing was not being forgotten if a higher numbered group was later
213        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
214        group 1 to "a", when in fact it should be unset. This applied to multi-
215        branched capturing and non-capturing groups, repeated or not, and also to
216        positive assertions (capturing in negative assertions does not happen
217        in PCRE) and also to nested atomic groups.
219    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
220        subject after a captured substring, to make it easier to tell which of a
221        number of identical substrings has been captured.
223    24. The way atomic groups are processed by pcre_exec() has been changed so that
224        if they are repeated, backtracking one repetition now resets captured
225        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
226        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
227        "aaa". Previously, it would have been "a". As part of this code
228        refactoring, the way recursive calls are handled has also been changed.
230    25. If an assertion condition captured any substrings, they were not passed
231        back unless some other capturing happened later. For example, if
232        (?(?=(a))a) was matched against "a", no capturing was returned.
234    26. When studying a pattern that contained subroutine calls or assertions,
235        the code for finding the minimum length of a possible match was handling
236        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
237        group 1 called group 2 while simultaneously a separate group 2 called group
238        1). A stack overflow occurred in this case. I have fixed this by limiting
239        the recursion depth to 10.
241    27. Updated RunTest.bat in the distribution to the version supplied by Tom
242        Fortmann. This supports explicit test numbers on the command line, and has
243        argument validation and error reporting.
245    28. An instance of \X with an unlimited repeat could fail if at any point the
246        first character it looked at was a mark character.
248    29. Some minor code refactoring concerning Unicode properties and scripts
249        should reduce the stack requirement of match() slightly.
251    30. Added the '=' option to pcretest to check the setting of unused capturing
252        slots at the end of the pattern, which are documented as being -1, but are
253        not included in the return count.
255    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
256        compiled something random. Now it gives a compile-time error (as does
257        Perl).
259    32. A *MARK encountered during the processing of a positive assertion is now
260        recorded and passed back (compatible with Perl).
262    33. If --only-matching or --colour was set on a pcregrep call whose pattern
263        had alternative anchored branches, the search for a second match in a line
264        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
265        matched the line "0102" twice. The same bug affected patterns that started
266        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
267        twice.
269    34. Previously, PCRE did not allow quantification of assertions. However, Perl
270        does, and because of capturing effects, quantifying parenthesized
271        assertions may at times be useful. Quantifiers are now allowed for
272        parenthesized assertions.
274    35. A minor code tidy in pcre_compile() when checking options for \R usage.
276    36. \g was being checked for fancy things in a character class, when it should
277        just be a literal "g".
279    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
280        appearance of a nested POSIX class supersedes an apparent external class.
281        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
282        unescaped square brackets may also appear as part of class names. For
283        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
284        more like Perl. (But see 8.20/1 above.)
286    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
287        was because it thought it was \N{name}, which is not supported).
289    39. Add minix to OS list not supporting the -S option in pcretest.
291    40. PCRE tries to detect cases of infinite recursion at compile time, but it
292        cannot analyze patterns in sufficient detail to catch mutual recursions
293        such as ((?1))((?2)). There is now a runtime test that gives an error if a
294        subgroup is called recursively as a subpattern for a second time at the
295        same position in the subject string. In previous releases this might have
296        been caught by the recursion limit, or it might have run out of stack.
298    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
299        happen only once. PCRE was, however incorrectly giving a compile time error
300        "recursive call could loop indefinitely" because it cannot analyze the
301        pattern in sufficient detail. The compile time test no longer happens when
302        PCRE is compiling a conditional subpattern, but actual runaway loops are
303        now caught at runtime (see 40 above).
305    42. It seems that Perl allows any characters other than a closing parenthesis
306        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
307        has been changed to be the same.
309    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
310        as not to get warnings when autogen.sh is called. Also changed
311        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
313    44. To help people who use pcregrep to scan files containing exceedingly long
314        lines, the following changes have been made:
316        (a) The default value of the buffer size parameter has been increased from
317            8K to 20K. (The actual buffer used is three times this size.)
319        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
320            PCRE is built.
322        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
323            to be set at run time.
325        (d) Numerical values in pcregrep options can be followed by K or M, for
326            example --buffer-size=50K.
328        (e) If a line being scanned overflows pcregrep's buffer, an error is now
329            given and the return code is set to 2.
331    45. Add a pointer to the latest mark to the callout data block.
333    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
334        partial match of an empty string instead of no match. This was specific to
335        the use of ".".
337    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
338        complete match instead of a partial match. This bug was dependent on both
339        the PCRE_UTF8 and PCRE_DOTALL options being set.
341    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
342        starting byte set, because \b was not being ignored.
345    Version 8.12 15-Jan-2011
346    ------------------------
348    1.  Fixed some typos in the markup of the man pages, and wrote a script that
349        checks for such things as part of the documentation building process.
351    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
352        --match-limit and --recursion-limit options (added for 8.11). In
353        particular, this made one of the standard tests fail. (The integer value
354        went into the wrong half of a long int.)
356    3.  If the --colour option was given to pcregrep with -v (invert match), it
357        did strange things, either producing crazy output, or crashing. It should,
358        of course, ignore a request for colour when reporting lines that do not
359        match.
361    4.  Another pcregrep bug caused similar problems if --colour was specified with
362        -M (multiline) and the pattern match finished with a line ending.
364    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
365        matched in multiline mode, the following line was shown as part of the
366        match. This seems wrong, so I have changed it.
368    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
369        the check for further matches in the same line (so they could be coloured)
370        to overrun the end of the current line. If another match was found, it was
371        incorrectly shown (and then shown again when found in the next line).
373    7.  If pcregrep was compiled under Windows, there was a reference to the
374        function pcregrep_exit() before it was defined. I am assuming this was
375        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
376        reported by a user. I've moved the definition above the reference.
379    Version 8.11 10-Dec-2010
380    ------------------------
382    1.  (*THEN) was not working properly if there were untried alternatives prior
383        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
384        backtracked to try for "b" instead of moving to the next alternative branch
385        at the same level (in this case, to look for "c"). The Perl documentation
386        is clear that when (*THEN) is backtracked onto, it goes to the "next
387        alternative in the innermost enclosing group".
389    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
390        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
391        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
392        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
393        (*THEN).
395    3.  If \s appeared in a character class, it removed the VT character from
396        the class, even if it had been included by some previous item, for example
397        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
398        of \s, but is part of the POSIX "space" class.)
400    4.  A partial match never returns an empty string (because you can always
401        match an empty string at the end of the subject); however the checking for
402        an empty string was starting at the "start of match" point. This has been
403        changed to the "earliest inspected character" point, because the returned
404        data for a partial match starts at this character. This means that, for
405        example, /(?<=abc)def/ gives a partial match for the subject "abc"
406        (previously it gave "no match").
408    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
409        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
410        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
411        has an implication that the given string is incomplete (because a partial
412        match is preferred over a full match). For this reason, these items now
413        give a partial match in this situation. [Aside: previously, the one case
414        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
415        match rather than a full match, which was wrong by the old rules, but is
416        now correct.]
418    6.  There was a bug in the handling of #-introduced comments, recognized when
419        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
420        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
421        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
422        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
423        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
424        places in pcre_compile().
426    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
427        comments when looking ahead for named forward references to subpatterns,
428        the only newline sequence it recognized was NL. It now handles newlines
429        according to the set newline convention.
431    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
432        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
433        cater for a lack of strerror(). These oversights have been fixed.
435    9.  Added --match-limit and --recursion-limit to pcregrep.
437    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
439    11. When the -o option was used, pcregrep was setting a return code of 1, even
440        when matches were found, and --line-buffered was not being honoured.
442    12. Added an optional parentheses number to the -o and --only-matching options
443        of pcregrep.
445    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
446        can match an empty string. The code to do it in pcretest and pcredemo
447        needed fixing:
449        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
450            only one byte after an empty string match just before CRLF (this case
451            just got forgotten; "any" and "anycrlf" were OK).
453        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
454            mode when an empty string match preceded an ASCII character followed by
455            a non-ASCII character. (The code for advancing by one character rather
456            than one byte was nonsense.)
458        (c) The pcredemo.c sample program did not have any code at all to handle
459            the cases when CRLF is a valid newline sequence.
461    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
462        as a starting offset was within the subject string. There is now a new
463        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
464        negative or greater than the length of the string. In order to test this,
465        pcretest is extended to allow the setting of negative starting offsets.
467    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
468        starting offset points to the beginning of a UTF-8 character was
469        unnecessarily clumsy. I tidied it up.
471    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
472        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
474    17. Nobody had reported that the --include_dir option, which was added in
475        release 7.7 should have been called --include-dir (hyphen, not underscore)
476        for compatibility with GNU grep. I have changed it to --include-dir, but
477        left --include_dir as an undocumented synonym, and the same for
478        --exclude-dir, though that is not available in GNU grep, at least as of
479        release 2.5.4.
481    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
482        characters from a string of bytes) have been redefined so as not to use
483        loops, in order to improve performance in some environments. At the same
484        time, I abstracted some of the common code into auxiliary macros to save
485        repetition (this should not affect the compiled code).
487    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
488        compile-time error is now given if \c is not followed by an ASCII
489        character, that is, a byte less than 128. (In EBCDIC mode, the code is
490        different, and any byte value is allowed.)
492    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
493        START_OPTIMIZE option, which is now allowed at compile time - but just
494        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
495        to pcregrep and other applications that have no direct access to PCRE
496        options. The new /Y option in pcretest sets this option when calling
497        pcre_compile().
499    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
500        back references. Groups containing recursive back references were forced to
501        be atomic by that change, but in the case of named groups, the amount of
502        memory required was incorrectly computed, leading to "Failed: internal
503        error: code overflow". This has been fixed.
505    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
506        pcretest.c, to avoid build problems in some Borland environments.
509    Version 8.10 25-Jun-2010
510    ------------------------
512    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
513        THEN.
515    2.  (*ACCEPT) was not working when inside an atomic group.
517    3.  Inside a character class, \B is treated as a literal by default, but
518        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
519        causes the error). The code is unchanged, but I tidied the documentation.
521    4.  Inside a character class, PCRE always treated \R and \X as literals,
522        whereas Perl faults them if its -w option is set. I have changed PCRE so
523        that it faults them when PCRE_EXTRA is set.
525    5.  Added support for \N, which always matches any character other than
526        newline. (It is the same as "." when PCRE_DOTALL is not set.)
528    6.  When compiling pcregrep with newer versions of gcc which may have
529        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
530        declared with attribute warn_unused_result" were given. Just casting the
531        result to (void) does not stop the warnings; a more elaborate fudge is
532        needed. I've used a macro to implement this.
534    7.  Minor change to pcretest.c to avoid a compiler warning.
536    8.  Added four artifical Unicode properties to help with an option to make
537        \s etc use properties (see next item). The new properties are: Xan
538        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
540    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
541        use Unicode properties. (*UCP) at the start of a pattern can be used to set
542        this option. Modified pcretest to add /W to test this facility. Added
543        REG_UCP to make it available via the POSIX interface.
545    10. Added --line-buffered to pcregrep.
547    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
548        studied, and the match started with a letter with a code point greater than
549        127 whose first byte was different to the first byte of the other case of
550        the letter, the other case of this starting letter was not recognized
551        (#976).
553    12. If a pattern that was studied started with a repeated Unicode property
554        test, for example, \p{Nd}+, there was the theoretical possibility of
555        setting up an incorrect bitmap of starting bytes, but fortunately it could
556        not have actually happened in practice until change 8 above was made (it
557        added property types that matched character-matching opcodes).
559    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
560        possible starting bytes for non-anchored patterns.
562    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
563        \R, and also a number of cases that involve Unicode properties, both
564        explicit and implicit when PCRE_UCP is set.
566    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
567        input, it could crash or give wrong results if characters with values
568        greater than 0xc0 were present in the subject string. (Detail: it assumed
569        UTF-8 input when processing these items.)
571    16. Added a lot of (int) casts to avoid compiler warnings in systems where
572        size_t is 64-bit (#991).
574    17. Added a check for running out of memory when PCRE is compiled with
575        --disable-stack-for-recursion (#990).
577    18. If the last data line in a file for pcretest does not have a newline on
578        the end, a newline was missing in the output.
580    19. The default pcre_chartables.c file recognizes only ASCII characters (values
581        less than 128) in its various bitmaps. However, there is a facility for
582        generating tables according to the current locale when PCRE is compiled. It
583        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
584        characters, are recognized by isspace() and therefore were getting set in
585        these tables, and indeed these tables seem to approximate to ISO 8859. This
586        caused a problem in UTF-8 mode when pcre_study() was used to create a list
587        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
588        which of course cannot start UTF-8 characters. I have changed the code so
589        that only real ASCII characters (less than 128) and the correct starting
590        bytes for UTF-8 encodings are set for characters greater than 127 when in
591        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
592        altogether.)
594    20. Added the /T option to pcretest so as to be able to run tests with non-
595        standard character tables, thus making it possible to include the tests
596        used for 19 above in the standard set of tests.
598    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
599        reference to a subpattern the other side of a comment that contains an
600        opening parenthesis caused either an internal compiling error, or a
601        reference to the wrong subpattern.
604    Version 8.02 19-Mar-2010
605    ------------------------
607    1.  The Unicode data tables have been updated to Unicode 5.2.0.
609    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
610        configured.
612    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
613        original author of that file, following a query about its status.
615    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
616        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
618    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
619        quantifier applied to a forward-referencing subroutine call, could compile
620        incorrect code or give the error "internal error: previously-checked
621        referenced subpattern not found".
623    6.  Both MS Visual Studio and Symbian OS have problems with initializing
624        variables to point to external functions. For these systems, therefore,
625        pcre_malloc etc. are now initialized to local functions that call the
626        relevant global functions.
628    7.  There were two entries missing in the vectors called coptable and poptable
629        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
630        I've fixed the data, and added a kludgy way of testing at compile time that
631        the lengths are correct (equal to the number of opcodes).
633    8.  Following on from 7, I added a similar kludge to check the length of the
634        eint vector in pcreposix.c.
636    9.  Error texts for pcre_compile() are held as one long string to avoid too
637        much relocation at load time. To find a text, the string is searched,
638        counting zeros. There was no check for running off the end of the string,
639        which could happen if a new error number was added without updating the
640        string.
642    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
644    11. \K was not working if it appeared in an atomic group or in a group that
645        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
646        \K is "not well defined" if used in an assertion. PCRE now accepts it if
647        the assertion is positive, but not if it is negative.
649    12. Change 11 fortuitously reduced the size of the stack frame used in the
650        "match()" function of pcre_exec.c by one pointer. Forthcoming
651        implementation of support for (*MARK) will need an extra pointer on the
652        stack; I have reserved it now, so that the stack frame size does not
653        decrease.
655    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
656        item in branch that calls a recursion is a subroutine call - as in the
657        second branch in the above example - was incorrectly given the compile-
658        time error "recursive call could loop indefinitely" because pcre_compile()
659        was not correctly checking the subroutine for matching a non-empty string.
661    14. The checks for overrunning compiling workspace could trigger after an
662        overrun had occurred. This is a "should never occur" error, but it can be
663        triggered by pathological patterns such as hundreds of nested parentheses.
664        The checks now trigger 100 bytes before the end of the workspace.
666    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
669    Version 8.01 19-Jan-2010
670    ------------------------
672    1.  If a pattern contained a conditional subpattern with only one branch (in
673        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
674        computed the wrong minimum data length (which is of course zero for such
675        subpatterns). This could cause incorrect "no match" results.
677    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
678        the pattern is reset in the first branch, pcre_compile() failed with
679        "internal error: code overflow at offset...". This happened only when
680        the reset was to the original external option setting. (An optimization
681        abstracts leading options settings into an external setting, which was the
682        cause of this.)
684    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
685        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
686        assertion pattern did not match (meaning that the assertion was true), it
687        was incorrectly treated as false if the SKIP had been reached during the
688        matching. This also applied to assertions used as conditions.
690    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
691        assertion subpattern, including such a pattern used as a condition,
692        unpredictable results occurred, instead of the error return
695    5.  The C++ GlobalReplace function was not working like Perl for the special
696        situation when an empty string is matched. It now does the fancy magic
697        stuff that is necessary.
699    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
700        removed. (These were left over from very, very early versions of PCRE.)
702    7.  Some cosmetic changes to the code to make life easier when compiling it
703        as part of something else:
705        (a) Change DEBUG to PCRE_DEBUG.
707        (b) In pcre_compile(), rename the member of the "branch_chain" structure
708            called "current" as "current_branch", to prevent a collision with the
709            Linux macro when compiled as a kernel module.
711        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
712            prevent a collision with the Linux macro when compiled as a kernel
713            module.
715    8.  In pcre_compile() there are some checks for integer overflows that used to
716        cast potentially large values to (double). This has been changed to that
717        when building, a check for int64_t is made, and if it is found, it is used
718        instead, thus avoiding the use of floating point arithmetic. (There is no
719        other use of FP in PCRE.) If int64_t is not found, the fallback is to
720        double.
722    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
723        2005 (difference between two addresses compared to an unsigned value).
725    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
726        custom one, because of the following reported problem in Windows:
728          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
729              under Win32.
730          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
731              therefore missing the function definition.
732          - The compiler thus generates a "C" signature for the test function.
733          - The linker fails to find the "C" function.
734          - PCRE fails to configure if asked to do so against libbz2.
736    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
737        messages were output:
739          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
740          rerunning libtoolize, to keep the correct libtool macros in-tree.
741          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
743        I have done both of these things.
745    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
746        most of the time, it *can* run out if it is given a pattern that contains a
747        runaway infinite recursion. I updated the discussion in the pcrestack man
748        page.
750    13. Now that we have gone to the x.xx style of version numbers, the minor
751        version may start with zero. Using 08 or 09 is a bad idea because users
752        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
753        interpreted as invalid octal numbers. I've updated the previous comment in
754        configure.ac, and also added a check that gives an error if 08 or 09 are
755        used.
757    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
758        causing partial matching to fail when the end of the subject matched \W
759        in a UTF-8 pattern where \W was quantified with a minimum of 3.
761    15. There were some discrepancies between the declarations in pcre_internal.h
762        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
763        their definitions. The declarations used "const uschar *" and the
764        definitions used USPTR. Even though USPTR is normally defined as "const
765        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
766        reported that: "This difference in casting confuses some C++ compilers, for
767        example, SunCC recognizes above declarations as different functions and
768        generates broken code for hbpcre." I have changed the declarations to use
769        USPTR.
771    16. GNU libtool is named differently on some systems. The autogen.sh script now
772        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
773        (FreeBSD).
775    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
776        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
777        comment: "Figure out how to create a longlong from a string: strtoll and
778        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
779        instance, but it only takes 2 args instead of 3!"
781    18. A subtle bug concerned with back references has been fixed by a change of
782        specification, with a corresponding code fix. A pattern such as
783        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
784        refers, was giving matches when it shouldn't. For example, xa=xaaa would
785        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
786        same bug. Such groups have to be quantified to be useful, or contained
787        inside another quantified group. (If there's no repetition, the reference
788        can never match.) The problem arises because, having left the group and
789        moved on to the rest of the pattern, a later failure that backtracks into
790        the group uses the captured value from the final iteration of the group
791        rather than the correct earlier one. I have fixed this in PCRE by forcing
792        any group that contains a reference to itself to be an atomic group; that
793        is, there cannot be any backtracking into it once it has completed. This is
794        similar to recursive and subroutine calls.
797    Version 8.00 19-Oct-09
798    ----------------------
800    1.  The table for translating pcre_compile() error codes into POSIX error codes
801        was out-of-date, and there was no check on the pcre_compile() error code
802        being within the table. This could lead to an OK return being given in
803        error.
805    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
806        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
807        in a Windows environment.
809    3.  The pcregrep --count option prints the count for each file even when it is
810        zero, as does GNU grep. However, pcregrep was also printing all files when
811        --files-with-matches was added. Now, when both options are given, it prints
812        counts only for those files that have at least one match. (GNU grep just
813        prints the file name in this circumstance, but including the count seems
814        more useful - otherwise, why use --count?) Also ensured that the
815        combination -clh just lists non-zero counts, with no names.
817    4.  The long form of the pcregrep -F option was incorrectly implemented as
818        --fixed_strings instead of --fixed-strings. This is an incompatible change,
819        but it seems right to fix it, and I didn't think it was worth preserving
820        the old behaviour.
822    5.  The command line items --regex=pattern and --regexp=pattern were not
823        recognized by pcregrep, which required --regex pattern or --regexp pattern
824        (with a space rather than an '='). The man page documented the '=' forms,
825        which are compatible with GNU grep; these now work.
827    6.  No libpcreposix.pc file was created for pkg-config; there was just
828        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
830    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
831        when UCP support is not needed, by modifying the Python script that
832        generates it from Unicode data files. This should not matter if the module
833        is correctly used as a library, but I received one complaint about 50K of
834        unwanted data. My guess is that the person linked everything into his
835        program rather than using a library. Anyway, it does no harm.
837    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
838        was a minimum greater than 1 for a wide character in a possessive
839        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
840        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
841        character. Chaos in the form of incorrect output or a compiling loop could
842        result.
844    9.  The restrictions on what a pattern can contain when partial matching is
845        requested for pcre_exec() have been removed. All patterns can now be
846        partially matched by this function. In addition, if there are at least two
847        slots in the offset vector, the offset of the earliest inspected character
848        for the match and the offset of the end of the subject are set in them when
849        PCRE_ERROR_PARTIAL is returned.
851    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
852        synonymous with PCRE_PARTIAL, for backwards compatibility, and
853        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
854        and may be more useful for multi-segment matching.
856    11. Partial matching with pcre_exec() is now more intuitive. A partial match
857        used to be given if ever the end of the subject was reached; now it is
858        given only if matching could not proceed because another character was
859        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
860        string "Z", which now yields "no match" instead of "partial match". In the
861        case of pcre_dfa_exec(), "no match" is given if every matching path for the
862        final character ended with (*FAIL).
864    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
865        if the pattern had a "must contain" character that was already found in the
866        earlier partial match, unless partial matching was again requested. For
867        example, with the pattern /dog.(body)?/, the "must contain" character is
868        "g". If the first part-match was for the string "dog", restarting with
869        "sbody" failed. This bug has been fixed.
871    13. The string returned by pcre_dfa_exec() after a partial match has been
872        changed so that it starts at the first inspected character rather than the
873        first character of the match. This makes a difference only if the pattern
874        starts with a lookbehind assertion or \b or \B (\K is not supported by
875        pcre_dfa_exec()). It's an incompatible change, but it makes the two
876        matching functions compatible, and I think it's the right thing to do.
878    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
879        so that the demonstration program is easily available in environments where
880        PCRE has not been installed from source.
882    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
883        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
884        library.
886    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
887        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
888        is not the first non-POSIX option to be added. Clearly some people find
889        these options useful.
891    17. If a caller to the POSIX matching function regexec() passes a non-zero
892        value for nmatch with a NULL value for pmatch, the value of
893        nmatch is forced to zero.
895    18. RunGrepTest did not have a test for the availability of the -u option of
896        the diff command, as RunTest does. It now checks in the same way as
897        RunTest, and also checks for the -b option.
899    19. If an odd number of negated classes containing just a single character
900        interposed, within parentheses, between a forward reference to a named
901        subpattern and the definition of the subpattern, compilation crashed with
902        an internal error, complaining that it could not find the referenced
903        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
904        [The bug was that it was starting one character too far in when skipping
905        over the character class, thus treating the ] as data rather than
906        terminating the class. This meant it could skip too much.]
908    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
909        /g option in pcretest when the pattern contains \K, which makes it possible
910        to have an empty string match not at the start, even when the pattern is
911        anchored. Updated pcretest and pcredemo to use this option.
913    21. If the maximum number of capturing subpatterns in a recursion was greater
914        than the maximum at the outer level, the higher number was returned, but
915        with unset values at the outer level. The correct (outer level) value is
916        now given.
918    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
919        PCRE did not set those parentheses (unlike Perl). I have now found a way to
920        make it do so. The string so far is captured, making this feature
921        compatible with Perl.
923    23. The tests have been re-organized, adding tests 11 and 12, to make it
924        possible to check the Perl 5.10 features against Perl 5.10.
926    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
927        pattern matches a fixed length string. PCRE did not allow this; now it
928        does. Neither allows recursion.
930    25. I finally figured out how to implement a request to provide the minimum
931        length of subject string that was needed in order to match a given pattern.
932        (It was back references and recursion that I had previously got hung up
933        on.) This code has now been added to pcre_study(); it finds a lower bound
934        to the length of subject needed. It is not necessarily the greatest lower
935        bound, but using it to avoid searching strings that are too short does give
936        some useful speed-ups. The value is available to calling programs via
937        pcre_fullinfo().
939    26. While implementing 25, I discovered to my embarrassment that pcretest had
940        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
941        study optimizations had never been tested with that matching function.
942        Oops. What is worse, even when it was passed study data, there was a bug in
943        pcre_dfa_exec() that meant it never actually used it. Double oops. There
944        were also very few tests of studied patterns with pcre_dfa_exec().
946    27. If (?| is used to create subpatterns with duplicate numbers, they are now
947        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
948        on the other side of the coin, they are no longer allowed to have different
949        names, because these cannot be distinguished in PCRE, and this has caused
950        confusion. (This is a difference from Perl.)
952    28. When duplicate subpattern names are present (necessarily with different
953        numbers, as required by 27 above), and a test is made by name in a
954        conditional pattern, either for a subpattern having been matched, or for
955        recursion in such a pattern, all the associated numbered subpatterns are
956        tested, and the overall condition is true if the condition is true for any
957        one of them. This is the way Perl works, and is also more like the way
958        testing by number works.
961    Version 7.9 11-Apr-09
962    ---------------------
964    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
965        (pcretest), all targets were linked against these libraries. This included
966        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
967        libraries. This caused unwanted dependencies to be created. This problem
968        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
969        pcretest is linked with readline.
971    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
972        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
973        moved outside it again, because FALSE and TRUE are already defined in AIX,
974        but BOOL is not.
976    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
977        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
979    4.  The pcregrep documentation said spaces were inserted as well as colons (or
980        hyphens) following file names and line numbers when outputting matching
981        lines. This is not true; no spaces are inserted. I have also clarified the
982        wording for the --colour (or --color) option.
984    5.  In pcregrep, when --colour was used with -o, the list of matching strings
985        was not coloured; this is different to GNU grep, so I have changed it to be
986        the same.
988    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
989        each matching line was coloured. Now it goes on to look for further matches
990        of any of the test patterns, which is the same behaviour as GNU grep.
992    7.  A pattern that could match an empty string could cause pcregrep to loop; it
993        doesn't make sense to accept an empty string match in pcregrep, so I have
994        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
995        seems to be how GNU grep behaves.
997    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
998        start or after a newline", because the conditional assertion was not being
999        correctly handled. The rule now is that both the assertion and what follows
1000        in the first alternative must satisfy the test.
1002    9.  If auto-callout was enabled in a pattern with a conditional group whose
1003        condition was an assertion, PCRE could crash during matching, both with
1004        pcre_exec() and pcre_dfa_exec().
1006    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1007        used for matching.
1009    11. Unicode property support in character classes was not working for
1010        characters (bytes) greater than 127 when not in UTF-8 mode.
1012    12. Added the -M command line option to pcretest.
1014    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1016    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1018    16. Added comments and documentation about mis-use of no_arg in the C++
1019        wrapper.
1021    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1022        from Martin Jerabek that uses macro names for all relevant character and
1023        string constants.
1025    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1026        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1027        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1028        these, but not everybody uses configure.
1030    19. A conditional group that had only one branch was not being correctly
1031        recognized as an item that could match an empty string. This meant that an
1032        enclosing group might also not be so recognized, causing infinite looping
1033        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1034        with the subject "ab", where knowledge that the repeated group can match
1035        nothing is needed in order to break the loop.
1037    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1038        exec(), but without supplying a callout function, matching went wrong.
1040    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1041        leak if the size of the offset vector was greater than 30. When the vector
1042        is smaller, the saved offsets during recursion go onto a local stack
1043        vector, but for larger vectors malloc() is used. It was failing to free
1044        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1045        error, in fact).
1047    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1048        heapframe that is used only when UTF-8 support is enabled. This caused no
1049        problem, but was untidy.
1051    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1052        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1053        included within another project.
1055    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1056        slightly modified by me:
1058          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1059              not building pcregrep.
1061          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1062              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1064    25. Forward references, both numeric and by name, in patterns that made use of
1065        duplicate group numbers, could behave incorrectly or give incorrect errors,
1066        because when scanning forward to find the reference group, PCRE was not
1067        taking into account the duplicate group numbers. A pattern such as
1068        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1070    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1071        the feature of a custom pointer more persuasive (as requested by a user).
1073    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1074        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1075        pre-defined.
1077    28. Added support for (*UTF8) at the start of a pattern.
1079    29. Arrange for flags added by the "release type" setting in CMake to be shown
1080        in the configuration summary.
1083    Version 7.8 05-Sep-08
1084    ---------------------
1086    1.  Replaced UCP searching code with optimized version as implemented for Ad
1087        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1088        stage table and inline lookup instead of a function, giving speed ups of 2
1089        to 5 times on some simple patterns that I tested. Permission was given to
1090        distribute the MultiStage2.py script that generates the tables (it's not in
1091        the tarball, but is in the Subversion repository).
1093    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1094        scripts.
1096    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1097        a group with a zero qualifier. The result of the study could be incorrect,
1098        or the function might crash, depending on the pattern.
1100    4.  Caseless matching was not working for non-ASCII characters in back
1101        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1102        It now works when Unicode Property Support is available.
1104    5.  In pcretest, an escape such as \x{de} in the data was always generating
1105        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1106        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1107        truncation.
1109    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1111    7.  Added two (int) casts to pcregrep when printing the difference of two
1112        pointers, in case they are 64-bit values.
1114    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1115        test 2 if it fails.
1117    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1118        and a #define of that name to empty if it is not externally set. This is to
1119        allow users of MSVC to set it if necessary.
1121    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1122        the convenience functions in the pcre_get.c source file.
1124    11. An option change at the start of a pattern that had top-level alternatives
1125        could cause overwriting and/or a crash. This command provoked a crash in
1126        some environments:
1128          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1130        This potential security problem was recorded as CVE-2008-2371.
1132    12. For a pattern where the match had to start at the beginning or immediately
1133        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1134        pcre_dfa_exec() could read past the end of the passed subject if there was
1135        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1136        pcretest so that it places the subject at the end of its malloc-ed buffer.
1138    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1139        exec() might read past the end of the data buffer in UTF-8 mode.
1141    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1142        the data contained the byte 0x85 as part of a UTF-8 character within its
1143        first line. This applied both to normal and DFA matching.
1145    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1146        /^[^d]*?$/8 failed to match "abc".
1148    16. Added a missing copyright notice to pcrecpp_internal.h.
1150    17. Make it more clear in the documentation that values returned from
1151        pcre_exec() in ovector are byte offsets, not character counts.
1153    18. Tidied a few places to stop certain compilers from issuing warnings.
1155    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1156        supplied by Stefan Weber. I made a further small update for 7.8 because
1157        there is a change of source arrangements: the pcre_searchfuncs.c module is
1158        replaced by pcre_ucd.c.
1161    Version 7.7 07-May-08
1162    ---------------------
1164    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1165        a string to a long long, pretend we don't even have a long long." This is
1166        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1168    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1169        pre-7.6 versions, which defined a global no_arg variable instead of putting
1170        it in the RE class. (See also #8 below.)
1172    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1173        Lopes.
1175    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1177        (1) The include/exclude patterns were being applied to the whole pathnames
1178            of files, instead of just to the final components.
1180        (2) If there was more than one level of directory, the subdirectories were
1181            skipped unless they satisfied the include/exclude conditions. This is
1182            inconsistent with GNU grep (and could even be seen as contrary to the
1183            pcregrep specification - which I improved to make it absolutely clear).
1184            The action now is always to scan all levels of directory, and just
1185            apply the include/exclude patterns to regular files.
1187    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1188        --exclude_dir in the tests to avoid scanning .svn directories.
1190    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1191        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1192        doesn't support NULs in patterns.
1194    7.  Added some missing "const"s to declarations of static tables in
1195        pcre_compile.c and pcre_dfa_exec.c.
1197    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1198        caused by fix #2  above. (Subsequently also a second patch to fix the
1199        first patch. And a third patch - this was a messy problem.)
1201    9.  Applied Craig's patch to remove the use of push_back().
1203    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1204        matching function regexec().
1206    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1207        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1208        references. PCRE supports relative numbers with this syntax (I don't think
1209        Oniguruma does).
1211    12. Previously, a group with a zero repeat such as (...){0} was completely
1212        omitted from the compiled regex. However, this means that if the group
1213        was called as a subroutine from elsewhere in the pattern, things went wrong
1214        (an internal error was given). Such groups are now left in the compiled
1215        pattern, with a new opcode that causes them to be skipped at execution
1216        time.
1218    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1219        to the way PCRE behaves:
1221        (a) A lone ] character is dis-allowed (Perl treats it as data).
1223        (b) A back reference to an unmatched subpattern matches an empty string
1224            (Perl fails the current match path).
1226        (c) A data ] in a character class must be notated as \] because if the
1227            first data character in a class is ], it defines an empty class. (In
1228            Perl it is not possible to have an empty class.) The empty class []
1229            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1230            The negative empty class [^] matches any one character, independently
1231            of the DOTALL setting.
1233    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1234        non-existent subpattern following a character class starting with ']' and
1235        containing () gave an internal compiling error instead of "reference to
1236        non-existent subpattern". Fortunately, when the pattern did exist, the
1237        compiled code was correct. (When scanning forwards to check for the
1238        existencd of the subpattern, it was treating the data ']' as terminating
1239        the class, so got the count wrong. When actually compiling, the reference
1240        was subsequently set up correctly.)
1242    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1243        it was being rejected as not supported by pcre_dfa_exec(), even though
1244        other assertions are supported. I have made pcre_dfa_exec() support
1245        (*FAIL).
1247    16. The implementation of 13c above involved the invention of a new opcode,
1248        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1249        cannot be changed at match time, I realized I could make a small
1250        improvement to matching performance by compiling OP_ALLANY instead of
1251        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1252        on the OP_ANY path.
1254    17. Compiling pcretest on Windows with readline support failed without the
1255        following two fixes: (1) Make the unistd.h include conditional on
1256        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1258    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1259        ncurses library to be included for pcretest when ReadLine support is
1260        requested, but also to allow for it to be overridden. This patch came from
1261        Daniel Bergström.
1263    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1264        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1265        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1266        spotting this.
1269    Version 7.6 28-Jan-08
1270    ---------------------
1272    1.  A character class containing a very large number of characters with
1273        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1274        overflow.
1276    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1277        HAVE_LONG_LONG is not defined.
1279    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1280        bring them up to date and include new features. This patch includes:
1282        - Fixed PH's badly added libz and libbz2 support.
1283        - Fixed a problem with static linking.
1284        - Added pcredemo. [But later removed - see 7 below.]
1285        - Fixed dftables problem and added an option.
1286        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1287            HAVE_LONG_LONG.
1288        - Added readline support for pcretest.
1289        - Added an listing of the option settings after cmake has run.
1291    4.  A user submitted a patch to Makefile that makes it easy to create
1292        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1293        Makefile.am that cause it to include this special target, without
1294        affecting anything else. Note that the same mingw target plus all
1295        the other distribution libraries and programs are now supported
1296        when configuring with CMake (see 6 below) instead of with
1297        Configure/Make.
1299    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1300        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1301        exported in the Windows port". It has not yet been confirmed that the patch
1302        solves the problem, but it does no harm.
1304    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1305        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1306        with CMake, and also correct the comment about stack recursion.
1308    7.  Remove the automatic building of pcredemo from the ./configure system and
1309        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1310        of a program that users should build themselves after PCRE is installed, so
1311        building it automatically is not really right. What is more, it gave
1312        trouble in some build environments.
1314    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1317    Version 7.5 10-Jan-08
1318    ---------------------
1320    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1321        values in parens when parsing an RE using the C++ wrapper."
1323    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1324        Characters greater than 255 were excluded from the class instead of being
1325        included.
1327    3.  The same bug as (2) above applied to negated POSIX classes such as
1328        [:^space:].
1330    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1331        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1332        I have changed it.
1334    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1335        first named subpattern) and a construct such as (?&a) would reference the
1336        first named subpattern whose name started with "a" (in other words, the
1337        length check was missing). Both these problems are fixed. "Subpattern name
1338        expected" is now given for (?&) (a zero-length name), and this patch also
1339        makes it give the same error for \k'' (previously it complained that that
1340        was a reference to a non-existent subpattern).
1342    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1343        this is right because (?- can be followed by option settings as well as by
1344        digits. I have, however, made the messages clearer.
1346    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1347        than the number used in the conditional) now cause a compile-time error.
1348        This is actually not compatible with Perl, which accepts such patterns, but
1349        treats the conditional as always being FALSE (as PCRE used to), but it
1350        seems to me that giving a diagnostic is better.
1352    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1353        and messages.
1355    9.  Fix two occurrences of "backslash" in comments that should have been
1356        "backspace".
1358    10. Remove two redundant lines of code that can never be obeyed (their function
1359        was moved elsewhere).
1361    11. The program that makes PCRE's Unicode character property table had a bug
1362        which caused it to generate incorrect table entries for sequences of
1363        characters that have the same character type, but are in different scripts.
1364        It amalgamated them into a single range, with the script of the first of
1365        them. In other words, some characters were in the wrong script. There were
1366        thirteen such cases, affecting characters in the following ranges:
1368          U+002b0 - U+002c1
1369          U+0060c - U+0060d
1370          U+0061e - U+00612
1371          U+0064b - U+0065e
1372          U+0074d - U+0076d
1373          U+01800 - U+01805
1374          U+01d00 - U+01d77
1375          U+01d9b - U+01dbf
1376          U+0200b - U+0200f
1377          U+030fc - U+030fe
1378          U+03260 - U+0327f
1379          U+0fb46 - U+0fbb1
1380          U+10450 - U+1049d
1382    12. The -o option (show only the matching part of a line) for pcregrep was not
1383        compatible with GNU grep in that, if there was more than one match in a
1384        line, it showed only the first of them. It now behaves in the same way as
1385        GNU grep.
1387    13. If the -o and -v options were combined for pcregrep, it printed a blank
1388        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1389        does the same. The return code can be used to tell if there were any
1390        non-matching lines.
1392    14. Added --file-offsets and --line-offsets to pcregrep.
1394    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1395        infinitely looping recursion. The bug was that positive lookaheads were not
1396        being skipped when checking for a possible empty match (negative lookaheads
1397        and both kinds of lookbehind were skipped).
1399    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1400        inclusion of <windows.h> to before rather than after the definition of
1401        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1403    17. Specifying a possessive quantifier with a specific limit for a Unicode
1404        character property caused pcre_compile() to compile bad code, which led at
1405        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1406        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1407        caused the error; without that there was no problem.
1409    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1411    19. Added --enable-pcretest-libreadline.
1413    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1414        RE::GlobalReplace(). As a result, the number of replacements returned was
1415        double what it should be. I removed one of the increments, but Craig sent a
1416        later patch that removed the other one (the right fix) and added unit tests
1417        that check the return values (which was not done before).
1419    21. Several CMake things:
1421        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1422            the names libpcre and libpcreposix, not just pcre and pcreposix.
1424        (2) The above change means that pcretest and pcregrep are now correctly
1425            linked with the newly-built libraries, not previously installed ones.
1429    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1430        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1431        UTF-8 newline character). The key issue is that the pattern starts .*;
1432        this means that the match must be either at the beginning, or after a
1433        newline. The bug was in the code for advancing after a failed match and
1434        checking that the new position followed a newline. It was not taking
1435        account of UTF-8 characters correctly.
1437    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1438        character classes. PCRE was not treating the sequence [:...:] as a
1439        character class unless the ... were all letters. Perl, however, seems to
1440        allow any characters between [: and :], though of course it rejects as
1441        unknown any "names" that contain non-letters, because all the known class
1442        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1443        for example, whereas PCRE did not - it did not recognize a POSIX character
1444        class. This seemed a bit dangerous, so the code has been changed to be
1445        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1446        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1447        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1448        Perl does, and where it didn't before.
1450    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1451        Windows environments %n is disabled by default.
1454    Version 7.4 21-Sep-07
1455    ---------------------
1457    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1458        means that a class such as [\s] counted as "explicit reference to CR or
1459        LF". That isn't really right - the whole point of the change was to try to
1460        help when there was an actual mention of one of the two characters. So now
1461        the change happens only if \r or \n (or a literal CR or LF) character is
1462        encountered.
1464    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1465        of both had grown to the point where there were only 3 bits left.
1466        Fortunately, there was spare space in the data structure, and so I have
1467        moved the internal flags into a new 16-bit field to free up more option
1468        bits.
1470    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1471        but did not set the internal JCHANGED flag - either of these is enough to
1472        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1473        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1474        start sets both bits.
1476    4.  Added options (at build time, compile time, exec time) to change \R from
1477        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1479    5.  doc/pcresyntax.html was missing from the distribution.
1481    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1482        compatibility, even though it is no longer used.
1484    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1485        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1486        windows.h file is present (where different names are used). [This was
1487        reversed later after testing - see 16 below.]
1489    8.  Changed all #include <config.h> to #include "config.h". There were also
1490        some further <pcre.h> cases that I changed to "pcre.h".
1492    9.  When pcregrep was used with the --colour option, it missed the line ending
1493        sequence off the lines that it output.
1495    10. It was pointed out to me that arrays of string pointers cause lots of
1496        relocations when a shared library is dynamically loaded. A technique of
1497        using a single long string with a table of offsets can drastically reduce
1498        these. I have refactored PCRE in four places to do this. The result is
1499        dramatic:
1501          Originally:                          290
1502          After changing UCP table:            187
1503          After changing error message table:   43
1504          After changing table of "verbs"       36
1505          After changing table of Posix names   22
1507        Thanks to the folks working on Gregex for glib for this insight.
1509    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1510        unicode-properties was also set.
1512    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1514    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1515        checked only for CRLF.
1517    14. Added casts to pcretest.c to avoid compiler warnings.
1519    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1521    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1522        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1523        entirely. This removes changes made in 7 above.
1525    17. The CMake files have been updated, and there is now more information about
1526        building with CMake in the NON-UNIX-USE document.
1529    Version 7.3 28-Aug-07
1530    ---------------------
1532     1. In the rejigging of the build system that eventually resulted in 7.1, the
1533        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1534        brackets there is not right, since it causes compilers to look for an
1535        installed pcre.h, not the version that is in the source that is being
1536        compiled (which of course may be different). I have changed it back to:
1538          #include "pcre.h"
1540        I have a vague recollection that the change was concerned with compiling in
1541        different directories, but in the new build system, that is taken care of
1542        by the VPATH setting the Makefile.
1544     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1545        when the subject happened to end in the byte 0x85 (e.g. if the last
1546        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1547        characters but of course it shouldn't be taken as a newline when it is part
1548        of another character. The bug was that, for an unlimited repeat of . in
1549        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1550        characters when looking for a newline.
1552     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1554     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1555        in debug output.
1557     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1558        long printing in the pcrecpp unittest when running under MinGW.
1560     6. ESC_K was left out of the EBCDIC table.
1562     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1563        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1564        limit also applies to "virtual nesting" when a pattern is recursive, and in
1565        this case 1000 isn't so big. I have been able to remove this limit at the
1566        expense of backing off one optimization in certain circumstances. Normally,
1567        when pcre_exec() would call its internal match() function recursively and
1568        immediately return the result unconditionally, it uses a "tail recursion"
1569        feature to save stack. However, when a subpattern that can match an empty
1570        string has an unlimited repetition quantifier, it no longer makes this
1571        optimization. That gives it a stack frame in which to save the data for
1572        checking that an empty string has been matched. Previously this was taken
1573        from the 1000-entry workspace that had been reserved. So now there is no
1574        explicit limit, but more stack is used.
1576     8. Applied Daniel's patches to solve problems with the import/export magic
1577        syntax that is required for Windows, and which was going wrong for the
1578        pcreposix and pcrecpp parts of the library. These were overlooked when this
1579        problem was solved for the main library.
1581     9. There were some crude static tests to avoid integer overflow when computing
1582        the size of patterns that contain repeated groups with explicit upper
1583        limits. As the maximum quantifier is 65535, the maximum group length was
1584        set at 30,000 so that the product of these two numbers did not overflow a
1585        32-bit integer. However, it turns out that people want to use groups that
1586        are longer than 30,000 bytes (though not repeat them that many times).
1587        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1588        made it possible to implement the integer overflow checks in a much more
1589        dynamic way, which I have now done. The artificial limitation on group
1590        length has been removed - we now have only the limit on the total length of
1591        the compiled pattern, which depends on the LINK_SIZE setting.
1593    10. Fixed a bug in the documentation for get/copy named substring when
1594        duplicate names are permitted. If none of the named substrings are set, the
1595        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1596        empty string.
1598    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1599        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1600        because the ] is interpreted as the first data character and the
1601        terminating ] is not found. PCRE has been made compatible with Perl in this
1602        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1603        cause memory overwriting.
1605    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1606        string has been matched (to stop an infinite loop). It was not recognizing
1607        a conditional subpattern that could match an empty string if that
1608        subpattern was within another subpattern. For example, it looped when
1609        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1610        condition was not nested. This bug has been fixed.
1612    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1613        past the start of the subject in the presence of bytes with the top bit
1614        set, for example "\x8aBCD".
1616    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1617        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1619    14. Optimized (?!) to (*FAIL).
1621    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1622        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1623        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1624        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1625        does: it's just the validity check that is more restrictive.
1627    16. Inserted checks for integer overflows during escape sequence (backslash)
1628        processing, and also fixed erroneous offset values for syntax errors during
1629        backslash processing.
1631    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1632        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1634    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1635        caused an overrun.
1637    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1638        something other than just ASCII characters) inside a group that had an
1639        unlimited repeat caused a loop at compile time (while checking to see
1640        whether the group could match an empty string).
1642    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1643        [\P{Any}] did so. (Error in the code for printing property names.)
1645    21. An orphan \E inside a character class could cause a crash.
1647    22. A repeated capturing bracket such as (A)? could cause a wild memory
1648        reference during compilation.
1650    23. There are several functions in pcre_compile() that scan along a compiled
1651        expression for various reasons (e.g. to see if it's fixed length for look
1652        behind). There were bugs in these functions when a repeated \p or \P was
1653        present in the pattern. These operators have additional parameters compared
1654        with \d, etc, and these were not being taken into account when moving along
1655        the compiled data. Specifically:
1657        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1658            length.
1660        (b) An item such as \pL+ within a repeated group could cause crashes or
1661            loops.
1663        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1664            "reference to non-existent subpattern" error.
1666        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1668    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1669        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1671    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1673    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1674        character were causing crashes (broken optimization).
1676    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1677        \p or \P) caused a compile-time loop.
1679    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1680        break. For example, the unstudied pattern [\r\n]A does not match the string
1681        "\r\nA" because change 7.0/46 below moves the current point on by two
1682        characters after failing to match at the start. However, the pattern \nA
1683        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1684        the same is true. There doesn't seem any very clean way out of this, but
1685        what I have chosen to do makes the common cases work: PCRE now takes note
1686        of whether there can be an explicit match for \r or \n anywhere in the
1687        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1688        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1689        pattern has explicit CR or LF references.
1691    29. Added (*CR) etc for changing newline setting at start of pattern.
1694    Version 7.2 19-Jun-07
1695    ---------------------
1697     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1698        which is apparently normally available under Windows.
1700     2. Re-jig the pcregrep tests with different newline settings in an attempt
1701        to make them independent of the local environment's newline setting.
1703     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1705     4. Some of the "internals" tests were previously cut out when the link size
1706        was not 2, because the output contained actual offsets. The recent new
1707        "Z" feature of pcretest means that these can be cut out, making the tests
1708        usable with all link sizes.
1710     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1711        stack recursion. This gives a massive performance boost under BSD, but just
1712        a small improvement under Linux. However, it saves one field in the frame
1713        in all cases.
1715     6. Added more features from the forthcoming Perl 5.10:
1717        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1718            recursion call. It refers to the nth most recently opened parentheses.
1720        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1721            to be opened parentheses.
1723        (c) Conditions that refer to capturing parentheses can be specified
1724            relatively, for example, (?(-2)... or (?(+3)...
1726        (d) \K resets the start of the current match so that everything before
1727            is not part of it.
1729        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1731        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1732            reference syntax.
1734        (g) (?| introduces a group in which the numbering of parentheses in each
1735            alternative starts with the same number.
1737        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1739     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1742     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1743        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1744        for detecting groups that can match an empty string.
1746     9. A pattern with a very large number of alternatives (more than several
1747        hundred) was running out of internal workspace during the pre-compile
1748        phase, where pcre_compile() figures out how much memory will be needed. A
1749        bit of new cunning has reduced the workspace needed for groups with
1750        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1751        workspace instead of running out of the 4096 that are available.
1753    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1755    11. Applied patch from Google to remove an optimization that didn't quite work.
1756        The report of the bug said:
1758          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1759          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1760          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1762    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1763        it matched the wrong number of bytes.
1766    Version 7.1 24-Apr-07
1767    ---------------------
1769     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1770        that is more "standard", making use of automake and other Autotools. There
1771        is some re-arrangement of the files and adjustment of comments consequent
1772        on this.
1774     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1775        for recursive directory scanning broke on some systems because the files
1776        are not scanned in any specific order and on different systems the order
1777        was different. A call to "sort" has been inserted into RunGrepTest for the
1778        approprate test as a short-term fix. In the longer term there may be an
1779        alternative.
1781     3. I had an email from Eric Raymond about problems translating some of PCRE's
1782        man pages to HTML (despite the fact that I distribute HTML pages, some
1783        people do their own conversions for various reasons). The problems
1784        concerned the use of low-level troff macros .br and .in. I have therefore
1785        removed all such uses from the man pages (some were redundant, some could
1786        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1787        HTML has been updated to handle .nf/.fi and to complain if it encounters
1788        .br or .in.
1790     4. Updated comments in configure.ac that get placed in config.h.in and also
1791        arranged for config.h to be included in the distribution, with the name
1792        config.h.generic, for the benefit of those who have to compile without
1793        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1795     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1796        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1797        makevp.bat for the current PCRE, using the additional files
1798        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1800     6. A Windows user reported a minor discrepancy with test 2, which turned out
1801        to be caused by a trailing space on an input line that had got lost in his
1802        copy. The trailing space was an accident, so I've just removed it.
1804     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1805        that is needed.
1807     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1808        as "const" (a) because they are and (b) because it helps the PHP
1809        maintainers who have recently made a script to detect big data structures
1810        in the php code that should be moved to the .rodata section. I remembered
1811        to update Builducptable as well, so it won't revert if ucptable.h is ever
1812        re-created.
1814     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1815        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1816        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1817        support is not required. This saves 1.5-2K of code, which is important in
1818        some applications.
1820        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1821        so as not to refer to the tables, even though these functions will never be
1822        called when UTF-8 support is disabled. Otherwise there are problems with a
1823        shared library.
1825    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1827        (a) It was defining its arguments as char * instead of void *.
1829        (b) It was assuming that all moves were upwards in memory; this was true
1830            a long time ago when I wrote it, but is no longer the case.
1832        The emulated memove() is provided for those environments that have neither
1833        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1834        is clearly not the case, as these two bugs were recently reported.
1836    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1837        and Detrail to create the HTML documentation, the .txt form of the man
1838        pages, and it removes trailing spaces from listed files. It also creates
1839        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1840        case, it wraps all the #defines with #ifndefs. This script should be run
1841        before "make dist".
1843    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1844        with Unicode property support.
1846        (a) For a maximizing quantifier, if the two different cases of the
1847            character were of different lengths in their UTF-8 codings (there are
1848            some cases like this - I found 11), and the matching function had to
1849            back up over a mixture of the two cases, it incorrectly assumed they
1850            were both the same length.
1852        (b) When PCRE was configured to use the heap rather than the stack for
1853            recursion during matching, it was not correctly preserving the data for
1854            the other case of a UTF-8 character when checking ahead for a match
1855            while processing a minimizing repeat. If the check also involved
1856            matching a wide character, but failed, corruption could cause an
1857            erroneous result when trying to check for a repeat of the original
1858            character.
1860    13. Some tidying changes to the testing mechanism:
1862        (a) The RunTest script now detects the internal link size and whether there
1863            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1864            values substituted by "configure". (The RunGrepTest script already did
1865            this for UTF-8.) The configure.ac script no longer substitutes the
1866            relevant variables.
1868        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1869            with length and offset values. This means that the output is different
1870            for different internal link sizes. Test 2 is skipped for link sizes
1871            other than 2 because of this, bypassing the problem. Unfortunately,
1872            there was also a test in test 3 (the locale tests) that used /B and
1873            failed for link sizes other than 2. Rather than cut the whole test out,
1874            I have added a new /Z option to pcretest that replaces the length and
1875            offset values with spaces. This is now used to make test 3 independent
1876            of link size. (Test 2 will be tidied up later.)
1878    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1879        segmentation fault instead of returning the appropriate error message.
1881    15. In multiline mode when the newline sequence was set to "any", the pattern
1882        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1883        This doesn't seem right; it now treats the CRLF combination as the line
1884        ending, and so does not match in that case. It's only a pattern such as ^$
1885        that would hit this one: something like ^ABC$ would have failed after \r
1886        and then tried again after \r\n.
1888    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1889        in an attempt to make files that differ only in their line terminators
1890        compare equal. This works on Linux.
1892    17. Under certain error circumstances pcregrep might try to free random memory
1893        as it exited. This is now fixed, thanks to valgrind.
1895    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1896        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1897        was because its rules for how to advance for /g after matching an empty
1898        string at the end of a line did not allow for this case. They now check for
1899        it specially.
1901    20. pcretest is supposed to handle patterns and data of any length, by
1902        extending its buffers when necessary. It was getting this wrong when the
1903        buffer for a data line had to be extended.
1905    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1906        CRLF as a newline sequence.
1908    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1909        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1910        I have nevertheless tidied it up.
1912    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1914    24. Added a man page for pcre-config.
1917    Version 7.0 19-Dec-06
1918    ---------------------
1920     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1921        moving to gcc 4.1.1.
1923     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1924        sys/time.h, which is documented as needed for this function. It doesn't
1925        seem to matter on Linux, but it showed up on some releases of OS X.
1927     3. It seems that there are systems where bytes whose values are greater than
1928        127 match isprint() in the "C" locale. The "C" locale should be the
1929        default when a C program starts up. In most systems, only ASCII printing
1930        characters match isprint(). This difference caused the output from pcretest
1931        to vary, making some of the tests fail. I have changed pcretest so that:
1933        (a) When it is outputting text in the compiled version of a pattern, bytes
1934            other than 32-126 are always shown as hex escapes.
1936        (b) When it is outputting text that is a matched part of a subject string,
1937            it does the same, unless a different locale has been set for the match
1938            (using the /L modifier). In this case, it uses isprint() to decide.
1940     4. Fixed a major bug that caused incorrect computation of the amount of memory
1941        required for a compiled pattern when options that changed within the
1942        pattern affected the logic of the preliminary scan that determines the
1943        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1944        that the computed length was too small. The symptoms of this bug were
1945        either the PCRE error "internal error: code overflow" from pcre_compile(),
1946        or a glibc crash with a message such as "pcretest: free(): invalid next
1947        size (fast)". Examples of patterns that provoked this bug (shown in
1948        pcretest format) are:
1950          /(?-x: )/x
1951          /(?x)(?-x: \s*#\s*)/
1952          /((?i)[\x{c0}])/8
1953          /(?i:[\x{c0}])/8
1955        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1956        is now done differently.
1958     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1959        wrapper classes; (b) implement a new function in the C++ scanner that is
1960        more efficient than the old way of doing things because it avoids levels of
1961        recursion in the regex matching; (c) add a paragraph to the documentation
1962        for the FullMatch() function.
1964     6. The escape sequence \n was being treated as whatever was defined as
1965        "newline". Not only was this contrary to the documentation, which states
1966        that \n is character 10 (hex 0A), but it also went horribly wrong when
1967        "newline" was defined as CRLF. This has been fixed.
1969     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1970        was being set to -1 for the "end of line" case (supposedly a value that no
1971        character can have). Though this value is never used (the check for end of
1972        line is "zero bytes in current character"), it caused compiler complaints.
1973        I've changed it to 0xffffffff.
1975     8. In pcre_version.c, the version string was being built by a sequence of
1976        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1977        string (as it is for production releases) called a macro with an empty
1978        argument. The C standard says the result of this is undefined. The gcc
1979        compiler treats it as an empty string (which was what was wanted) but it is
1980        reported that Visual C gives an error. The source has been hacked around to
1981        avoid this problem.
1983     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1984        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1985        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1986        of them did).
1988    10. Originally, pcretest opened its input and output without "b"; then I was
1989        told that "b" was needed in some environments, so it was added for release
1990        5.0 to both the input and output. (It makes no difference on Unix-like
1991        systems.) Later I was told that it is wrong for the input on Windows. I've
1992        now abstracted the modes into two macros, to make it easier to fiddle with
1993        them, and removed "b" from the input mode under Windows.
1995    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1997    12. Added -help and --help to pcretest as an official way of being reminded
1998        of the options.
2000    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2001        and pcrecpp.cc because they annoy compilers at high warning levels.
2003    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2005    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2006        scripts are not C programs :-) and which was not noticed because it works
2007        on Linux.
2009    16. pcretest is supposed to handle any length of pattern and data line (as one
2010        line or as a continued sequence of lines) by extending its input buffer if
2011        necessary. This feature was broken for very long pattern lines, leading to
2012        a string of junk being passed to pcre_compile() if the pattern was longer
2013        than about 50K.
2015    17. I have done a major re-factoring of the way pcre_compile() computes the
2016        amount of memory needed for a compiled pattern. Previously, there was code
2017        that made a preliminary scan of the pattern in order to do this. That was
2018        OK when PCRE was new, but as the facilities have expanded, it has become
2019        harder and harder to keep it in step with the real compile phase, and there
2020        have been a number of bugs (see for example, 4 above). I have now found a
2021        cunning way of running the real compile function in a "fake" mode that
2022        enables it to compute how much memory it would need, while actually only
2023        ever using a few hundred bytes of working memory and without too many
2024        tests of the mode. This should make future maintenance and development
2025        easier. A side effect of this work is that the limit of 200 on the nesting
2026        depth of parentheses has been removed (though this was never a serious
2027        limitation, I suspect). However, there is a downside: pcre_compile() now
2028        runs more slowly than before (30% or more, depending on the pattern). I
2029        hope this isn't a big issue. There is no effect on runtime performance.
2031    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2032        newline (only possible for the last line of a file) and it was a
2033        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2035    19. Added additional timing features to pcretest. (1) The -tm option now times
2036        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2037        separate command line item, by a number that specifies the number of
2038        repeats to use when timing. The default is 50000; this gives better
2039        precision, but takes uncomfortably long for very large patterns.
2041    20. Extended pcre_study() to be more clever in cases where a branch of a
2042        subpattern has no definite first character. For example, (a*|b*)[cd] would
2043        previously give no result from pcre_study(). Now it recognizes that the
2044        first character must be a, b, c, or d.
2046    21. There was an incorrect error "recursive call could loop indefinitely" if
2047        a subpattern (or the entire pattern) that was being tested for matching an
2048        empty string contained only one non-empty item after a nested subpattern.
2049        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2050        incorrectly, because the \d was being skipped in the check.
2052    22. The pcretest program now has a new pattern option /B and a command line
2053        option -b, which is equivalent to adding /B to every pattern. This causes
2054        it to show the compiled bytecode, without the additional information that
2055        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2056        is the same as /B/I).
2058    23. A new optimization is now able automatically to treat some sequences such
2059        as a*b as a*+b. More specifically, if something simple (such as a character
2060        or a simple class like \d) has an unlimited quantifier, and is followed by
2061        something that cannot possibly match the quantified thing, the quantifier
2062        is automatically "possessified".
2064    24. A recursive reference to a subpattern whose number was greater than 39
2065        went wrong under certain circumstances in UTF-8 mode. This bug could also
2066        have affected the operation of pcre_study().
2068    25. Realized that a little bit of performance could be had by replacing
2069        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2071    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2073    27. Possessive quantifiers such as a++ were previously implemented by turning
2074        them into atomic groups such as ($>a+). Now they have their own opcodes,
2075        which improves performance. This includes the automatically created ones
2076        from 23 above.
2078    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2079        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2080        the first matched character to be a colon. This applied both to named and
2081        numbered groups.
2083    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2085    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2086        building PCRE on Windows easier, so I have included it in the distribution.
2088    31. There is now a check in pcretest against a ridiculously large number being
2089        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2090        loop, the loop is abandoned.
2092    32. Forward references to subpatterns in conditions such as (?(2)...) where
2093        subpattern 2 is defined later cause pcre_compile() to search forwards in
2094        the pattern for the relevant set of parentheses. This search went wrong
2095        when there were unescaped parentheses in a character class, parentheses
2096        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2098    33. "Subroutine" calls and backreferences were previously restricted to
2099        referencing subpatterns earlier in the regex. This restriction has now
2100        been removed.
2102    34. Added a number of extra features that are going to be in Perl 5.10. On the
2103        whole, these are just syntactic alternatives for features that PCRE had
2104        previously implemented using the Python syntax or my own invention. The
2105        other formats are all retained for compatibility.
2107        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2108            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2109            also .NET compatible.
2111        (b) A recursion or subroutine call to a named group can now be defined as
2112            (?&name) as well as (?P>name).
2114        (c) A backreference to a named group can now be defined as \k<name> or
2115            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2116            5.10, are also .NET compatible.
2118        (d) A conditional reference to a named group can now use the syntax
2119            (?(<name>) or (?('name') as well as (?(name).
2121        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2122            groups (named and numbered) that are never evaluated inline, but can be
2123            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2124            is always false. There may be only one alternative in such a group.
2126        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2127            as the simple (?(R). The condition is true only if the most recent
2128            recursion is that of the given number or name. It does not search out
2129            through the entire recursion stack.
2131        (g) The escape \gN or \g{N} has been added, where N is a positive or
2132            negative number, specifying an absolute or relative reference.
2134    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2135        some "unreachable code" warnings.
2137    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2138        things, this adds five new scripts.
2140    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2141        There were also incompatibilities regarding the handling of \Q..\E inside
2142        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2143        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2145    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2146        matches an empty string, and forcibly breaks the loop. There were bugs in
2147        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2148        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2149        separate and independent bugs (that affected different cases) have been
2150        fixed.
2152    39. Refactored the code to abolish the use of different opcodes for small
2153        capturing bracket numbers. This is a tidy that I avoided doing when I
2154        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2155        The new approach is not only tidier, it makes it possible to reduce the
2156        memory needed to fix the previous bug (38).
2158    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2159        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2160        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2161        mode.
2163    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2164        report.
2166    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2167        copy construction and assignment in the C++ wrapper.
2169    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2170        couple of bugs that could have given wrong results in the "--newline=crlf"
2171        case.
2173    44. Added a number of casts and did some reorganization of signed/unsigned int
2174        variables following suggestions from Dair Grant. Also renamed the variable
2175        "this" as "item" because it is a C++ keyword.
2177    45. Arranged for dftables to add
2179          #include "pcre_internal.h"
2181        to pcre_chartables.c because without it, gcc 4.x may remove the array
2182        definition from the final binary if PCRE is built into a static library and
2183        dead code stripping is activated.
2185    46. For an unanchored pattern, if a match attempt fails at the start of a
2186        newline sequence, and the newline setting is CRLF or ANY, and the next two
2187        characters are CRLF, advance by two characters instead of one.
2190    Version 6.7 04-Jul-06
2191    ---------------------
2193     1. In order to handle tests when input lines are enormously long, pcretest has
2194        been re-factored so that it automatically extends its buffers when
2195        necessary. The code is crude, but this _is_ just a test program. The
2196        default size has been increased from 32K to 50K.
2198     2. The code in pcre_study() was using the value of the re argument before
2199        testing it for NULL. (Of course, in any sensible call of the function, it
2200        won't be NULL.)
2202     3. The memmove() emulation function in pcre_internal.h, which is used on
2203        systems that lack both memmove() and bcopy() - that is, hardly ever -
2204        was missing a "static" storage class specifier.
2206     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2207        containing an extended class (one that cannot be represented by a bitmap
2208        because it contains high-valued characters or Unicode property items, e.g.
2209        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2210        pattern, but PCRE should not loop if you do not (it no longer does).
2211        [Detail: two cases were found: (a) a repeated subpattern containing an
2212        extended class; (b) a recursive reference to a subpattern that followed a
2213        previous extended class. It wasn't skipping over the extended class
2214        correctly when UTF-8 mode was not set.]
2216     5. A negated single-character class was not being recognized as fixed-length
2217        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2218        compile error "lookbehind assertion is not fixed length".
2220     6. The RunPerlTest auxiliary script was showing an unexpected difference
2221        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2222        write a Perl script that can interpret lines of an input file either as
2223        byte characters or as UTF-8, which is what "perltest" was being required to
2224        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2225        can't do is switch easily at run time between having the "use utf8;" pragma
2226        or not. In the end, I fudged it by using the RunPerlTest script to insert
2227        "use utf8;" explicitly for the UTF-8 tests.
2229     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2230        the end of the subject string, contrary to the documentation and to what
2231        Perl does. This was true of both matching functions. Now it matches only at
2232        the start of the subject and immediately after *internal* newlines.
2234     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2235        a pointer to an int instead of a pointer to an unsigned long int. This
2236        caused problems on 64-bit systems.
2238     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2239        instance of the 'standard' template library not being so standard".
2241    10. There was no check on the number of named subpatterns nor the maximum
2242        length of a subpattern name. The product of these values is used to compute
2243        the size of the memory block for a compiled pattern. By supplying a very
2244        long subpattern name and a large number of named subpatterns, the size
2245        computation could be caused to overflow. This is now prevented by limiting
2246        the length of names to 32 characters, and the number of named subpatterns
2247        to 10,000.
2249    11. Subpatterns that are repeated with specific counts have to be replicated in
2250        the compiled pattern. The size of memory for this was computed from the
2251        length of the subpattern and the repeat count. The latter is limited to
2252        65535, but there was no limit on the former, meaning that integer overflow
2253        could in principle occur. The compiled length of a repeated subpattern is
2254        now limited to 30,000 bytes in order to prevent this.
2256    12. Added the optional facility to have named substrings with the same name.
2258    13. Added the ability to use a named substring as a condition, using the
2259        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2260        are numbers (not recommended). Forward references are permitted.
2262    14. Added forward references in named backreferences (if you see what I mean).
2264    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2265        pattern could run off the end of the subject. For example, the pattern
2266        "(?s)(.{1,5})"8 did this with the subject "ab".
2268    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2269        PCRE_CASELESS was set when matching characters that were quantified with ?
2270        or *.
2272    17. A character class other than a single negated character that had a minimum
2273        but no maximum quantifier - for example [ab]{6,} - was not handled
2274        correctly by pce_dfa_exec(). It would match only one character.
2276    18. A valid (though odd) pattern that looked like a POSIX character
2277        class but used an invalid character after [ (for example [[,abc,]]) caused
2278        pcre_compile() to give the error "Failed: internal error: code overflow" or
2279        in some cases to crash with a glibc free() error. This could even happen if
2280        the pattern terminated after [[ but there just happened to be a sequence of
2281        letters, a binary zero, and a closing ] in the memory that followed.
2283    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2284        over the years. Originally (before any Unicode support), just the bottom 8
2285        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2286        output from "man perlunicode" includes this:
2288          The regular expression compiler produces polymorphic opcodes.  That
2289          is, the pattern adapts to the data and automatically switches to
2290          the Unicode character scheme when presented with Unicode data--or
2291          instead uses a traditional byte scheme when presented with byte
2292          data.
2294        Sadly, a wide octal escape does not cause a switch, and in a string with
2295        no other multibyte characters, these octal escapes are treated as before.
2296        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2297        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2298        Unicode string.
2300        I have not perpetrated such confusion in PCRE. Up till now, it took just
2301        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2302        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2303        translate to the appropriate multibyte character.
2305    29. Applied some refactoring to reduce the number of warnings from Microsoft
2306        and Borland compilers. This has included removing the fudge introduced
2307        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2308        a warning about an unused variable.
2310    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2311        characters since release 4.0, because Perl (from release 5.004) does not.
2312        [Or at least, is documented not to: some releases seem to be in conflict
2313        with the documentation.] However, when a pattern was studied with
2314        pcre_study() and all its branches started with \s, PCRE still included VT
2315        as a possible starting character. Of course, this did no harm; it just
2316        caused an unnecessary match attempt.
2318    22. Removed a now-redundant internal flag bit that recorded the fact that case
2319        dependency changed within the pattern. This was once needed for "required
2320        byte" processing, but is no longer used. This recovers a now-scarce options
2321        bit. Also moved the least significant internal flag bit to the most-
2322        significant bit of the word, which was not previously used (hangover from
2323        the days when it was an int rather than a uint) to free up another bit for
2324        the future.
2326    23. Added support for CRLF line endings as well as CR and LF. As well as the
2327        default being selectable at build time, it can now be changed at runtime
2328        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2329        specify that it is scanning data with non-default line endings.
2331    24. Changed the definition of CXXLINK to make it agree with the definition of
2332        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2334    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2335        recursions. This makes a big different to stack usage for some patterns.
2337    26. If a subpattern containing a named recursion or subroutine reference such
2338        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2339        the space required for the compiled pattern went wrong and gave too small a
2340        value. Depending on the environment, this could lead to "Failed: internal
2341        error: code overflow at offset 49" or "glibc detected double free or
2342        corruption" errors.
2344    27. Applied patches from Google (a) to support the new newline modes and (b) to
2345        advance over multibyte UTF-8 characters in GlobalReplace.
2347    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2348        difference for some implementation of PCRE in some Windows version.
2350    29. Added some extra testing facilities to pcretest:
2352        \q<number>   in a data line sets the "match limit" value
2353        \Q<number>   in a data line sets the "match recursion limt" value
2354        -S <number>  sets the stack size, where <number> is in megabytes
2356        The -S option isn't available for Windows.
2359    Version 6.6 06-Feb-06
2360    ---------------------
2362     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2363        in pcreposix.h. I have copied the definition from pcre.h.
2365     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2366        because pcre.h is no longer a built file.
2368     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2369        not normally included in the compiled code.
2372    Version 6.5 01-Feb-06
2373    ---------------------
2375     1. When using the partial match feature with pcre_dfa_exec(), it was not
2376        anchoring the second and subsequent partial matches at the new starting
2377        point. This could lead to incorrect results. For example, with the pattern
2378        /1234/, partially matching against "123" and then "a4" gave a match.
2380     2. Changes to pcregrep:
2382        (a) All non-match returns from pcre_exec() were being treated as failures
2383            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2384            error message is output. Some extra information is given for the
2385            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2386            probably the only errors that are likely to be caused by users (by
2387            specifying a regex that has nested indefinite repeats, for instance).
2388            If there are more than 20 of these errors, pcregrep is abandoned.
2390        (b) A binary zero was treated as data while matching, but terminated the
2391            output line if it was written out. This has been fixed: binary zeroes
2392            are now no different to any other data bytes.
2394        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2395            used to set a locale for matching. The --locale=xxxx long option has
2396            been added (no short equivalent) to specify a locale explicitly on the
2397            pcregrep command, overriding the environment variables.
2399        (d) When -B was used with -n, some line numbers in the output were one less
2400            than they should have been.
2402        (e) Added the -o (--only-matching) option.
2404        (f) If -A or -C was used with -c (count only), some lines of context were
2405            accidentally printed for the final match.
2407        (g) Added the -H (--with-filename) option.
2409        (h) The combination of options -rh failed to suppress file names for files
2410            that were found from directory arguments.
2412        (i) Added the -D (--devices) and -d (--directories) options.
2414        (j) Added the -F (--fixed-strings) option.
2416        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2418        (l) Added the --colo(u)r option.
2420        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2421            is not present by default.
2423     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2424        items such as (?R) or (?1), when the recursion could match a number of
2425        alternatives. If it matched one of the alternatives, but subsequently,
2426        outside the recursion, there was a failure, the code tried to back up into
2427        the recursion. However, because of the way PCRE is implemented, this is not
2428        possible, and the result was an incorrect result from the match.
2430        In order to prevent this happening, the specification of recursion has
2431        been changed so that all such subpatterns are automatically treated as
2432        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2434     4. I had overlooked the fact that, in some locales, there are characters for
2435        which isalpha() is true but neither isupper() nor islower() are true. In
2436        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2437        and ordfeminine) are like this. This affected the treatment of \w and \W
2438        when they appeared in character classes, but not when they appeared outside
2439        a character class. The bit map for "word" characters is now created
2440        separately from the results of isalnum() instead of just taking it from the
2441        upper, lower, and digit maps. (Plus the underscore character, of course.)
2443     5. The above bug also affected the handling of POSIX character classes such as
2444        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2445        permanent tables. Instead, the bit maps for such a class were previously
2446        created as the appropriate unions of the upper, lower, and digit bitmaps.
2447        Now they are created by subtraction from the [[:word:]] class, which has
2448        its own bitmap.
2450     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2451        It is created by subtracting the vertical space characters (\x09, \x0a,
2452        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2453        subtraction was done in the overall bitmap for a character class, meaning
2454        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2455        be recognized. This bug has been fixed.
2457     7. Patches from the folks at Google:
2459          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2460          real life, but is still worth protecting against".
2462          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2463          regular expressions".
2465          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2466          have it.
2468          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2469          "configure" and the latter not, in order to fix a problem somebody had
2470          with compiling the Arg class on HP-UX.
2472          (e) Improve the error-handling of the C++ wrapper a little bit.
2474          (f) New tests for checking recursion limiting.
2476     8. The pcre_memmove() function, which is used only if the environment does not
2477        have a standard memmove() function (and is therefore rarely compiled),
2478        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2479        returning a result (though PCRE never actually uses the result).
2481     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2482        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2483        returned instead of calling malloc() with an overflowing number that would
2484        most likely cause subsequent chaos.
2486    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2488    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2489        with this option is matched, the nmatch and pmatch options of regexec() are
2490        ignored.
2492    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2493        provided in case anyone wants to the the POSIX interface with UTF-8
2494        strings.
2496    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2497        C++ linking (needed for some HP-UX environments).
2499    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2500        (unused parameter) and in the pcre_printint() function (omitted "default"
2501        switch label when the default is to do nothing).
2503    15. Added some code to make it possible, when PCRE is compiled as a C++
2504        library, to replace subject pointers for pcre_exec() with a smart pointer
2505        class, thus making it possible to process discontinuous strings.
2507    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2508        much the same function. They were added by different people who were trying
2509        to make PCRE easy to compile on non-Unix systems. It has been suggested
2510        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2511        for compiling on Windows systems. I have therefore replaced it with
2512        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2513        defaults to "extern" for C or "extern C" for C++, which works fine on
2514        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2515        SCOPE with something explicit in config.h. In addition:
2517        (a) pcreposix.h still had just "extern" instead of either of these macros;
2518            I have replaced it with PCRE_DATA_SCOPE.
2520        (b) Functions such as _pcre_xclass(), which are internal to the library,
2521            but external in the C sense, all had PCRE_EXPORT in their definitions.
2522            This is apparently wrong for the Windows case, so I have removed it.
2523            (It makes no difference on Unix-like systems.)
2525    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2526        of recursive calls to match(). This is different to MATCH_LIMIT because
2527        that limits the total number of calls to match(), not all of which increase
2528        the depth of recursion. Limiting the recursion depth limits the amount of
2529        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2530        when PCRE is compiled, and changed at run time. A patch from Google adds
2531        this functionality to the C++ interface.
2533    18. Changes to the handling of Unicode character properties:
2535        (a) Updated the table to Unicode 4.1.0.
2537        (b) Recognize characters that are not in the table as "Cn" (undefined).
2539        (c) I revised the way the table is implemented to a much improved format
2540            which includes recognition of ranges. It now supports the ranges that
2541            are defined in UnicodeData.txt, and it also amalgamates other
2542            characters into ranges. This has reduced the number of entries in the
2543            table from around 16,000 to around 3,000, thus reducing its size
2544            considerably. I realized I did not need to use a tree structure after
2545            all - a binary chop search is just as efficient. Having reduced the
2546            number of entries, I extended their size from 6 bytes to 8 bytes to
2547            allow for more data.
2549        (d) Added support for Unicode script names via properties such as \p{Han}.
2551    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2552        matching that character.
2554    20. When matching a repeated Unicode property with a minimum greater than zero,
2555        (for example \pL{2,}), PCRE could look past the end of the subject if it
2556        reached it while seeking the minimum number of characters. This could
2557        happen only if some of the characters were more than one byte long, because
2558        there is a check for at least the minimum number of bytes.
2560    21. Refactored the implementation of \p and \P so as to be more general, to
2561        allow for more different types of property in future. This has changed the
2562        compiled form incompatibly. Anybody with saved compiled patterns that use
2563        \p or \P will have to recompile them.
2565    22. Added "Any" and "L&" to the supported property types.
2567    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2568        but give a compile time error if the value is greater than 0xff.
2570    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2571        accidentally not being installed or uninstalled.
2573    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2574        made were to insert the current release number. This seemed silly, because
2575        it made things harder for people building PCRE on systems that don't run
2576        "configure". I have turned pcre.h into a distributed file, no longer built
2577        by "configure", with the version identification directly included. There is
2578        no longer a pcre.h.in file.
2580        However, this change necessitated a change to the pcre-config script as
2581        well. It is built from pcre-config.in, and one of the substitutions was the
2582        release number. I have updated configure.ac so that ./configure now finds
2583        the release number by grepping pcre.h.
2585    26. Added the ability to run the tests under valgrind.
2588    Version 6.4 05-Sep-05
2589    ---------------------
2591     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2592        "--" to be printed when multiple files were scanned, even when none of the
2593        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2594        consider it to be a bug, and have restored the previous behaviour.
2596     2. A couple of code tidies to get rid of compiler warnings.
2598     3. The pcretest program used to cheat by referring to symbols in the library
2599        whose names begin with _pcre_. These are internal symbols that are not
2600        really supposed to be visible externally, and in some environments it is
2601        possible to suppress them. The cheating is now confined to including
2602        certain files from the library's source, which is a bit cleaner.
2604     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2605        file's purpose clearer.
2607     5. Reorganized pcre_ucp_findchar().
2610    Version 6.3 15-Aug-05
2611    ---------------------
2613     1. The file libpcre.pc.in did not have general read permission in the tarball.
2615     2. There were some problems when building without C++ support:
2617        (a) If C++ support was not built, "make install" and "make test" still
2618            tried to test it.
2620        (b) There were problems when the value of CXX was explicitly set. Some
2621            changes have been made to try to fix these, and ...
2623        (c) --disable-cpp can now be used to explicitly disable C++ support.
2625        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2626            backslash in a target when C++ was disabled. This confuses some
2627            versions of "make", apparently. Using an intermediate variable solves
2628            this. (Same for CPP_LOBJ.)
2630     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2631        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2632        necessary on certain architectures.
2634     4. Added a setting of -export-symbols-regex to the link command to remove
2635        those symbols that are exported in the C sense, but actually are local
2636        within the library, and not documented. Their names all begin with
2637        "_pcre_". This is not a perfect job, because (a) we have to except some
2638        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2639        available (and never for static libraries). I have made a note to try to
2640        find a way round (a) in the future.
2643    Version 6.2 01-Aug-05
2644    ---------------------
2646     1. There was no test for integer overflow of quantifier values. A construction
2647        such as {1111111111111111} would give undefined results. What is worse, if
2648        a minimum quantifier for a parenthesized subpattern overflowed and became
2649        negative, the calculation of the memory size went wrong. This could have
2650        led to memory overwriting.
2652     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2654     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2655        operating environments where this matters.
2657     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2658        PCRE options from within the C++ wrapper.
2660     5. Named capturing subpatterns were not being correctly counted when a pattern
2661        was compiled. This caused two problems: (a) If there were more than 100
2662        such subpatterns, the calculation of the memory needed for the whole
2663        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2664        back references of the form \12, where the number was greater than 9, were
2665        not recognized as back references, even though there were sufficient
2666        previous subpatterns.
2668     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2669        versions of gcc, e.g. 2.95.4.
2672    Version 6.1 21-Jun-05
2673    ---------------------
2675     1. There was one reference to the variable "posix" in pcretest.c that was not
2676        surrounded by "#if !defined NOPOSIX".
2678     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2679        the cross-check on the old pcre_info() function, for the benefit of the
2680        cut-down version of PCRE that is currently imported into Exim.
2682     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2683        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2684        patterns that start (?i) but I don't think that matters. The use of (?i) is
2685        just an example; this all applies to the other options as well.
2687     4. Since libtool seems to echo the compile commands it is issuing, the output
2688        from "make" can be reduced a bit by putting "@" in front of each libtool
2689        compile command.
2691     5. Patch from the folks at Google for configure.in to be a bit more thorough
2692        in checking for a suitable C++ installation before trying to compile the
2693        C++ stuff. This should fix a reported problem when a compiler was present,
2694        but no suitable headers.
2696     6. The man pages all had just "PCRE" as their title. I have changed them to
2697        be the relevant file name. I have also arranged that these names are
2698        retained in the file doc/pcre.txt, which is a concatenation in text format
2699        of all the man pages except the little individual ones for each function.
2701     7. The NON-UNIX-USE file had not been updated for the different set of source
2702        files that come with release 6. I also added a few comments about the C++
2703        wrapper.
2706    Version 6.0 07-Jun-05
2707    ---------------------
2709     1. Some minor internal re-organization to help with my DFA experiments.
2711     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2712        didn't matter for the library itself when fully configured, but did matter
2713        when compiling without UCP support, or within Exim, where the ucp files are
2714        not imported.
2716     3. Refactoring of the library code to split up the various functions into
2717        different source modules. The addition of the new DFA matching code (see
2718        below) to a single monolithic source would have made it really too
2719        unwieldy, quite apart from causing all the code to be include in a
2720        statically linked application, when only some functions are used. This is
2721        relevant even without the DFA addition now that patterns can be compiled in
2722        one application and matched in another.
2724        The downside of splitting up is that there have to be some external
2725        functions and data tables that are used internally in different modules of
2726        the library but which are not part of the API. These have all had their
2727        names changed to start with "_pcre_" so that they are unlikely to clash
2728        with other external names.
2730     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2731        a different (DFA) algorithm. Although it is slower than the original
2732        function, it does have some advantages for certain types of matching
2733        problem.
2735     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2736        including restarting after a partial match.
2738     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2739        defined when compiling for Windows was sent to me. I have put it into the
2740        code, though I have no means of testing or verifying it.
2742     7. Added the pcre_refcount() auxiliary function.
2744     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2745        match before or at the first newline in the subject string. In pcretest,
2746        the /f option on a pattern can be used to set this.
2748     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2749        would behave wrongly. This has been present in PCRE since release 4.0.
2751    10. A number of changes to the pcregrep command:
2753        (a) Refactored how -x works; insert ^(...)$ instead of setting
2754            PCRE_ANCHORED and checking the length, in preparation for adding
2755            something similar for -w.
2757        (b) Added the -w (match as a word) option.
2759        (c) Refactored the way lines are read and buffered so as to have more
2760            than one at a time available.
2762        (d) Implemented a pcregrep test script.
2764        (e) Added the -M (multiline match) option. This allows patterns to match
2765            over several lines of the subject. The buffering ensures that at least
2766            8K, or the rest of the document (whichever is the shorter) is available
2767            for matching (and similarly the previous 8K for lookbehind assertions).
2769        (f) Changed the --help output so that it now says
2771              -w, --word-regex(p)
2773            instead of two lines, one with "regex" and the other with "regexp"
2774            because that confused at least one person since the short forms are the
2775            same. (This required a bit of code, as the output is generated
2776            automatically from a table. It wasn't just a text change.)
2778        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2779            option but starts with a hyphen. Could be a pattern or a path name
2780            starting with a hyphen, for instance.
2782        (h) "-" can be given as a file name to represent stdin.
2784        (i) When file names are being printed, "(standard input)" is used for
2785            the standard input, for compatibility with GNU grep. Previously
2786            "<stdin>" was used.
2788        (j) The option --label=xxx can be used to supply a name to be used for
2789            stdin when file names are being printed. There is no short form.
2791        (k) Re-factored the options decoding logic because we are going to add
2792            two more options that take data. Such options can now be given in four
2793            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2795        (l) Added the -A, -B, and -C options for requesting that lines of context
2796            around matches be printed.
2798        (m) Added the -L option to print the names of files that do not contain
2799            any matching lines, that is, the complement of -l.
2801        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2802            continue to scan other files.
2804        (o) The -s option was incorrectly implemented. For compatibility with other
2805            greps, it now suppresses the error message for a non-existent or non-
2806            accessible file (but not the return code). There is a new option called
2807            -q that suppresses the output of matching lines, which was what -s was
2808            previously doing.
2810        (p) Added --include and --exclude options to specify files for inclusion
2811            and exclusion when recursing.
2813    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2814        Hopefully, it now does.
2816    12. Missing cast in pcre_study().
2818    13. Added an "uninstall" target to the makefile.
2820    14. Replaced "extern" in the function prototypes in Makefile.in with
2821        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2822        world, but is set differently for Windows.
2824    15. Added a second compiling function called pcre_compile2(). The only
2825        difference is that it has an extra argument, which is a pointer to an
2826        integer error code. When there is a compile-time failure, this is set
2827        non-zero, in addition to the error test pointer being set to point to an
2828        error message. The new argument may be NULL if no error number is required
2829        (but then you may as well call pcre_compile(), which is now just a
2830        wrapper). This facility is provided because some applications need a
2831        numeric error indication, but it has also enabled me to tidy up the way
2832        compile-time errors are handled in the POSIX wrapper.
2834    16. Added VPATH=.libs to the makefile; this should help when building with one
2835        prefix path and installing with another. (Or so I'm told by someone who
2836        knows more about this stuff than I do.)
2838    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2839        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2840        match everything, including newlines. This is not POSIX-compatible, but
2841        somebody wanted the feature. From pcretest it can be activated by using
2842        both the P and the s flags.
2844    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2846    19. libpcre.pc was being incorrectly installed as executable.
2848    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2849        it now also looks for '\r' so that it will work unmodified on Windows.
2851    21. Added Google's contributed C++ wrapper to the distribution.
2853    22. Added some untidy missing memory free() calls in pcretest, to keep
2854        Electric Fence happy when testing.
2858    Version 5.0 13-Sep-04
2859    ---------------------
2861     1. Internal change: literal characters are no longer packed up into items
2862        containing multiple characters in a single byte-string. Each character
2863        is now matched using a separate opcode. However, there may be more than one
2864        byte in the character in UTF-8 mode.
2866     2. The pcre_callout_block structure has two new fields: pattern_position and
2867        next_item_length. These contain the offset in the pattern to the next match
2868        item, and its length, respectively.
2870     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2871        insertion of callouts before each pattern item. Added the /C option to
2872        pcretest to make use of this.
2874     4. On the advice of a Windows user, the lines
2876          #if defined(_WIN32) || defined(WIN32)
2877          _setmode( _fileno( stdout ), 0x8000 );
2878          #endif  /* defined(_WIN32) || defined(WIN32) */
2880        have been added to the source of pcretest. This apparently does useful
2881        magic in relation to line terminators.
2883     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2884        for the benefit of those environments where the "b" makes a difference.
2886     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2887        to know about it. I have put a hack into configure.in that adds in code
2888        to set GCC=yes if CC=icc. This seems to end up at a point in the
2889        generated configure script that is early enough to affect the setting of
2890        compiler options, which is what is needed, but I have no means of testing
2891        whether it really works. (The user who reported this had patched the
2892        generated configure script, which of course I cannot do.)
2894        LATER: After change 22 below (new libtool files), the configure script
2895        seems to know about icc (and also ecc). Therefore, I have commented out
2896        this hack in configure.in.
2898     7. Added support for pkg-config (2 patches were sent in).
2900     8. Negated POSIX character classes that used a combination of internal tables
2901        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2902        [[:^ascii]]. Typically, they would match almost any characters. The other
2903        POSIX classes were not broken in this way.
2905     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2906        to find the match, as PCRE was deluded into thinking that the match had to
2907        start at the start point or following a newline. The same bug applied to
2908        patterns with negative forward assertions or any backward assertions
2909        preceding ".*" at the start, unless the pattern required a fixed first
2910        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2912    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2913        starting at the last subject character, bytes beyond the end of the subject
2914        string were read.
2916    11. Renamed the variable "class" as "classbits" to make life easier for C++
2917        users. (Previously there was a macro definition, but it apparently wasn't
2918        enough.)
2920    12. Added the new field "tables" to the extra data so that tables can be passed
2921        in at exec time, or the internal tables can be re-selected. This allows
2922        a compiled regex to be saved and re-used at a later time by a different
2923        program that might have everything at different addresses.
2925    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2926        -R library as well as a -L library.
2928    14. The debugging options of pcretest (-d on the command line or D on a
2929        pattern) showed incorrect output for anything following an extended class
2930        that contained multibyte characters and which was followed by a quantifier.
2932    15. Added optional support for general category Unicode character properties
2933        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2934        support. It adds about 90K to the size of the library. The meanings of the
2935        inbuilt class escapes such as \d and \s have NOT been changed.
2937    16. Updated pcredemo.c to include calls to free() to release the memory for the
2938        compiled pattern.
2940    17. The generated file chartables.c was being created in the source directory
2941        instead of in the building directory. This caused the build to fail if the
2942        source directory was different from the building directory, and was
2943        read-only.
2945    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2946        file. No doubt somebody will tell me if they don't make sense... Also added
2947        Dan Mooney's comments about building on OpenVMS.
2949    19. Added support for partial matching via the PCRE_PARTIAL option for
2950        pcre_exec() and the \P data escape in pcretest.
2952    20. Extended pcretest with 3 new pattern features:
2954        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2955              write the compiled pattern to the file whose name is "rest-of-line".
2956              This is a straight binary dump of the data, with the saved pointer to
2957              the character tables forced to be NULL. The study data, if any, is
2958              written too. After writing, pcretest reads a new pattern.
2960        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2961              compiled pattern from the given file. There must not be any
2962              occurrences of "<" in the file name (pretty unlikely); if there are,
2963              pcretest will instead treat the initial "<" as a pattern delimiter.
2964              After reading in the pattern, pcretest goes on to read data lines as
2965              usual.
2967        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2968              and 16-bit fields in a compiled pattern, to simulate a pattern that
2969              was compiled on a host of opposite endianness.
2971    21. The pcre-exec() function can now cope with patterns that were compiled on
2972        hosts of opposite endianness, with this restriction:
2974          As for any compiled expression that is saved and used later, the tables
2975          pointer field cannot be preserved; the extra_data field in the arguments
2976          to pcre_exec() should be used to pass in a tables address if a value
2977          other than the default internal tables were used at compile time.
2979    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2980        now diagnosed as an error. Previously, most of the time, a negative number
2981        would have been treated as zero, but if in addition "ovector" was passed as
2982        NULL, a crash could occur.
2984    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2985        new versions from the libtool 1.5 distribution (the last one is a copy of
2986        a file called libtool.m4). This seems to have fixed the need to patch
2987        "configure" to support Darwin 1.3 (which I used to do). However, I still
2988        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2989        workstation).
2991    24. Changed the PCRE licence to be the more standard "BSD" licence.
2994    Version 4.5 01-Dec-03
2995    ---------------------
2997     1. There has been some re-arrangement of the code for the match() function so
2998        that it can be compiled in a version that does not call itself recursively.
2999        Instead, it keeps those local variables that need separate instances for
3000        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3001        needs to "recurse". Keeping track of where control must go is done by means
3002        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3003        hide most of the details from the main code, and operates only if
3004        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3005        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3006        operating.
3008        To make it easier for callers to provide specially tailored get/free
3009        functions for this usage, two new functions, pcre_stack_malloc, and
3010        pcre_stack_free, are used. They are always called in strict stacking order,
3011        and the size of block requested is always the same.
3013        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3014        PCRE has been compiled to use the stack or the heap for recursion. The
3015        -C option of pcretest uses this to show which version is compiled.
3017        A new data escape \S, is added to pcretest; it causes the amounts of store
3018        obtained and freed by both kinds of malloc/free at match time to be added
3019        to the output.
3021     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3022        what's available on my current Linux desktop machine.
3024     3. When matching a UTF-8 string, the test for a valid string at the start has
3025        been extended. If start_offset is not zero, PCRE now checks that it points
3026        to a byte that is the start of a UTF-8 character. If not, it returns
3027        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3028        this is necessary because there may be backward assertions in the pattern.
3029        When matching the same subject several times, it may save resources to use
3030        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3032     4. The code for checking the validity of UTF-8 strings has been tightened so
3033        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3034        containing "overlong sequences".
3036     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3037        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3038        should have been "&", but it just so happened that all the cases this let
3039        through by mistake were picked up later in the function.
3041     6. I had used a variable called "isblank" - this is a C99 function, causing
3042        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3044     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3045        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3046        so the version is shown; (c) stop "make test" echoing "./RunTest".
3048     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3050     9. The prototype for memmove() for systems that don't have it was using
3051        size_t, but the inclusion of the header that defines size_t was later. I've
3052        moved the #includes for the C headers earlier to avoid this.
3054    10. Added some adjustments to the code to make it easier to compiler on certain
3055        special systems:
3057          (a) Some "const" qualifiers were missing.
3058          (b) Added the macro EXPORT before all exported functions; by default this
3059              is defined to be empty.
3060          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3061              that it reads its output file name as an argument instead of writing
3062              to the standard output and assuming this can be redirected.
3064    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3065        class containing characters with values greater than 255, PCRE compilation
3066        went into a loop.
3068    12. A recursive reference to a subpattern that was within another subpattern
3069        that had a minimum quantifier of zero caused PCRE to crash. For example,
3070        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3071        recursion. If the recursively-called subpattern itself had a zero repeat,
3072        that was OK.
3074    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3075        buffer into which it was copied (for escape processing) was still set at
3076        1024, so long lines caused crashes.
3078    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3079        "internal error: code overflow...". This applied to any character class
3080        that was followed by a possessive quantifier.
3082    15. Modified the Makefile to add libpcre.la as a prerequisite for
3083        libpcreposix.la because I was told this is needed for a parallel build to
3084        work.
3086    16. If a pattern that contained .* following optional items at the start was
3087        studied, the wrong optimizing data was generated, leading to matching
3088        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3089        matching string must start with a or b or c. The correct conclusion for
3090        this pattern is that a match can start with any character.
3093    Version 4.4 13-Aug-03
3094    ---------------------
3096     1. In UTF-8 mode, a character class containing characters with values between
3097        127 and 255 was not handled correctly if the compiled pattern was studied.
3098        In fixing this, I have also improved the studying algorithm for such
3099        classes (slightly).
3101     2. Three internal functions had redundant arguments passed to them. Removal
3102        might give a very teeny performance improvement.
3104     3. Documentation bug: the value of the capture_top field in a callout is *one
3105        more than* the number of the hightest numbered captured substring.
3107     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3108        in incorrectly linking with a previously installed version. They now link
3109        explicitly with libpcre.la.
3111     5. configure.in no longer needs to recognize Cygwin specially.
3113     6. A problem in pcre.in for Windows platforms is fixed.
3115     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3116        pcretest, it used to include the size of the study block as part of its
3117        output. Unfortunately, the structure contains a field that has a different
3118        size on different hardware architectures. This meant that the tests that
3119        showed this size failed. As the block is currently always of a fixed size,
3120        this information isn't actually particularly useful in pcretest output, so
3121        I have just removed it.
3123     8. Three pre-processor statements accidentally did not start in column 1.
3124        Sadly, there are *still* compilers around that complain, even though
3125        standard C has not required this for well over a decade. Sigh.
3127     9. In pcretest, the code for checking callouts passed small integers in the
3128        callout_data field, which is a void * field. However, some picky compilers
3129        complained about the casts involved for this on 64-bit systems. Now
3130        pcretest passes the address of the small integer instead, which should get
3131        rid of the warnings.
3133    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3134        both compile and run time, and gives an error if an invalid UTF-8 sequence
3135        is found. There is a option for disabling this check in cases where the
3136        string is known to be correct and/or the maximum performance is wanted.
3138    11. In response to a bug report, I changed one line in Makefile.in from
3140            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3141        to
3142            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3144        to look similar to other lines, but I have no way of telling whether this
3145        is the right thing to do, as I do not use Windows. No doubt I'll get told
3146        if it's wrong...
3149    Version 4.3 21-May-03
3150    ---------------------
3152    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3153       Makefile.
3155    2. Some refactoring to improve the quality of the code:
3157       (i)   The utf8_table... variables are now declared "const".
3159       (ii)  The code for \cx, which used the "case flipping" table to upper case
3160             lower case letters, now just substracts 32. This is ASCII-specific,
3161             but the whole concept of \cx is ASCII-specific, so it seems
3162             reasonable.
3164       (iii) PCRE was using its character types table to recognize decimal and
3165             hexadecimal digits in the pattern. This is silly, because it handles
3166             only 0-9, a-f, and A-F, but the character types table is locale-
3167             specific, which means strange things might happen. A private
3168             table is now used for this - though it costs 256 bytes, a table is
3169             much faster than multiple explicit tests. Of course, the standard
3170             character types table is still used for matching digits in subject
3171             strings against \d.
3173       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3174             ending in _t are). So I've renamed it as ESC_tee.
3176    3. The first argument for regexec() in the POSIX wrapper should have been
3177       defined as "const".
3179    4. Changed pcretest to use malloc() for its buffers so that they can be
3180       Electric Fenced for debugging.
3182    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3183       to read one or more bytes before the start of the subject string. Often this
3184       had no effect on PCRE's behaviour, but in some circumstances it could
3185       provoke a segmentation fault.
3187    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3188       to try to read one or more bytes before the start of the subject string.
3190    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3191       UTF-8 support could misbehave in various ways if the subject string
3192       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3193       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3194       back over UTF-8 characters.)
3197    Version 4.2 14-Apr-03
3198    ---------------------
3200    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3202    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3203         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3204         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3205         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3206         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3207           and BUILD_EXEEXT
3208         Note: automatic setting of the BUILD variables is not yet working
3209         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3210           compile-time but not at link-time
3211         [LINK]: use for linking executables only
3212         make different versions for Windows and non-Windows
3213         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3214           libraries
3215         [LINK_FOR_BUILD]: new variable
3216         [OBJEXT]: use throughout
3217         [EXEEXT]: use throughout
3218         <winshared>: new target
3219         <wininstall>: new target
3220         <dftables.o>: use native compiler
3221         <dftables>: use native linker
3222         <install>: handle Windows platform correctly
3223         <clean>: ditto
3224         <check>: ditto
3225         copy DLL to top builddir before testing
3227       As part of these changes, -no-undefined was removed again. This was reported
3228       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3229       in any case.
3231    3. Some tidies to get rid of compiler warnings:
3233       . In the match_data structure, match_limit was an unsigned long int, whereas
3234         match_call_count was an int. I've made them both unsigned long ints.
3236       . In pcretest the fact that a const uschar * doesn't automatically cast to
3237         a void * provoked a warning.
3239       . Turning on some more compiler warnings threw up some "shadow" variables
3240         and a few more missing casts.
3242    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3243       option, a class that contained a single character with a value between 128
3244       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3246    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3247       option, a class that contained several characters, but with at least one
3248       whose value was between 128 and 255 caused PCRE to crash.
3251    Version 4.1 12-Mar-03
3252    ---------------------
3254    1. Compiling with gcc -pedantic found a couple of places where casts were
3255    needed, and a string in dftables.c that was longer than standard compilers are
3256    required to support.
3258    2. Compiling with Sun's compiler found a few more places where the code could
3259    be tidied up in order to avoid warnings.
3261    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3262    first of these names is deprecated in the latest Autoconf in favour of the name
3263    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3264    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3265    analogy I have changed it to CFLAGS_FOR_BUILD.
3267    4. Added -no-undefined to the linking command in the Makefile, because this is
3268    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3269    linking step for the pcreposix library.
3271    5. PCRE was failing to diagnose the case of two named groups with the same
3272    name.
3274    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3275    literal character that is needed in the subject for a match, and scans along to
3276    ensure that it is present before embarking on the full matching process. This
3277    saves time in cases of nested unlimited repeats that are never going to match.
3278    Problem: the scan can take a lot of time if the subject is very long (e.g.
3279    megabytes), thus penalizing straightforward matches. It is now done only if the
3280    amount of subject to be scanned is less than 1000 bytes.
3282    7. A lesser problem with the same optimization is that it was recording the
3283    first character of an anchored pattern as "needed", thus provoking a search
3284    right along the subject, even when the first match of the pattern was going to
3285    fail. The "needed" character is now not set for anchored patterns, unless it
3286    follows something in the pattern that is of non-fixed length. Thus, it still
3287    fulfils its original purpose of finding quick non-matches in cases of nested
3288    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3291    Version 4.0 17-Feb-03
3292    ---------------------
3294    1. If a comment in an extended regex that started immediately after a meta-item
3295    extended to the end of string, PCRE compiled incorrect data. This could lead to
3296    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3298    2. Moved to autoconf 2.53 and libtool 1.4.2.
3300    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3301    the special perltest8 script is no longer needed - all the tests can be run
3302    from a single perltest script.
3304    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3305    by \s. It has now been removed in PCRE. This means it isn't recognized as
3306    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3307    class [:space:] *does* include VT, thereby creating a mess.
3309    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3310    space and tab.
3312    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3313    its new features into the main test script, reducing the number of scripts.
3315    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3316    were backward compatible, and made the (?i) apply to the whole pattern, as if
3317    /i were given. Now it behaves more logically, and applies the option setting
3318    only to what follows. PCRE has been changed to follow suit. However, if it
3319    finds options settings right at the start of the pattern, it extracts them into
3320    the global options, as before. Thus, they show up in the info data.
3322    8. Added support for the \Q...\E escape sequence. Characters in between are
3323    treated as literals. This is slightly different from Perl in that $ and @ are
3324    also handled as literals inside the quotes. In Perl, they will cause variable
3325    interpolation. Note the following examples:
3327        Pattern            PCRE matches      Perl matches
3329        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3330        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3331        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3333    For compatibility with Perl, \Q...\E sequences are recognized inside character
3334    classes as well as outside them.
3336    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3337    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3338    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3339    signed/unsigned warnings.
3341    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3342    option for pcretest, so I've replaced it by a simple function that does just
3343    that job.
3345    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3346    "pcregrep -".
3348    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3349    Java package. This provides some syntactic sugar for simple cases of what my
3350    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3351    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3352    item, you can use this simplified notation. Note that only makes sense with
3353    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3354    greediness, whatever the setting of the PCRE_UNGREEDY option.
3356    13. A change of greediness default within a pattern was not taking effect at
3357    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3358    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3359    was abstracted outside.
3361    14. PCRE now supports the \G assertion. It is true when the current matching
3362    position is at the start point of the match. This differs from \A when the
3363    starting offset is non-zero. Used with the /g option of pcretest (or similar
3364    code), it works in the same way as it does for Perl's /g option. If all
3365    alternatives of a regex begin with \G, the expression is anchored to the start
3366    match position, and the "anchored" flag is set in the compiled expression.
3368    15. Some bugs concerning the handling of certain option changes within patterns
3369    have been fixed. These applied to options other than (?ims). For example,
3370    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3371    been the other way round. Some of this was related to change 7 above.
3373    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3374    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3375    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3376    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3378    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3379    mode. Unlike ".", it always matches newline, whatever the setting of
3380    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3381    assertions. Perl allows it, but it doesn't (in general) work because it can't
3382    calculate the length of the lookbehind. At least, that's the case for Perl
3383    5.8.0 - I've been told they are going to document that it doesn't work in
3384    future.
3386    18. Added an error diagnosis for escapes that PCRE does not support: these are
3387    \L, \l, \N, \P, \p, \U, \u, and \X.
3389    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3390    reading past the end of the pattern in cases such as /[abcd/.
3392    20. PCRE was getting more memory than necessary for patterns with classes that
3393    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3395    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3396    compiling PCRE for use with Virtual Pascal.
3398    22. Small fix to the Makefile to make it work properly if the build is done
3399    outside the source tree.
3401    23. Added a new extension: a condition to go with recursion. If a conditional
3402    subpattern starts with (?(R) the "true" branch is used if recursion has
3403    happened, whereas the "false" branch is used only at the top level.
3405    24. When there was a very long string of literal characters (over 255 bytes
3406    without UTF support, over 250 bytes with UTF support), the computation of how
3407    much memory was required could be incorrect, leading to segfaults or other
3408    strange effects.
3410    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3411    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3412    there was a subsequent back reference to those brackets. This meant that, for
3413    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3414    possible to check for precisely this case. All we can do is abandon the
3415    optimization if .* occurs inside capturing brackets when there are any back
3416    references whatsoever. (See below for a better fix that came later.)
3418    26. The handling of the optimization for finding the first character of a
3419    non-anchored pattern, and for finding a character that is required later in the
3420    match were failing in some cases. This didn't break the matching; it just
3421    failed to optimize when it could. The way this is done has been re-implemented.
3423    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3425    28. Added a new feature that provides some of the functionality that Perl
3426    provides with (?{...}). The facility is termed a "callout". The way it is done
3427    in PCRE is for the caller to provide an optional function, by setting
3428    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3429    global variable. By default it is unset, which disables all calling out. To get
3430    the function called, the regex must include (?C) at appropriate points. This
3431    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3432    This provides a means of identifying different callout points. When PCRE
3433    reaches such a point in the regex, if pcre_callout has been set, the external
3434    function is called. It is provided with data in a structure called
3435    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3436    matching continues; if it returns a non-zero value, the match at the current
3437    point fails. However, backtracking will occur if possible. [This was changed
3438    later and other features added - see item 49 below.]
3440    29. pcretest is upgraded to test the callout functionality. It provides a
3441    callout function that displays information. By default, it shows the start of
3442    the match and the current position in the text. There are some new data escapes
3443    to vary what happens:
3445        \C+         in addition, show current contents of captured substrings
3446        \C-         do not supply a callout function
3447        \C!n        return 1 when callout number n is reached
3448        \C!n!m      return 1 when callout number n is reached for the mth time
3450    30. If pcregrep was called with the -l option and just a single file name, it
3451    output "<stdin>" if a match was found, instead of the file name.
3453    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3454    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3455    pcre_exec(). This saves a malloc/free per call. The default value of
3456    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3457    when configuring.
3459    32. The default maximum size of a compiled pattern is 64K. There have been a
3460    few cases of people hitting this limit. The code now uses macros to handle the
3461    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3462    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3463    configuring. Tests 2 and 5 work only with 2-byte links because they output
3464    debugging information about compiled patterns.
3466    33. Internal code re-arrangements:
3468    (a) Moved the debugging function for printing out a compiled regex into
3469        its own source file (printint.c) and used #include to pull it into
3470        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3471        separate copies.
3473    (b) Defined the list of op-code names for debugging as a macro in
3474        internal.h so that it is next to the definition of the opcodes.
3476    (c) Defined a table of op-code lengths for simpler skipping along compiled
3477        code. This is again a macro in internal.h so that it is next to the
3478        definition of the opcodes.
3480    34. Added support for recursive calls to individual subpatterns, along the
3481    lines of Robin Houston's patch (but implemented somewhat differently).
3483    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3484    allow it to read and process whole directories in Win32. This code was
3485    contributed by Lionel Fourquaux; it has not been tested by me.
3487    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3488    used to name a group. Names consist of alphanumerics and underscores, and must
3489    be unique. Back references use the syntax (?P=name) and recursive calls use
3490    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3491    numbers. The function pcre_fullinfo() can be used after compilation to extract
3492    a name/number map. There are three relevant calls:
3494      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3495      PCRE_INFO_NAMECOUNT            yields the number of entries
3496      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3498    The map is a vector of fixed-size entries. The size of each entry depends on
3499    the length of the longest name used. The first two bytes of each entry are the
3500    group number, most significant byte first. There follows the corresponding
3501    name, zero terminated. The names are in alphabetical order.
3503    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3504    case instead of 255. Making it the same both with and without UTF-8 support
3505    means that the same test output works with both.
3507    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3508    calling malloc() with a zero argument.
3510    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3511    optimization. I've improved things by keeping a bitmap of backreferences with
3512    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3513    fact referenced, the optimization can be applied. It is unlikely that a
3514    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3515    the match to follow \n) will appear inside brackets with a number greater than
3516    31, but if it does, any back reference > 31 suppresses the optimization.
3518    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3519    of disabling numbered capturing parentheses. Any opening parenthesis that is
3520    not followed by ? behaves as if it were followed by ?: but named parentheses
3521    can still be used for capturing (and they will acquire numbers in the usual
3522    way).
3524    41. Redesigned the return codes from the match() function into yes/no/error so
3525    that errors can be passed back from deep inside the nested calls. A malloc
3526    failure while inside a recursive subpattern call now causes the
3527    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3529    42. It is now possible to set a limit on the number of times the match()
3530    function is called in a call to pcre_exec(). This facility makes it possible to
3531    limit the amount of recursion and backtracking, though not in a directly
3532    obvious way, because the match() function is used in a number of different
3533    circumstances. The count starts from zero for each position in the subject
3534    string (for non-anchored patterns). The default limit is, for compatibility, a
3535    large number, namely 10 000 000. You can change this in two ways:
3537    (a) When configuring PCRE before making, you can use --with-match-limit=n
3538        to set a default value for the compiled library.
3540    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3541        a different value is set. See 45 below.
3543    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3545    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3546    of things that can be changed at compile time. The first argument specifies
3547    what is wanted and the second points to where the information is to be placed.
3548    The current list of available information is:
3550      PCRE_CONFIG_UTF8
3552    The output is an integer that is set to one if UTF-8 support is available;
3553    otherwise it is set to zero.
3557    The output is an integer that it set to the value of the code that is used for
3558    newline. It is either LF (10) or CR (13).
3562    The output is an integer that contains the number of bytes used for internal
3563    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3567    The output is an integer that contains the threshold above which the POSIX
3568    interface uses malloc() for output vectors. See item 31 above.
3572    The output is an unsigned integer that contains the default limit of the number
3573    of match() calls in a pcre_exec() execution. See 42 above.
3575    44. pcretest has been upgraded by the addition of the -C option. This causes it
3576    to extract all the available output from the new pcre_config() function, and to
3577    output it. The program then exits immediately.
3579    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3580    order to support additional features. One way would have been to define
3581    pcre_exec2() (for example) with extra arguments, but this would not have been
3582    extensible, and would also have required all calls to the original function to
3583    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3584    is used for passing in "extra" data from pcre_study().
3586    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3587    contains the following fields:
3589      flags         a bitmap indicating which of the following fields are set
3590      study_data    opaque data from pcre_study()
3591      match_limit   a way of specifying a limit on match() calls for a specific
3592                      call to pcre_exec()
3593      callout_data  data for callouts (see 49 below)
3595    The flag bits are also defined in pcre.h, and are
3601    The pcre_study() function now returns one of these new pcre_extra blocks, with
3602    the actual study data pointed to by the study_data field, and the
3603    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3604    before. That is, this change is entirely upwards-compatible and requires no
3605    change to existing code.
3607    If you want to pass in additional data to pcre_exec(), you can either place it
3608    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3609    block.
3611    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a