ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 33 by nigel, Sat Feb 24 21:39:01 2007 UTC revision 762 by ph10, Tue Nov 22 13:36:51 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.21
5    ------------
7    1.  Updating the JIT compiler.
9    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
10        are added as well.
12    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
13        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
14        calling _pcre_jit_exec. Some extra comments are added.
16    4.  Mark settings inside atomic groups that do not contain any capturing
17        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
18        was introduced by change 18 for 8.20.
20    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
21        ECMA-262 standard.
23    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
24        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
25        This bug was probably introduced by change 9 of 8.13.
27    7.  While fixing 6 above, I noticed that a number of other items were being
28        incorrectly rejected as "not fixed length". This arose partly because newer
29        opcodes had not been added to the fixed-length checking code. I have (a)
30        corrected the bug and added tests for these items, and (b) arranged for an
31        error to occur if an unknown opcode is encountered while checking for fixed
32        length instead of just assuming "not fixed length". The items that were
33        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
34        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
35        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
37    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
38        being incorrectly compiled and would have given unpredicatble results.
40    9.  A possessively repeated subpattern with minimum repeat count greater than
41        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
42        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
43        the first (A) could occur when it should not.
45    10. Add a cast and remove a redundant test from the code.
47    11. JIT should use pcre_malloc/pcre_free for allocation.
49    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
50        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
51        is anything other than /usr, -L is still shown).
53    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
55    14. Perl does not support \N without a following name in a [] class; PCRE now
56        also gives an error.
58    15. If a forward reference was repeated with an upper limit of around 2000,
59        it caused the error "internal error: overran compiling workspace". This
60        is now checked, and causes "too many forward references" instead.
62    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
63        incorrectly expecting the subject to contain another "a" after the start.
66    Version 8.20 21-Oct-2011
67    ------------------------
69    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
70        a POSIX class. After further experiments with Perl, which convinced me that
71        Perl has bugs and confusions, a closing square bracket is no longer allowed
72        in a POSIX name. This bug also affected patterns with classes that started
73        with full stops.
75    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
76        captured substring, but while checking the failing first alternative,
77        substring 1 is temporarily captured. If the output vector supplied to
78        pcre_exec() was not big enough for this capture, the yield of the function
79        was still zero ("insufficient space for captured substrings"). This cannot
80        be totally fixed without adding another stack variable, which seems a lot
81        of expense for a edge case. However, I have improved the situation in cases
82        such as /(a)(b)x|abc/ matched against "abc", where the return code
83        indicates that fewer than the maximum number of slots in the ovector have
84        been set.
86    3.  Related to (2) above: when there are more back references in a pattern than
87        slots in the output vector, pcre_exec() uses temporary memory during
88        matching, and copies in the captures as far as possible afterwards. It was
89        using the entire output vector, but this conflicts with the specification
90        that only 2/3 is used for passing back captured substrings. Now it uses
91        only the first 2/3, for compatibility. This is, of course, another edge
92        case.
94    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
95        main code base, and can be used by building with --enable-jit. When this is
96        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
97        runtime --no-jit option is given.
99    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
100        ovector, the return from the function was zero, implying that there were
101        other matches that did not fit. The correct "exactly full" value is now
102        returned.
104    6.  If a subpattern that was called recursively or as a subroutine contained
105        (*PRUNE) or any other control that caused it to give a non-standard return,
106        invalid errors such as "Error -26 (nested recursion at the same subject
107        position)" or even infinite loops could occur.
109    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
110        computing the minimum length on reaching *ACCEPT, and so ended up with the
111        wrong value of 1 rather than 0. Further investigation indicates that
112        computing a minimum subject length in the presence of *ACCEPT is difficult
113        (think back references, subroutine calls), and so I have changed the code
114        so that no minimum is registered for a pattern that contains *ACCEPT.
116    8.  If (*THEN) was present in the first (true) branch of a conditional group,
117        it was not handled as intended. [But see 16 below.]
119    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
120        Sheri Pierce.
122    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
123        the first byte in a match must be "a".
125    11. Change 17 for 8.13 increased the recursion depth for patterns like
126        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
127        pattern contains any instances of (*THEN). If it does not, the old
128        optimizations are restored. It would be nice to do this on a per-group
129        basis, but at the moment that is not feasible.
131    12. In some environments, the output of pcretest -C is CRLF terminated. This
132        broke RunTest's code that checks for the link size. A single white space
133        character after the value is now allowed for.
135    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
136        For "fr", it uses the Windows-specific input and output files.
138    14. If (*THEN) appeared in a group that was called recursively or as a
139        subroutine, it did not work as intended. [But see next item.]
141    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
142        pattern fragments (but not containing any | characters). If A and B are
143        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
144        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
145        D. In other words, Perl considers parentheses that do not contain any |
146        characters to be part of a surrounding alternative, whereas PCRE was
147        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
148        differently. PCRE now behaves in the same way as Perl, except in the case
149        of subroutine/recursion calls such as (?1) which have in any case always
150        been different (but PCRE had them first :-).
152    16. Related to 15 above: Perl does not treat the | in a conditional group as
153        creating alternatives. Such a group is treated in the same way as an
154        ordinary group without any | characters when processing (*THEN). PCRE has
155        been changed to match Perl's behaviour.
157    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
158        RunGrepTest script failed.
160    18. Change 22 for version 13 caused atomic groups to use more stack. This is
161        inevitable for groups that contain captures, but it can lead to a lot of
162        stack use in large patterns. The old behaviour has been restored for atomic
163        groups that do not contain any capturing parentheses.
165    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
166        suppress the check for a minimum subject length at run time. (If it was
167        given to pcre_exec() or pcre_dfa_exec() it did work.)
169    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
170        fail to work when decoding hex characters in data strings in EBCDIC
171        environments.
173    21. It appears that in at least one Mac OS environment, the isxdigit() function
174        is implemented as a macro that evaluates to its argument more than once,
175        contravening the C 90 Standard (I haven't checked a later standard). There
176        was an instance in pcretest which caused it to go wrong when processing
177        \x{...} escapes in subject strings. The has been rewritten to avoid using
178        things like p++ in the argument of isxdigit().
181    Version 8.13 16-Aug-2011
182    ------------------------
184    1.  The Unicode data tables have been updated to Unicode 6.0.0.
186    2.  Two minor typos in pcre_internal.h have been fixed.
188    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
189        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
190        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
192    4.  There were a number of related bugs in the code for matching backrefences
193        caselessly in UTF-8 mode when codes for the characters concerned were
194        different numbers of bytes. For example, U+023A and U+2C65 are an upper
195        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
196        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
197        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
198        2-byte code at the end of the subject (it thought there wasn't enough data
199        left).
201    5.  Comprehensive information about what went wrong is now returned by
202        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
203        as the output vector has at least 2 elements. The offset of the start of
204        the failing character and a reason code are placed in the vector.
206    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
207        now returned is for the first byte of the failing character, instead of the
208        last byte inspected. This is an incompatible change, but I hope it is small
209        enough not to be a problem. It makes the returned offset consistent with
210        pcre_exec() and pcre_dfa_exec().
212    7.  pcretest now gives a text phrase as well as the error number when
213        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
214        failure, the offset and reason code are output.
216    8.  When \R was used with a maximizing quantifier it failed to skip backwards
217        over a \r\n pair if the subsequent match failed. Instead, it just skipped
218        back over a single character (\n). This seems wrong (because it treated the
219        two characters as a single entity when going forwards), conflicts with the
220        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
221        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
222        has been changed.
224    9.  Some internal refactoring has changed the processing so that the handling
225        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
226        time (the PCRE_DOTALL option was changed this way some time ago: version
227        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
228        which was always a bit of a fudge. It also means that there is one less
229        argument for the match() function, which reduces its stack requirements
230        slightly. This change also fixes an incompatibility with Perl: the pattern
231        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
233    10. More internal refactoring has drastically reduced the number of recursive
234        calls to match() for possessively repeated groups such as (abc)++ when
235        using pcre_exec().
237    11. While implementing 10, a number of bugs in the handling of groups were
238        discovered and fixed:
240        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
241        (a|)*(?1) gave a compile-time internal error.
242        ((a|)+)+  did not notice that the outer group could match an empty string.
243        (^a|^)+   was not marked as anchored.
244        (.*a|.*)+ was not marked as matching at start or after a newline.
246    12. Yet more internal refactoring has removed another argument from the match()
247        function. Special calls to this function are now indicated by setting a
248        value in a variable in the "match data" data block.
250    13. Be more explicit in pcre_study() instead of relying on "default" for
251        opcodes that mean there is no starting character; this means that when new
252        ones are added and accidentally left out of pcre_study(), testing should
253        pick them up.
255    14. The -s option of pcretest has been documented for ages as being an old
256        synonym of -m (show memory usage). I have changed it to mean "force study
257        for every regex", that is, assume /S for every regex. This is similar to -i
258        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
259        using it. It makes it easier to run collections of tests with and without
260        study enabled, and thereby test pcre_study() more easily. All the standard
261        tests are now run with and without -s (but some patterns can be marked as
262        "never study" - see 20 below).
264    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
265        restoration of the capturing data to the outer values was not happening
266        correctly.
268    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
269        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
270        pattern had matched an empty string, and so incorrectly returned a no
271        match.
273    17. There was optimizing code for the last branch of non-capturing parentheses,
274        and also for the obeyed branch of a conditional subexpression, which used
275        tail recursion to cut down on stack usage. Unfortunately, now that there is
276        the possibility of (*THEN) occurring in these branches, tail recursion is
277        no longer possible because the return has to be checked for (*THEN). These
278        two optimizations have therefore been removed. [But see 8.20/11 above.]
280    18. If a pattern containing \R was studied, it was assumed that \R always
281        matched two bytes, thus causing the minimum subject length to be
282        incorrectly computed because \R can also match just one byte.
284    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
285        was incorrectly computed.
287    20. If /S is present twice on a test pattern in pcretest input, it now
288        *disables* studying, thereby overriding the use of -s on the command line
289        (see 14 above). This is necessary for one or two tests to keep the output
290        identical in both cases.
292    21. When (*ACCEPT) was used in an assertion that matched an empty string and
293        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
295    22. When an atomic group that contained a capturing parenthesis was
296        successfully matched, but the branch in which it appeared failed, the
297        capturing was not being forgotten if a higher numbered group was later
298        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
299        group 1 to "a", when in fact it should be unset. This applied to multi-
300        branched capturing and non-capturing groups, repeated or not, and also to
301        positive assertions (capturing in negative assertions does not happen
302        in PCRE) and also to nested atomic groups.
304    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
305        subject after a captured substring, to make it easier to tell which of a
306        number of identical substrings has been captured.
308    24. The way atomic groups are processed by pcre_exec() has been changed so that
309        if they are repeated, backtracking one repetition now resets captured
310        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
311        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
312        "aaa". Previously, it would have been "a". As part of this code
313        refactoring, the way recursive calls are handled has also been changed.
315    25. If an assertion condition captured any substrings, they were not passed
316        back unless some other capturing happened later. For example, if
317        (?(?=(a))a) was matched against "a", no capturing was returned.
319    26. When studying a pattern that contained subroutine calls or assertions,
320        the code for finding the minimum length of a possible match was handling
321        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
322        group 1 called group 2 while simultaneously a separate group 2 called group
323        1). A stack overflow occurred in this case. I have fixed this by limiting
324        the recursion depth to 10.
326    27. Updated RunTest.bat in the distribution to the version supplied by Tom
327        Fortmann. This supports explicit test numbers on the command line, and has
328        argument validation and error reporting.
330    28. An instance of \X with an unlimited repeat could fail if at any point the
331        first character it looked at was a mark character.
333    29. Some minor code refactoring concerning Unicode properties and scripts
334        should reduce the stack requirement of match() slightly.
336    30. Added the '=' option to pcretest to check the setting of unused capturing
337        slots at the end of the pattern, which are documented as being -1, but are
338        not included in the return count.
340    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
341        compiled something random. Now it gives a compile-time error (as does
342        Perl).
344    32. A *MARK encountered during the processing of a positive assertion is now
345        recorded and passed back (compatible with Perl).
347    33. If --only-matching or --colour was set on a pcregrep call whose pattern
348        had alternative anchored branches, the search for a second match in a line
349        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
350        matched the line "0102" twice. The same bug affected patterns that started
351        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
352        twice.
354    34. Previously, PCRE did not allow quantification of assertions. However, Perl
355        does, and because of capturing effects, quantifying parenthesized
356        assertions may at times be useful. Quantifiers are now allowed for
357        parenthesized assertions.
359    35. A minor code tidy in pcre_compile() when checking options for \R usage.
361    36. \g was being checked for fancy things in a character class, when it should
362        just be a literal "g".
364    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
365        appearance of a nested POSIX class supersedes an apparent external class.
366        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
367        unescaped square brackets may also appear as part of class names. For
368        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
369        more like Perl. (But see 8.20/1 above.)
371    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
372        was because it thought it was \N{name}, which is not supported).
374    39. Add minix to OS list not supporting the -S option in pcretest.
376    40. PCRE tries to detect cases of infinite recursion at compile time, but it
377        cannot analyze patterns in sufficient detail to catch mutual recursions
378        such as ((?1))((?2)). There is now a runtime test that gives an error if a
379        subgroup is called recursively as a subpattern for a second time at the
380        same position in the subject string. In previous releases this might have
381        been caught by the recursion limit, or it might have run out of stack.
383    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
384        happen only once. PCRE was, however incorrectly giving a compile time error
385        "recursive call could loop indefinitely" because it cannot analyze the
386        pattern in sufficient detail. The compile time test no longer happens when
387        PCRE is compiling a conditional subpattern, but actual runaway loops are
388        now caught at runtime (see 40 above).
390    42. It seems that Perl allows any characters other than a closing parenthesis
391        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
392        has been changed to be the same.
394    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
395        as not to get warnings when autogen.sh is called. Also changed
396        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
398    44. To help people who use pcregrep to scan files containing exceedingly long
399        lines, the following changes have been made:
401        (a) The default value of the buffer size parameter has been increased from
402            8K to 20K. (The actual buffer used is three times this size.)
404        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
405            PCRE is built.
407        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
408            to be set at run time.
410        (d) Numerical values in pcregrep options can be followed by K or M, for
411            example --buffer-size=50K.
413        (e) If a line being scanned overflows pcregrep's buffer, an error is now
414            given and the return code is set to 2.
416    45. Add a pointer to the latest mark to the callout data block.
418    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
419        partial match of an empty string instead of no match. This was specific to
420        the use of ".".
422    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
423        complete match instead of a partial match. This bug was dependent on both
424        the PCRE_UTF8 and PCRE_DOTALL options being set.
426    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
427        starting byte set, because \b was not being ignored.
430    Version 8.12 15-Jan-2011
431    ------------------------
433    1.  Fixed some typos in the markup of the man pages, and wrote a script that
434        checks for such things as part of the documentation building process.
436    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
437        --match-limit and --recursion-limit options (added for 8.11). In
438        particular, this made one of the standard tests fail. (The integer value
439        went into the wrong half of a long int.)
441    3.  If the --colour option was given to pcregrep with -v (invert match), it
442        did strange things, either producing crazy output, or crashing. It should,
443        of course, ignore a request for colour when reporting lines that do not
444        match.
446    4.  Another pcregrep bug caused similar problems if --colour was specified with
447        -M (multiline) and the pattern match finished with a line ending.
449    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
450        matched in multiline mode, the following line was shown as part of the
451        match. This seems wrong, so I have changed it.
453    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
454        the check for further matches in the same line (so they could be coloured)
455        to overrun the end of the current line. If another match was found, it was
456        incorrectly shown (and then shown again when found in the next line).
458    7.  If pcregrep was compiled under Windows, there was a reference to the
459        function pcregrep_exit() before it was defined. I am assuming this was
460        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
461        reported by a user. I've moved the definition above the reference.
464    Version 8.11 10-Dec-2010
465    ------------------------
467    1.  (*THEN) was not working properly if there were untried alternatives prior
468        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
469        backtracked to try for "b" instead of moving to the next alternative branch
470        at the same level (in this case, to look for "c"). The Perl documentation
471        is clear that when (*THEN) is backtracked onto, it goes to the "next
472        alternative in the innermost enclosing group".
474    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
475        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
476        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
477        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
478        (*THEN).
480    3.  If \s appeared in a character class, it removed the VT character from
481        the class, even if it had been included by some previous item, for example
482        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
483        of \s, but is part of the POSIX "space" class.)
485    4.  A partial match never returns an empty string (because you can always
486        match an empty string at the end of the subject); however the checking for
487        an empty string was starting at the "start of match" point. This has been
488        changed to the "earliest inspected character" point, because the returned
489        data for a partial match starts at this character. This means that, for
490        example, /(?<=abc)def/ gives a partial match for the subject "abc"
491        (previously it gave "no match").
493    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
494        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
495        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
496        has an implication that the given string is incomplete (because a partial
497        match is preferred over a full match). For this reason, these items now
498        give a partial match in this situation. [Aside: previously, the one case
499        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
500        match rather than a full match, which was wrong by the old rules, but is
501        now correct.]
503    6.  There was a bug in the handling of #-introduced comments, recognized when
504        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
505        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
506        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
507        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
508        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
509        places in pcre_compile().
511    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
512        comments when looking ahead for named forward references to subpatterns,
513        the only newline sequence it recognized was NL. It now handles newlines
514        according to the set newline convention.
516    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
517        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
518        cater for a lack of strerror(). These oversights have been fixed.
520    9.  Added --match-limit and --recursion-limit to pcregrep.
522    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
524    11. When the -o option was used, pcregrep was setting a return code of 1, even
525        when matches were found, and --line-buffered was not being honoured.
527    12. Added an optional parentheses number to the -o and --only-matching options
528        of pcregrep.
530    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
531        can match an empty string. The code to do it in pcretest and pcredemo
532        needed fixing:
534        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
535            only one byte after an empty string match just before CRLF (this case
536            just got forgotten; "any" and "anycrlf" were OK).
538        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
539            mode when an empty string match preceded an ASCII character followed by
540            a non-ASCII character. (The code for advancing by one character rather
541            than one byte was nonsense.)
543        (c) The pcredemo.c sample program did not have any code at all to handle
544            the cases when CRLF is a valid newline sequence.
546    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
547        as a starting offset was within the subject string. There is now a new
548        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
549        negative or greater than the length of the string. In order to test this,
550        pcretest is extended to allow the setting of negative starting offsets.
552    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
553        starting offset points to the beginning of a UTF-8 character was
554        unnecessarily clumsy. I tidied it up.
556    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
557        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
559    17. Nobody had reported that the --include_dir option, which was added in
560        release 7.7 should have been called --include-dir (hyphen, not underscore)
561        for compatibility with GNU grep. I have changed it to --include-dir, but
562        left --include_dir as an undocumented synonym, and the same for
563        --exclude-dir, though that is not available in GNU grep, at least as of
564        release 2.5.4.
566    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
567        characters from a string of bytes) have been redefined so as not to use
568        loops, in order to improve performance in some environments. At the same
569        time, I abstracted some of the common code into auxiliary macros to save
570        repetition (this should not affect the compiled code).
572    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
573        compile-time error is now given if \c is not followed by an ASCII
574        character, that is, a byte less than 128. (In EBCDIC mode, the code is
575        different, and any byte value is allowed.)
577    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
578        START_OPTIMIZE option, which is now allowed at compile time - but just
579        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
580        to pcregrep and other applications that have no direct access to PCRE
581        options. The new /Y option in pcretest sets this option when calling
582        pcre_compile().
584    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
585        back references. Groups containing recursive back references were forced to
586        be atomic by that change, but in the case of named groups, the amount of
587        memory required was incorrectly computed, leading to "Failed: internal
588        error: code overflow". This has been fixed.
590    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
591        pcretest.c, to avoid build problems in some Borland environments.
594    Version 8.10 25-Jun-2010
595    ------------------------
597    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
598        THEN.
600    2.  (*ACCEPT) was not working when inside an atomic group.
602    3.  Inside a character class, \B is treated as a literal by default, but
603        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
604        causes the error). The code is unchanged, but I tidied the documentation.
606    4.  Inside a character class, PCRE always treated \R and \X as literals,
607        whereas Perl faults them if its -w option is set. I have changed PCRE so
608        that it faults them when PCRE_EXTRA is set.
610    5.  Added support for \N, which always matches any character other than
611        newline. (It is the same as "." when PCRE_DOTALL is not set.)
613    6.  When compiling pcregrep with newer versions of gcc which may have
614        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
615        declared with attribute warn_unused_result" were given. Just casting the
616        result to (void) does not stop the warnings; a more elaborate fudge is
617        needed. I've used a macro to implement this.
619    7.  Minor change to pcretest.c to avoid a compiler warning.
621    8.  Added four artifical Unicode properties to help with an option to make
622        \s etc use properties (see next item). The new properties are: Xan
623        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
625    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
626        use Unicode properties. (*UCP) at the start of a pattern can be used to set
627        this option. Modified pcretest to add /W to test this facility. Added
628        REG_UCP to make it available via the POSIX interface.
630    10. Added --line-buffered to pcregrep.
632    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
633        studied, and the match started with a letter with a code point greater than
634        127 whose first byte was different to the first byte of the other case of
635        the letter, the other case of this starting letter was not recognized
636        (#976).
638    12. If a pattern that was studied started with a repeated Unicode property
639        test, for example, \p{Nd}+, there was the theoretical possibility of
640        setting up an incorrect bitmap of starting bytes, but fortunately it could
641        not have actually happened in practice until change 8 above was made (it
642        added property types that matched character-matching opcodes).
644    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
645        possible starting bytes for non-anchored patterns.
647    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
648        \R, and also a number of cases that involve Unicode properties, both
649        explicit and implicit when PCRE_UCP is set.
651    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
652        input, it could crash or give wrong results if characters with values
653        greater than 0xc0 were present in the subject string. (Detail: it assumed
654        UTF-8 input when processing these items.)
656    16. Added a lot of (int) casts to avoid compiler warnings in systems where
657        size_t is 64-bit (#991).
659    17. Added a check for running out of memory when PCRE is compiled with
660        --disable-stack-for-recursion (#990).
662    18. If the last data line in a file for pcretest does not have a newline on
663        the end, a newline was missing in the output.
665    19. The default pcre_chartables.c file recognizes only ASCII characters (values
666        less than 128) in its various bitmaps. However, there is a facility for
667        generating tables according to the current locale when PCRE is compiled. It
668        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
669        characters, are recognized by isspace() and therefore were getting set in
670        these tables, and indeed these tables seem to approximate to ISO 8859. This
671        caused a problem in UTF-8 mode when pcre_study() was used to create a list
672        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
673        which of course cannot start UTF-8 characters. I have changed the code so
674        that only real ASCII characters (less than 128) and the correct starting
675        bytes for UTF-8 encodings are set for characters greater than 127 when in
676        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
677        altogether.)
679    20. Added the /T option to pcretest so as to be able to run tests with non-
680        standard character tables, thus making it possible to include the tests
681        used for 19 above in the standard set of tests.
683    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
684        reference to a subpattern the other side of a comment that contains an
685        opening parenthesis caused either an internal compiling error, or a
686        reference to the wrong subpattern.
689    Version 8.02 19-Mar-2010
690    ------------------------
692    1.  The Unicode data tables have been updated to Unicode 5.2.0.
694    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
695        configured.
697    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
698        original author of that file, following a query about its status.
700    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
701        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
703    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
704        quantifier applied to a forward-referencing subroutine call, could compile
705        incorrect code or give the error "internal error: previously-checked
706        referenced subpattern not found".
708    6.  Both MS Visual Studio and Symbian OS have problems with initializing
709        variables to point to external functions. For these systems, therefore,
710        pcre_malloc etc. are now initialized to local functions that call the
711        relevant global functions.
713    7.  There were two entries missing in the vectors called coptable and poptable
714        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
715        I've fixed the data, and added a kludgy way of testing at compile time that
716        the lengths are correct (equal to the number of opcodes).
718    8.  Following on from 7, I added a similar kludge to check the length of the
719        eint vector in pcreposix.c.
721    9.  Error texts for pcre_compile() are held as one long string to avoid too
722        much relocation at load time. To find a text, the string is searched,
723        counting zeros. There was no check for running off the end of the string,
724        which could happen if a new error number was added without updating the
725        string.
727    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
729    11. \K was not working if it appeared in an atomic group or in a group that
730        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
731        \K is "not well defined" if used in an assertion. PCRE now accepts it if
732        the assertion is positive, but not if it is negative.
734    12. Change 11 fortuitously reduced the size of the stack frame used in the
735        "match()" function of pcre_exec.c by one pointer. Forthcoming
736        implementation of support for (*MARK) will need an extra pointer on the
737        stack; I have reserved it now, so that the stack frame size does not
738        decrease.
740    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
741        item in branch that calls a recursion is a subroutine call - as in the
742        second branch in the above example - was incorrectly given the compile-
743        time error "recursive call could loop indefinitely" because pcre_compile()
744        was not correctly checking the subroutine for matching a non-empty string.
746    14. The checks for overrunning compiling workspace could trigger after an
747        overrun had occurred. This is a "should never occur" error, but it can be
748        triggered by pathological patterns such as hundreds of nested parentheses.
749        The checks now trigger 100 bytes before the end of the workspace.
751    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
754    Version 8.01 19-Jan-2010
755    ------------------------
757    1.  If a pattern contained a conditional subpattern with only one branch (in
758        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
759        computed the wrong minimum data length (which is of course zero for such
760        subpatterns). This could cause incorrect "no match" results.
762    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
763        the pattern is reset in the first branch, pcre_compile() failed with
764        "internal error: code overflow at offset...". This happened only when
765        the reset was to the original external option setting. (An optimization
766        abstracts leading options settings into an external setting, which was the
767        cause of this.)
769    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
770        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
771        assertion pattern did not match (meaning that the assertion was true), it
772        was incorrectly treated as false if the SKIP had been reached during the
773        matching. This also applied to assertions used as conditions.
775    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
776        assertion subpattern, including such a pattern used as a condition,
777        unpredictable results occurred, instead of the error return
780    5.  The C++ GlobalReplace function was not working like Perl for the special
781        situation when an empty string is matched. It now does the fancy magic
782        stuff that is necessary.
784    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
785        removed. (These were left over from very, very early versions of PCRE.)
787    7.  Some cosmetic changes to the code to make life easier when compiling it
788        as part of something else:
790        (a) Change DEBUG to PCRE_DEBUG.
792        (b) In pcre_compile(), rename the member of the "branch_chain" structure
793            called "current" as "current_branch", to prevent a collision with the
794            Linux macro when compiled as a kernel module.
796        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
797            prevent a collision with the Linux macro when compiled as a kernel
798            module.
800    8.  In pcre_compile() there are some checks for integer overflows that used to
801        cast potentially large values to (double). This has been changed to that
802        when building, a check for int64_t is made, and if it is found, it is used
803        instead, thus avoiding the use of floating point arithmetic. (There is no
804        other use of FP in PCRE.) If int64_t is not found, the fallback is to
805        double.
807    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
808        2005 (difference between two addresses compared to an unsigned value).
810    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
811        custom one, because of the following reported problem in Windows:
813          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
814              under Win32.
815          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
816              therefore missing the function definition.
817          - The compiler thus generates a "C" signature for the test function.
818          - The linker fails to find the "C" function.
819          - PCRE fails to configure if asked to do so against libbz2.
821    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
822        messages were output:
824          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
825          rerunning libtoolize, to keep the correct libtool macros in-tree.
826          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
828        I have done both of these things.
830    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
831        most of the time, it *can* run out if it is given a pattern that contains a
832        runaway infinite recursion. I updated the discussion in the pcrestack man
833        page.
835    13. Now that we have gone to the x.xx style of version numbers, the minor
836        version may start with zero. Using 08 or 09 is a bad idea because users
837        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
838        interpreted as invalid octal numbers. I've updated the previous comment in
839        configure.ac, and also added a check that gives an error if 08 or 09 are
840        used.
842    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
843        causing partial matching to fail when the end of the subject matched \W
844        in a UTF-8 pattern where \W was quantified with a minimum of 3.
846    15. There were some discrepancies between the declarations in pcre_internal.h
847        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
848        their definitions. The declarations used "const uschar *" and the
849        definitions used USPTR. Even though USPTR is normally defined as "const
850        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
851        reported that: "This difference in casting confuses some C++ compilers, for
852        example, SunCC recognizes above declarations as different functions and
853        generates broken code for hbpcre." I have changed the declarations to use
854        USPTR.
856    16. GNU libtool is named differently on some systems. The autogen.sh script now
857        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
858        (FreeBSD).
860    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
861        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
862        comment: "Figure out how to create a longlong from a string: strtoll and
863        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
864        instance, but it only takes 2 args instead of 3!"
866    18. A subtle bug concerned with back references has been fixed by a change of
867        specification, with a corresponding code fix. A pattern such as
868        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
869        refers, was giving matches when it shouldn't. For example, xa=xaaa would
870        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
871        same bug. Such groups have to be quantified to be useful, or contained
872        inside another quantified group. (If there's no repetition, the reference
873        can never match.) The problem arises because, having left the group and
874        moved on to the rest of the pattern, a later failure that backtracks into
875        the group uses the captured value from the final iteration of the group
876        rather than the correct earlier one. I have fixed this in PCRE by forcing
877        any group that contains a reference to itself to be an atomic group; that
878        is, there cannot be any backtracking into it once it has completed. This is
879        similar to recursive and subroutine calls.
882    Version 8.00 19-Oct-09
883    ----------------------
885    1.  The table for translating pcre_compile() error codes into POSIX error codes
886        was out-of-date, and there was no check on the pcre_compile() error code
887        being within the table. This could lead to an OK return being given in
888        error.
890    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
891        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
892        in a Windows environment.
894    3.  The pcregrep --count option prints the count for each file even when it is
895        zero, as does GNU grep. However, pcregrep was also printing all files when
896        --files-with-matches was added. Now, when both options are given, it prints
897        counts only for those files that have at least one match. (GNU grep just
898        prints the file name in this circumstance, but including the count seems
899        more useful - otherwise, why use --count?) Also ensured that the
900        combination -clh just lists non-zero counts, with no names.
902    4.  The long form of the pcregrep -F option was incorrectly implemented as
903        --fixed_strings instead of --fixed-strings. This is an incompatible change,
904        but it seems right to fix it, and I didn't think it was worth preserving
905        the old behaviour.
907    5.  The command line items --regex=pattern and --regexp=pattern were not
908        recognized by pcregrep, which required --regex pattern or --regexp pattern
909        (with a space rather than an '='). The man page documented the '=' forms,
910        which are compatible with GNU grep; these now work.
912    6.  No libpcreposix.pc file was created for pkg-config; there was just
913        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
915    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
916        when UCP support is not needed, by modifying the Python script that
917        generates it from Unicode data files. This should not matter if the module
918        is correctly used as a library, but I received one complaint about 50K of
919        unwanted data. My guess is that the person linked everything into his
920        program rather than using a library. Anyway, it does no harm.
922    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
923        was a minimum greater than 1 for a wide character in a possessive
924        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
925        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
926        character. Chaos in the form of incorrect output or a compiling loop could
927        result.
929    9.  The restrictions on what a pattern can contain when partial matching is
930        requested for pcre_exec() have been removed. All patterns can now be
931        partially matched by this function. In addition, if there are at least two
932        slots in the offset vector, the offset of the earliest inspected character
933        for the match and the offset of the end of the subject are set in them when
934        PCRE_ERROR_PARTIAL is returned.
936    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
937        synonymous with PCRE_PARTIAL, for backwards compatibility, and
938        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
939        and may be more useful for multi-segment matching.
941    11. Partial matching with pcre_exec() is now more intuitive. A partial match
942        used to be given if ever the end of the subject was reached; now it is
943        given only if matching could not proceed because another character was
944        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
945        string "Z", which now yields "no match" instead of "partial match". In the
946        case of pcre_dfa_exec(), "no match" is given if every matching path for the
947        final character ended with (*FAIL).
949    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
950        if the pattern had a "must contain" character that was already found in the
951        earlier partial match, unless partial matching was again requested. For
952        example, with the pattern /dog.(body)?/, the "must contain" character is
953        "g". If the first part-match was for the string "dog", restarting with
954        "sbody" failed. This bug has been fixed.
956    13. The string returned by pcre_dfa_exec() after a partial match has been
957        changed so that it starts at the first inspected character rather than the
958        first character of the match. This makes a difference only if the pattern
959        starts with a lookbehind assertion or \b or \B (\K is not supported by
960        pcre_dfa_exec()). It's an incompatible change, but it makes the two
961        matching functions compatible, and I think it's the right thing to do.
963    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
964        so that the demonstration program is easily available in environments where
965        PCRE has not been installed from source.
967    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
968        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
969        library.
971    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
972        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
973        is not the first non-POSIX option to be added. Clearly some people find
974        these options useful.
976    17. If a caller to the POSIX matching function regexec() passes a non-zero
977        value for nmatch with a NULL value for pmatch, the value of
978        nmatch is forced to zero.
980    18. RunGrepTest did not have a test for the availability of the -u option of
981        the diff command, as RunTest does. It now checks in the same way as
982        RunTest, and also checks for the -b option.
984    19. If an odd number of negated classes containing just a single character
985        interposed, within parentheses, between a forward reference to a named
986        subpattern and the definition of the subpattern, compilation crashed with
987        an internal error, complaining that it could not find the referenced
988        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
989        [The bug was that it was starting one character too far in when skipping
990        over the character class, thus treating the ] as data rather than
991        terminating the class. This meant it could skip too much.]
993    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
994        /g option in pcretest when the pattern contains \K, which makes it possible
995        to have an empty string match not at the start, even when the pattern is
996        anchored. Updated pcretest and pcredemo to use this option.
998    21. If the maximum number of capturing subpatterns in a recursion was greater
999        than the maximum at the outer level, the higher number was returned, but
1000        with unset values at the outer level. The correct (outer level) value is
1001        now given.
1003    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1004        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1005        make it do so. The string so far is captured, making this feature
1006        compatible with Perl.
1008    23. The tests have been re-organized, adding tests 11 and 12, to make it
1009        possible to check the Perl 5.10 features against Perl 5.10.
1011    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1012        pattern matches a fixed length string. PCRE did not allow this; now it
1013        does. Neither allows recursion.
1015    25. I finally figured out how to implement a request to provide the minimum
1016        length of subject string that was needed in order to match a given pattern.
1017        (It was back references and recursion that I had previously got hung up
1018        on.) This code has now been added to pcre_study(); it finds a lower bound
1019        to the length of subject needed. It is not necessarily the greatest lower
1020        bound, but using it to avoid searching strings that are too short does give
1021        some useful speed-ups. The value is available to calling programs via
1022        pcre_fullinfo().
1024    26. While implementing 25, I discovered to my embarrassment that pcretest had
1025        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1026        study optimizations had never been tested with that matching function.
1027        Oops. What is worse, even when it was passed study data, there was a bug in
1028        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1029        were also very few tests of studied patterns with pcre_dfa_exec().
1031    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1032        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1033        on the other side of the coin, they are no longer allowed to have different
1034        names, because these cannot be distinguished in PCRE, and this has caused
1035        confusion. (This is a difference from Perl.)
1037    28. When duplicate subpattern names are present (necessarily with different
1038        numbers, as required by 27 above), and a test is made by name in a
1039        conditional pattern, either for a subpattern having been matched, or for
1040        recursion in such a pattern, all the associated numbered subpatterns are
1041        tested, and the overall condition is true if the condition is true for any
1042        one of them. This is the way Perl works, and is also more like the way
1043        testing by number works.
1046    Version 7.9 11-Apr-09
1047    ---------------------
1049    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1050        (pcretest), all targets were linked against these libraries. This included
1051        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1052        libraries. This caused unwanted dependencies to be created. This problem
1053        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1054        pcretest is linked with readline.
1056    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1057        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1058        moved outside it again, because FALSE and TRUE are already defined in AIX,
1059        but BOOL is not.
1061    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1062        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1064    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1065        hyphens) following file names and line numbers when outputting matching
1066        lines. This is not true; no spaces are inserted. I have also clarified the
1067        wording for the --colour (or --color) option.
1069    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1070        was not coloured; this is different to GNU grep, so I have changed it to be
1071        the same.
1073    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1074        each matching line was coloured. Now it goes on to look for further matches
1075        of any of the test patterns, which is the same behaviour as GNU grep.
1077    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1078        doesn't make sense to accept an empty string match in pcregrep, so I have
1079        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1080        seems to be how GNU grep behaves.
1082    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1083        start or after a newline", because the conditional assertion was not being
1084        correctly handled. The rule now is that both the assertion and what follows
1085        in the first alternative must satisfy the test.
1087    9.  If auto-callout was enabled in a pattern with a conditional group whose
1088        condition was an assertion, PCRE could crash during matching, both with
1089        pcre_exec() and pcre_dfa_exec().
1091    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1092        used for matching.
1094    11. Unicode property support in character classes was not working for
1095        characters (bytes) greater than 127 when not in UTF-8 mode.
1097    12. Added the -M command line option to pcretest.
1099    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1101    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1103    16. Added comments and documentation about mis-use of no_arg in the C++
1104        wrapper.
1106    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1107        from Martin Jerabek that uses macro names for all relevant character and
1108        string constants.
1110    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1111        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1112        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1113        these, but not everybody uses configure.
1115    19. A conditional group that had only one branch was not being correctly
1116        recognized as an item that could match an empty string. This meant that an
1117        enclosing group might also not be so recognized, causing infinite looping
1118        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1119        with the subject "ab", where knowledge that the repeated group can match
1120        nothing is needed in order to break the loop.
1122    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1123        exec(), but without supplying a callout function, matching went wrong.
1125    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1126        leak if the size of the offset vector was greater than 30. When the vector
1127        is smaller, the saved offsets during recursion go onto a local stack
1128        vector, but for larger vectors malloc() is used. It was failing to free
1129        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1130        error, in fact).
1132    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1133        heapframe that is used only when UTF-8 support is enabled. This caused no
1134        problem, but was untidy.
1136    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1137        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1138        included within another project.
1140    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1141        slightly modified by me:
1143          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1144              not building pcregrep.
1146          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1147              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1149    25. Forward references, both numeric and by name, in patterns that made use of
1150        duplicate group numbers, could behave incorrectly or give incorrect errors,
1151        because when scanning forward to find the reference group, PCRE was not
1152        taking into account the duplicate group numbers. A pattern such as
1153        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1155    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1156        the feature of a custom pointer more persuasive (as requested by a user).
1158    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1159        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1160        pre-defined.
1162    28. Added support for (*UTF8) at the start of a pattern.
1164    29. Arrange for flags added by the "release type" setting in CMake to be shown
1165        in the configuration summary.
1168    Version 7.8 05-Sep-08
1169    ---------------------
1171    1.  Replaced UCP searching code with optimized version as implemented for Ad
1172        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1173        stage table and inline lookup instead of a function, giving speed ups of 2
1174        to 5 times on some simple patterns that I tested. Permission was given to
1175        distribute the MultiStage2.py script that generates the tables (it's not in
1176        the tarball, but is in the Subversion repository).
1178    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1179        scripts.
1181    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1182        a group with a zero qualifier. The result of the study could be incorrect,
1183        or the function might crash, depending on the pattern.
1185    4.  Caseless matching was not working for non-ASCII characters in back
1186        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1187        It now works when Unicode Property Support is available.
1189    5.  In pcretest, an escape such as \x{de} in the data was always generating
1190        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1191        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1192        truncation.
1194    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1196    7.  Added two (int) casts to pcregrep when printing the difference of two
1197        pointers, in case they are 64-bit values.
1199    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1200        test 2 if it fails.
1202    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1203        and a #define of that name to empty if it is not externally set. This is to
1204        allow users of MSVC to set it if necessary.
1206    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1207        the convenience functions in the pcre_get.c source file.
1209    11. An option change at the start of a pattern that had top-level alternatives
1210        could cause overwriting and/or a crash. This command provoked a crash in
1211        some environments:
1213          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1215        This potential security problem was recorded as CVE-2008-2371.
1217    12. For a pattern where the match had to start at the beginning or immediately
1218        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1219        pcre_dfa_exec() could read past the end of the passed subject if there was
1220        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1221        pcretest so that it places the subject at the end of its malloc-ed buffer.
1223    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1224        exec() might read past the end of the data buffer in UTF-8 mode.
1226    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1227        the data contained the byte 0x85 as part of a UTF-8 character within its
1228        first line. This applied both to normal and DFA matching.
1230    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1231        /^[^d]*?$/8 failed to match "abc".
1233    16. Added a missing copyright notice to pcrecpp_internal.h.
1235    17. Make it more clear in the documentation that values returned from
1236        pcre_exec() in ovector are byte offsets, not character counts.
1238    18. Tidied a few places to stop certain compilers from issuing warnings.
1240    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1241        supplied by Stefan Weber. I made a further small update for 7.8 because
1242        there is a change of source arrangements: the pcre_searchfuncs.c module is
1243        replaced by pcre_ucd.c.
1246    Version 7.7 07-May-08
1247    ---------------------
1249    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1250        a string to a long long, pretend we don't even have a long long." This is
1251        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1253    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1254        pre-7.6 versions, which defined a global no_arg variable instead of putting
1255        it in the RE class. (See also #8 below.)
1257    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1258        Lopes.
1260    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1262        (1) The include/exclude patterns were being applied to the whole pathnames
1263            of files, instead of just to the final components.
1265        (2) If there was more than one level of directory, the subdirectories were
1266            skipped unless they satisfied the include/exclude conditions. This is
1267            inconsistent with GNU grep (and could even be seen as contrary to the
1268            pcregrep specification - which I improved to make it absolutely clear).
1269            The action now is always to scan all levels of directory, and just
1270            apply the include/exclude patterns to regular files.
1272    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1273        --exclude_dir in the tests to avoid scanning .svn directories.
1275    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1276        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1277        doesn't support NULs in patterns.
1279    7.  Added some missing "const"s to declarations of static tables in
1280        pcre_compile.c and pcre_dfa_exec.c.
1282    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1283        caused by fix #2  above. (Subsequently also a second patch to fix the
1284        first patch. And a third patch - this was a messy problem.)
1286    9.  Applied Craig's patch to remove the use of push_back().
1288    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1289        matching function regexec().
1291    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1292        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1293        references. PCRE supports relative numbers with this syntax (I don't think
1294        Oniguruma does).
1296    12. Previously, a group with a zero repeat such as (...){0} was completely
1297        omitted from the compiled regex. However, this means that if the group
1298        was called as a subroutine from elsewhere in the pattern, things went wrong
1299        (an internal error was given). Such groups are now left in the compiled
1300        pattern, with a new opcode that causes them to be skipped at execution
1301        time.
1303    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1304        to the way PCRE behaves:
1306        (a) A lone ] character is dis-allowed (Perl treats it as data).
1308        (b) A back reference to an unmatched subpattern matches an empty string
1309            (Perl fails the current match path).
1311        (c) A data ] in a character class must be notated as \] because if the
1312            first data character in a class is ], it defines an empty class. (In
1313            Perl it is not possible to have an empty class.) The empty class []
1314            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1315            The negative empty class [^] matches any one character, independently
1316            of the DOTALL setting.
1318    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1319        non-existent subpattern following a character class starting with ']' and
1320        containing () gave an internal compiling error instead of "reference to
1321        non-existent subpattern". Fortunately, when the pattern did exist, the
1322        compiled code was correct. (When scanning forwards to check for the
1323        existencd of the subpattern, it was treating the data ']' as terminating
1324        the class, so got the count wrong. When actually compiling, the reference
1325        was subsequently set up correctly.)
1327    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1328        it was being rejected as not supported by pcre_dfa_exec(), even though
1329        other assertions are supported. I have made pcre_dfa_exec() support
1330        (*FAIL).
1332    16. The implementation of 13c above involved the invention of a new opcode,
1333        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1334        cannot be changed at match time, I realized I could make a small
1335        improvement to matching performance by compiling OP_ALLANY instead of
1336        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1337        on the OP_ANY path.
1339    17. Compiling pcretest on Windows with readline support failed without the
1340        following two fixes: (1) Make the unistd.h include conditional on
1341        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1343    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1344        ncurses library to be included for pcretest when ReadLine support is
1345        requested, but also to allow for it to be overridden. This patch came from
1346        Daniel Bergström.
1348    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1349        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1350        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1351        spotting this.
1354    Version 7.6 28-Jan-08
1355    ---------------------
1357    1.  A character class containing a very large number of characters with
1358        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1359        overflow.
1361    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1362        HAVE_LONG_LONG is not defined.
1364    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1365        bring them up to date and include new features. This patch includes:
1367        - Fixed PH's badly added libz and libbz2 support.
1368        - Fixed a problem with static linking.
1369        - Added pcredemo. [But later removed - see 7 below.]
1370        - Fixed dftables problem and added an option.
1371        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1372            HAVE_LONG_LONG.
1373        - Added readline support for pcretest.
1374        - Added an listing of the option settings after cmake has run.
1376    4.  A user submitted a patch to Makefile that makes it easy to create
1377        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1378        Makefile.am that cause it to include this special target, without
1379        affecting anything else. Note that the same mingw target plus all
1380        the other distribution libraries and programs are now supported
1381        when configuring with CMake (see 6 below) instead of with
1382        Configure/Make.
1384    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1385        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1386        exported in the Windows port". It has not yet been confirmed that the patch
1387        solves the problem, but it does no harm.
1389    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1390        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1391        with CMake, and also correct the comment about stack recursion.
1393    7.  Remove the automatic building of pcredemo from the ./configure system and
1394        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1395        of a program that users should build themselves after PCRE is installed, so
1396        building it automatically is not really right. What is more, it gave
1397        trouble in some build environments.
1399    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1402    Version 7.5 10-Jan-08
1403    ---------------------
1405    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1406        values in parens when parsing an RE using the C++ wrapper."
1408    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1409        Characters greater than 255 were excluded from the class instead of being
1410        included.
1412    3.  The same bug as (2) above applied to negated POSIX classes such as
1413        [:^space:].
1415    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1416        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1417        I have changed it.
1419    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1420        first named subpattern) and a construct such as (?&a) would reference the
1421        first named subpattern whose name started with "a" (in other words, the
1422        length check was missing). Both these problems are fixed. "Subpattern name
1423        expected" is now given for (?&) (a zero-length name), and this patch also
1424        makes it give the same error for \k'' (previously it complained that that
1425        was a reference to a non-existent subpattern).
1427    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1428        this is right because (?- can be followed by option settings as well as by
1429        digits. I have, however, made the messages clearer.
1431    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1432        than the number used in the conditional) now cause a compile-time error.
1433        This is actually not compatible with Perl, which accepts such patterns, but
1434        treats the conditional as always being FALSE (as PCRE used to), but it
1435        seems to me that giving a diagnostic is better.
1437    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1438        and messages.
1440    9.  Fix two occurrences of "backslash" in comments that should have been
1441        "backspace".
1443    10. Remove two redundant lines of code that can never be obeyed (their function
1444        was moved elsewhere).
1446    11. The program that makes PCRE's Unicode character property table had a bug
1447        which caused it to generate incorrect table entries for sequences of
1448        characters that have the same character type, but are in different scripts.
1449        It amalgamated them into a single range, with the script of the first of
1450        them. In other words, some characters were in the wrong script. There were
1451        thirteen such cases, affecting characters in the following ranges:
1453          U+002b0 - U+002c1
1454          U+0060c - U+0060d
1455          U+0061e - U+00612
1456          U+0064b - U+0065e
1457          U+0074d - U+0076d
1458          U+01800 - U+01805
1459          U+01d00 - U+01d77
1460          U+01d9b - U+01dbf
1461          U+0200b - U+0200f
1462          U+030fc - U+030fe
1463          U+03260 - U+0327f
1464          U+0fb46 - U+0fbb1
1465          U+10450 - U+1049d
1467    12. The -o option (show only the matching part of a line) for pcregrep was not
1468        compatible with GNU grep in that, if there was more than one match in a
1469        line, it showed only the first of them. It now behaves in the same way as
1470        GNU grep.
1472    13. If the -o and -v options were combined for pcregrep, it printed a blank
1473        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1474        does the same. The return code can be used to tell if there were any
1475        non-matching lines.
1477    14. Added --file-offsets and --line-offsets to pcregrep.
1479    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1480        infinitely looping recursion. The bug was that positive lookaheads were not
1481        being skipped when checking for a possible empty match (negative lookaheads
1482        and both kinds of lookbehind were skipped).
1484    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1485        inclusion of <windows.h> to before rather than after the definition of
1486        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1488    17. Specifying a possessive quantifier with a specific limit for a Unicode
1489        character property caused pcre_compile() to compile bad code, which led at
1490        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1491        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1492        caused the error; without that there was no problem.
1494    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1496    19. Added --enable-pcretest-libreadline.
1498    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1499        RE::GlobalReplace(). As a result, the number of replacements returned was
1500        double what it should be. I removed one of the increments, but Craig sent a
1501        later patch that removed the other one (the right fix) and added unit tests
1502        that check the return values (which was not done before).
1504    21. Several CMake things:
1506        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1507            the names libpcre and libpcreposix, not just pcre and pcreposix.
1509        (2) The above change means that pcretest and pcregrep are now correctly
1510            linked with the newly-built libraries, not previously installed ones.
1514    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1515        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1516        UTF-8 newline character). The key issue is that the pattern starts .*;
1517        this means that the match must be either at the beginning, or after a
1518        newline. The bug was in the code for advancing after a failed match and
1519        checking that the new position followed a newline. It was not taking
1520        account of UTF-8 characters correctly.
1522    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1523        character classes. PCRE was not treating the sequence [:...:] as a
1524        character class unless the ... were all letters. Perl, however, seems to
1525        allow any characters between [: and :], though of course it rejects as
1526        unknown any "names" that contain non-letters, because all the known class
1527        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1528        for example, whereas PCRE did not - it did not recognize a POSIX character
1529        class. This seemed a bit dangerous, so the code has been changed to be
1530        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1531        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1532        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1533        Perl does, and where it didn't before.
1535    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1536        Windows environments %n is disabled by default.
1539    Version 7.4 21-Sep-07
1540    ---------------------
1542    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1543        means that a class such as [\s] counted as "explicit reference to CR or
1544        LF". That isn't really right - the whole point of the change was to try to
1545        help when there was an actual mention of one of the two characters. So now
1546        the change happens only if \r or \n (or a literal CR or LF) character is
1547        encountered.
1549    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1550        of both had grown to the point where there were only 3 bits left.
1551        Fortunately, there was spare space in the data structure, and so I have
1552        moved the internal flags into a new 16-bit field to free up more option
1553        bits.
1555    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1556        but did not set the internal JCHANGED flag - either of these is enough to
1557        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1558        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1559        start sets both bits.
1561    4.  Added options (at build time, compile time, exec time) to change \R from
1562        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1564    5.  doc/pcresyntax.html was missing from the distribution.
1566    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1567        compatibility, even though it is no longer used.
1569    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1570        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1571        windows.h file is present (where different names are used). [This was
1572        reversed later after testing - see 16 below.]
1574    8.  Changed all #include <config.h> to #include "config.h". There were also
1575        some further <pcre.h> cases that I changed to "pcre.h".
1577    9.  When pcregrep was used with the --colour option, it missed the line ending
1578        sequence off the lines that it output.
1580    10. It was pointed out to me that arrays of string pointers cause lots of
1581        relocations when a shared library is dynamically loaded. A technique of
1582        using a single long string with a table of offsets can drastically reduce
1583        these. I have refactored PCRE in four places to do this. The result is
1584        dramatic:
1586          Originally:                          290
1587          After changing UCP table:            187
1588          After changing error message table:   43
1589          After changing table of "verbs"       36
1590          After changing table of Posix names   22
1592        Thanks to the folks working on Gregex for glib for this insight.
1594    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1595        unicode-properties was also set.
1597    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1599    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1600        checked only for CRLF.
1602    14. Added casts to pcretest.c to avoid compiler warnings.
1604    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1606    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1607        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1608        entirely. This removes changes made in 7 above.
1610    17. The CMake files have been updated, and there is now more information about
1611        building with CMake in the NON-UNIX-USE document.
1614    Version 7.3 28-Aug-07
1615    ---------------------
1617     1. In the rejigging of the build system that eventually resulted in 7.1, the
1618        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1619        brackets there is not right, since it causes compilers to look for an
1620        installed pcre.h, not the version that is in the source that is being
1621        compiled (which of course may be different). I have changed it back to:
1623          #include "pcre.h"
1625        I have a vague recollection that the change was concerned with compiling in
1626        different directories, but in the new build system, that is taken care of
1627        by the VPATH setting the Makefile.
1629     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1630        when the subject happened to end in the byte 0x85 (e.g. if the last
1631        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1632        characters but of course it shouldn't be taken as a newline when it is part
1633        of another character. The bug was that, for an unlimited repeat of . in
1634        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1635        characters when looking for a newline.
1637     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1639     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1640        in debug output.
1642     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1643        long printing in the pcrecpp unittest when running under MinGW.
1645     6. ESC_K was left out of the EBCDIC table.
1647     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1648        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1649        limit also applies to "virtual nesting" when a pattern is recursive, and in
1650        this case 1000 isn't so big. I have been able to remove this limit at the
1651        expense of backing off one optimization in certain circumstances. Normally,
1652        when pcre_exec() would call its internal match() function recursively and
1653        immediately return the result unconditionally, it uses a "tail recursion"
1654        feature to save stack. However, when a subpattern that can match an empty
1655        string has an unlimited repetition quantifier, it no longer makes this
1656        optimization. That gives it a stack frame in which to save the data for
1657        checking that an empty string has been matched. Previously this was taken
1658        from the 1000-entry workspace that had been reserved. So now there is no
1659        explicit limit, but more stack is used.
1661     8. Applied Daniel's patches to solve problems with the import/export magic
1662        syntax that is required for Windows, and which was going wrong for the
1663        pcreposix and pcrecpp parts of the library. These were overlooked when this
1664        problem was solved for the main library.
1666     9. There were some crude static tests to avoid integer overflow when computing
1667        the size of patterns that contain repeated groups with explicit upper
1668        limits. As the maximum quantifier is 65535, the maximum group length was
1669        set at 30,000 so that the product of these two numbers did not overflow a
1670        32-bit integer. However, it turns out that people want to use groups that
1671        are longer than 30,000 bytes (though not repeat them that many times).
1672        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1673        made it possible to implement the integer overflow checks in a much more
1674        dynamic way, which I have now done. The artificial limitation on group
1675        length has been removed - we now have only the limit on the total length of
1676        the compiled pattern, which depends on the LINK_SIZE setting.
1678    10. Fixed a bug in the documentation for get/copy named substring when
1679        duplicate names are permitted. If none of the named substrings are set, the
1680        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1681        empty string.
1683    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1684        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1685        because the ] is interpreted as the first data character and the
1686        terminating ] is not found. PCRE has been made compatible with Perl in this
1687        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1688        cause memory overwriting.
1690    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1691        string has been matched (to stop an infinite loop). It was not recognizing
1692        a conditional subpattern that could match an empty string if that
1693        subpattern was within another subpattern. For example, it looped when
1694        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1695        condition was not nested. This bug has been fixed.
1697    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1698        past the start of the subject in the presence of bytes with the top bit
1699        set, for example "\x8aBCD".
1701    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1702        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1704    14. Optimized (?!) to (*FAIL).
1706    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1707        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1708        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1709        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1710        does: it's just the validity check that is more restrictive.
1712    16. Inserted checks for integer overflows during escape sequence (backslash)
1713        processing, and also fixed erroneous offset values for syntax errors during
1714        backslash processing.
1716    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1717        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1719    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1720        caused an overrun.
1722    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1723        something other than just ASCII characters) inside a group that had an
1724        unlimited repeat caused a loop at compile time (while checking to see
1725        whether the group could match an empty string).
1727    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1728        [\P{Any}] did so. (Error in the code for printing property names.)
1730    21. An orphan \E inside a character class could cause a crash.
1732    22. A repeated capturing bracket such as (A)? could cause a wild memory
1733        reference during compilation.
1735    23. There are several functions in pcre_compile() that scan along a compiled
1736        expression for various reasons (e.g. to see if it's fixed length for look
1737        behind). There were bugs in these functions when a repeated \p or \P was
1738        present in the pattern. These operators have additional parameters compared
1739        with \d, etc, and these were not being taken into account when moving along
1740        the compiled data. Specifically:
1742        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1743            length.
1745        (b) An item such as \pL+ within a repeated group could cause crashes or
1746            loops.
1748        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1749            "reference to non-existent subpattern" error.
1751        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1753    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1754        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1756    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1758    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1759        character were causing crashes (broken optimization).
1761    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1762        \p or \P) caused a compile-time loop.
1764    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1765        break. For example, the unstudied pattern [\r\n]A does not match the string
1766        "\r\nA" because change 7.0/46 below moves the current point on by two
1767        characters after failing to match at the start. However, the pattern \nA
1768        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1769        the same is true. There doesn't seem any very clean way out of this, but
1770        what I have chosen to do makes the common cases work: PCRE now takes note
1771        of whether there can be an explicit match for \r or \n anywhere in the
1772        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1773        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1774        pattern has explicit CR or LF references.
1776    29. Added (*CR) etc for changing newline setting at start of pattern.
1779    Version 7.2 19-Jun-07
1780    ---------------------
1782     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1783        which is apparently normally available under Windows.
1785     2. Re-jig the pcregrep tests with different newline settings in an attempt
1786        to make them independent of the local environment's newline setting.
1788     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1790     4. Some of the "internals" tests were previously cut out when the link size
1791        was not 2, because the output contained actual offsets. The recent new
1792        "Z" feature of pcretest means that these can be cut out, making the tests
1793        usable with all link sizes.
1795     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1796        stack recursion. This gives a massive performance boost under BSD, but just
1797        a small improvement under Linux. However, it saves one field in the frame
1798        in all cases.
1800     6. Added more features from the forthcoming Perl 5.10:
1802        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1803            recursion call. It refers to the nth most recently opened parentheses.
1805        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1806            to be opened parentheses.
1808        (c) Conditions that refer to capturing parentheses can be specified
1809            relatively, for example, (?(-2)... or (?(+3)...
1811        (d) \K resets the start of the current match so that everything before
1812            is not part of it.
1814        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1816        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1817            reference syntax.
1819        (g) (?| introduces a group in which the numbering of parentheses in each
1820            alternative starts with the same number.
1822        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1824     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1827     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1828        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1829        for detecting groups that can match an empty string.
1831     9. A pattern with a very large number of alternatives (more than several
1832        hundred) was running out of internal workspace during the pre-compile
1833        phase, where pcre_compile() figures out how much memory will be needed. A
1834        bit of new cunning has reduced the workspace needed for groups with
1835        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1836        workspace instead of running out of the 4096 that are available.
1838    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1840    11. Applied patch from Google to remove an optimization that didn't quite work.
1841        The report of the bug said:
1843          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1844          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1845          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1847    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1848        it matched the wrong number of bytes.
1851    Version 7.1 24-Apr-07
1852    ---------------------
1854     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1855        that is more "standard", making use of automake and other Autotools. There
1856        is some re-arrangement of the files and adjustment of comments consequent
1857        on this.
1859     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1860        for recursive directory scanning broke on some systems because the files
1861        are not scanned in any specific order and on different systems the order
1862        was different. A call to "sort" has been inserted into RunGrepTest for the
1863        approprate test as a short-term fix. In the longer term there may be an
1864        alternative.
1866     3. I had an email from Eric Raymond about problems translating some of PCRE's
1867        man pages to HTML (despite the fact that I distribute HTML pages, some
1868        people do their own conversions for various reasons). The problems
1869        concerned the use of low-level troff macros .br and .in. I have therefore
1870        removed all such uses from the man pages (some were redundant, some could
1871        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1872        HTML has been updated to handle .nf/.fi and to complain if it encounters
1873        .br or .in.
1875     4. Updated comments in configure.ac that get placed in config.h.in and also
1876        arranged for config.h to be included in the distribution, with the name
1877        config.h.generic, for the benefit of those who have to compile without
1878        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1880     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1881        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1882        makevp.bat for the current PCRE, using the additional files
1883        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1885     6. A Windows user reported a minor discrepancy with test 2, which turned out
1886        to be caused by a trailing space on an input line that had got lost in his
1887        copy. The trailing space was an accident, so I've just removed it.
1889     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1890        that is needed.
1892     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1893        as "const" (a) because they are and (b) because it helps the PHP
1894        maintainers who have recently made a script to detect big data structures
1895        in the php code that should be moved to the .rodata section. I remembered
1896        to update Builducptable as well, so it won't revert if ucptable.h is ever
1897        re-created.
1899     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1900        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1901        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1902        support is not required. This saves 1.5-2K of code, which is important in
1903        some applications.
1905        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1906        so as not to refer to the tables, even though these functions will never be
1907        called when UTF-8 support is disabled. Otherwise there are problems with a
1908        shared library.
1910    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1912        (a) It was defining its arguments as char * instead of void *.
1914        (b) It was assuming that all moves were upwards in memory; this was true
1915            a long time ago when I wrote it, but is no longer the case.
1917        The emulated memove() is provided for those environments that have neither
1918        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1919        is clearly not the case, as these two bugs were recently reported.
1921    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1922        and Detrail to create the HTML documentation, the .txt form of the man
1923        pages, and it removes trailing spaces from listed files. It also creates
1924        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1925        case, it wraps all the #defines with #ifndefs. This script should be run
1926        before "make dist".
1928    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1929        with Unicode property support.
1931        (a) For a maximizing quantifier, if the two different cases of the
1932            character were of different lengths in their UTF-8 codings (there are
1933            some cases like this - I found 11), and the matching function had to
1934            back up over a mixture of the two cases, it incorrectly assumed they
1935            were both the same length.
1937        (b) When PCRE was configured to use the heap rather than the stack for
1938            recursion during matching, it was not correctly preserving the data for
1939            the other case of a UTF-8 character when checking ahead for a match
1940            while processing a minimizing repeat. If the check also involved
1941            matching a wide character, but failed, corruption could cause an
1942            erroneous result when trying to check for a repeat of the original
1943            character.
1945    13. Some tidying changes to the testing mechanism:
1947        (a) The RunTest script now detects the internal link size and whether there
1948            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1949            values substituted by "configure". (The RunGrepTest script already did
1950            this for UTF-8.) The configure.ac script no longer substitutes the
1951            relevant variables.
1953        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1954            with length and offset values. This means that the output is different
1955            for different internal link sizes. Test 2 is skipped for link sizes
1956            other than 2 because of this, bypassing the problem. Unfortunately,
1957            there was also a test in test 3 (the locale tests) that used /B and
1958            failed for link sizes other than 2. Rather than cut the whole test out,
1959            I have added a new /Z option to pcretest that replaces the length and
1960            offset values with spaces. This is now used to make test 3 independent
1961            of link size. (Test 2 will be tidied up later.)
1963    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1964        segmentation fault instead of returning the appropriate error message.
1966    15. In multiline mode when the newline sequence was set to "any", the pattern
1967        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1968        This doesn't seem right; it now treats the CRLF combination as the line
1969        ending, and so does not match in that case. It's only a pattern such as ^$
1970        that would hit this one: something like ^ABC$ would have failed after \r
1971        and then tried again after \r\n.
1973    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1974        in an attempt to make files that differ only in their line terminators
1975        compare equal. This works on Linux.
1977    17. Under certain error circumstances pcregrep might try to free random memory
1978        as it exited. This is now fixed, thanks to valgrind.
1980    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1981        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1982        was because its rules for how to advance for /g after matching an empty
1983        string at the end of a line did not allow for this case. They now check for
1984        it specially.
1986    20. pcretest is supposed to handle patterns and data of any length, by
1987        extending its buffers when necessary. It was getting this wrong when the
1988        buffer for a data line had to be extended.
1990    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1991        CRLF as a newline sequence.
1993    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1994        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1995        I have nevertheless tidied it up.
1997    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1999    24. Added a man page for pcre-config.
2002    Version 7.0 19-Dec-06
2003    ---------------------
2005     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2006        moving to gcc 4.1.1.
2008     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2009        sys/time.h, which is documented as needed for this function. It doesn't
2010        seem to matter on Linux, but it showed up on some releases of OS X.
2012     3. It seems that there are systems where bytes whose values are greater than
2013        127 match isprint() in the "C" locale. The "C" locale should be the
2014        default when a C program starts up. In most systems, only ASCII printing
2015        characters match isprint(). This difference caused the output from pcretest
2016        to vary, making some of the tests fail. I have changed pcretest so that:
2018        (a) When it is outputting text in the compiled version of a pattern, bytes
2019            other than 32-126 are always shown as hex escapes.
2021        (b) When it is outputting text that is a matched part of a subject string,
2022            it does the same, unless a different locale has been set for the match
2023            (using the /L modifier). In this case, it uses isprint() to decide.
2025     4. Fixed a major bug that caused incorrect computation of the amount of memory
2026        required for a compiled pattern when options that changed within the
2027        pattern affected the logic of the preliminary scan that determines the
2028        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2029        that the computed length was too small. The symptoms of this bug were
2030        either the PCRE error "internal error: code overflow" from pcre_compile(),
2031        or a glibc crash with a message such as "pcretest: free(): invalid next
2032        size (fast)". Examples of patterns that provoked this bug (shown in
2033        pcretest format) are:
2035          /(?-x: )/x
2036          /(?x)(?-x: \s*#\s*)/
2037          /((?i)[\x{c0}])/8
2038          /(?i:[\x{c0}])/8
2040        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2041        is now done differently.
2043     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2044        wrapper classes; (b) implement a new function in the C++ scanner that is
2045        more efficient than the old way of doing things because it avoids levels of
2046        recursion in the regex matching; (c) add a paragraph to the documentation
2047        for the FullMatch() function.
2049     6. The escape sequence \n was being treated as whatever was defined as
2050        "newline". Not only was this contrary to the documentation, which states
2051        that \n is character 10 (hex 0A), but it also went horribly wrong when
2052        "newline" was defined as CRLF. This has been fixed.
2054     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2055        was being set to -1 for the "end of line" case (supposedly a value that no
2056        character can have). Though this value is never used (the check for end of
2057        line is "zero bytes in current character"), it caused compiler complaints.
2058        I've changed it to 0xffffffff.
2060     8. In pcre_version.c, the version string was being built by a sequence of
2061        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2062        string (as it is for production releases) called a macro with an empty
2063        argument. The C standard says the result of this is undefined. The gcc
2064        compiler treats it as an empty string (which was what was wanted) but it is
2065        reported that Visual C gives an error. The source has been hacked around to
2066        avoid this problem.
2068     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2069        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2070        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2071        of them did).
2073    10. Originally, pcretest opened its input and output without "b"; then I was
2074        told that "b" was needed in some environments, so it was added for release
2075        5.0 to both the input and output. (It makes no difference on Unix-like
2076        systems.) Later I was told that it is wrong for the input on Windows. I've
2077        now abstracted the modes into two macros, to make it easier to fiddle with
2078        them, and removed "b" from the input mode under Windows.
2080    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2082    12. Added -help and --help to pcretest as an official way of being reminded
2083        of the options.
2085    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2086        and pcrecpp.cc because they annoy compilers at high warning levels.
2088    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2090    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2091        scripts are not C programs :-) and which was not noticed because it works
2092        on Linux.
2094    16. pcretest is supposed to handle any length of pattern and data line (as one
2095        line or as a continued sequence of lines) by extending its input buffer if
2096        necessary. This feature was broken for very long pattern lines, leading to
2097        a string of junk being passed to pcre_compile() if the pattern was longer
2098        than about 50K.
2100    17. I have done a major re-factoring of the way pcre_compile() computes the
2101        amount of memory needed for a compiled pattern. Previously, there was code
2102        that made a preliminary scan of the pattern in order to do this. That was
2103        OK when PCRE was new, but as the facilities have expanded, it has become
2104        harder and harder to keep it in step with the real compile phase, and there
2105        have been a number of bugs (see for example, 4 above). I have now found a
2106        cunning way of running the real compile function in a "fake" mode that
2107        enables it to compute how much memory it would need, while actually only
2108        ever using a few hundred bytes of working memory and without too many
2109        tests of the mode. This should make future maintenance and development
2110        easier. A side effect of this work is that the limit of 200 on the nesting
2111        depth of parentheses has been removed (though this was never a serious
2112        limitation, I suspect). However, there is a downside: pcre_compile() now
2113        runs more slowly than before (30% or more, depending on the pattern). I
2114        hope this isn't a big issue. There is no effect on runtime performance.
2116    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2117        newline (only possible for the last line of a file) and it was a
2118        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2120    19. Added additional timing features to pcretest. (1) The -tm option now times
2121        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2122        separate command line item, by a number that specifies the number of
2123        repeats to use when timing. The default is 50000; this gives better
2124        precision, but takes uncomfortably long for very large patterns.
2126    20. Extended pcre_study() to be more clever in cases where a branch of a
2127        subpattern has no definite first character. For example, (a*|b*)[cd] would
2128        previously give no result from pcre_study(). Now it recognizes that the
2129        first character must be a, b, c, or d.
2131    21. There was an incorrect error "recursive call could loop indefinitely" if
2132        a subpattern (or the entire pattern) that was being tested for matching an
2133        empty string contained only one non-empty item after a nested subpattern.
2134        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2135        incorrectly, because the \d was being skipped in the check.
2137    22. The pcretest program now has a new pattern option /B and a command line
2138        option -b, which is equivalent to adding /B to every pattern. This causes
2139        it to show the compiled bytecode, without the additional information that
2140        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2141        is the same as /B/I).
2143    23. A new optimization is now able automatically to treat some sequences such
2144        as a*b as a*+b. More specifically, if something simple (such as a character
2145        or a simple class like \d) has an unlimited quantifier, and is followed by
2146        something that cannot possibly match the quantified thing, the quantifier
2147        is automatically "possessified".
2149    24. A recursive reference to a subpattern whose number was greater than 39
2150        went wrong under certain circumstances in UTF-8 mode. This bug could also
2151        have affected the operation of pcre_study().
2153    25. Realized that a little bit of performance could be had by replacing
2154        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2156    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2158    27. Possessive quantifiers such as a++ were previously implemented by turning
2159        them into atomic groups such as ($>a+). Now they have their own opcodes,
2160        which improves performance. This includes the automatically created ones
2161        from 23 above.
2163    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2164        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2165        the first matched character to be a colon. This applied both to named and
2166        numbered groups.
2168    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2170    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2171        building PCRE on Windows easier, so I have included it in the distribution.
2173    31. There is now a check in pcretest against a ridiculously large number being
2174        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2175        loop, the loop is abandoned.
2177    32. Forward references to subpatterns in conditions such as (?(2)...) where
2178        subpattern 2 is defined later cause pcre_compile() to search forwards in
2179        the pattern for the relevant set of parentheses. This search went wrong
2180        when there were unescaped parentheses in a character class, parentheses
2181        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2183    33. "Subroutine" calls and backreferences were previously restricted to
2184        referencing subpatterns earlier in the regex. This restriction has now
2185        been removed.
2187    34. Added a number of extra features that are going to be in Perl 5.10. On the
2188        whole, these are just syntactic alternatives for features that PCRE had
2189        previously implemented using the Python syntax or my own invention. The
2190        other formats are all retained for compatibility.
2192        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2193            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2194            also .NET compatible.
2196        (b) A recursion or subroutine call to a named group can now be defined as
2197            (?&name) as well as (?P>name).
2199        (c) A backreference to a named group can now be defined as \k<name> or
2200            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2201            5.10, are also .NET compatible.
2203        (d) A conditional reference to a named group can now use the syntax
2204            (?(<name>) or (?('name') as well as (?(name).
2206        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2207            groups (named and numbered) that are never evaluated inline, but can be
2208            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2209            is always false. There may be only one alternative in such a group.
2211        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2212            as the simple (?(R). The condition is true only if the most recent
2213            recursion is that of the given number or name. It does not search out
2214            through the entire recursion stack.
2216        (g) The escape \gN or \g{N} has been added, where N is a positive or
2217            negative number, specifying an absolute or relative reference.
2219    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2220        some "unreachable code" warnings.
2222    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2223        things, this adds five new scripts.
2225    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2226        There were also incompatibilities regarding the handling of \Q..\E inside
2227        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2228        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2230    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2231        matches an empty string, and forcibly breaks the loop. There were bugs in
2232        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2233        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2234        separate and independent bugs (that affected different cases) have been
2235        fixed.
2237    39. Refactored the code to abolish the use of different opcodes for small
2238        capturing bracket numbers. This is a tidy that I avoided doing when I
2239        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2240        The new approach is not only tidier, it makes it possible to reduce the
2241        memory needed to fix the previous bug (38).
2243    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2244        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2245        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2246        mode.
2248    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2249        report.
2251    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2252        copy construction and assignment in the C++ wrapper.
2254    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2255        couple of bugs that could have given wrong results in the "--newline=crlf"
2256        case.
2258    44. Added a number of casts and did some reorganization of signed/unsigned int
2259        variables following suggestions from Dair Grant. Also renamed the variable
2260        "this" as "item" because it is a C++ keyword.
2262    45. Arranged for dftables to add
2264          #include "pcre_internal.h"
2266        to pcre_chartables.c because without it, gcc 4.x may remove the array
2267        definition from the final binary if PCRE is built into a static library and
2268        dead code stripping is activated.
2270    46. For an unanchored pattern, if a match attempt fails at the start of a
2271        newline sequence, and the newline setting is CRLF or ANY, and the next two
2272        characters are CRLF, advance by two characters instead of one.
2275    Version 6.7 04-Jul-06
2276    ---------------------
2278     1. In order to handle tests when input lines are enormously long, pcretest has
2279        been re-factored so that it automatically extends its buffers when
2280        necessary. The code is crude, but this _is_ just a test program. The
2281        default size has been increased from 32K to 50K.
2283     2. The code in pcre_study() was using the value of the re argument before
2284        testing it for NULL. (Of course, in any sensible call of the function, it
2285        won't be NULL.)
2287     3. The memmove() emulation function in pcre_internal.h, which is used on
2288        systems that lack both memmove() and bcopy() - that is, hardly ever -
2289        was missing a "static" storage class specifier.
2291     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2292        containing an extended class (one that cannot be represented by a bitmap
2293        because it contains high-valued characters or Unicode property items, e.g.
2294        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2295        pattern, but PCRE should not loop if you do not (it no longer does).
2296        [Detail: two cases were found: (a) a repeated subpattern containing an
2297        extended class; (b) a recursive reference to a subpattern that followed a
2298        previous extended class. It wasn't skipping over the extended class
2299        correctly when UTF-8 mode was not set.]
2301     5. A negated single-character class was not being recognized as fixed-length
2302        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2303        compile error "lookbehind assertion is not fixed length".
2305     6. The RunPerlTest auxiliary script was showing an unexpected difference
2306        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2307        write a Perl script that can interpret lines of an input file either as
2308        byte characters or as UTF-8, which is what "perltest" was being required to
2309        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2310        can't do is switch easily at run time between having the "use utf8;" pragma
2311        or not. In the end, I fudged it by using the RunPerlTest script to insert
2312        "use utf8;" explicitly for the UTF-8 tests.
2314     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2315        the end of the subject string, contrary to the documentation and to what
2316        Perl does. This was true of both matching functions. Now it matches only at
2317        the start of the subject and immediately after *internal* newlines.
2319     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2320        a pointer to an int instead of a pointer to an unsigned long int. This
2321        caused problems on 64-bit systems.
2323     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2324        instance of the 'standard' template library not being so standard".
2326    10. There was no check on the number of named subpatterns nor the maximum
2327        length of a subpattern name. The product of these values is used to compute
2328        the size of the memory block for a compiled pattern. By supplying a very
2329        long subpattern name and a large number of named subpatterns, the size
2330        computation could be caused to overflow. This is now prevented by limiting
2331        the length of names to 32 characters, and the number of named subpatterns
2332        to 10,000.
2334    11. Subpatterns that are repeated with specific counts have to be replicated in
2335        the compiled pattern. The size of memory for this was computed from the
2336        length of the subpattern and the repeat count. The latter is limited to
2337        65535, but there was no limit on the former, meaning that integer overflow
2338        could in principle occur. The compiled length of a repeated subpattern is
2339        now limited to 30,000 bytes in order to prevent this.
2341    12. Added the optional facility to have named substrings with the same name.
2343    13. Added the ability to use a named substring as a condition, using the
2344        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2345        are numbers (not recommended). Forward references are permitted.
2347    14. Added forward references in named backreferences (if you see what I mean).
2349    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2350        pattern could run off the end of the subject. For example, the pattern
2351        "(?s)(.{1,5})"8 did this with the subject "ab".
2353    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2354        PCRE_CASELESS was set when matching characters that were quantified with ?
2355        or *.
2357    17. A character class other than a single negated character that had a minimum
2358        but no maximum quantifier - for example [ab]{6,} - was not handled
2359        correctly by pce_dfa_exec(). It would match only one character.
2361    18. A valid (though odd) pattern that looked like a POSIX character
2362        class but used an invalid character after [ (for example [[,abc,]]) caused
2363        pcre_compile() to give the error "Failed: internal error: code overflow" or
2364        in some cases to crash with a glibc free() error. This could even happen if
2365        the pattern terminated after [[ but there just happened to be a sequence of
2366        letters, a binary zero, and a closing ] in the memory that followed.
2368    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2369        over the years. Originally (before any Unicode support), just the bottom 8
2370        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2371        output from "man perlunicode" includes this:
2373          The regular expression compiler produces polymorphic opcodes.  That
2374          is, the pattern adapts to the data and automatically switches to
2375          the Unicode character scheme when presented with Unicode data--or
2376          instead uses a traditional byte scheme when presented with byte
2377          data.
2379        Sadly, a wide octal escape does not cause a switch, and in a string with
2380        no other multibyte characters, these octal escapes are treated as before.
2381        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2382        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2383        Unicode string.
2385        I have not perpetrated such confusion in PCRE. Up till now, it took just
2386        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2387        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2388        translate to the appropriate multibyte character.
2390    29. Applied some refactoring to reduce the number of warnings from Microsoft
2391        and Borland compilers. This has included removing the fudge introduced
2392        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2393        a warning about an unused variable.
2395    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2396        characters since release 4.0, because Perl (from release 5.004) does not.
2397        [Or at least, is documented not to: some releases seem to be in conflict
2398        with the documentation.] However, when a pattern was studied with
2399        pcre_study() and all its branches started with \s, PCRE still included VT
2400        as a possible starting character. Of course, this did no harm; it just
2401        caused an unnecessary match attempt.
2403    22. Removed a now-redundant internal flag bit that recorded the fact that case
2404        dependency changed within the pattern. This was once needed for "required
2405        byte" processing, but is no longer used. This recovers a now-scarce options
2406        bit. Also moved the least significant internal flag bit to the most-
2407        significant bit of the word, which was not previously used (hangover from
2408        the days when it was an int rather than a uint) to free up another bit for
2409        the future.
2411    23. Added support for CRLF line endings as well as CR and LF. As well as the
2412        default being selectable at build time, it can now be changed at runtime
2413        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2414        specify that it is scanning data with non-default line endings.
2416    24. Changed the definition of CXXLINK to make it agree with the definition of
2417        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2419    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2420        recursions. This makes a big different to stack usage for some patterns.
2422    26. If a subpattern containing a named recursion or subroutine reference such
2423        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2424        the space required for the compiled pattern went wrong and gave too small a
2425        value. Depending on the environment, this could lead to "Failed: internal
2426        error: code overflow at offset 49" or "glibc detected double free or
2427        corruption" errors.
2429    27. Applied patches from Google (a) to support the new newline modes and (b) to
2430        advance over multibyte UTF-8 characters in GlobalReplace.
2432    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2433        difference for some implementation of PCRE in some Windows version.
2435    29. Added some extra testing facilities to pcretest:
2437        \q<number>   in a data line sets the "match limit" value
2438        \Q<number>   in a data line sets the "match recursion limt" value
2439        -S <number>  sets the stack size, where <number> is in megabytes
2441        The -S option isn't available for Windows.
2444    Version 6.6 06-Feb-06
2445    ---------------------
2447     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2448        in pcreposix.h. I have copied the definition from pcre.h.
2450     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2451        because pcre.h is no longer a built file.
2453     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2454        not normally included in the compiled code.
2457    Version 6.5 01-Feb-06
2458    ---------------------
2460     1. When using the partial match feature with pcre_dfa_exec(), it was not
2461        anchoring the second and subsequent partial matches at the new starting
2462        point. This could lead to incorrect results. For example, with the pattern
2463        /1234/, partially matching against "123" and then "a4" gave a match.
2465     2. Changes to pcregrep:
2467        (a) All non-match returns from pcre_exec() were being treated as failures
2468            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2469            error message is output. Some extra information is given for the
2470            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2471            probably the only errors that are likely to be caused by users (by
2472            specifying a regex that has nested indefinite repeats, for instance).
2473            If there are more than 20 of these errors, pcregrep is abandoned.
2475        (b) A binary zero was treated as data while matching, but terminated the
2476            output line if it was written out. This has been fixed: binary zeroes
2477            are now no different to any other data bytes.
2479        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2480            used to set a locale for matching. The --locale=xxxx long option has
2481            been added (no short equivalent) to specify a locale explicitly on the
2482            pcregrep command, overriding the environment variables.
2484        (d) When -B was used with -n, some line numbers in the output were one less
2485            than they should have been.
2487        (e) Added the -o (--only-matching) option.
2489        (f) If -A or -C was used with -c (count only), some lines of context were
2490            accidentally printed for the final match.
2492        (g) Added the -H (--with-filename) option.
2494        (h) The combination of options -rh failed to suppress file names for files
2495            that were found from directory arguments.
2497        (i) Added the -D (--devices) and -d (--directories) options.
2499        (j) Added the -F (--fixed-strings) option.
2501        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2503        (l) Added the --colo(u)r option.
2505        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2506            is not present by default.
2508     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2509        items such as (?R) or (?1), when the recursion could match a number of
2510        alternatives. If it matched one of the alternatives, but subsequently,
2511        outside the recursion, there was a failure, the code tried to back up into
2512        the recursion. However, because of the way PCRE is implemented, this is not
2513        possible, and the result was an incorrect result from the match.
2515        In order to prevent this happening, the specification of recursion has
2516        been changed so that all such subpatterns are automatically treated as
2517        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2519     4. I had overlooked the fact that, in some locales, there are characters for
2520        which isalpha() is true but neither isupper() nor islower() are true. In
2521        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2522        and ordfeminine) are like this. This affected the treatment of \w and \W
2523        when they appeared in character classes, but not when they appeared outside
2524        a character class. The bit map for "word" characters is now created
2525        separately from the results of isalnum() instead of just taking it from the
2526        upper, lower, and digit maps. (Plus the underscore character, of course.)
2528     5. The above bug also affected the handling of POSIX character classes such as
2529        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2530        permanent tables. Instead, the bit maps for such a class were previously
2531        created as the appropriate unions of the upper, lower, and digit bitmaps.
2532        Now they are created by subtraction from the [[:word:]] class, which has
2533        its own bitmap.
2535     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2536        It is created by subtracting the vertical space characters (\x09, \x0a,
2537        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2538        subtraction was done in the overall bitmap for a character class, meaning
2539        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2540        be recognized. This bug has been fixed.
2542     7. Patches from the folks at Google:
2544          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2545          real life, but is still worth protecting against".
2547          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2548          regular expressions".
2550          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2551          have it.
2553          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2554          "configure" and the latter not, in order to fix a problem somebody had
2555          with compiling the Arg class on HP-UX.
2557          (e) Improve the error-handling of the C++ wrapper a little bit.
2559          (f) New tests for checking recursion limiting.
2561     8. The pcre_memmove() function, which is used only if the environment does not
2562        have a standard memmove() function (and is therefore rarely compiled),
2563        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2564        returning a result (though PCRE never actually uses the result).
2566     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2567        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2568        returned instead of calling malloc() with an overflowing number that would
2569        most likely cause subsequent chaos.
2571    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2573    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2574        with this option is matched, the nmatch and pmatch options of regexec() are
2575        ignored.
2577    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2578        provided in case anyone wants to the the POSIX interface with UTF-8
2579        strings.
2581    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2582        C++ linking (needed for some HP-UX environments).
2584    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2585        (unused parameter) and in the pcre_printint() function (omitted "default"
2586        switch label when the default is to do nothing).
2588    15. Added some code to make it possible, when PCRE is compiled as a C++
2589        library, to replace subject pointers for pcre_exec() with a smart pointer
2590        class, thus making it possible to process discontinuous strings.
2592    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2593        much the same function. They were added by different people who were trying
2594        to make PCRE easy to compile on non-Unix systems. It has been suggested
2595        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2596        for compiling on Windows systems. I have therefore replaced it with
2597        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2598        defaults to "extern" for C or "extern C" for C++, which works fine on
2599        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2600        SCOPE with something explicit in config.h. In addition:
2602        (a) pcreposix.h still had just "extern" instead of either of these macros;
2603            I have replaced it with PCRE_DATA_SCOPE.
2605        (b) Functions such as _pcre_xclass(), which are internal to the library,
2606            but external in the C sense, all had PCRE_EXPORT in their definitions.
2607            This is apparently wrong for the Windows case, so I have removed it.
2608            (It makes no difference on Unix-like systems.)
2610    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2611        of recursive calls to match(). This is different to MATCH_LIMIT because
2612        that limits the total number of calls to match(), not all of which increase
2613        the depth of recursion. Limiting the recursion depth limits the amount of
2614        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2615        when PCRE is compiled, and changed at run time. A patch from Google adds
2616        this functionality to the C++ interface.
2618    18. Changes to the handling of Unicode character properties:
2620        (a) Updated the table to Unicode 4.1.0.
2622        (b) Recognize characters that are not in the table as "Cn" (undefined).
2624        (c) I revised the way the table is implemented to a much improved format
2625            which includes recognition of ranges. It now supports the ranges that
2626            are defined in UnicodeData.txt, and it also amalgamates other
2627            characters into ranges. This has reduced the number of entries in the
2628            table from around 16,000 to around 3,000, thus reducing its size
2629            considerably. I realized I did not need to use a tree structure after
2630            all - a binary chop search is just as efficient. Having reduced the
2631            number of entries, I extended their size from 6 bytes to 8 bytes to
2632            allow for more data.
2634        (d) Added support for Unicode script names via properties such as \p{Han}.
2636    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2637        matching that character.
2639    20. When matching a repeated Unicode property with a minimum greater than zero,
2640        (for example \pL{2,}), PCRE could look past the end of the subject if it
2641        reached it while seeking the minimum number of characters. This could
2642        happen only if some of the characters were more than one byte long, because
2643        there is a check for at least the minimum number of bytes.
2645    21. Refactored the implementation of \p and \P so as to be more general, to
2646        allow for more different types of property in future. This has changed the
2647        compiled form incompatibly. Anybody with saved compiled patterns that use
2648        \p or \P will have to recompile them.
2650    22. Added "Any" and "L&" to the supported property types.
2652    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2653        but give a compile time error if the value is greater than 0xff.
2655    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2656        accidentally not being installed or uninstalled.
2658    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2659        made were to insert the current release number. This seemed silly, because
2660        it made things harder for people building PCRE on systems that don't run
2661        "configure". I have turned pcre.h into a distributed file, no longer built
2662        by "configure", with the version identification directly included. There is
2663        no longer a pcre.h.in file.
2665        However, this change necessitated a change to the pcre-config script as
2666        well. It is built from pcre-config.in, and one of the substitutions was the
2667        release number. I have updated configure.ac so that ./configure now finds
2668        the release number by grepping pcre.h.
2670    26. Added the ability to run the tests under valgrind.
2673    Version 6.4 05-Sep-05
2674    ---------------------
2676     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2677        "--" to be printed when multiple files were scanned, even when none of the
2678        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2679        consider it to be a bug, and have restored the previous behaviour.
2681     2. A couple of code tidies to get rid of compiler warnings.
2683     3. The pcretest program used to cheat by referring to symbols in the library
2684        whose names begin with _pcre_. These are internal symbols that are not
2685        really supposed to be visible externally, and in some environments it is
2686        possible to suppress them. The cheating is now confined to including
2687        certain files from the library's source, which is a bit cleaner.
2689     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2690        file's purpose clearer.
2692     5. Reorganized pcre_ucp_findchar().
2695    Version 6.3 15-Aug-05
2696    ---------------------
2698     1. The file libpcre.pc.in did not have general read permission in the tarball.
2700     2. There were some problems when building without C++ support:
2702        (a) If C++ support was not built, "make install" and "make test" still
2703            tried to test it.
2705        (b) There were problems when the value of CXX was explicitly set. Some
2706            changes have been made to try to fix these, and ...
2708        (c) --disable-cpp can now be used to explicitly disable C++ support.
2710        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2711            backslash in a target when C++ was disabled. This confuses some
2712            versions of "make", apparently. Using an intermediate variable solves
2713            this. (Same for CPP_LOBJ.)
2715     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2716        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2717        necessary on certain architectures.
2719     4. Added a setting of -export-symbols-regex to the link command to remove
2720        those symbols that are exported in the C sense, but actually are local
2721        within the library, and not documented. Their names all begin with
2722        "_pcre_". This is not a perfect job, because (a) we have to except some
2723        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2724        available (and never for static libraries). I have made a note to try to
2725        find a way round (a) in the future.
2728    Version 6.2 01-Aug-05
2729    ---------------------
2731     1. There was no test for integer overflow of quantifier values. A construction
2732        such as {1111111111111111} would give undefined results. What is worse, if
2733        a minimum quantifier for a parenthesized subpattern overflowed and became
2734        negative, the calculation of the memory size went wrong. This could have
2735        led to memory overwriting.
2737     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2739     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2740        operating environments where this matters.
2742     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2743        PCRE options from within the C++ wrapper.
2745     5. Named capturing subpatterns were not being correctly counted when a pattern
2746        was compiled. This caused two problems: (a) If there were more than 100
2747        such subpatterns, the calculation of the memory needed for the whole
2748        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2749        back references of the form \12, where the number was greater than 9, were
2750        not recognized as back references, even though there were sufficient
2751        previous subpatterns.
2753     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2754        versions of gcc, e.g. 2.95.4.
2757    Version 6.1 21-Jun-05
2758    ---------------------
2760     1. There was one reference to the variable "posix" in pcretest.c that was not
2761        surrounded by "#if !defined NOPOSIX".
2763     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2764        the cross-check on the old pcre_info() function, for the benefit of the
2765        cut-down version of PCRE that is currently imported into Exim.
2767     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2768        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2769        patterns that start (?i) but I don't think that matters. The use of (?i) is
2770        just an example; this all applies to the other options as well.
2772     4. Since libtool seems to echo the compile commands it is issuing, the output
2773        from "make" can be reduced a bit by putting "@" in front of each libtool
2774        compile command.
2776     5. Patch from the folks at Google for configure.in to be a bit more thorough
2777        in checking for a suitable C++ installation before trying to compile the
2778        C++ stuff. This should fix a reported problem when a compiler was present,
2779        but no suitable headers.
2781     6. The man pages all had just "PCRE" as their title. I have changed them to
2782        be the relevant file name. I have also arranged that these names are
2783        retained in the file doc/pcre.txt, which is a concatenation in text format
2784        of all the man pages except the little individual ones for each function.
2786     7. The NON-UNIX-USE file had not been updated for the different set of source
2787        files that come with release 6. I also added a few comments about the C++
2788        wrapper.
2791    Version 6.0 07-Jun-05
2792    ---------------------
2794     1. Some minor internal re-organization to help with my DFA experiments.
2796     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2797        didn't matter for the library itself when fully configured, but did matter
2798        when compiling without UCP support, or within Exim, where the ucp files are
2799        not imported.
2801     3. Refactoring of the library code to split up the various functions into
2802        different source modules. The addition of the new DFA matching code (see
2803        below) to a single monolithic source would have made it really too
2804        unwieldy, quite apart from causing all the code to be include in a
2805        statically linked application, when only some functions are used. This is
2806        relevant even without the DFA addition now that patterns can be compiled in
2807        one application and matched in another.
2809        The downside of splitting up is that there have to be some external
2810        functions and data tables that are used internally in different modules of
2811        the library but which are not part of the API. These have all had their
2812        names changed to start with "_pcre_" so that they are unlikely to clash
2813        with other external names.
2815     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2816        a different (DFA) algorithm. Although it is slower than the original
2817        function, it does have some advantages for certain types of matching
2818        problem.
2820     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2821        including restarting after a partial match.
2823     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2824        defined when compiling for Windows was sent to me. I have put it into the
2825        code, though I have no means of testing or verifying it.
2827     7. Added the pcre_refcount() auxiliary function.
2829     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2830        match before or at the first newline in the subject string. In pcretest,
2831        the /f option on a pattern can be used to set this.
2833     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2834        would behave wrongly. This has been present in PCRE since release 4.0.
2836    10. A number of changes to the pcregrep command:
2838        (a) Refactored how -x works; insert ^(...)$ instead of setting
2839            PCRE_ANCHORED and checking the length, in preparation for adding
2840            something similar for -w.
2842        (b) Added the -w (match as a word) option.
2844        (c) Refactored the way lines are read and buffered so as to have more
2845            than one at a time available.
2847        (d) Implemented a pcregrep test script.
2849        (e) Added the -M (multiline match) option. This allows patterns to match
2850            over several lines of the subject. The buffering ensures that at least
2851            8K, or the rest of the document (whichever is the shorter) is available
2852            for matching (and similarly the previous 8K for lookbehind assertions).
2854        (f) Changed the --help output so that it now says
2856              -w, --word-regex(p)
2858            instead of two lines, one with "regex" and the other with "regexp"
2859            because that confused at least one person since the short forms are the
2860            same. (This required a bit of code, as the output is generated
2861            automatically from a table. It wasn't just a text change.)
2863        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2864            option but starts with a hyphen. Could be a pattern or a path name
2865            starting with a hyphen, for instance.
2867        (h) "-" can be given as a file name to represent stdin.
2869        (i) When file names are being printed, "(standard input)" is used for
2870            the standard input, for compatibility with GNU grep. Previously
2871            "<stdin>" was used.
2873        (j) The option --label=xxx can be used to supply a name to be used for
2874            stdin when file names are being printed. There is no short form.
2876        (k) Re-factored the options decoding logic because we are going to add
2877            two more options that take data. Such options can now be given in four
2878            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2880        (l) Added the -A, -B, and -C options for requesting that lines of context
2881            around matches be printed.
2883        (m) Added the -L option to print the names of files that do not contain
2884            any matching lines, that is, the complement of -l.
2886        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2887            continue to scan other files.
2889        (o) The -s option was incorrectly implemented. For compatibility with other
2890            greps, it now suppresses the error message for a non-existent or non-
2891            accessible file (but not the return code). There is a new option called
2892            -q that suppresses the output of matching lines, which was what -s was
2893            previously doing.
2895        (p) Added --include and --exclude options to specify files for inclusion
2896            and exclusion when recursing.
2898    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2899        Hopefully, it now does.
2901    12. Missing cast in pcre_study().
2903    13. Added an "uninstall" target to the makefile.
2905    14. Replaced "extern" in the function prototypes in Makefile.in with
2906        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2907        world, but is set differently for Windows.
2909    15. Added a second compiling function called pcre_compile2(). The only
2910        difference is that it has an extra argument, which is a pointer to an
2911        integer error code. When there is a compile-time failure, this is set
2912        non-zero, in addition to the error test pointer being set to point to an
2913        error message. The new argument may be NULL if no error number is required
2914        (but then you may as well call pcre_compile(), which is now just a
2915        wrapper). This facility is provided because some applications need a
2916        numeric error indication, but it has also enabled me to tidy up the way
2917        compile-time errors are handled in the POSIX wrapper.
2919    16. Added VPATH=.libs to the makefile; this should help when building with one
2920        prefix path and installing with another. (Or so I'm told by someone who
2921        knows more about this stuff than I do.)
2923    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2924        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2925        match everything, including newlines. This is not POSIX-compatible, but
2926        somebody wanted the feature. From pcretest it can be activated by using
2927        both the P and the s flags.
2929    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2931    19. libpcre.pc was being incorrectly installed as executable.
2933    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2934        it now also looks for '\r' so that it will work unmodified on Windows.
2936    21. Added Google's contributed C++ wrapper to the distribution.
2938    22. Added some untidy missing memory free() calls in pcretest, to keep
2939        Electric Fence happy when testing.
2943    Version 5.0 13-Sep-04
2944    ---------------------
2946     1. Internal change: literal characters are no longer packed up into items
2947        containing multiple characters in a single byte-string. Each character
2948        is now matched using a separate opcode. However, there may be more than one
2949        byte in the character in UTF-8 mode.
2951     2. The pcre_callout_block structure has two new fields: pattern_position and
2952        next_item_length. These contain the offset in the pattern to the next match
2953        item, and its length, respectively.
2955     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2956        insertion of callouts before each pattern item. Added the /C option to
2957        pcretest to make use of this.
2959     4. On the advice of a Windows user, the lines
2961          #if defined(_WIN32) || defined(WIN32)
2962          _setmode( _fileno( stdout ), 0x8000 );
2963          #endif  /* defined(_WIN32) || defined(WIN32) */
2965        have been added to the source of pcretest. This apparently does useful
2966        magic in relation to line terminators.
2968     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2969        for the benefit of those environments where the "b" makes a difference.
2971     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2972        to know about it. I have put a hack into configure.in that adds in code
2973        to set GCC=yes if CC=icc. This seems to end up at a point in the
2974        generated configure script that is early enough to affect the setting of
2975        compiler options, which is what is needed, but I have no means of testing
2976        whether it really works. (The user who reported this had patched the
2977        generated configure script, which of course I cannot do.)
2979        LATER: After change 22 below (new libtool files), the configure script
2980        seems to know about icc (and also ecc). Therefore, I have commented out
2981        this hack in configure.in.
2983     7. Added support for pkg-config (2 patches were sent in).
2985     8. Negated POSIX character classes that used a combination of internal tables
2986        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2987        [[:^ascii]]. Typically, they would match almost any characters. The other
2988        POSIX classes were not broken in this way.
2990     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2991        to find the match, as PCRE was deluded into thinking that the match had to
2992        start at the start point or following a newline. The same bug applied to
2993        patterns with negative forward assertions or any backward assertions
2994        preceding ".*" at the start, unless the pattern required a fixed first
2995        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2997    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2998        starting at the last subject character, bytes beyond the end of the subject
2999        string were read.
3001    11. Renamed the variable "class" as "classbits" to make life easier for C++
3002        users. (Previously there was a macro definition, but it apparently wasn't
3003        enough.)
3005    12. Added the new field "tables" to the extra data so that tables can be passed
3006        in at exec time, or the internal tables can be re-selected. This allows
3007        a compiled regex to be saved and re-used at a later time by a different
3008        program that might have everything at different addresses.
3010    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3011        -R library as well as a -L library.
3013    14. The debugging options of pcretest (-d on the command line or D on a
3014        pattern) showed incorrect output for anything following an extended class
3015        that contained multibyte characters and which was followed by a quantifier.
3017    15. Added optional support for general category Unicode character properties
3018        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3019        support. It adds about 90K to the size of the library. The meanings of the
3020        inbuilt class escapes such as \d and \s have NOT been changed.
3022    16. Updated pcredemo.c to include calls to free() to release the memory for the
3023        compiled pattern.
3025    17. The generated file chartables.c was being created in the source directory
3026        instead of in the building directory. This caused the build to fail if the
3027        source directory was different from the building directory, and was
3028        read-only.
3030    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3031        file. No doubt somebody will tell me if they don't make sense... Also added
3032        Dan Mooney's comments about building on OpenVMS.
3034    19. Added support for partial matching via the PCRE_PARTIAL option for
3035        pcre_exec() and the \P data escape in pcretest.
3037    20. Extended pcretest with 3 new pattern features:
3039        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3040              write the compiled pattern to the file whose name is "rest-of-line".
3041              This is a straight binary dump of the data, with the saved pointer to
3042              the character tables forced to be NULL. The study data, if any, is
3043              written too. After writing, pcretest reads a new pattern.
3045        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3046              compiled pattern from the given file. There must not be any
3047              occurrences of "<" in the file name (pretty unlikely); if there are,
3048              pcretest will instead treat the initial "<" as a pattern delimiter.
3049              After reading in the pattern, pcretest goes on to read data lines as
3050              usual.
3052        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3053              and 16-bit fields in a compiled pattern, to simulate a pattern that
3054              was compiled on a host of opposite endianness.
3056    21. The pcre-exec() function can now cope with patterns that were compiled on
3057        hosts of opposite endianness, with this restriction:
3059          As for any compiled expression that is saved and used later, the tables
3060          pointer field cannot be preserved; the extra_data field in the arguments
3061          to pcre_exec() should be used to pass in a tables address if a value
3062          other than the default internal tables were used at compile time.
3064    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3065        now diagnosed as an error. Previously, most of the time, a negative number
3066        would have been treated as zero, but if in addition "ovector" was passed as
3067        NULL, a crash could occur.
3069    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3070        new versions from the libtool 1.5 distribution (the last one is a copy of
3071        a file called libtool.m4). This seems to have fixed the need to patch
3072        "configure" to support Darwin 1.3 (which I used to do). However, I still
3073        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3074        workstation).
3076    24. Changed the PCRE licence to be the more standard "BSD" licence.
3079    Version 4.5 01-Dec-03
3080    ---------------------
3082     1. There has been some re-arrangement of the code for the match() function so
3083        that it can be compiled in a version that does not call itself recursively.
3084        Instead, it keeps those local variables that need separate instances for
3085        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3086        needs to "recurse". Keeping track of where control must go is done by means
3087        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3088        hide most of the details from the main code, and operates only if
3089        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3090        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3091        operating.
3093        To make it easier for callers to provide specially tailored get/free
3094        functions for this usage, two new functions, pcre_stack_malloc, and
3095        pcre_stack_free, are used. They are always called in strict stacking order,
3096        and the size of block requested is always the same.
3098        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3099        PCRE has been compiled to use the stack or the heap for recursion. The
3100        -C option of pcretest uses this to show which version is compiled.
3102        A new data escape \S, is added to pcretest; it causes the amounts of store
3103        obtained and freed by both kinds of malloc/free at match time to be added
3104        to the output.
3106     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3107        what's available on my current Linux desktop machine.
3109     3. When matching a UTF-8 string, the test for a valid string at the start has
3110        been extended. If start_offset is not zero, PCRE now checks that it points
3111        to a byte that is the start of a UTF-8 character. If not, it returns
3112        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3113        this is necessary because there may be backward assertions in the pattern.
3114        When matching the same subject several times, it may save resources to use
3115        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3117     4. The code for checking the validity of UTF-8 strings has been tightened so
3118        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3119        containing "overlong sequences".
3121     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3122        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3123        should have been "&", but it just so happened that all the cases this let
3124        through by mistake were picked up later in the function.
3126     6. I had used a variable called "isblank" - this is a C99 function, causing
3127        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3129     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3130        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3131        so the version is shown; (c) stop "make test" echoing "./RunTest".
3133     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3135     9. The prototype for memmove() for systems that don't have it was using
3136        size_t, but the inclusion of the header that defines size_t was later. I've
3137        moved the #includes for the C headers earlier to avoid this.
3139    10. Added some adjustments to the code to make it easier to compiler on certain
3140        special systems:
3142          (a) Some "const" qualifiers were missing.
3143          (b) Added the macro EXPORT before all exported functions; by default this
3144              is defined to be empty.
3145          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3146              that it reads its output file name as an argument instead of writing
3147              to the standard output and assuming this can be redirected.
3149    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3150        class containing characters with values greater than 255, PCRE compilation
3151        went into a loop.
3153    12. A recursive reference to a subpattern that was within another subpattern
3154        that had a minimum quantifier of zero caused PCRE to crash. For example,
3155        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3156        recursion. If the recursively-called subpattern itself had a zero repeat,
3157        that was OK.
3159    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3160        buffer into which it was copied (for escape processing) was still set at
3161        1024, so long lines caused crashes.
3163    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3164        "internal error: code overflow...". This applied to any character class
3165        that was followed by a possessive quantifier.
3167    15. Modified the Makefile to add libpcre.la as a prerequisite for
3168        libpcreposix.la because I was told this is needed for a parallel build to
3169        work.
3171    16. If a pattern that contained .* following optional items at the start was
3172        studied, the wrong optimizing data was generated, leading to matching
3173        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3174        matching string must start with a or b or c. The correct conclusion for
3175        this pattern is that a match can start with any character.
3178    Version 4.4 13-Aug-03
3179    ---------------------
3181     1. In UTF-8 mode, a character class containing characters with values between
3182        127 and 255 was not handled correctly if the compiled pattern was studied.
3183        In fixing this, I have also improved the studying algorithm for such
3184        classes (slightly).
3186     2. Three internal functions had redundant arguments passed to them. Removal
3187        might give a very teeny performance improvement.
3189     3. Documentation bug: the value of the capture_top field in a callout is *one
3190        more than* the number of the hightest numbered captured substring.
3192     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3193        in incorrectly linking with a previously installed version. They now link
3194        explicitly with libpcre.la.
3196     5. configure.in no longer needs to recognize Cygwin specially.
3198     6. A problem in pcre.in for Windows platforms is fixed.
3200     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3201        pcretest, it used to include the size of the study block as part of its
3202        output. Unfortunately, the structure contains a field that has a different
3203        size on different hardware architectures. This meant that the tests that
3204        showed this size failed. As the block is currently always of a fixed size,
3205        this information isn't actually particularly useful in pcretest output, so
3206        I have just removed it.
3208     8. Three pre-processor statements accidentally did not start in column 1.
3209        Sadly, there are *still* compilers around that complain, even though
3210        standard C has not required this for well over a decade. Sigh.
3212     9. In pcretest, the code for checking callouts passed small integers in the
3213        callout_data field, which is a void * field. However, some picky compilers
3214        complained about the casts involved for this on 64-bit systems. Now
3215        pcretest passes the address of the small integer instead, which should get
3216        rid of the warnings.
3218    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3219        both compile and run time, and gives an error if an invalid UTF-8 sequence
3220        is found. There is a option for disabling this check in cases where the
3221        string is known to be correct and/or the maximum performance is wanted.
3223    11. In response to a bug report, I changed one line in Makefile.in from
3225            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3226        to
3227            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3229        to look similar to other lines, but I have no way of telling whether this
3230        is the right thing to do, as I do not use Windows. No doubt I'll get told
3231        if it's wrong...
3234    Version 4.3 21-May-03
3235    ---------------------
3237    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3238       Makefile.
3240    2. Some refactoring to improve the quality of the code:
3242       (i)   The utf8_table... variables are now declared "const".
3244       (ii)  The code for \cx, which used the "case flipping" table to upper case
3245             lower case letters, now just substracts 32. This is ASCII-specific,
3246             but the whole concept of \cx is ASCII-specific, so it seems
3247             reasonable.
3249       (iii) PCRE was using its character types table to recognize decimal and
3250             hexadecimal digits in the pattern. This is silly, because it handles
3251             only 0-9, a-f, and A-F, but the character types table is locale-
3252             specific, which means strange things might happen. A private
3253             table is now used for this - though it costs 256 bytes, a table is
3254             much faster than multiple explicit tests. Of course, the standard
3255             character types table is still used for matching digits in subject
3256             strings against \d.
3258       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3259             ending in _t are). So I've renamed it as ESC_tee.
3261    3. The first argument for regexec() in the POSIX wrapper should have been
3262       defined as "const".
3264    4. Changed pcretest to use malloc() for its buffers so that they can be
3265       Electric Fenced for debugging.
3267    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3268       to read one or more bytes before the start of the subject string. Often this
3269       had no effect on PCRE's behaviour, but in some circumstances it could
3270       provoke a segmentation fault.
3272    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3273       to try to read one or more bytes before the start of the subject string.
3275    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3276       UTF-8 support could misbehave in various ways if the subject string
3277       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3278       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3279       back over UTF-8 characters.)
3282    Version 4.2 14-Apr-03
3283    ---------------------
3285    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3287    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3288         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3289         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3290         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3291         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3292           and BUILD_EXEEXT
3293         Note: automatic setting of the BUILD variables is not yet working
3294         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3295           compile-time but not at link-time
3296         [LINK]: use for linking executables only
3297         make different versions for Windows and non-Windows
3298         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3299           libraries
3300         [LINK_FOR_BUILD]: new variable
3301         [OBJEXT]: use throughout
3302         [EXEEXT]: use throughout
3303         <winshared>: new target
3304         <wininstall>: new target
3305         <dftables.o>: use native compiler
3306         <dftables>: use native linker
3307         <install>: handle Windows platform correctly
3308         <clean>: ditto
3309         <check>: ditto
3310         copy DLL to top builddir before testing
3312       As part of these changes, -no-undefined was removed again. This was reported
3313       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3314       in any case.
3316    3. Some tidies to get rid of compiler warnings:
3318       . In the match_data structure, match_limit was an unsigned long int, whereas
3319         match_call_count was an int. I've made them both unsigned long ints.
3321       . In pcretest the fact that a const uschar * doesn't automatically cast to
3322         a void * provoked a warning.
3324       . Turning on some more compiler warnings threw up some "shadow" variables
3325         and a few more missing casts.
3327    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3328       option, a class that contained a single character with a value between 128
3329       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3331    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3332       option, a class that contained several characters, but with at least one
3333       whose value was between 128 and 255 caused PCRE to crash.
3336    Version 4.1 12-Mar-03
3337    ---------------------
3339    1. Compiling with gcc -pedantic found a couple of places where casts were
3340    needed, and a string in dftables.c that was longer than standard compilers are
3341    required to support.
3343    2. Compiling with Sun's compiler found a few more places where the code could
3344    be tidied up in order to avoid warnings.
3346    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3347    first of these names is deprecated in the latest Autoconf in favour of the name
3348    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3349    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3350    analogy I have changed it to CFLAGS_FOR_BUILD.
3352    4. Added -no-undefined to the linking command in the Makefile, because this is
3353    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3354    linking step for the pcreposix library.
3356    5. PCRE was failing to diagnose the case of two named groups with the same
3357    name.
3359    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3360    literal character that is needed in the subject for a match, and scans along to
3361    ensure that it is present before embarking on the full matching process. This
3362    saves time in cases of nested unlimited repeats that are never going to match.
3363    Problem: the scan can take a lot of time if the subject is very long (e.g.
3364    megabytes), thus penalizing straightforward matches. It is now done only if the
3365    amount of subject to be scanned is less than 1000 bytes.
3367    7. A lesser problem with the same optimization is that it was recording the
3368    first character of an anchored pattern as "needed", thus provoking a search
3369    right along the subject, even when the first match of the pattern was going to
3370    fail. The "needed" character is now not set for anchored patterns, unless it
3371    follows something in the pattern that is of non-fixed length. Thus, it still
3372    fulfils its original purpose of finding quick non-matches in cases of nested
3373    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3376    Version 4.0 17-Feb-03
3377    ---------------------
3379    1. If a comment in an extended regex that started immediately after a meta-item
3380    extended to the end of string, PCRE compiled incorrect data. This could lead to
3381    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3383    2. Moved to autoconf 2.53 and libtool 1.4.2.
3385    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3386    the special perltest8 script is no longer needed - all the tests can be run
3387    from a single perltest script.
3389    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3390    by \s. It has now been removed in PCRE. This means it isn't recognized as
3391    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3392    class [:space:] *does* include VT, thereby creating a mess.
3394    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3395    space and tab.
3397    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3398    its new features into the main test script, reducing the number of scripts.
3400    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3401    were backward compatible, and made the (?i) apply to the whole pattern, as if
3402    /i were given. Now it behaves more logically, and applies the option setting
3403    only to what follows. PCRE has been changed to follow suit. However, if it
3404    finds options settings right at the start of the pattern, it extracts them into
3405    the global options, as before. Thus, they show up in the info data.
3407    8. Added support for the \Q...\E escape sequence. Characters in between are
3408    treated as literals. This is slightly different from Perl in that $ and @ are
3409    also handled as literals inside the quotes. In Perl, they will cause variable
3410    interpolation. Note the following examples:
3412        Pattern            PCRE matches      Perl matches
3414        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3415        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3416        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3418    For compatibility with Perl, \Q...\E sequences are recognized inside character
3419    classes as well as outside them.
3421    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3422    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3423    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3424    signed/unsigned warnings.
3426    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3427    option for pcretest, so I've replaced it by a simple function that does just
3428    that job.
3430    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3431    "pcregrep -".
3433    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3434    Java package. This provides some syntactic sugar for simple cases of what my
3435    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3436    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3437    item, you can use this simplified notation. Note that only makes sense with
3438    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3439    greediness, whatever the setting of the PCRE_UNGREEDY option.
3441    13. A change of greediness default within a pattern was not taking effect at
3442    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3443    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3444    was abstracted outside.
3446    14. PCRE now supports the \G assertion. It is true when the current matching
3447    position is at the start point of the match. This differs from \A when the
3448    starting offset is non-zero. Used with the /g option of pcretest (or similar
3449    code), it works in the same way as it does for Perl's /g option. If all
3450    alternatives of a regex begin with \G, the expression is anchored to the start
3451    match position, and the "anchored" flag is set in the compiled expression.
3453    15. Some bugs concerning the handling of certain option changes within patterns
3454    have been fixed. These applied to options other than (?ims). For example,
3455    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3456    been the other way round. Some of this was related to change 7 above.
3458    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3459    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3460    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3461    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3463    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3464    mode. Unlike ".", it always matches newline, whatever the setting of
3465    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3466    assertions. Perl allows it, but it doesn't (in general) work because it can't
3467    calculate the length of the lookbehind. At least, that's the case for Perl
3468    5.8.0 - I've been told they are going to document that it doesn't work in
3469    future.
3471    18. Added an error diagnosis for escapes that PCRE does not support: these are
3472    \L, \l, \N, \P, \p, \U, \u, and \X.
3474    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3475    reading past the end of the pattern in cases such as /[abcd/.
3477    20. PCRE was getting more memory than necessary for patterns with classes that
3478    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3480    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3481    compiling PCRE for use with Virtual Pascal.
3483    22. Small fix to the Makefile to make it work properly if the build is done
3484    outside the source tree.
3486    23. Added a new extension: a condition to go with recursion. If a conditional
3487    subpattern starts with (?(R) the "true" branch is used if recursion has
3488    happened, whereas the "false" branch is used only at the top level.
3490    24. When there was a very long string of literal characters (over 255 bytes
3491    without UTF support, over 250 bytes with UTF support), the computation of how
3492    much memory was required could be incorrect, leading to segfaults or other
3493    strange effects.
3495    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3496    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3497    there was a subsequent back reference to those brackets. This meant that, for
3498    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3499    possible to check for precisely this case. All we can do is abandon the
3500    optimization if .* occurs inside capturing brackets when there are any back
3501    references whatsoever. (See below for a better fix that came later.)
3503    26. The handling of the optimization for finding the first character of a
3504    non-anchored pattern, and for finding a character that is required later in the
3505    match were failing in some cases. This didn't break the matching; it just
3506    failed to optimize when it could. The way this is done has been re-implemented.
3508    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3510    28. Added a new feature that provides some of the functionality that Perl
3511    provides with (?{...}). The facility is termed a "callout". The way it is done
3512    in PCRE is for the caller to provide an optional function, by setting
3513    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3514    global variable. By default it is unset, which disables all calling out. To get
3515    the function called, the regex must include (?C) at appropriate points. This
3516    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3517    This provides a means of identifying different callout points. When PCRE
3518    reaches such a point in the regex, if pcre_callout has been set, the external
3519    function is called. It is provided with data in a structure called
3520    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3521    matching continues; if it returns a non-zero value, the match at the current
3522    point fails. However, backtracking will occur if possible. [This was changed
3523    later and other features added - see item 49 below.]
3525    29. pcretest is upgraded to test the callout functionality. It provides a
3526    callout function that displays information. By default, it shows the start of
3527    the match and the current position in the text. There are some new data escapes
3528    to vary what happens:
3530        \C+         in addition, show current contents of captured substrings
3531        \C-         do not supply a callout function
3532        \C!n        return 1 when callout number n is reached
3533        \C!n!m      return 1 when callout number n is reached for the mth time
3535    30. If pcregrep was called with the -l option and just a single file name, it
3536    output "<stdin>" if a match was found, instead of the file name.
3538    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3539    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3540    pcre_exec(). This saves a malloc/free per call. The default value of
3541    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3542    when configuring.
3544    32. The default maximum size of a compiled pattern is 64K. There have been a
3545    few cases of people hitting this limit. The code now uses macros to handle the
3546    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3547    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3548    configuring. Tests 2 and 5 work only with 2-byte links because they output
3549    debugging information about compiled patterns.
3551    33. Internal code re-arrangements:
3553    (a) Moved the debugging function for printing out a compiled regex into
3554        its own source file (printint.c) and used #include to pull it into
3555        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3556        separate copies.
3558    (b) Defined the list of op-code names for debugging as a macro in
3559        internal.h so that it is next to the definition of the opcodes.
3561    (c) Defined a table of op-code lengths for simpler skipping along compiled
3562        code. This is again a macro in internal.h so that it is next to the
3563        definition of the opcodes.
3565    34. Added support for recursive calls to individual subpatterns, along the
3566    lines of Robin Houston's patch (but implemented somewhat differently).
3568    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3569    allow it to read and process whole directories in Win32. This code was
3570    contributed by Lionel Fourquaux; it has not been tested by me.
3572    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3573    used to name a group. Names consist of alphanumerics and underscores, and must
3574    be unique. Back references use the syntax (?P=name) and recursive calls use
3575    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3576    numbers. The function pcre_fullinfo() can be used after compilation to extract
3577    a name/number map. There are three relevant calls:
3579      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3580      PCRE_INFO_NAMECOUNT            yields the number of entries
3581      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3583    The map is a vector of fixed-size entries. The size of each entry depends on
3584    the length of the longest name used. The first two bytes of each entry are the
3585    group number, most significant byte first. There follows the corresponding
3586    name, zero terminated. The names are in alphabetical order.
3588    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3589    case instead of 255. Making it the same both with and without UTF-8 support
3590    means that the same test output works with both.
3592    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3593    calling malloc() with a zero argument.
3595    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3596    optimization. I've improved things by keeping a bitmap of backreferences with
3597    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3598    fact referenced, the optimization can be applied. It is unlikely that a
3599    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3600    the match to follow \n) will appear inside brackets with a number greater than
3601    31, but if it does, any back reference > 31 suppresses the optimization.
3603    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3604    of disabling numbered capturing parentheses. Any opening parenthesis that is
3605    not followed by ? behaves as if it were followed by ?: but named parentheses
3606    can still be used for capturing (and they will acquire numbers in the usual
3607    way).
3609    41. Redesigned the return codes from the match() function into yes/no/error so