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

Diff of /code/trunk/ChangeLog

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

revision 143 by ph10, Mon Apr 2 10:08:14 2007 UTC revision 760 by ph10, Tue Nov 22 11:23:43 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 12-Mar-07  Version 8.21
5    ------------
6    
7    1.  Updating the JIT compiler.
8    
9    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
10        are added as well.
11    
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.
15    
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.
19    
20    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
21        ECMA-262 standard.
22    
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.
26    
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.
36    
37    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
38        being incorrectly compiled and would have given unpredicatble results.
39    
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.
44    
45    10. Add a cast and remove a redundant test from the code.
46    
47    11. JIT should use pcre_malloc/pcre_free for allocation.
48    
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).
52    
53    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
54    
55    14. Perl does not support \N without a following name in a [] class; PCRE now
56        also gives an error.
57    
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.
61    
62    
63    Version 8.20 21-Oct-2011
64    ------------------------
65    
66    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
67        a POSIX class. After further experiments with Perl, which convinced me that
68        Perl has bugs and confusions, a closing square bracket is no longer allowed
69        in a POSIX name. This bug also affected patterns with classes that started
70        with full stops.
71    
72    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
73        captured substring, but while checking the failing first alternative,
74        substring 1 is temporarily captured. If the output vector supplied to
75        pcre_exec() was not big enough for this capture, the yield of the function
76        was still zero ("insufficient space for captured substrings"). This cannot
77        be totally fixed without adding another stack variable, which seems a lot
78        of expense for a edge case. However, I have improved the situation in cases
79        such as /(a)(b)x|abc/ matched against "abc", where the return code
80        indicates that fewer than the maximum number of slots in the ovector have
81        been set.
82    
83    3.  Related to (2) above: when there are more back references in a pattern than
84        slots in the output vector, pcre_exec() uses temporary memory during
85        matching, and copies in the captures as far as possible afterwards. It was
86        using the entire output vector, but this conflicts with the specification
87        that only 2/3 is used for passing back captured substrings. Now it uses
88        only the first 2/3, for compatibility. This is, of course, another edge
89        case.
90    
91    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
92        main code base, and can be used by building with --enable-jit. When this is
93        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
94        runtime --no-jit option is given.
95    
96    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
97        ovector, the return from the function was zero, implying that there were
98        other matches that did not fit. The correct "exactly full" value is now
99        returned.
100    
101    6.  If a subpattern that was called recursively or as a subroutine contained
102        (*PRUNE) or any other control that caused it to give a non-standard return,
103        invalid errors such as "Error -26 (nested recursion at the same subject
104        position)" or even infinite loops could occur.
105    
106    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
107        computing the minimum length on reaching *ACCEPT, and so ended up with the
108        wrong value of 1 rather than 0. Further investigation indicates that
109        computing a minimum subject length in the presence of *ACCEPT is difficult
110        (think back references, subroutine calls), and so I have changed the code
111        so that no minimum is registered for a pattern that contains *ACCEPT.
112    
113    8.  If (*THEN) was present in the first (true) branch of a conditional group,
114        it was not handled as intended. [But see 16 below.]
115    
116    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
117        Sheri Pierce.
118    
119    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
120        the first byte in a match must be "a".
121    
122    11. Change 17 for 8.13 increased the recursion depth for patterns like
123        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
124        pattern contains any instances of (*THEN). If it does not, the old
125        optimizations are restored. It would be nice to do this on a per-group
126        basis, but at the moment that is not feasible.
127    
128    12. In some environments, the output of pcretest -C is CRLF terminated. This
129        broke RunTest's code that checks for the link size. A single white space
130        character after the value is now allowed for.
131    
132    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
133        For "fr", it uses the Windows-specific input and output files.
134    
135    14. If (*THEN) appeared in a group that was called recursively or as a
136        subroutine, it did not work as intended. [But see next item.]
137    
138    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
139        pattern fragments (but not containing any | characters). If A and B are
140        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
141        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
142        D. In other words, Perl considers parentheses that do not contain any |
143        characters to be part of a surrounding alternative, whereas PCRE was
144        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
145        differently. PCRE now behaves in the same way as Perl, except in the case
146        of subroutine/recursion calls such as (?1) which have in any case always
147        been different (but PCRE had them first :-).
148    
149    16. Related to 15 above: Perl does not treat the | in a conditional group as
150        creating alternatives. Such a group is treated in the same way as an
151        ordinary group without any | characters when processing (*THEN). PCRE has
152        been changed to match Perl's behaviour.
153    
154    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
155        RunGrepTest script failed.
156    
157    18. Change 22 for version 13 caused atomic groups to use more stack. This is
158        inevitable for groups that contain captures, but it can lead to a lot of
159        stack use in large patterns. The old behaviour has been restored for atomic
160        groups that do not contain any capturing parentheses.
161    
162    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
163        suppress the check for a minimum subject length at run time. (If it was
164        given to pcre_exec() or pcre_dfa_exec() it did work.)
165    
166    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
167        fail to work when decoding hex characters in data strings in EBCDIC
168        environments.
169    
170    21. It appears that in at least one Mac OS environment, the isxdigit() function
171        is implemented as a macro that evaluates to its argument more than once,
172        contravening the C 90 Standard (I haven't checked a later standard). There
173        was an instance in pcretest which caused it to go wrong when processing
174        \x{...} escapes in subject strings. The has been rewritten to avoid using
175        things like p++ in the argument of isxdigit().
176    
177    
178    Version 8.13 16-Aug-2011
179    ------------------------
180    
181    1.  The Unicode data tables have been updated to Unicode 6.0.0.
182    
183    2.  Two minor typos in pcre_internal.h have been fixed.
184    
185    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
186        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
187        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
188    
189    4.  There were a number of related bugs in the code for matching backrefences
190        caselessly in UTF-8 mode when codes for the characters concerned were
191        different numbers of bytes. For example, U+023A and U+2C65 are an upper
192        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
193        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
194        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
195        2-byte code at the end of the subject (it thought there wasn't enough data
196        left).
197    
198    5.  Comprehensive information about what went wrong is now returned by
199        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
200        as the output vector has at least 2 elements. The offset of the start of
201        the failing character and a reason code are placed in the vector.
202    
203    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
204        now returned is for the first byte of the failing character, instead of the
205        last byte inspected. This is an incompatible change, but I hope it is small
206        enough not to be a problem. It makes the returned offset consistent with
207        pcre_exec() and pcre_dfa_exec().
208    
209    7.  pcretest now gives a text phrase as well as the error number when
210        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
211        failure, the offset and reason code are output.
212    
213    8.  When \R was used with a maximizing quantifier it failed to skip backwards
214        over a \r\n pair if the subsequent match failed. Instead, it just skipped
215        back over a single character (\n). This seems wrong (because it treated the
216        two characters as a single entity when going forwards), conflicts with the
217        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
218        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
219        has been changed.
220    
221    9.  Some internal refactoring has changed the processing so that the handling
222        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
223        time (the PCRE_DOTALL option was changed this way some time ago: version
224        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
225        which was always a bit of a fudge. It also means that there is one less
226        argument for the match() function, which reduces its stack requirements
227        slightly. This change also fixes an incompatibility with Perl: the pattern
228        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
229    
230    10. More internal refactoring has drastically reduced the number of recursive
231        calls to match() for possessively repeated groups such as (abc)++ when
232        using pcre_exec().
233    
234    11. While implementing 10, a number of bugs in the handling of groups were
235        discovered and fixed:
236    
237        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
238        (a|)*(?1) gave a compile-time internal error.
239        ((a|)+)+  did not notice that the outer group could match an empty string.
240        (^a|^)+   was not marked as anchored.
241        (.*a|.*)+ was not marked as matching at start or after a newline.
242    
243    12. Yet more internal refactoring has removed another argument from the match()
244        function. Special calls to this function are now indicated by setting a
245        value in a variable in the "match data" data block.
246    
247    13. Be more explicit in pcre_study() instead of relying on "default" for
248        opcodes that mean there is no starting character; this means that when new
249        ones are added and accidentally left out of pcre_study(), testing should
250        pick them up.
251    
252    14. The -s option of pcretest has been documented for ages as being an old
253        synonym of -m (show memory usage). I have changed it to mean "force study
254        for every regex", that is, assume /S for every regex. This is similar to -i
255        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
256        using it. It makes it easier to run collections of tests with and without
257        study enabled, and thereby test pcre_study() more easily. All the standard
258        tests are now run with and without -s (but some patterns can be marked as
259        "never study" - see 20 below).
260    
261    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
262        restoration of the capturing data to the outer values was not happening
263        correctly.
264    
265    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
266        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
267        pattern had matched an empty string, and so incorrectly returned a no
268        match.
269    
270    17. There was optimizing code for the last branch of non-capturing parentheses,
271        and also for the obeyed branch of a conditional subexpression, which used
272        tail recursion to cut down on stack usage. Unfortunately, now that there is
273        the possibility of (*THEN) occurring in these branches, tail recursion is
274        no longer possible because the return has to be checked for (*THEN). These
275        two optimizations have therefore been removed. [But see 8.20/11 above.]
276    
277    18. If a pattern containing \R was studied, it was assumed that \R always
278        matched two bytes, thus causing the minimum subject length to be
279        incorrectly computed because \R can also match just one byte.
280    
281    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
282        was incorrectly computed.
283    
284    20. If /S is present twice on a test pattern in pcretest input, it now
285        *disables* studying, thereby overriding the use of -s on the command line
286        (see 14 above). This is necessary for one or two tests to keep the output
287        identical in both cases.
288    
289    21. When (*ACCEPT) was used in an assertion that matched an empty string and
290        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
291    
292    22. When an atomic group that contained a capturing parenthesis was
293        successfully matched, but the branch in which it appeared failed, the
294        capturing was not being forgotten if a higher numbered group was later
295        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
296        group 1 to "a", when in fact it should be unset. This applied to multi-
297        branched capturing and non-capturing groups, repeated or not, and also to
298        positive assertions (capturing in negative assertions does not happen
299        in PCRE) and also to nested atomic groups.
300    
301    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
302        subject after a captured substring, to make it easier to tell which of a
303        number of identical substrings has been captured.
304    
305    24. The way atomic groups are processed by pcre_exec() has been changed so that
306        if they are repeated, backtracking one repetition now resets captured
307        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
308        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
309        "aaa". Previously, it would have been "a". As part of this code
310        refactoring, the way recursive calls are handled has also been changed.
311    
312    25. If an assertion condition captured any substrings, they were not passed
313        back unless some other capturing happened later. For example, if
314        (?(?=(a))a) was matched against "a", no capturing was returned.
315    
316    26. When studying a pattern that contained subroutine calls or assertions,
317        the code for finding the minimum length of a possible match was handling
318        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
319        group 1 called group 2 while simultaneously a separate group 2 called group
320        1). A stack overflow occurred in this case. I have fixed this by limiting
321        the recursion depth to 10.
322    
323    27. Updated RunTest.bat in the distribution to the version supplied by Tom
324        Fortmann. This supports explicit test numbers on the command line, and has
325        argument validation and error reporting.
326    
327    28. An instance of \X with an unlimited repeat could fail if at any point the
328        first character it looked at was a mark character.
329    
330    29. Some minor code refactoring concerning Unicode properties and scripts
331        should reduce the stack requirement of match() slightly.
332    
333    30. Added the '=' option to pcretest to check the setting of unused capturing
334        slots at the end of the pattern, which are documented as being -1, but are
335        not included in the return count.
336    
337    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
338        compiled something random. Now it gives a compile-time error (as does
339        Perl).
340    
341    32. A *MARK encountered during the processing of a positive assertion is now
342        recorded and passed back (compatible with Perl).
343    
344    33. If --only-matching or --colour was set on a pcregrep call whose pattern
345        had alternative anchored branches, the search for a second match in a line
346        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
347        matched the line "0102" twice. The same bug affected patterns that started
348        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
349        twice.
350    
351    34. Previously, PCRE did not allow quantification of assertions. However, Perl
352        does, and because of capturing effects, quantifying parenthesized
353        assertions may at times be useful. Quantifiers are now allowed for
354        parenthesized assertions.
355    
356    35. A minor code tidy in pcre_compile() when checking options for \R usage.
357    
358    36. \g was being checked for fancy things in a character class, when it should
359        just be a literal "g".
360    
361    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
362        appearance of a nested POSIX class supersedes an apparent external class.
363        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
364        unescaped square brackets may also appear as part of class names. For
365        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
366        more like Perl. (But see 8.20/1 above.)
367    
368    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
369        was because it thought it was \N{name}, which is not supported).
370    
371    39. Add minix to OS list not supporting the -S option in pcretest.
372    
373    40. PCRE tries to detect cases of infinite recursion at compile time, but it
374        cannot analyze patterns in sufficient detail to catch mutual recursions
375        such as ((?1))((?2)). There is now a runtime test that gives an error if a
376        subgroup is called recursively as a subpattern for a second time at the
377        same position in the subject string. In previous releases this might have
378        been caught by the recursion limit, or it might have run out of stack.
379    
380    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
381        happen only once. PCRE was, however incorrectly giving a compile time error
382        "recursive call could loop indefinitely" because it cannot analyze the
383        pattern in sufficient detail. The compile time test no longer happens when
384        PCRE is compiling a conditional subpattern, but actual runaway loops are
385        now caught at runtime (see 40 above).
386    
387    42. It seems that Perl allows any characters other than a closing parenthesis
388        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
389        has been changed to be the same.
390    
391    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
392        as not to get warnings when autogen.sh is called. Also changed
393        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
394    
395    44. To help people who use pcregrep to scan files containing exceedingly long
396        lines, the following changes have been made:
397    
398        (a) The default value of the buffer size parameter has been increased from
399            8K to 20K. (The actual buffer used is three times this size.)
400    
401        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
402            PCRE is built.
403    
404        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
405            to be set at run time.
406    
407        (d) Numerical values in pcregrep options can be followed by K or M, for
408            example --buffer-size=50K.
409    
410        (e) If a line being scanned overflows pcregrep's buffer, an error is now
411            given and the return code is set to 2.
412    
413    45. Add a pointer to the latest mark to the callout data block.
414    
415    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
416        partial match of an empty string instead of no match. This was specific to
417        the use of ".".
418    
419    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
420        complete match instead of a partial match. This bug was dependent on both
421        the PCRE_UTF8 and PCRE_DOTALL options being set.
422    
423    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
424        starting byte set, because \b was not being ignored.
425    
426    
427    Version 8.12 15-Jan-2011
428    ------------------------
429    
430    1.  Fixed some typos in the markup of the man pages, and wrote a script that
431        checks for such things as part of the documentation building process.
432    
433    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
434        --match-limit and --recursion-limit options (added for 8.11). In
435        particular, this made one of the standard tests fail. (The integer value
436        went into the wrong half of a long int.)
437    
438    3.  If the --colour option was given to pcregrep with -v (invert match), it
439        did strange things, either producing crazy output, or crashing. It should,
440        of course, ignore a request for colour when reporting lines that do not
441        match.
442    
443    4.  Another pcregrep bug caused similar problems if --colour was specified with
444        -M (multiline) and the pattern match finished with a line ending.
445    
446    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
447        matched in multiline mode, the following line was shown as part of the
448        match. This seems wrong, so I have changed it.
449    
450    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
451        the check for further matches in the same line (so they could be coloured)
452        to overrun the end of the current line. If another match was found, it was
453        incorrectly shown (and then shown again when found in the next line).
454    
455    7.  If pcregrep was compiled under Windows, there was a reference to the
456        function pcregrep_exit() before it was defined. I am assuming this was
457        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
458        reported by a user. I've moved the definition above the reference.
459    
460    
461    Version 8.11 10-Dec-2010
462    ------------------------
463    
464    1.  (*THEN) was not working properly if there were untried alternatives prior
465        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
466        backtracked to try for "b" instead of moving to the next alternative branch
467        at the same level (in this case, to look for "c"). The Perl documentation
468        is clear that when (*THEN) is backtracked onto, it goes to the "next
469        alternative in the innermost enclosing group".
470    
471    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
472        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
473        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
474        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
475        (*THEN).
476    
477    3.  If \s appeared in a character class, it removed the VT character from
478        the class, even if it had been included by some previous item, for example
479        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
480        of \s, but is part of the POSIX "space" class.)
481    
482    4.  A partial match never returns an empty string (because you can always
483        match an empty string at the end of the subject); however the checking for
484        an empty string was starting at the "start of match" point. This has been
485        changed to the "earliest inspected character" point, because the returned
486        data for a partial match starts at this character. This means that, for
487        example, /(?<=abc)def/ gives a partial match for the subject "abc"
488        (previously it gave "no match").
489    
490    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
491        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
492        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
493        has an implication that the given string is incomplete (because a partial
494        match is preferred over a full match). For this reason, these items now
495        give a partial match in this situation. [Aside: previously, the one case
496        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
497        match rather than a full match, which was wrong by the old rules, but is
498        now correct.]
499    
500    6.  There was a bug in the handling of #-introduced comments, recognized when
501        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
502        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
503        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
504        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
505        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
506        places in pcre_compile().
507    
508    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
509        comments when looking ahead for named forward references to subpatterns,
510        the only newline sequence it recognized was NL. It now handles newlines
511        according to the set newline convention.
512    
513    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
514        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
515        cater for a lack of strerror(). These oversights have been fixed.
516    
517    9.  Added --match-limit and --recursion-limit to pcregrep.
518    
519    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
520    
521    11. When the -o option was used, pcregrep was setting a return code of 1, even
522        when matches were found, and --line-buffered was not being honoured.
523    
524    12. Added an optional parentheses number to the -o and --only-matching options
525        of pcregrep.
526    
527    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
528        can match an empty string. The code to do it in pcretest and pcredemo
529        needed fixing:
530    
531        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
532            only one byte after an empty string match just before CRLF (this case
533            just got forgotten; "any" and "anycrlf" were OK).
534    
535        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
536            mode when an empty string match preceded an ASCII character followed by
537            a non-ASCII character. (The code for advancing by one character rather
538            than one byte was nonsense.)
539    
540        (c) The pcredemo.c sample program did not have any code at all to handle
541            the cases when CRLF is a valid newline sequence.
542    
543    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
544        as a starting offset was within the subject string. There is now a new
545        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
546        negative or greater than the length of the string. In order to test this,
547        pcretest is extended to allow the setting of negative starting offsets.
548    
549    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
550        starting offset points to the beginning of a UTF-8 character was
551        unnecessarily clumsy. I tidied it up.
552    
553    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
554        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
555    
556    17. Nobody had reported that the --include_dir option, which was added in
557        release 7.7 should have been called --include-dir (hyphen, not underscore)
558        for compatibility with GNU grep. I have changed it to --include-dir, but
559        left --include_dir as an undocumented synonym, and the same for
560        --exclude-dir, though that is not available in GNU grep, at least as of
561        release 2.5.4.
562    
563    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
564        characters from a string of bytes) have been redefined so as not to use
565        loops, in order to improve performance in some environments. At the same
566        time, I abstracted some of the common code into auxiliary macros to save
567        repetition (this should not affect the compiled code).
568    
569    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
570        compile-time error is now given if \c is not followed by an ASCII
571        character, that is, a byte less than 128. (In EBCDIC mode, the code is
572        different, and any byte value is allowed.)
573    
574    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
575        START_OPTIMIZE option, which is now allowed at compile time - but just
576        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
577        to pcregrep and other applications that have no direct access to PCRE
578        options. The new /Y option in pcretest sets this option when calling
579        pcre_compile().
580    
581    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
582        back references. Groups containing recursive back references were forced to
583        be atomic by that change, but in the case of named groups, the amount of
584        memory required was incorrectly computed, leading to "Failed: internal
585        error: code overflow". This has been fixed.
586    
587    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
588        pcretest.c, to avoid build problems in some Borland environments.
589    
590    
591    Version 8.10 25-Jun-2010
592    ------------------------
593    
594    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
595        THEN.
596    
597    2.  (*ACCEPT) was not working when inside an atomic group.
598    
599    3.  Inside a character class, \B is treated as a literal by default, but
600        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
601        causes the error). The code is unchanged, but I tidied the documentation.
602    
603    4.  Inside a character class, PCRE always treated \R and \X as literals,
604        whereas Perl faults them if its -w option is set. I have changed PCRE so
605        that it faults them when PCRE_EXTRA is set.
606    
607    5.  Added support for \N, which always matches any character other than
608        newline. (It is the same as "." when PCRE_DOTALL is not set.)
609    
610    6.  When compiling pcregrep with newer versions of gcc which may have
611        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
612        declared with attribute warn_unused_result" were given. Just casting the
613        result to (void) does not stop the warnings; a more elaborate fudge is
614        needed. I've used a macro to implement this.
615    
616    7.  Minor change to pcretest.c to avoid a compiler warning.
617    
618    8.  Added four artifical Unicode properties to help with an option to make
619        \s etc use properties (see next item). The new properties are: Xan
620        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
621    
622    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
623        use Unicode properties. (*UCP) at the start of a pattern can be used to set
624        this option. Modified pcretest to add /W to test this facility. Added
625        REG_UCP to make it available via the POSIX interface.
626    
627    10. Added --line-buffered to pcregrep.
628    
629    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
630        studied, and the match started with a letter with a code point greater than
631        127 whose first byte was different to the first byte of the other case of
632        the letter, the other case of this starting letter was not recognized
633        (#976).
634    
635    12. If a pattern that was studied started with a repeated Unicode property
636        test, for example, \p{Nd}+, there was the theoretical possibility of
637        setting up an incorrect bitmap of starting bytes, but fortunately it could
638        not have actually happened in practice until change 8 above was made (it
639        added property types that matched character-matching opcodes).
640    
641    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
642        possible starting bytes for non-anchored patterns.
643    
644    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
645        \R, and also a number of cases that involve Unicode properties, both
646        explicit and implicit when PCRE_UCP is set.
647    
648    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
649        input, it could crash or give wrong results if characters with values
650        greater than 0xc0 were present in the subject string. (Detail: it assumed
651        UTF-8 input when processing these items.)
652    
653    16. Added a lot of (int) casts to avoid compiler warnings in systems where
654        size_t is 64-bit (#991).
655    
656    17. Added a check for running out of memory when PCRE is compiled with
657        --disable-stack-for-recursion (#990).
658    
659    18. If the last data line in a file for pcretest does not have a newline on
660        the end, a newline was missing in the output.
661    
662    19. The default pcre_chartables.c file recognizes only ASCII characters (values
663        less than 128) in its various bitmaps. However, there is a facility for
664        generating tables according to the current locale when PCRE is compiled. It
665        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
666        characters, are recognized by isspace() and therefore were getting set in
667        these tables, and indeed these tables seem to approximate to ISO 8859. This
668        caused a problem in UTF-8 mode when pcre_study() was used to create a list
669        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
670        which of course cannot start UTF-8 characters. I have changed the code so
671        that only real ASCII characters (less than 128) and the correct starting
672        bytes for UTF-8 encodings are set for characters greater than 127 when in
673        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
674        altogether.)
675    
676    20. Added the /T option to pcretest so as to be able to run tests with non-
677        standard character tables, thus making it possible to include the tests
678        used for 19 above in the standard set of tests.
679    
680    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
681        reference to a subpattern the other side of a comment that contains an
682        opening parenthesis caused either an internal compiling error, or a
683        reference to the wrong subpattern.
684    
685    
686    Version 8.02 19-Mar-2010
687    ------------------------
688    
689    1.  The Unicode data tables have been updated to Unicode 5.2.0.
690    
691    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
692        configured.
693    
694    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
695        original author of that file, following a query about its status.
696    
697    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
698        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
699    
700    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
701        quantifier applied to a forward-referencing subroutine call, could compile
702        incorrect code or give the error "internal error: previously-checked
703        referenced subpattern not found".
704    
705    6.  Both MS Visual Studio and Symbian OS have problems with initializing
706        variables to point to external functions. For these systems, therefore,
707        pcre_malloc etc. are now initialized to local functions that call the
708        relevant global functions.
709    
710    7.  There were two entries missing in the vectors called coptable and poptable
711        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
712        I've fixed the data, and added a kludgy way of testing at compile time that
713        the lengths are correct (equal to the number of opcodes).
714    
715    8.  Following on from 7, I added a similar kludge to check the length of the
716        eint vector in pcreposix.c.
717    
718    9.  Error texts for pcre_compile() are held as one long string to avoid too
719        much relocation at load time. To find a text, the string is searched,
720        counting zeros. There was no check for running off the end of the string,
721        which could happen if a new error number was added without updating the
722        string.
723    
724    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
725    
726    11. \K was not working if it appeared in an atomic group or in a group that
727        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
728        \K is "not well defined" if used in an assertion. PCRE now accepts it if
729        the assertion is positive, but not if it is negative.
730    
731    12. Change 11 fortuitously reduced the size of the stack frame used in the
732        "match()" function of pcre_exec.c by one pointer. Forthcoming
733        implementation of support for (*MARK) will need an extra pointer on the
734        stack; I have reserved it now, so that the stack frame size does not
735        decrease.
736    
737    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
738        item in branch that calls a recursion is a subroutine call - as in the
739        second branch in the above example - was incorrectly given the compile-
740        time error "recursive call could loop indefinitely" because pcre_compile()
741        was not correctly checking the subroutine for matching a non-empty string.
742    
743    14. The checks for overrunning compiling workspace could trigger after an
744        overrun had occurred. This is a "should never occur" error, but it can be
745        triggered by pathological patterns such as hundreds of nested parentheses.
746        The checks now trigger 100 bytes before the end of the workspace.
747    
748    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
749    
750    
751    Version 8.01 19-Jan-2010
752    ------------------------
753    
754    1.  If a pattern contained a conditional subpattern with only one branch (in
755        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
756        computed the wrong minimum data length (which is of course zero for such
757        subpatterns). This could cause incorrect "no match" results.
758    
759    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
760        the pattern is reset in the first branch, pcre_compile() failed with
761        "internal error: code overflow at offset...". This happened only when
762        the reset was to the original external option setting. (An optimization
763        abstracts leading options settings into an external setting, which was the
764        cause of this.)
765    
766    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
767        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
768        assertion pattern did not match (meaning that the assertion was true), it
769        was incorrectly treated as false if the SKIP had been reached during the
770        matching. This also applied to assertions used as conditions.
771    
772    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
773        assertion subpattern, including such a pattern used as a condition,
774        unpredictable results occurred, instead of the error return
775        PCRE_ERROR_DFA_UITEM.
776    
777    5.  The C++ GlobalReplace function was not working like Perl for the special
778        situation when an empty string is matched. It now does the fancy magic
779        stuff that is necessary.
780    
781    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
782        removed. (These were left over from very, very early versions of PCRE.)
783    
784    7.  Some cosmetic changes to the code to make life easier when compiling it
785        as part of something else:
786    
787        (a) Change DEBUG to PCRE_DEBUG.
788    
789        (b) In pcre_compile(), rename the member of the "branch_chain" structure
790            called "current" as "current_branch", to prevent a collision with the
791            Linux macro when compiled as a kernel module.
792    
793        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
794            prevent a collision with the Linux macro when compiled as a kernel
795            module.
796    
797    8.  In pcre_compile() there are some checks for integer overflows that used to
798        cast potentially large values to (double). This has been changed to that
799        when building, a check for int64_t is made, and if it is found, it is used
800        instead, thus avoiding the use of floating point arithmetic. (There is no
801        other use of FP in PCRE.) If int64_t is not found, the fallback is to
802        double.
803    
804    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
805        2005 (difference between two addresses compared to an unsigned value).
806    
807    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
808        custom one, because of the following reported problem in Windows:
809    
810          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
811              under Win32.
812          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
813              therefore missing the function definition.
814          - The compiler thus generates a "C" signature for the test function.
815          - The linker fails to find the "C" function.
816          - PCRE fails to configure if asked to do so against libbz2.
817    
818    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
819        messages were output:
820    
821          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
822          rerunning libtoolize, to keep the correct libtool macros in-tree.
823          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
824    
825        I have done both of these things.
826    
827    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
828        most of the time, it *can* run out if it is given a pattern that contains a
829        runaway infinite recursion. I updated the discussion in the pcrestack man
830        page.
831    
832    13. Now that we have gone to the x.xx style of version numbers, the minor
833        version may start with zero. Using 08 or 09 is a bad idea because users
834        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
835        interpreted as invalid octal numbers. I've updated the previous comment in
836        configure.ac, and also added a check that gives an error if 08 or 09 are
837        used.
838    
839    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
840        causing partial matching to fail when the end of the subject matched \W
841        in a UTF-8 pattern where \W was quantified with a minimum of 3.
842    
843    15. There were some discrepancies between the declarations in pcre_internal.h
844        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
845        their definitions. The declarations used "const uschar *" and the
846        definitions used USPTR. Even though USPTR is normally defined as "const
847        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
848        reported that: "This difference in casting confuses some C++ compilers, for
849        example, SunCC recognizes above declarations as different functions and
850        generates broken code for hbpcre." I have changed the declarations to use
851        USPTR.
852    
853    16. GNU libtool is named differently on some systems. The autogen.sh script now
854        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
855        (FreeBSD).
856    
857    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
858        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
859        comment: "Figure out how to create a longlong from a string: strtoll and
860        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
861        instance, but it only takes 2 args instead of 3!"
862    
863    18. A subtle bug concerned with back references has been fixed by a change of
864        specification, with a corresponding code fix. A pattern such as
865        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
866        refers, was giving matches when it shouldn't. For example, xa=xaaa would
867        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
868        same bug. Such groups have to be quantified to be useful, or contained
869        inside another quantified group. (If there's no repetition, the reference
870        can never match.) The problem arises because, having left the group and
871        moved on to the rest of the pattern, a later failure that backtracks into
872        the group uses the captured value from the final iteration of the group
873        rather than the correct earlier one. I have fixed this in PCRE by forcing
874        any group that contains a reference to itself to be an atomic group; that
875        is, there cannot be any backtracking into it once it has completed. This is
876        similar to recursive and subroutine calls.
877    
878    
879    Version 8.00 19-Oct-09
880    ----------------------
881    
882    1.  The table for translating pcre_compile() error codes into POSIX error codes
883        was out-of-date, and there was no check on the pcre_compile() error code
884        being within the table. This could lead to an OK return being given in
885        error.
886    
887    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
888        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
889        in a Windows environment.
890    
891    3.  The pcregrep --count option prints the count for each file even when it is
892        zero, as does GNU grep. However, pcregrep was also printing all files when
893        --files-with-matches was added. Now, when both options are given, it prints
894        counts only for those files that have at least one match. (GNU grep just
895        prints the file name in this circumstance, but including the count seems
896        more useful - otherwise, why use --count?) Also ensured that the
897        combination -clh just lists non-zero counts, with no names.
898    
899    4.  The long form of the pcregrep -F option was incorrectly implemented as
900        --fixed_strings instead of --fixed-strings. This is an incompatible change,
901        but it seems right to fix it, and I didn't think it was worth preserving
902        the old behaviour.
903    
904    5.  The command line items --regex=pattern and --regexp=pattern were not
905        recognized by pcregrep, which required --regex pattern or --regexp pattern
906        (with a space rather than an '='). The man page documented the '=' forms,
907        which are compatible with GNU grep; these now work.
908    
909    6.  No libpcreposix.pc file was created for pkg-config; there was just
910        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
911    
912    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
913        when UCP support is not needed, by modifying the Python script that
914        generates it from Unicode data files. This should not matter if the module
915        is correctly used as a library, but I received one complaint about 50K of
916        unwanted data. My guess is that the person linked everything into his
917        program rather than using a library. Anyway, it does no harm.
918    
919    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
920        was a minimum greater than 1 for a wide character in a possessive
921        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
922        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
923        character. Chaos in the form of incorrect output or a compiling loop could
924        result.
925    
926    9.  The restrictions on what a pattern can contain when partial matching is
927        requested for pcre_exec() have been removed. All patterns can now be
928        partially matched by this function. In addition, if there are at least two
929        slots in the offset vector, the offset of the earliest inspected character
930        for the match and the offset of the end of the subject are set in them when
931        PCRE_ERROR_PARTIAL is returned.
932    
933    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
934        synonymous with PCRE_PARTIAL, for backwards compatibility, and
935        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
936        and may be more useful for multi-segment matching.
937    
938    11. Partial matching with pcre_exec() is now more intuitive. A partial match
939        used to be given if ever the end of the subject was reached; now it is
940        given only if matching could not proceed because another character was
941        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
942        string "Z", which now yields "no match" instead of "partial match". In the
943        case of pcre_dfa_exec(), "no match" is given if every matching path for the
944        final character ended with (*FAIL).
945    
946    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
947        if the pattern had a "must contain" character that was already found in the
948        earlier partial match, unless partial matching was again requested. For
949        example, with the pattern /dog.(body)?/, the "must contain" character is
950        "g". If the first part-match was for the string "dog", restarting with
951        "sbody" failed. This bug has been fixed.
952    
953    13. The string returned by pcre_dfa_exec() after a partial match has been
954        changed so that it starts at the first inspected character rather than the
955        first character of the match. This makes a difference only if the pattern
956        starts with a lookbehind assertion or \b or \B (\K is not supported by
957        pcre_dfa_exec()). It's an incompatible change, but it makes the two
958        matching functions compatible, and I think it's the right thing to do.
959    
960    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
961        so that the demonstration program is easily available in environments where
962        PCRE has not been installed from source.
963    
964    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
965        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
966        library.
967    
968    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
969        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
970        is not the first non-POSIX option to be added. Clearly some people find
971        these options useful.
972    
973    17. If a caller to the POSIX matching function regexec() passes a non-zero
974        value for nmatch with a NULL value for pmatch, the value of
975        nmatch is forced to zero.
976    
977    18. RunGrepTest did not have a test for the availability of the -u option of
978        the diff command, as RunTest does. It now checks in the same way as
979        RunTest, and also checks for the -b option.
980    
981    19. If an odd number of negated classes containing just a single character
982        interposed, within parentheses, between a forward reference to a named
983        subpattern and the definition of the subpattern, compilation crashed with
984        an internal error, complaining that it could not find the referenced
985        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
986        [The bug was that it was starting one character too far in when skipping
987        over the character class, thus treating the ] as data rather than
988        terminating the class. This meant it could skip too much.]
989    
990    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
991        /g option in pcretest when the pattern contains \K, which makes it possible
992        to have an empty string match not at the start, even when the pattern is
993        anchored. Updated pcretest and pcredemo to use this option.
994    
995    21. If the maximum number of capturing subpatterns in a recursion was greater
996        than the maximum at the outer level, the higher number was returned, but
997        with unset values at the outer level. The correct (outer level) value is
998        now given.
999    
1000    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1001        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1002        make it do so. The string so far is captured, making this feature
1003        compatible with Perl.
1004    
1005    23. The tests have been re-organized, adding tests 11 and 12, to make it
1006        possible to check the Perl 5.10 features against Perl 5.10.
1007    
1008    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1009        pattern matches a fixed length string. PCRE did not allow this; now it
1010        does. Neither allows recursion.
1011    
1012    25. I finally figured out how to implement a request to provide the minimum
1013        length of subject string that was needed in order to match a given pattern.
1014        (It was back references and recursion that I had previously got hung up
1015        on.) This code has now been added to pcre_study(); it finds a lower bound
1016        to the length of subject needed. It is not necessarily the greatest lower
1017        bound, but using it to avoid searching strings that are too short does give
1018        some useful speed-ups. The value is available to calling programs via
1019        pcre_fullinfo().
1020    
1021    26. While implementing 25, I discovered to my embarrassment that pcretest had
1022        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1023        study optimizations had never been tested with that matching function.
1024        Oops. What is worse, even when it was passed study data, there was a bug in
1025        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1026        were also very few tests of studied patterns with pcre_dfa_exec().
1027    
1028    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1029        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1030        on the other side of the coin, they are no longer allowed to have different
1031        names, because these cannot be distinguished in PCRE, and this has caused
1032        confusion. (This is a difference from Perl.)
1033    
1034    28. When duplicate subpattern names are present (necessarily with different
1035        numbers, as required by 27 above), and a test is made by name in a
1036        conditional pattern, either for a subpattern having been matched, or for
1037        recursion in such a pattern, all the associated numbered subpatterns are
1038        tested, and the overall condition is true if the condition is true for any
1039        one of them. This is the way Perl works, and is also more like the way
1040        testing by number works.
1041    
1042    
1043    Version 7.9 11-Apr-09
1044    ---------------------
1045    
1046    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1047        (pcretest), all targets were linked against these libraries. This included
1048        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1049        libraries. This caused unwanted dependencies to be created. This problem
1050        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1051        pcretest is linked with readline.
1052    
1053    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1054        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1055        moved outside it again, because FALSE and TRUE are already defined in AIX,
1056        but BOOL is not.
1057    
1058    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1059        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1060    
1061    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1062        hyphens) following file names and line numbers when outputting matching
1063        lines. This is not true; no spaces are inserted. I have also clarified the
1064        wording for the --colour (or --color) option.
1065    
1066    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1067        was not coloured; this is different to GNU grep, so I have changed it to be
1068        the same.
1069    
1070    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1071        each matching line was coloured. Now it goes on to look for further matches
1072        of any of the test patterns, which is the same behaviour as GNU grep.
1073    
1074    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1075        doesn't make sense to accept an empty string match in pcregrep, so I have
1076        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1077        seems to be how GNU grep behaves.
1078    
1079    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1080        start or after a newline", because the conditional assertion was not being
1081        correctly handled. The rule now is that both the assertion and what follows
1082        in the first alternative must satisfy the test.
1083    
1084    9.  If auto-callout was enabled in a pattern with a conditional group whose
1085        condition was an assertion, PCRE could crash during matching, both with
1086        pcre_exec() and pcre_dfa_exec().
1087    
1088    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1089        used for matching.
1090    
1091    11. Unicode property support in character classes was not working for
1092        characters (bytes) greater than 127 when not in UTF-8 mode.
1093    
1094    12. Added the -M command line option to pcretest.
1095    
1096    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1097    
1098    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1099    
1100    16. Added comments and documentation about mis-use of no_arg in the C++
1101        wrapper.
1102    
1103    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1104        from Martin Jerabek that uses macro names for all relevant character and
1105        string constants.
1106    
1107    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1108        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1109        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1110        these, but not everybody uses configure.
1111    
1112    19. A conditional group that had only one branch was not being correctly
1113        recognized as an item that could match an empty string. This meant that an
1114        enclosing group might also not be so recognized, causing infinite looping
1115        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1116        with the subject "ab", where knowledge that the repeated group can match
1117        nothing is needed in order to break the loop.
1118    
1119    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1120        exec(), but without supplying a callout function, matching went wrong.
1121    
1122    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1123        leak if the size of the offset vector was greater than 30. When the vector
1124        is smaller, the saved offsets during recursion go onto a local stack
1125        vector, but for larger vectors malloc() is used. It was failing to free
1126        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1127        error, in fact).
1128    
1129    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1130        heapframe that is used only when UTF-8 support is enabled. This caused no
1131        problem, but was untidy.
1132    
1133    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1134        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1135        included within another project.
1136    
1137    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1138        slightly modified by me:
1139    
1140          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1141              not building pcregrep.
1142    
1143          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1144              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1145    
1146    25. Forward references, both numeric and by name, in patterns that made use of
1147        duplicate group numbers, could behave incorrectly or give incorrect errors,
1148        because when scanning forward to find the reference group, PCRE was not
1149        taking into account the duplicate group numbers. A pattern such as
1150        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1151    
1152    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1153        the feature of a custom pointer more persuasive (as requested by a user).
1154    
1155    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1156        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1157        pre-defined.
1158    
1159    28. Added support for (*UTF8) at the start of a pattern.
1160    
1161    29. Arrange for flags added by the "release type" setting in CMake to be shown
1162        in the configuration summary.
1163    
1164    
1165    Version 7.8 05-Sep-08
1166    ---------------------
1167    
1168    1.  Replaced UCP searching code with optimized version as implemented for Ad
1169        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1170        stage table and inline lookup instead of a function, giving speed ups of 2
1171        to 5 times on some simple patterns that I tested. Permission was given to
1172        distribute the MultiStage2.py script that generates the tables (it's not in
1173        the tarball, but is in the Subversion repository).
1174    
1175    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1176        scripts.
1177    
1178    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1179        a group with a zero qualifier. The result of the study could be incorrect,
1180        or the function might crash, depending on the pattern.
1181    
1182    4.  Caseless matching was not working for non-ASCII characters in back
1183        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1184        It now works when Unicode Property Support is available.
1185    
1186    5.  In pcretest, an escape such as \x{de} in the data was always generating
1187        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1188        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1189        truncation.
1190    
1191    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1192    
1193    7.  Added two (int) casts to pcregrep when printing the difference of two
1194        pointers, in case they are 64-bit values.
1195    
1196    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1197        test 2 if it fails.
1198    
1199    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1200        and a #define of that name to empty if it is not externally set. This is to
1201        allow users of MSVC to set it if necessary.
1202    
1203    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1204        the convenience functions in the pcre_get.c source file.
1205    
1206    11. An option change at the start of a pattern that had top-level alternatives
1207        could cause overwriting and/or a crash. This command provoked a crash in
1208        some environments:
1209    
1210          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1211    
1212        This potential security problem was recorded as CVE-2008-2371.
1213    
1214    12. For a pattern where the match had to start at the beginning or immediately
1215        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1216        pcre_dfa_exec() could read past the end of the passed subject if there was
1217        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1218        pcretest so that it places the subject at the end of its malloc-ed buffer.
1219    
1220    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1221        exec() might read past the end of the data buffer in UTF-8 mode.
1222    
1223    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1224        the data contained the byte 0x85 as part of a UTF-8 character within its
1225        first line. This applied both to normal and DFA matching.
1226    
1227    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1228        /^[^d]*?$/8 failed to match "abc".
1229    
1230    16. Added a missing copyright notice to pcrecpp_internal.h.
1231    
1232    17. Make it more clear in the documentation that values returned from
1233        pcre_exec() in ovector are byte offsets, not character counts.
1234    
1235    18. Tidied a few places to stop certain compilers from issuing warnings.
1236    
1237    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1238        supplied by Stefan Weber. I made a further small update for 7.8 because
1239        there is a change of source arrangements: the pcre_searchfuncs.c module is
1240        replaced by pcre_ucd.c.
1241    
1242    
1243    Version 7.7 07-May-08
1244    ---------------------
1245    
1246    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1247        a string to a long long, pretend we don't even have a long long." This is
1248        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1249    
1250    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1251        pre-7.6 versions, which defined a global no_arg variable instead of putting
1252        it in the RE class. (See also #8 below.)
1253    
1254    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1255        Lopes.
1256    
1257    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1258    
1259        (1) The include/exclude patterns were being applied to the whole pathnames
1260            of files, instead of just to the final components.
1261    
1262        (2) If there was more than one level of directory, the subdirectories were
1263            skipped unless they satisfied the include/exclude conditions. This is
1264            inconsistent with GNU grep (and could even be seen as contrary to the
1265            pcregrep specification - which I improved to make it absolutely clear).
1266            The action now is always to scan all levels of directory, and just
1267            apply the include/exclude patterns to regular files.
1268    
1269    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1270        --exclude_dir in the tests to avoid scanning .svn directories.
1271    
1272    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1273        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1274        doesn't support NULs in patterns.
1275    
1276    7.  Added some missing "const"s to declarations of static tables in
1277        pcre_compile.c and pcre_dfa_exec.c.
1278    
1279    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1280        caused by fix #2  above. (Subsequently also a second patch to fix the
1281        first patch. And a third patch - this was a messy problem.)
1282    
1283    9.  Applied Craig's patch to remove the use of push_back().
1284    
1285    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1286        matching function regexec().
1287    
1288    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1289        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1290        references. PCRE supports relative numbers with this syntax (I don't think
1291        Oniguruma does).
1292    
1293    12. Previously, a group with a zero repeat such as (...){0} was completely
1294        omitted from the compiled regex. However, this means that if the group
1295        was called as a subroutine from elsewhere in the pattern, things went wrong
1296        (an internal error was given). Such groups are now left in the compiled
1297        pattern, with a new opcode that causes them to be skipped at execution
1298        time.
1299    
1300    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1301        to the way PCRE behaves:
1302    
1303        (a) A lone ] character is dis-allowed (Perl treats it as data).
1304    
1305        (b) A back reference to an unmatched subpattern matches an empty string
1306            (Perl fails the current match path).
1307    
1308        (c) A data ] in a character class must be notated as \] because if the
1309            first data character in a class is ], it defines an empty class. (In
1310            Perl it is not possible to have an empty class.) The empty class []
1311            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1312            The negative empty class [^] matches any one character, independently
1313            of the DOTALL setting.
1314    
1315    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1316        non-existent subpattern following a character class starting with ']' and
1317        containing () gave an internal compiling error instead of "reference to
1318        non-existent subpattern". Fortunately, when the pattern did exist, the
1319        compiled code was correct. (When scanning forwards to check for the
1320        existencd of the subpattern, it was treating the data ']' as terminating
1321        the class, so got the count wrong. When actually compiling, the reference
1322        was subsequently set up correctly.)
1323    
1324    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1325        it was being rejected as not supported by pcre_dfa_exec(), even though
1326        other assertions are supported. I have made pcre_dfa_exec() support
1327        (*FAIL).
1328    
1329    16. The implementation of 13c above involved the invention of a new opcode,
1330        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1331        cannot be changed at match time, I realized I could make a small
1332        improvement to matching performance by compiling OP_ALLANY instead of
1333        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1334        on the OP_ANY path.
1335    
1336    17. Compiling pcretest on Windows with readline support failed without the
1337        following two fixes: (1) Make the unistd.h include conditional on
1338        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1339    
1340    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1341        ncurses library to be included for pcretest when ReadLine support is
1342        requested, but also to allow for it to be overridden. This patch came from
1343        Daniel Bergström.
1344    
1345    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1346        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1347        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1348        spotting this.
1349    
1350    
1351    Version 7.6 28-Jan-08
1352    ---------------------
1353    
1354    1.  A character class containing a very large number of characters with
1355        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1356        overflow.
1357    
1358    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1359        HAVE_LONG_LONG is not defined.
1360    
1361    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1362        bring them up to date and include new features. This patch includes:
1363    
1364        - Fixed PH's badly added libz and libbz2 support.
1365        - Fixed a problem with static linking.
1366        - Added pcredemo. [But later removed - see 7 below.]
1367        - Fixed dftables problem and added an option.
1368        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1369            HAVE_LONG_LONG.
1370        - Added readline support for pcretest.
1371        - Added an listing of the option settings after cmake has run.
1372    
1373    4.  A user submitted a patch to Makefile that makes it easy to create
1374        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1375        Makefile.am that cause it to include this special target, without
1376        affecting anything else. Note that the same mingw target plus all
1377        the other distribution libraries and programs are now supported
1378        when configuring with CMake (see 6 below) instead of with
1379        Configure/Make.
1380    
1381    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1382        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1383        exported in the Windows port". It has not yet been confirmed that the patch
1384        solves the problem, but it does no harm.
1385    
1386    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1387        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1388        with CMake, and also correct the comment about stack recursion.
1389    
1390    7.  Remove the automatic building of pcredemo from the ./configure system and
1391        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1392        of a program that users should build themselves after PCRE is installed, so
1393        building it automatically is not really right. What is more, it gave
1394        trouble in some build environments.
1395    
1396    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1397    
1398    
1399    Version 7.5 10-Jan-08
1400    ---------------------
1401    
1402    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1403        values in parens when parsing an RE using the C++ wrapper."
1404    
1405    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1406        Characters greater than 255 were excluded from the class instead of being
1407        included.
1408    
1409    3.  The same bug as (2) above applied to negated POSIX classes such as
1410        [:^space:].
1411    
1412    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1413        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1414        I have changed it.
1415    
1416    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1417        first named subpattern) and a construct such as (?&a) would reference the
1418        first named subpattern whose name started with "a" (in other words, the
1419        length check was missing). Both these problems are fixed. "Subpattern name
1420        expected" is now given for (?&) (a zero-length name), and this patch also
1421        makes it give the same error for \k'' (previously it complained that that
1422        was a reference to a non-existent subpattern).
1423    
1424    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1425        this is right because (?- can be followed by option settings as well as by
1426        digits. I have, however, made the messages clearer.
1427    
1428    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1429        than the number used in the conditional) now cause a compile-time error.
1430        This is actually not compatible with Perl, which accepts such patterns, but
1431        treats the conditional as always being FALSE (as PCRE used to), but it
1432        seems to me that giving a diagnostic is better.
1433    
1434    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1435        and messages.
1436    
1437    9.  Fix two occurrences of "backslash" in comments that should have been
1438        "backspace".
1439    
1440    10. Remove two redundant lines of code that can never be obeyed (their function
1441        was moved elsewhere).
1442    
1443    11. The program that makes PCRE's Unicode character property table had a bug
1444        which caused it to generate incorrect table entries for sequences of
1445        characters that have the same character type, but are in different scripts.
1446        It amalgamated them into a single range, with the script of the first of
1447        them. In other words, some characters were in the wrong script. There were
1448        thirteen such cases, affecting characters in the following ranges:
1449    
1450          U+002b0 - U+002c1
1451          U+0060c - U+0060d
1452          U+0061e - U+00612
1453          U+0064b - U+0065e
1454          U+0074d - U+0076d
1455          U+01800 - U+01805
1456          U+01d00 - U+01d77
1457          U+01d9b - U+01dbf
1458          U+0200b - U+0200f
1459          U+030fc - U+030fe
1460          U+03260 - U+0327f
1461          U+0fb46 - U+0fbb1
1462          U+10450 - U+1049d
1463    
1464    12. The -o option (show only the matching part of a line) for pcregrep was not
1465        compatible with GNU grep in that, if there was more than one match in a
1466        line, it showed only the first of them. It now behaves in the same way as
1467        GNU grep.
1468    
1469    13. If the -o and -v options were combined for pcregrep, it printed a blank
1470        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1471        does the same. The return code can be used to tell if there were any
1472        non-matching lines.
1473    
1474    14. Added --file-offsets and --line-offsets to pcregrep.
1475    
1476    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1477        infinitely looping recursion. The bug was that positive lookaheads were not
1478        being skipped when checking for a possible empty match (negative lookaheads
1479        and both kinds of lookbehind were skipped).
1480    
1481    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1482        inclusion of <windows.h> to before rather than after the definition of
1483        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1484    
1485    17. Specifying a possessive quantifier with a specific limit for a Unicode
1486        character property caused pcre_compile() to compile bad code, which led at
1487        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1488        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1489        caused the error; without that there was no problem.
1490    
1491    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1492    
1493    19. Added --enable-pcretest-libreadline.
1494    
1495    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1496        RE::GlobalReplace(). As a result, the number of replacements returned was
1497        double what it should be. I removed one of the increments, but Craig sent a
1498        later patch that removed the other one (the right fix) and added unit tests
1499        that check the return values (which was not done before).
1500    
1501    21. Several CMake things:
1502    
1503        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1504            the names libpcre and libpcreposix, not just pcre and pcreposix.
1505    
1506        (2) The above change means that pcretest and pcregrep are now correctly
1507            linked with the newly-built libraries, not previously installed ones.
1508    
1509        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1510    
1511    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1512        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1513        UTF-8 newline character). The key issue is that the pattern starts .*;
1514        this means that the match must be either at the beginning, or after a
1515        newline. The bug was in the code for advancing after a failed match and
1516        checking that the new position followed a newline. It was not taking
1517        account of UTF-8 characters correctly.
1518    
1519    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1520        character classes. PCRE was not treating the sequence [:...:] as a
1521        character class unless the ... were all letters. Perl, however, seems to
1522        allow any characters between [: and :], though of course it rejects as
1523        unknown any "names" that contain non-letters, because all the known class
1524        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1525        for example, whereas PCRE did not - it did not recognize a POSIX character
1526        class. This seemed a bit dangerous, so the code has been changed to be
1527        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1528        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1529        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1530        Perl does, and where it didn't before.
1531    
1532    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1533        Windows environments %n is disabled by default.
1534    
1535    
1536    Version 7.4 21-Sep-07
1537    ---------------------
1538    
1539    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1540        means that a class such as [\s] counted as "explicit reference to CR or
1541        LF". That isn't really right - the whole point of the change was to try to
1542        help when there was an actual mention of one of the two characters. So now
1543        the change happens only if \r or \n (or a literal CR or LF) character is
1544        encountered.
1545    
1546    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1547        of both had grown to the point where there were only 3 bits left.
1548        Fortunately, there was spare space in the data structure, and so I have
1549        moved the internal flags into a new 16-bit field to free up more option
1550        bits.
1551    
1552    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1553        but did not set the internal JCHANGED flag - either of these is enough to
1554        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1555        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1556        start sets both bits.
1557    
1558    4.  Added options (at build time, compile time, exec time) to change \R from
1559        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1560    
1561    5.  doc/pcresyntax.html was missing from the distribution.
1562    
1563    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1564        compatibility, even though it is no longer used.
1565    
1566    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1567        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1568        windows.h file is present (where different names are used). [This was
1569        reversed later after testing - see 16 below.]
1570    
1571    8.  Changed all #include <config.h> to #include "config.h". There were also
1572        some further <pcre.h> cases that I changed to "pcre.h".
1573    
1574    9.  When pcregrep was used with the --colour option, it missed the line ending
1575        sequence off the lines that it output.
1576    
1577    10. It was pointed out to me that arrays of string pointers cause lots of
1578        relocations when a shared library is dynamically loaded. A technique of
1579        using a single long string with a table of offsets can drastically reduce
1580        these. I have refactored PCRE in four places to do this. The result is
1581        dramatic:
1582    
1583          Originally:                          290
1584          After changing UCP table:            187
1585          After changing error message table:   43
1586          After changing table of "verbs"       36
1587          After changing table of Posix names   22
1588    
1589        Thanks to the folks working on Gregex for glib for this insight.
1590    
1591    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1592        unicode-properties was also set.
1593    
1594    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1595    
1596    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1597        checked only for CRLF.
1598    
1599    14. Added casts to pcretest.c to avoid compiler warnings.
1600    
1601    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1602    
1603    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1604        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1605        entirely. This removes changes made in 7 above.
1606    
1607    17. The CMake files have been updated, and there is now more information about
1608        building with CMake in the NON-UNIX-USE document.
1609    
1610    
1611    Version 7.3 28-Aug-07
1612    ---------------------
1613    
1614     1. In the rejigging of the build system that eventually resulted in 7.1, the
1615        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1616        brackets there is not right, since it causes compilers to look for an
1617        installed pcre.h, not the version that is in the source that is being
1618        compiled (which of course may be different). I have changed it back to:
1619    
1620          #include "pcre.h"
1621    
1622        I have a vague recollection that the change was concerned with compiling in
1623        different directories, but in the new build system, that is taken care of
1624        by the VPATH setting the Makefile.
1625    
1626     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1627        when the subject happened to end in the byte 0x85 (e.g. if the last
1628        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1629        characters but of course it shouldn't be taken as a newline when it is part
1630        of another character. The bug was that, for an unlimited repeat of . in
1631        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1632        characters when looking for a newline.
1633    
1634     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1635    
1636     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1637        in debug output.
1638    
1639     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1640        long printing in the pcrecpp unittest when running under MinGW.
1641    
1642     6. ESC_K was left out of the EBCDIC table.
1643    
1644     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1645        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1646        limit also applies to "virtual nesting" when a pattern is recursive, and in
1647        this case 1000 isn't so big. I have been able to remove this limit at the
1648        expense of backing off one optimization in certain circumstances. Normally,
1649        when pcre_exec() would call its internal match() function recursively and
1650        immediately return the result unconditionally, it uses a "tail recursion"
1651        feature to save stack. However, when a subpattern that can match an empty
1652        string has an unlimited repetition quantifier, it no longer makes this
1653        optimization. That gives it a stack frame in which to save the data for
1654        checking that an empty string has been matched. Previously this was taken
1655        from the 1000-entry workspace that had been reserved. So now there is no
1656        explicit limit, but more stack is used.
1657    
1658     8. Applied Daniel's patches to solve problems with the import/export magic
1659        syntax that is required for Windows, and which was going wrong for the
1660        pcreposix and pcrecpp parts of the library. These were overlooked when this
1661        problem was solved for the main library.
1662    
1663     9. There were some crude static tests to avoid integer overflow when computing
1664        the size of patterns that contain repeated groups with explicit upper
1665        limits. As the maximum quantifier is 65535, the maximum group length was
1666        set at 30,000 so that the product of these two numbers did not overflow a
1667        32-bit integer. However, it turns out that people want to use groups that
1668        are longer than 30,000 bytes (though not repeat them that many times).
1669        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1670        made it possible to implement the integer overflow checks in a much more
1671        dynamic way, which I have now done. The artificial limitation on group
1672        length has been removed - we now have only the limit on the total length of
1673        the compiled pattern, which depends on the LINK_SIZE setting.
1674    
1675    10. Fixed a bug in the documentation for get/copy named substring when
1676        duplicate names are permitted. If none of the named substrings are set, the
1677        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1678        empty string.
1679    
1680    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1681        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1682        because the ] is interpreted as the first data character and the
1683        terminating ] is not found. PCRE has been made compatible with Perl in this
1684        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1685        cause memory overwriting.
1686    
1687    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1688        string has been matched (to stop an infinite loop). It was not recognizing
1689        a conditional subpattern that could match an empty string if that
1690        subpattern was within another subpattern. For example, it looped when
1691        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1692        condition was not nested. This bug has been fixed.
1693    
1694    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1695        past the start of the subject in the presence of bytes with the top bit
1696        set, for example "\x8aBCD".
1697    
1698    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1699        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1700    
1701    14. Optimized (?!) to (*FAIL).
1702    
1703    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1704        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1705        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1706        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1707        does: it's just the validity check that is more restrictive.
1708    
1709    16. Inserted checks for integer overflows during escape sequence (backslash)
1710        processing, and also fixed erroneous offset values for syntax errors during
1711        backslash processing.
1712    
1713    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1714        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1715    
1716    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1717        caused an overrun.
1718    
1719    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1720        something other than just ASCII characters) inside a group that had an
1721        unlimited repeat caused a loop at compile time (while checking to see
1722        whether the group could match an empty string).
1723    
1724    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1725        [\P{Any}] did so. (Error in the code for printing property names.)
1726    
1727    21. An orphan \E inside a character class could cause a crash.
1728    
1729    22. A repeated capturing bracket such as (A)? could cause a wild memory
1730        reference during compilation.
1731    
1732    23. There are several functions in pcre_compile() that scan along a compiled
1733        expression for various reasons (e.g. to see if it's fixed length for look
1734        behind). There were bugs in these functions when a repeated \p or \P was
1735        present in the pattern. These operators have additional parameters compared
1736        with \d, etc, and these were not being taken into account when moving along
1737        the compiled data. Specifically:
1738    
1739        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1740            length.
1741    
1742        (b) An item such as \pL+ within a repeated group could cause crashes or
1743            loops.
1744    
1745        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1746            "reference to non-existent subpattern" error.
1747    
1748        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1749    
1750    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1751        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1752    
1753    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1754    
1755    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1756        character were causing crashes (broken optimization).
1757    
1758    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1759        \p or \P) caused a compile-time loop.
1760    
1761    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1762        break. For example, the unstudied pattern [\r\n]A does not match the string
1763        "\r\nA" because change 7.0/46 below moves the current point on by two
1764        characters after failing to match at the start. However, the pattern \nA
1765        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1766        the same is true. There doesn't seem any very clean way out of this, but
1767        what I have chosen to do makes the common cases work: PCRE now takes note
1768        of whether there can be an explicit match for \r or \n anywhere in the
1769        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1770        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1771        pattern has explicit CR or LF references.
1772    
1773    29. Added (*CR) etc for changing newline setting at start of pattern.
1774    
1775    
1776    Version 7.2 19-Jun-07
1777    ---------------------
1778    
1779     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1780        which is apparently normally available under Windows.
1781    
1782     2. Re-jig the pcregrep tests with different newline settings in an attempt
1783        to make them independent of the local environment's newline setting.
1784    
1785     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1786    
1787     4. Some of the "internals" tests were previously cut out when the link size
1788        was not 2, because the output contained actual offsets. The recent new
1789        "Z" feature of pcretest means that these can be cut out, making the tests
1790        usable with all link sizes.
1791    
1792     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1793        stack recursion. This gives a massive performance boost under BSD, but just
1794        a small improvement under Linux. However, it saves one field in the frame
1795        in all cases.
1796    
1797     6. Added more features from the forthcoming Perl 5.10:
1798    
1799        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1800            recursion call. It refers to the nth most recently opened parentheses.
1801    
1802        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1803            to be opened parentheses.
1804    
1805        (c) Conditions that refer to capturing parentheses can be specified
1806            relatively, for example, (?(-2)... or (?(+3)...
1807    
1808        (d) \K resets the start of the current match so that everything before
1809            is not part of it.
1810    
1811        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1812    
1813        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1814            reference syntax.
1815    
1816        (g) (?| introduces a group in which the numbering of parentheses in each
1817            alternative starts with the same number.
1818    
1819        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1820    
1821     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1822        PCRE_INFO_JCHANGED.
1823    
1824     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1825        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1826        for detecting groups that can match an empty string.
1827    
1828     9. A pattern with a very large number of alternatives (more than several
1829        hundred) was running out of internal workspace during the pre-compile
1830        phase, where pcre_compile() figures out how much memory will be needed. A
1831        bit of new cunning has reduced the workspace needed for groups with
1832        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1833        workspace instead of running out of the 4096 that are available.
1834    
1835    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1836    
1837    11. Applied patch from Google to remove an optimization that didn't quite work.
1838        The report of the bug said:
1839    
1840          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1841          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1842          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1843    
1844    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1845        it matched the wrong number of bytes.
1846    
1847    
1848    Version 7.1 24-Apr-07
1849  ---------------------  ---------------------
1850    
1851   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
# Line 136  Version 7.1 12-Mar-07 Line 1980  Version 7.1 12-Mar-07
1980      string at the end of a line did not allow for this case. They now check for      string at the end of a line did not allow for this case. They now check for
1981      it specially.      it specially.
1982    
1983    20. pcretest is supposed to handle patterns and data of any length, by
1984        extending its buffers when necessary. It was getting this wrong when the
1985        buffer for a data line had to be extended.
1986    
1987    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1988        CRLF as a newline sequence.
1989    
1990    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1991        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1992        I have nevertheless tidied it up.
1993    
1994    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1995    
1996    24. Added a man page for pcre-config.
1997    
1998    
1999  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06
2000  ---------------------  ---------------------

Legend:
Removed from v.143  
changed lines
  Added in v.760

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12