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

Diff of /code/trunk/ChangeLog

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

revision 25 by nigel, Sat Feb 24 21:38:45 2007 UTC revision 888 by ph10, Tue Jan 17 14:43:23 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.30
5    ------------
6    
7    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
8        name is defined in ctype.h.
9    
10    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
11        only in quite long subpatterns.
12    
13    3.  Removed the function pcre_info(), which has been obsolete and deprecated
14        since it was replaced by pcre_fullinfo() in February 2000.
15    
16    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
17        match a (*MARK), and the match failed at the start of the subject, a
18        reference to memory before the start of the subject could occur. This bug
19        was introduced by fix 17 of release 8.21.
20    
21    5.  A reference to an unset group with zero minimum repetition was giving
22        totally wrong answers (in non-JavaScript-compatibility mode). For example,
23        /(another)?(\1?)test/ matched against "hello world test". This bug was
24        introduced in release 8.13.
25    
26    6.  Add support for 16-bit character strings (a large amount of work involving
27        many changes and refactorings).
28    
29    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
30        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
31        from a file.
32    
33    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
34        rounding is not applied in this particular case).
35    
36    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
37        if they appear, or are escaped, in patterns.
38    
39    10. Get rid of a number of -Wunused-but-set-variable warnings.
40    
41    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
42        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
43        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
44        also returns the mark "x". This bug applied to capturing parentheses,
45        non-capturing parentheses, and atomic parentheses. It also applied to some
46        assertions.
47    
48    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
49        information out of configure.ac instead of relying on pcre.h.generic, which
50        is not stored in the repository.
51    
52    
53    Version 8.21 12-Dec-2011
54    ------------------------
55    
56    1.  Updating the JIT compiler.
57    
58    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
59        are added as well.
60    
61    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
62        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
63        calling _pcre_jit_exec. Some extra comments are added.
64    
65    4.  (*MARK) settings inside atomic groups that do not contain any capturing
66        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
67        was introduced by change 18 for 8.20.
68    
69    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
70        ECMA-262 standard.
71    
72    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
73        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
74        This bug was probably introduced by change 9 of 8.13.
75    
76    7.  While fixing 6 above, I noticed that a number of other items were being
77        incorrectly rejected as "not fixed length". This arose partly because newer
78        opcodes had not been added to the fixed-length checking code. I have (a)
79        corrected the bug and added tests for these items, and (b) arranged for an
80        error to occur if an unknown opcode is encountered while checking for fixed
81        length instead of just assuming "not fixed length". The items that were
82        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
83        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
84        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
85    
86    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
87        being incorrectly compiled and would have given unpredicatble results.
88    
89    9.  A possessively repeated subpattern with minimum repeat count greater than
90        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
91        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
92        the first (A) could occur when it should not.
93    
94    10. Add a cast and remove a redundant test from the code.
95    
96    11. JIT should use pcre_malloc/pcre_free for allocation.
97    
98    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
99        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
100        is anything other than /usr, -L is still shown).
101    
102    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
103    
104    14. Perl does not support \N without a following name in a [] class; PCRE now
105        also gives an error.
106    
107    15. If a forward reference was repeated with an upper limit of around 2000,
108        it caused the error "internal error: overran compiling workspace". The
109        maximum number of forward references (including repeats) was limited by the
110        internal workspace, and dependent on the LINK_SIZE. The code has been
111        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
112        the default depends on LINK_SIZE. There is a new upper limit (for safety)
113        of around 200,000 forward references. While doing this, I also speeded up
114        the filling in of repeated forward references.
115    
116    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
117        incorrectly expecting the subject to contain another "a" after the start.
118    
119    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
120        in the match, the SKIP should be ignored. This was not happening; instead
121        the SKIP was being treated as NOMATCH. For patterns such as
122        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
123        tested.
124    
125    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
126        now much more compatible with Perl, in particular in cases where the result
127        is a non-match for a non-anchored pattern. For example, if
128        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
129        "m", where previously it did not return a name. A side effect of this
130        change is that for partial matches, the last encountered mark name is
131        returned, as for non matches. A number of tests that were previously not
132        Perl-compatible have been moved into the Perl-compatible test files. The
133        refactoring has had the pleasing side effect of removing one argument from
134        the match() function, thus reducing its stack requirements.
135    
136    19. If the /S+ option was used in pcretest to study a pattern using JIT,
137        subsequent uses of /S (without +) incorrectly behaved like /S+.
138    
139    21. Retrieve executable code size support for the JIT compiler and fixing
140        some warnings.
141    
142    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
143        not work when the shorter character appeared right at the end of the
144        subject string.
145    
146    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
147        systems.
148    
149    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
150        output it when the /M option is used in pcretest.
151    
152    25. The CheckMan script was not being included in the distribution. Also, added
153        an explicit "perl" to run Perl scripts from the PrepareRelease script
154        because this is reportedly needed in Windows.
155    
156    26. If study data was being save in a file and studying had not found a set of
157        "starts with" bytes for the pattern, the data written to the file (though
158        never used) was taken from uninitialized memory and so caused valgrind to
159        complain.
160    
161    27. Updated RunTest.bat as provided by Sheri Pierce.
162    
163    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
164    
165    29. Computation of memory usage for the table of capturing group names was
166        giving an unnecessarily large value.
167    
168    
169    Version 8.20 21-Oct-2011
170    ------------------------
171    
172    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
173        a POSIX class. After further experiments with Perl, which convinced me that
174        Perl has bugs and confusions, a closing square bracket is no longer allowed
175        in a POSIX name. This bug also affected patterns with classes that started
176        with full stops.
177    
178    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
179        captured substring, but while checking the failing first alternative,
180        substring 1 is temporarily captured. If the output vector supplied to
181        pcre_exec() was not big enough for this capture, the yield of the function
182        was still zero ("insufficient space for captured substrings"). This cannot
183        be totally fixed without adding another stack variable, which seems a lot
184        of expense for a edge case. However, I have improved the situation in cases
185        such as /(a)(b)x|abc/ matched against "abc", where the return code
186        indicates that fewer than the maximum number of slots in the ovector have
187        been set.
188    
189    3.  Related to (2) above: when there are more back references in a pattern than
190        slots in the output vector, pcre_exec() uses temporary memory during
191        matching, and copies in the captures as far as possible afterwards. It was
192        using the entire output vector, but this conflicts with the specification
193        that only 2/3 is used for passing back captured substrings. Now it uses
194        only the first 2/3, for compatibility. This is, of course, another edge
195        case.
196    
197    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
198        main code base, and can be used by building with --enable-jit. When this is
199        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
200        runtime --no-jit option is given.
201    
202    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
203        ovector, the return from the function was zero, implying that there were
204        other matches that did not fit. The correct "exactly full" value is now
205        returned.
206    
207    6.  If a subpattern that was called recursively or as a subroutine contained
208        (*PRUNE) or any other control that caused it to give a non-standard return,
209        invalid errors such as "Error -26 (nested recursion at the same subject
210        position)" or even infinite loops could occur.
211    
212    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
213        computing the minimum length on reaching *ACCEPT, and so ended up with the
214        wrong value of 1 rather than 0. Further investigation indicates that
215        computing a minimum subject length in the presence of *ACCEPT is difficult
216        (think back references, subroutine calls), and so I have changed the code
217        so that no minimum is registered for a pattern that contains *ACCEPT.
218    
219    8.  If (*THEN) was present in the first (true) branch of a conditional group,
220        it was not handled as intended. [But see 16 below.]
221    
222    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
223        Sheri Pierce.
224    
225    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
226        the first byte in a match must be "a".
227    
228    11. Change 17 for 8.13 increased the recursion depth for patterns like
229        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
230        pattern contains any instances of (*THEN). If it does not, the old
231        optimizations are restored. It would be nice to do this on a per-group
232        basis, but at the moment that is not feasible.
233    
234    12. In some environments, the output of pcretest -C is CRLF terminated. This
235        broke RunTest's code that checks for the link size. A single white space
236        character after the value is now allowed for.
237    
238    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
239        For "fr", it uses the Windows-specific input and output files.
240    
241    14. If (*THEN) appeared in a group that was called recursively or as a
242        subroutine, it did not work as intended. [But see next item.]
243    
244    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
245        pattern fragments (but not containing any | characters). If A and B are
246        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
247        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
248        D. In other words, Perl considers parentheses that do not contain any |
249        characters to be part of a surrounding alternative, whereas PCRE was
250        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
251        differently. PCRE now behaves in the same way as Perl, except in the case
252        of subroutine/recursion calls such as (?1) which have in any case always
253        been different (but PCRE had them first :-).
254    
255    16. Related to 15 above: Perl does not treat the | in a conditional group as
256        creating alternatives. Such a group is treated in the same way as an
257        ordinary group without any | characters when processing (*THEN). PCRE has
258        been changed to match Perl's behaviour.
259    
260    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
261        RunGrepTest script failed.
262    
263    18. Change 22 for version 13 caused atomic groups to use more stack. This is
264        inevitable for groups that contain captures, but it can lead to a lot of
265        stack use in large patterns. The old behaviour has been restored for atomic
266        groups that do not contain any capturing parentheses.
267    
268    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
269        suppress the check for a minimum subject length at run time. (If it was
270        given to pcre_exec() or pcre_dfa_exec() it did work.)
271    
272    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
273        fail to work when decoding hex characters in data strings in EBCDIC
274        environments.
275    
276    21. It appears that in at least one Mac OS environment, the isxdigit() function
277        is implemented as a macro that evaluates to its argument more than once,
278        contravening the C 90 Standard (I haven't checked a later standard). There
279        was an instance in pcretest which caused it to go wrong when processing
280        \x{...} escapes in subject strings. The has been rewritten to avoid using
281        things like p++ in the argument of isxdigit().
282    
283    
284    Version 8.13 16-Aug-2011
285    ------------------------
286    
287    1.  The Unicode data tables have been updated to Unicode 6.0.0.
288    
289    2.  Two minor typos in pcre_internal.h have been fixed.
290    
291    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
292        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
293        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
294    
295    4.  There were a number of related bugs in the code for matching backrefences
296        caselessly in UTF-8 mode when codes for the characters concerned were
297        different numbers of bytes. For example, U+023A and U+2C65 are an upper
298        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
299        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
300        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
301        2-byte code at the end of the subject (it thought there wasn't enough data
302        left).
303    
304    5.  Comprehensive information about what went wrong is now returned by
305        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
306        as the output vector has at least 2 elements. The offset of the start of
307        the failing character and a reason code are placed in the vector.
308    
309    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
310        now returned is for the first byte of the failing character, instead of the
311        last byte inspected. This is an incompatible change, but I hope it is small
312        enough not to be a problem. It makes the returned offset consistent with
313        pcre_exec() and pcre_dfa_exec().
314    
315    7.  pcretest now gives a text phrase as well as the error number when
316        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
317        failure, the offset and reason code are output.
318    
319    8.  When \R was used with a maximizing quantifier it failed to skip backwards
320        over a \r\n pair if the subsequent match failed. Instead, it just skipped
321        back over a single character (\n). This seems wrong (because it treated the
322        two characters as a single entity when going forwards), conflicts with the
323        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
324        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
325        has been changed.
326    
327    9.  Some internal refactoring has changed the processing so that the handling
328        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
329        time (the PCRE_DOTALL option was changed this way some time ago: version
330        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
331        which was always a bit of a fudge. It also means that there is one less
332        argument for the match() function, which reduces its stack requirements
333        slightly. This change also fixes an incompatibility with Perl: the pattern
334        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
335    
336    10. More internal refactoring has drastically reduced the number of recursive
337        calls to match() for possessively repeated groups such as (abc)++ when
338        using pcre_exec().
339    
340    11. While implementing 10, a number of bugs in the handling of groups were
341        discovered and fixed:
342    
343        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
344        (a|)*(?1) gave a compile-time internal error.
345        ((a|)+)+  did not notice that the outer group could match an empty string.
346        (^a|^)+   was not marked as anchored.
347        (.*a|.*)+ was not marked as matching at start or after a newline.
348    
349    12. Yet more internal refactoring has removed another argument from the match()
350        function. Special calls to this function are now indicated by setting a
351        value in a variable in the "match data" data block.
352    
353    13. Be more explicit in pcre_study() instead of relying on "default" for
354        opcodes that mean there is no starting character; this means that when new
355        ones are added and accidentally left out of pcre_study(), testing should
356        pick them up.
357    
358    14. The -s option of pcretest has been documented for ages as being an old
359        synonym of -m (show memory usage). I have changed it to mean "force study
360        for every regex", that is, assume /S for every regex. This is similar to -i
361        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
362        using it. It makes it easier to run collections of tests with and without
363        study enabled, and thereby test pcre_study() more easily. All the standard
364        tests are now run with and without -s (but some patterns can be marked as
365        "never study" - see 20 below).
366    
367    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
368        restoration of the capturing data to the outer values was not happening
369        correctly.
370    
371    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
372        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
373        pattern had matched an empty string, and so incorrectly returned a no
374        match.
375    
376    17. There was optimizing code for the last branch of non-capturing parentheses,
377        and also for the obeyed branch of a conditional subexpression, which used
378        tail recursion to cut down on stack usage. Unfortunately, now that there is
379        the possibility of (*THEN) occurring in these branches, tail recursion is
380        no longer possible because the return has to be checked for (*THEN). These
381        two optimizations have therefore been removed. [But see 8.20/11 above.]
382    
383    18. If a pattern containing \R was studied, it was assumed that \R always
384        matched two bytes, thus causing the minimum subject length to be
385        incorrectly computed because \R can also match just one byte.
386    
387    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
388        was incorrectly computed.
389    
390    20. If /S is present twice on a test pattern in pcretest input, it now
391        *disables* studying, thereby overriding the use of -s on the command line
392        (see 14 above). This is necessary for one or two tests to keep the output
393        identical in both cases.
394    
395    21. When (*ACCEPT) was used in an assertion that matched an empty string and
396        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
397    
398    22. When an atomic group that contained a capturing parenthesis was
399        successfully matched, but the branch in which it appeared failed, the
400        capturing was not being forgotten if a higher numbered group was later
401        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
402        group 1 to "a", when in fact it should be unset. This applied to multi-
403        branched capturing and non-capturing groups, repeated or not, and also to
404        positive assertions (capturing in negative assertions does not happen
405        in PCRE) and also to nested atomic groups.
406    
407    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
408        subject after a captured substring, to make it easier to tell which of a
409        number of identical substrings has been captured.
410    
411    24. The way atomic groups are processed by pcre_exec() has been changed so that
412        if they are repeated, backtracking one repetition now resets captured
413        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
414        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
415        "aaa". Previously, it would have been "a". As part of this code
416        refactoring, the way recursive calls are handled has also been changed.
417    
418    25. If an assertion condition captured any substrings, they were not passed
419        back unless some other capturing happened later. For example, if
420        (?(?=(a))a) was matched against "a", no capturing was returned.
421    
422    26. When studying a pattern that contained subroutine calls or assertions,
423        the code for finding the minimum length of a possible match was handling
424        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
425        group 1 called group 2 while simultaneously a separate group 2 called group
426        1). A stack overflow occurred in this case. I have fixed this by limiting
427        the recursion depth to 10.
428    
429    27. Updated RunTest.bat in the distribution to the version supplied by Tom
430        Fortmann. This supports explicit test numbers on the command line, and has
431        argument validation and error reporting.
432    
433    28. An instance of \X with an unlimited repeat could fail if at any point the
434        first character it looked at was a mark character.
435    
436    29. Some minor code refactoring concerning Unicode properties and scripts
437        should reduce the stack requirement of match() slightly.
438    
439    30. Added the '=' option to pcretest to check the setting of unused capturing
440        slots at the end of the pattern, which are documented as being -1, but are
441        not included in the return count.
442    
443    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
444        compiled something random. Now it gives a compile-time error (as does
445        Perl).
446    
447    32. A *MARK encountered during the processing of a positive assertion is now
448        recorded and passed back (compatible with Perl).
449    
450    33. If --only-matching or --colour was set on a pcregrep call whose pattern
451        had alternative anchored branches, the search for a second match in a line
452        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
453        matched the line "0102" twice. The same bug affected patterns that started
454        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
455        twice.
456    
457    34. Previously, PCRE did not allow quantification of assertions. However, Perl
458        does, and because of capturing effects, quantifying parenthesized
459        assertions may at times be useful. Quantifiers are now allowed for
460        parenthesized assertions.
461    
462    35. A minor code tidy in pcre_compile() when checking options for \R usage.
463    
464    36. \g was being checked for fancy things in a character class, when it should
465        just be a literal "g".
466    
467    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
468        appearance of a nested POSIX class supersedes an apparent external class.
469        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
470        unescaped square brackets may also appear as part of class names. For
471        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
472        more like Perl. (But see 8.20/1 above.)
473    
474    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
475        was because it thought it was \N{name}, which is not supported).
476    
477    39. Add minix to OS list not supporting the -S option in pcretest.
478    
479    40. PCRE tries to detect cases of infinite recursion at compile time, but it
480        cannot analyze patterns in sufficient detail to catch mutual recursions
481        such as ((?1))((?2)). There is now a runtime test that gives an error if a
482        subgroup is called recursively as a subpattern for a second time at the
483        same position in the subject string. In previous releases this might have
484        been caught by the recursion limit, or it might have run out of stack.
485    
486    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
487        happen only once. PCRE was, however incorrectly giving a compile time error
488        "recursive call could loop indefinitely" because it cannot analyze the
489        pattern in sufficient detail. The compile time test no longer happens when
490        PCRE is compiling a conditional subpattern, but actual runaway loops are
491        now caught at runtime (see 40 above).
492    
493    42. It seems that Perl allows any characters other than a closing parenthesis
494        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
495        has been changed to be the same.
496    
497    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
498        as not to get warnings when autogen.sh is called. Also changed
499        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
500    
501    44. To help people who use pcregrep to scan files containing exceedingly long
502        lines, the following changes have been made:
503    
504        (a) The default value of the buffer size parameter has been increased from
505            8K to 20K. (The actual buffer used is three times this size.)
506    
507        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
508            PCRE is built.
509    
510        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
511            to be set at run time.
512    
513        (d) Numerical values in pcregrep options can be followed by K or M, for
514            example --buffer-size=50K.
515    
516        (e) If a line being scanned overflows pcregrep's buffer, an error is now
517            given and the return code is set to 2.
518    
519    45. Add a pointer to the latest mark to the callout data block.
520    
521    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
522        partial match of an empty string instead of no match. This was specific to
523        the use of ".".
524    
525    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
526        complete match instead of a partial match. This bug was dependent on both
527        the PCRE_UTF8 and PCRE_DOTALL options being set.
528    
529    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
530        starting byte set, because \b was not being ignored.
531    
532    
533    Version 8.12 15-Jan-2011
534    ------------------------
535    
536    1.  Fixed some typos in the markup of the man pages, and wrote a script that
537        checks for such things as part of the documentation building process.
538    
539    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
540        --match-limit and --recursion-limit options (added for 8.11). In
541        particular, this made one of the standard tests fail. (The integer value
542        went into the wrong half of a long int.)
543    
544    3.  If the --colour option was given to pcregrep with -v (invert match), it
545        did strange things, either producing crazy output, or crashing. It should,
546        of course, ignore a request for colour when reporting lines that do not
547        match.
548    
549    4.  Another pcregrep bug caused similar problems if --colour was specified with
550        -M (multiline) and the pattern match finished with a line ending.
551    
552    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
553        matched in multiline mode, the following line was shown as part of the
554        match. This seems wrong, so I have changed it.
555    
556    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
557        the check for further matches in the same line (so they could be coloured)
558        to overrun the end of the current line. If another match was found, it was
559        incorrectly shown (and then shown again when found in the next line).
560    
561    7.  If pcregrep was compiled under Windows, there was a reference to the
562        function pcregrep_exit() before it was defined. I am assuming this was
563        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
564        reported by a user. I've moved the definition above the reference.
565    
566    
567    Version 8.11 10-Dec-2010
568    ------------------------
569    
570    1.  (*THEN) was not working properly if there were untried alternatives prior
571        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
572        backtracked to try for "b" instead of moving to the next alternative branch
573        at the same level (in this case, to look for "c"). The Perl documentation
574        is clear that when (*THEN) is backtracked onto, it goes to the "next
575        alternative in the innermost enclosing group".
576    
577    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
578        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
579        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
580        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
581        (*THEN).
582    
583    3.  If \s appeared in a character class, it removed the VT character from
584        the class, even if it had been included by some previous item, for example
585        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
586        of \s, but is part of the POSIX "space" class.)
587    
588    4.  A partial match never returns an empty string (because you can always
589        match an empty string at the end of the subject); however the checking for
590        an empty string was starting at the "start of match" point. This has been
591        changed to the "earliest inspected character" point, because the returned
592        data for a partial match starts at this character. This means that, for
593        example, /(?<=abc)def/ gives a partial match for the subject "abc"
594        (previously it gave "no match").
595    
596    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
597        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
598        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
599        has an implication that the given string is incomplete (because a partial
600        match is preferred over a full match). For this reason, these items now
601        give a partial match in this situation. [Aside: previously, the one case
602        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
603        match rather than a full match, which was wrong by the old rules, but is
604        now correct.]
605    
606    6.  There was a bug in the handling of #-introduced comments, recognized when
607        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
608        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
609        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
610        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
611        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
612        places in pcre_compile().
613    
614    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
615        comments when looking ahead for named forward references to subpatterns,
616        the only newline sequence it recognized was NL. It now handles newlines
617        according to the set newline convention.
618    
619    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
620        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
621        cater for a lack of strerror(). These oversights have been fixed.
622    
623    9.  Added --match-limit and --recursion-limit to pcregrep.
624    
625    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
626    
627    11. When the -o option was used, pcregrep was setting a return code of 1, even
628        when matches were found, and --line-buffered was not being honoured.
629    
630    12. Added an optional parentheses number to the -o and --only-matching options
631        of pcregrep.
632    
633    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
634        can match an empty string. The code to do it in pcretest and pcredemo
635        needed fixing:
636    
637        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
638            only one byte after an empty string match just before CRLF (this case
639            just got forgotten; "any" and "anycrlf" were OK).
640    
641        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
642            mode when an empty string match preceded an ASCII character followed by
643            a non-ASCII character. (The code for advancing by one character rather
644            than one byte was nonsense.)
645    
646        (c) The pcredemo.c sample program did not have any code at all to handle
647            the cases when CRLF is a valid newline sequence.
648    
649    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
650        as a starting offset was within the subject string. There is now a new
651        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
652        negative or greater than the length of the string. In order to test this,
653        pcretest is extended to allow the setting of negative starting offsets.
654    
655    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
656        starting offset points to the beginning of a UTF-8 character was
657        unnecessarily clumsy. I tidied it up.
658    
659    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
660        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
661    
662    17. Nobody had reported that the --include_dir option, which was added in
663        release 7.7 should have been called --include-dir (hyphen, not underscore)
664        for compatibility with GNU grep. I have changed it to --include-dir, but
665        left --include_dir as an undocumented synonym, and the same for
666        --exclude-dir, though that is not available in GNU grep, at least as of
667        release 2.5.4.
668    
669    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
670        characters from a string of bytes) have been redefined so as not to use
671        loops, in order to improve performance in some environments. At the same
672        time, I abstracted some of the common code into auxiliary macros to save
673        repetition (this should not affect the compiled code).
674    
675    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
676        compile-time error is now given if \c is not followed by an ASCII
677        character, that is, a byte less than 128. (In EBCDIC mode, the code is
678        different, and any byte value is allowed.)
679    
680    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
681        START_OPTIMIZE option, which is now allowed at compile time - but just
682        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
683        to pcregrep and other applications that have no direct access to PCRE
684        options. The new /Y option in pcretest sets this option when calling
685        pcre_compile().
686    
687    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
688        back references. Groups containing recursive back references were forced to
689        be atomic by that change, but in the case of named groups, the amount of
690        memory required was incorrectly computed, leading to "Failed: internal
691        error: code overflow". This has been fixed.
692    
693    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
694        pcretest.c, to avoid build problems in some Borland environments.
695    
696    
697    Version 8.10 25-Jun-2010
698    ------------------------
699    
700    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
701        THEN.
702    
703    2.  (*ACCEPT) was not working when inside an atomic group.
704    
705    3.  Inside a character class, \B is treated as a literal by default, but
706        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
707        causes the error). The code is unchanged, but I tidied the documentation.
708    
709    4.  Inside a character class, PCRE always treated \R and \X as literals,
710        whereas Perl faults them if its -w option is set. I have changed PCRE so
711        that it faults them when PCRE_EXTRA is set.
712    
713    5.  Added support for \N, which always matches any character other than
714        newline. (It is the same as "." when PCRE_DOTALL is not set.)
715    
716    6.  When compiling pcregrep with newer versions of gcc which may have
717        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
718        declared with attribute warn_unused_result" were given. Just casting the
719        result to (void) does not stop the warnings; a more elaborate fudge is
720        needed. I've used a macro to implement this.
721    
722    7.  Minor change to pcretest.c to avoid a compiler warning.
723    
724    8.  Added four artifical Unicode properties to help with an option to make
725        \s etc use properties (see next item). The new properties are: Xan
726        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
727    
728    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
729        use Unicode properties. (*UCP) at the start of a pattern can be used to set
730        this option. Modified pcretest to add /W to test this facility. Added
731        REG_UCP to make it available via the POSIX interface.
732    
733    10. Added --line-buffered to pcregrep.
734    
735    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
736        studied, and the match started with a letter with a code point greater than
737        127 whose first byte was different to the first byte of the other case of
738        the letter, the other case of this starting letter was not recognized
739        (#976).
740    
741    12. If a pattern that was studied started with a repeated Unicode property
742        test, for example, \p{Nd}+, there was the theoretical possibility of
743        setting up an incorrect bitmap of starting bytes, but fortunately it could
744        not have actually happened in practice until change 8 above was made (it
745        added property types that matched character-matching opcodes).
746    
747    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
748        possible starting bytes for non-anchored patterns.
749    
750    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
751        \R, and also a number of cases that involve Unicode properties, both
752        explicit and implicit when PCRE_UCP is set.
753    
754    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
755        input, it could crash or give wrong results if characters with values
756        greater than 0xc0 were present in the subject string. (Detail: it assumed
757        UTF-8 input when processing these items.)
758    
759    16. Added a lot of (int) casts to avoid compiler warnings in systems where
760        size_t is 64-bit (#991).
761    
762    17. Added a check for running out of memory when PCRE is compiled with
763        --disable-stack-for-recursion (#990).
764    
765    18. If the last data line in a file for pcretest does not have a newline on
766        the end, a newline was missing in the output.
767    
768    19. The default pcre_chartables.c file recognizes only ASCII characters (values
769        less than 128) in its various bitmaps. However, there is a facility for
770        generating tables according to the current locale when PCRE is compiled. It
771        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
772        characters, are recognized by isspace() and therefore were getting set in
773        these tables, and indeed these tables seem to approximate to ISO 8859. This
774        caused a problem in UTF-8 mode when pcre_study() was used to create a list
775        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
776        which of course cannot start UTF-8 characters. I have changed the code so
777        that only real ASCII characters (less than 128) and the correct starting
778        bytes for UTF-8 encodings are set for characters greater than 127 when in
779        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
780        altogether.)
781    
782    20. Added the /T option to pcretest so as to be able to run tests with non-
783        standard character tables, thus making it possible to include the tests
784        used for 19 above in the standard set of tests.
785    
786    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
787        reference to a subpattern the other side of a comment that contains an
788        opening parenthesis caused either an internal compiling error, or a
789        reference to the wrong subpattern.
790    
791    
792    Version 8.02 19-Mar-2010
793    ------------------------
794    
795    1.  The Unicode data tables have been updated to Unicode 5.2.0.
796    
797    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
798        configured.
799    
800    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
801        original author of that file, following a query about its status.
802    
803    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
804        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
805    
806    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
807        quantifier applied to a forward-referencing subroutine call, could compile
808        incorrect code or give the error "internal error: previously-checked
809        referenced subpattern not found".
810    
811    6.  Both MS Visual Studio and Symbian OS have problems with initializing
812        variables to point to external functions. For these systems, therefore,
813        pcre_malloc etc. are now initialized to local functions that call the
814        relevant global functions.
815    
816    7.  There were two entries missing in the vectors called coptable and poptable
817        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
818        I've fixed the data, and added a kludgy way of testing at compile time that
819        the lengths are correct (equal to the number of opcodes).
820    
821    8.  Following on from 7, I added a similar kludge to check the length of the
822        eint vector in pcreposix.c.
823    
824    9.  Error texts for pcre_compile() are held as one long string to avoid too
825        much relocation at load time. To find a text, the string is searched,
826        counting zeros. There was no check for running off the end of the string,
827        which could happen if a new error number was added without updating the
828        string.
829    
830    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
831    
832    11. \K was not working if it appeared in an atomic group or in a group that
833        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
834        \K is "not well defined" if used in an assertion. PCRE now accepts it if
835        the assertion is positive, but not if it is negative.
836    
837    12. Change 11 fortuitously reduced the size of the stack frame used in the
838        "match()" function of pcre_exec.c by one pointer. Forthcoming
839        implementation of support for (*MARK) will need an extra pointer on the
840        stack; I have reserved it now, so that the stack frame size does not
841        decrease.
842    
843    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
844        item in branch that calls a recursion is a subroutine call - as in the
845        second branch in the above example - was incorrectly given the compile-
846        time error "recursive call could loop indefinitely" because pcre_compile()
847        was not correctly checking the subroutine for matching a non-empty string.
848    
849    14. The checks for overrunning compiling workspace could trigger after an
850        overrun had occurred. This is a "should never occur" error, but it can be
851        triggered by pathological patterns such as hundreds of nested parentheses.
852        The checks now trigger 100 bytes before the end of the workspace.
853    
854    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
855    
856    
857    Version 8.01 19-Jan-2010
858    ------------------------
859    
860    1.  If a pattern contained a conditional subpattern with only one branch (in
861        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
862        computed the wrong minimum data length (which is of course zero for such
863        subpatterns). This could cause incorrect "no match" results.
864    
865    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
866        the pattern is reset in the first branch, pcre_compile() failed with
867        "internal error: code overflow at offset...". This happened only when
868        the reset was to the original external option setting. (An optimization
869        abstracts leading options settings into an external setting, which was the
870        cause of this.)
871    
872    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
873        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
874        assertion pattern did not match (meaning that the assertion was true), it
875        was incorrectly treated as false if the SKIP had been reached during the
876        matching. This also applied to assertions used as conditions.
877    
878    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
879        assertion subpattern, including such a pattern used as a condition,
880        unpredictable results occurred, instead of the error return
881        PCRE_ERROR_DFA_UITEM.
882    
883    5.  The C++ GlobalReplace function was not working like Perl for the special
884        situation when an empty string is matched. It now does the fancy magic
885        stuff that is necessary.
886    
887    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
888        removed. (These were left over from very, very early versions of PCRE.)
889    
890    7.  Some cosmetic changes to the code to make life easier when compiling it
891        as part of something else:
892    
893        (a) Change DEBUG to PCRE_DEBUG.
894    
895        (b) In pcre_compile(), rename the member of the "branch_chain" structure
896            called "current" as "current_branch", to prevent a collision with the
897            Linux macro when compiled as a kernel module.
898    
899        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
900            prevent a collision with the Linux macro when compiled as a kernel
901            module.
902    
903    8.  In pcre_compile() there are some checks for integer overflows that used to
904        cast potentially large values to (double). This has been changed to that
905        when building, a check for int64_t is made, and if it is found, it is used
906        instead, thus avoiding the use of floating point arithmetic. (There is no
907        other use of FP in PCRE.) If int64_t is not found, the fallback is to
908        double.
909    
910    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
911        2005 (difference between two addresses compared to an unsigned value).
912    
913    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
914        custom one, because of the following reported problem in Windows:
915    
916          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
917              under Win32.
918          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
919              therefore missing the function definition.
920          - The compiler thus generates a "C" signature for the test function.
921          - The linker fails to find the "C" function.
922          - PCRE fails to configure if asked to do so against libbz2.
923    
924    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
925        messages were output:
926    
927          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
928          rerunning libtoolize, to keep the correct libtool macros in-tree.
929          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
930    
931        I have done both of these things.
932    
933    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
934        most of the time, it *can* run out if it is given a pattern that contains a
935        runaway infinite recursion. I updated the discussion in the pcrestack man
936        page.
937    
938    13. Now that we have gone to the x.xx style of version numbers, the minor
939        version may start with zero. Using 08 or 09 is a bad idea because users
940        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
941        interpreted as invalid octal numbers. I've updated the previous comment in
942        configure.ac, and also added a check that gives an error if 08 or 09 are
943        used.
944    
945    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
946        causing partial matching to fail when the end of the subject matched \W
947        in a UTF-8 pattern where \W was quantified with a minimum of 3.
948    
949    15. There were some discrepancies between the declarations in pcre_internal.h
950        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
951        their definitions. The declarations used "const uschar *" and the
952        definitions used USPTR. Even though USPTR is normally defined as "const
953        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
954        reported that: "This difference in casting confuses some C++ compilers, for
955        example, SunCC recognizes above declarations as different functions and
956        generates broken code for hbpcre." I have changed the declarations to use
957        USPTR.
958    
959    16. GNU libtool is named differently on some systems. The autogen.sh script now
960        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
961        (FreeBSD).
962    
963    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
964        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
965        comment: "Figure out how to create a longlong from a string: strtoll and
966        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
967        instance, but it only takes 2 args instead of 3!"
968    
969    18. A subtle bug concerned with back references has been fixed by a change of
970        specification, with a corresponding code fix. A pattern such as
971        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
972        refers, was giving matches when it shouldn't. For example, xa=xaaa would
973        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
974        same bug. Such groups have to be quantified to be useful, or contained
975        inside another quantified group. (If there's no repetition, the reference
976        can never match.) The problem arises because, having left the group and
977        moved on to the rest of the pattern, a later failure that backtracks into
978        the group uses the captured value from the final iteration of the group
979        rather than the correct earlier one. I have fixed this in PCRE by forcing
980        any group that contains a reference to itself to be an atomic group; that
981        is, there cannot be any backtracking into it once it has completed. This is
982        similar to recursive and subroutine calls.
983    
984    
985    Version 8.00 19-Oct-09
986    ----------------------
987    
988    1.  The table for translating pcre_compile() error codes into POSIX error codes
989        was out-of-date, and there was no check on the pcre_compile() error code
990        being within the table. This could lead to an OK return being given in
991        error.
992    
993    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
994        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
995        in a Windows environment.
996    
997    3.  The pcregrep --count option prints the count for each file even when it is
998        zero, as does GNU grep. However, pcregrep was also printing all files when
999        --files-with-matches was added. Now, when both options are given, it prints
1000        counts only for those files that have at least one match. (GNU grep just
1001        prints the file name in this circumstance, but including the count seems
1002        more useful - otherwise, why use --count?) Also ensured that the
1003        combination -clh just lists non-zero counts, with no names.
1004    
1005    4.  The long form of the pcregrep -F option was incorrectly implemented as
1006        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1007        but it seems right to fix it, and I didn't think it was worth preserving
1008        the old behaviour.
1009    
1010    5.  The command line items --regex=pattern and --regexp=pattern were not
1011        recognized by pcregrep, which required --regex pattern or --regexp pattern
1012        (with a space rather than an '='). The man page documented the '=' forms,
1013        which are compatible with GNU grep; these now work.
1014    
1015    6.  No libpcreposix.pc file was created for pkg-config; there was just
1016        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1017    
1018    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1019        when UCP support is not needed, by modifying the Python script that
1020        generates it from Unicode data files. This should not matter if the module
1021        is correctly used as a library, but I received one complaint about 50K of
1022        unwanted data. My guess is that the person linked everything into his
1023        program rather than using a library. Anyway, it does no harm.
1024    
1025    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1026        was a minimum greater than 1 for a wide character in a possessive
1027        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1028        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1029        character. Chaos in the form of incorrect output or a compiling loop could
1030        result.
1031    
1032    9.  The restrictions on what a pattern can contain when partial matching is
1033        requested for pcre_exec() have been removed. All patterns can now be
1034        partially matched by this function. In addition, if there are at least two
1035        slots in the offset vector, the offset of the earliest inspected character
1036        for the match and the offset of the end of the subject are set in them when
1037        PCRE_ERROR_PARTIAL is returned.
1038    
1039    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1040        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1041        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1042        and may be more useful for multi-segment matching.
1043    
1044    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1045        used to be given if ever the end of the subject was reached; now it is
1046        given only if matching could not proceed because another character was
1047        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1048        string "Z", which now yields "no match" instead of "partial match". In the
1049        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1050        final character ended with (*FAIL).
1051    
1052    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1053        if the pattern had a "must contain" character that was already found in the
1054        earlier partial match, unless partial matching was again requested. For
1055        example, with the pattern /dog.(body)?/, the "must contain" character is
1056        "g". If the first part-match was for the string "dog", restarting with
1057        "sbody" failed. This bug has been fixed.
1058    
1059    13. The string returned by pcre_dfa_exec() after a partial match has been
1060        changed so that it starts at the first inspected character rather than the
1061        first character of the match. This makes a difference only if the pattern
1062        starts with a lookbehind assertion or \b or \B (\K is not supported by
1063        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1064        matching functions compatible, and I think it's the right thing to do.
1065    
1066    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1067        so that the demonstration program is easily available in environments where
1068        PCRE has not been installed from source.
1069    
1070    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1071        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1072        library.
1073    
1074    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1075        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1076        is not the first non-POSIX option to be added. Clearly some people find
1077        these options useful.
1078    
1079    17. If a caller to the POSIX matching function regexec() passes a non-zero
1080        value for nmatch with a NULL value for pmatch, the value of
1081        nmatch is forced to zero.
1082    
1083    18. RunGrepTest did not have a test for the availability of the -u option of
1084        the diff command, as RunTest does. It now checks in the same way as
1085        RunTest, and also checks for the -b option.
1086    
1087    19. If an odd number of negated classes containing just a single character
1088        interposed, within parentheses, between a forward reference to a named
1089        subpattern and the definition of the subpattern, compilation crashed with
1090        an internal error, complaining that it could not find the referenced
1091        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1092        [The bug was that it was starting one character too far in when skipping
1093        over the character class, thus treating the ] as data rather than
1094        terminating the class. This meant it could skip too much.]
1095    
1096    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1097        /g option in pcretest when the pattern contains \K, which makes it possible
1098        to have an empty string match not at the start, even when the pattern is
1099        anchored. Updated pcretest and pcredemo to use this option.
1100    
1101    21. If the maximum number of capturing subpatterns in a recursion was greater
1102        than the maximum at the outer level, the higher number was returned, but
1103        with unset values at the outer level. The correct (outer level) value is
1104        now given.
1105    
1106    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1107        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1108        make it do so. The string so far is captured, making this feature
1109        compatible with Perl.
1110    
1111    23. The tests have been re-organized, adding tests 11 and 12, to make it
1112        possible to check the Perl 5.10 features against Perl 5.10.
1113    
1114    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1115        pattern matches a fixed length string. PCRE did not allow this; now it
1116        does. Neither allows recursion.
1117    
1118    25. I finally figured out how to implement a request to provide the minimum
1119        length of subject string that was needed in order to match a given pattern.
1120        (It was back references and recursion that I had previously got hung up
1121        on.) This code has now been added to pcre_study(); it finds a lower bound
1122        to the length of subject needed. It is not necessarily the greatest lower
1123        bound, but using it to avoid searching strings that are too short does give
1124        some useful speed-ups. The value is available to calling programs via
1125        pcre_fullinfo().
1126    
1127    26. While implementing 25, I discovered to my embarrassment that pcretest had
1128        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1129        study optimizations had never been tested with that matching function.
1130        Oops. What is worse, even when it was passed study data, there was a bug in
1131        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1132        were also very few tests of studied patterns with pcre_dfa_exec().
1133    
1134    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1135        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1136        on the other side of the coin, they are no longer allowed to have different
1137        names, because these cannot be distinguished in PCRE, and this has caused
1138        confusion. (This is a difference from Perl.)
1139    
1140    28. When duplicate subpattern names are present (necessarily with different
1141        numbers, as required by 27 above), and a test is made by name in a
1142        conditional pattern, either for a subpattern having been matched, or for
1143        recursion in such a pattern, all the associated numbered subpatterns are
1144        tested, and the overall condition is true if the condition is true for any
1145        one of them. This is the way Perl works, and is also more like the way
1146        testing by number works.
1147    
1148    
1149    Version 7.9 11-Apr-09
1150    ---------------------
1151    
1152    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1153        (pcretest), all targets were linked against these libraries. This included
1154        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1155        libraries. This caused unwanted dependencies to be created. This problem
1156        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1157        pcretest is linked with readline.
1158    
1159    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1160        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1161        moved outside it again, because FALSE and TRUE are already defined in AIX,
1162        but BOOL is not.
1163    
1164    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1165        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1166    
1167    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1168        hyphens) following file names and line numbers when outputting matching
1169        lines. This is not true; no spaces are inserted. I have also clarified the
1170        wording for the --colour (or --color) option.
1171    
1172    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1173        was not coloured; this is different to GNU grep, so I have changed it to be
1174        the same.
1175    
1176    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1177        each matching line was coloured. Now it goes on to look for further matches
1178        of any of the test patterns, which is the same behaviour as GNU grep.
1179    
1180    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1181        doesn't make sense to accept an empty string match in pcregrep, so I have
1182        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1183        seems to be how GNU grep behaves.
1184    
1185    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1186        start or after a newline", because the conditional assertion was not being
1187        correctly handled. The rule now is that both the assertion and what follows
1188        in the first alternative must satisfy the test.
1189    
1190    9.  If auto-callout was enabled in a pattern with a conditional group whose
1191        condition was an assertion, PCRE could crash during matching, both with
1192        pcre_exec() and pcre_dfa_exec().
1193    
1194    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1195        used for matching.
1196    
1197    11. Unicode property support in character classes was not working for
1198        characters (bytes) greater than 127 when not in UTF-8 mode.
1199    
1200    12. Added the -M command line option to pcretest.
1201    
1202    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1203    
1204    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1205    
1206    16. Added comments and documentation about mis-use of no_arg in the C++
1207        wrapper.
1208    
1209    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1210        from Martin Jerabek that uses macro names for all relevant character and
1211        string constants.
1212    
1213    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1214        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1215        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1216        these, but not everybody uses configure.
1217    
1218    19. A conditional group that had only one branch was not being correctly
1219        recognized as an item that could match an empty string. This meant that an
1220        enclosing group might also not be so recognized, causing infinite looping
1221        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1222        with the subject "ab", where knowledge that the repeated group can match
1223        nothing is needed in order to break the loop.
1224    
1225    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1226        exec(), but without supplying a callout function, matching went wrong.
1227    
1228    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1229        leak if the size of the offset vector was greater than 30. When the vector
1230        is smaller, the saved offsets during recursion go onto a local stack
1231        vector, but for larger vectors malloc() is used. It was failing to free
1232        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1233        error, in fact).
1234    
1235    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1236        heapframe that is used only when UTF-8 support is enabled. This caused no
1237        problem, but was untidy.
1238    
1239    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1240        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1241        included within another project.
1242    
1243    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1244        slightly modified by me:
1245    
1246          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1247              not building pcregrep.
1248    
1249          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1250              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1251    
1252    25. Forward references, both numeric and by name, in patterns that made use of
1253        duplicate group numbers, could behave incorrectly or give incorrect errors,
1254        because when scanning forward to find the reference group, PCRE was not
1255        taking into account the duplicate group numbers. A pattern such as
1256        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1257    
1258    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1259        the feature of a custom pointer more persuasive (as requested by a user).
1260    
1261    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1262        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1263        pre-defined.
1264    
1265    28. Added support for (*UTF8) at the start of a pattern.
1266    
1267    29. Arrange for flags added by the "release type" setting in CMake to be shown
1268        in the configuration summary.
1269    
1270    
1271    Version 7.8 05-Sep-08
1272    ---------------------
1273    
1274    1.  Replaced UCP searching code with optimized version as implemented for Ad
1275        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1276        stage table and inline lookup instead of a function, giving speed ups of 2
1277        to 5 times on some simple patterns that I tested. Permission was given to
1278        distribute the MultiStage2.py script that generates the tables (it's not in
1279        the tarball, but is in the Subversion repository).
1280    
1281    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1282        scripts.
1283    
1284    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1285        a group with a zero qualifier. The result of the study could be incorrect,
1286        or the function might crash, depending on the pattern.
1287    
1288    4.  Caseless matching was not working for non-ASCII characters in back
1289        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1290        It now works when Unicode Property Support is available.
1291    
1292    5.  In pcretest, an escape such as \x{de} in the data was always generating
1293        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1294        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1295        truncation.
1296    
1297    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1298    
1299    7.  Added two (int) casts to pcregrep when printing the difference of two
1300        pointers, in case they are 64-bit values.
1301    
1302    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1303        test 2 if it fails.
1304    
1305    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1306        and a #define of that name to empty if it is not externally set. This is to
1307        allow users of MSVC to set it if necessary.
1308    
1309    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1310        the convenience functions in the pcre_get.c source file.
1311    
1312    11. An option change at the start of a pattern that had top-level alternatives
1313        could cause overwriting and/or a crash. This command provoked a crash in
1314        some environments:
1315    
1316          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1317    
1318        This potential security problem was recorded as CVE-2008-2371.
1319    
1320    12. For a pattern where the match had to start at the beginning or immediately
1321        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1322        pcre_dfa_exec() could read past the end of the passed subject if there was
1323        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1324        pcretest so that it places the subject at the end of its malloc-ed buffer.
1325    
1326    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1327        exec() might read past the end of the data buffer in UTF-8 mode.
1328    
1329    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1330        the data contained the byte 0x85 as part of a UTF-8 character within its
1331        first line. This applied both to normal and DFA matching.
1332    
1333    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1334        /^[^d]*?$/8 failed to match "abc".
1335    
1336    16. Added a missing copyright notice to pcrecpp_internal.h.
1337    
1338    17. Make it more clear in the documentation that values returned from
1339        pcre_exec() in ovector are byte offsets, not character counts.
1340    
1341    18. Tidied a few places to stop certain compilers from issuing warnings.
1342    
1343    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1344        supplied by Stefan Weber. I made a further small update for 7.8 because
1345        there is a change of source arrangements: the pcre_searchfuncs.c module is
1346        replaced by pcre_ucd.c.
1347    
1348    
1349    Version 7.7 07-May-08
1350    ---------------------
1351    
1352    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1353        a string to a long long, pretend we don't even have a long long." This is
1354        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1355    
1356    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1357        pre-7.6 versions, which defined a global no_arg variable instead of putting
1358        it in the RE class. (See also #8 below.)
1359    
1360    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1361        Lopes.
1362    
1363    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1364    
1365        (1) The include/exclude patterns were being applied to the whole pathnames
1366            of files, instead of just to the final components.
1367    
1368        (2) If there was more than one level of directory, the subdirectories were
1369            skipped unless they satisfied the include/exclude conditions. This is
1370            inconsistent with GNU grep (and could even be seen as contrary to the
1371            pcregrep specification - which I improved to make it absolutely clear).
1372            The action now is always to scan all levels of directory, and just
1373            apply the include/exclude patterns to regular files.
1374    
1375    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1376        --exclude_dir in the tests to avoid scanning .svn directories.
1377    
1378    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1379        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1380        doesn't support NULs in patterns.
1381    
1382    7.  Added some missing "const"s to declarations of static tables in
1383        pcre_compile.c and pcre_dfa_exec.c.
1384    
1385    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1386        caused by fix #2  above. (Subsequently also a second patch to fix the
1387        first patch. And a third patch - this was a messy problem.)
1388    
1389    9.  Applied Craig's patch to remove the use of push_back().
1390    
1391    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1392        matching function regexec().
1393    
1394    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1395        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1396        references. PCRE supports relative numbers with this syntax (I don't think
1397        Oniguruma does).
1398    
1399    12. Previously, a group with a zero repeat such as (...){0} was completely
1400        omitted from the compiled regex. However, this means that if the group
1401        was called as a subroutine from elsewhere in the pattern, things went wrong
1402        (an internal error was given). Such groups are now left in the compiled
1403        pattern, with a new opcode that causes them to be skipped at execution
1404        time.
1405    
1406    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1407        to the way PCRE behaves:
1408    
1409        (a) A lone ] character is dis-allowed (Perl treats it as data).
1410    
1411        (b) A back reference to an unmatched subpattern matches an empty string
1412            (Perl fails the current match path).
1413    
1414        (c) A data ] in a character class must be notated as \] because if the
1415            first data character in a class is ], it defines an empty class. (In
1416            Perl it is not possible to have an empty class.) The empty class []
1417            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1418            The negative empty class [^] matches any one character, independently
1419            of the DOTALL setting.
1420    
1421    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1422        non-existent subpattern following a character class starting with ']' and
1423        containing () gave an internal compiling error instead of "reference to
1424        non-existent subpattern". Fortunately, when the pattern did exist, the
1425        compiled code was correct. (When scanning forwards to check for the
1426        existencd of the subpattern, it was treating the data ']' as terminating
1427        the class, so got the count wrong. When actually compiling, the reference
1428        was subsequently set up correctly.)
1429    
1430    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1431        it was being rejected as not supported by pcre_dfa_exec(), even though
1432        other assertions are supported. I have made pcre_dfa_exec() support
1433        (*FAIL).
1434    
1435    16. The implementation of 13c above involved the invention of a new opcode,
1436        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1437        cannot be changed at match time, I realized I could make a small
1438        improvement to matching performance by compiling OP_ALLANY instead of
1439        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1440        on the OP_ANY path.
1441    
1442    17. Compiling pcretest on Windows with readline support failed without the
1443        following two fixes: (1) Make the unistd.h include conditional on
1444        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1445    
1446    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1447        ncurses library to be included for pcretest when ReadLine support is
1448        requested, but also to allow for it to be overridden. This patch came from
1449        Daniel Bergström.
1450    
1451    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1452        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1453        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1454        spotting this.
1455    
1456    
1457    Version 7.6 28-Jan-08
1458    ---------------------
1459    
1460    1.  A character class containing a very large number of characters with
1461        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1462        overflow.
1463    
1464    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1465        HAVE_LONG_LONG is not defined.
1466    
1467    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1468        bring them up to date and include new features. This patch includes:
1469    
1470        - Fixed PH's badly added libz and libbz2 support.
1471        - Fixed a problem with static linking.
1472        - Added pcredemo. [But later removed - see 7 below.]
1473        - Fixed dftables problem and added an option.
1474        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1475            HAVE_LONG_LONG.
1476        - Added readline support for pcretest.
1477        - Added an listing of the option settings after cmake has run.
1478    
1479    4.  A user submitted a patch to Makefile that makes it easy to create
1480        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1481        Makefile.am that cause it to include this special target, without
1482        affecting anything else. Note that the same mingw target plus all
1483        the other distribution libraries and programs are now supported
1484        when configuring with CMake (see 6 below) instead of with
1485        Configure/Make.
1486    
1487    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1488        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1489        exported in the Windows port". It has not yet been confirmed that the patch
1490        solves the problem, but it does no harm.
1491    
1492    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1493        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1494        with CMake, and also correct the comment about stack recursion.
1495    
1496    7.  Remove the automatic building of pcredemo from the ./configure system and
1497        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1498        of a program that users should build themselves after PCRE is installed, so
1499        building it automatically is not really right. What is more, it gave
1500        trouble in some build environments.
1501    
1502    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1503    
1504    
1505    Version 7.5 10-Jan-08
1506    ---------------------
1507    
1508    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1509        values in parens when parsing an RE using the C++ wrapper."
1510    
1511    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1512        Characters greater than 255 were excluded from the class instead of being
1513        included.
1514    
1515    3.  The same bug as (2) above applied to negated POSIX classes such as
1516        [:^space:].
1517    
1518    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1519        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1520        I have changed it.
1521    
1522    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1523        first named subpattern) and a construct such as (?&a) would reference the
1524        first named subpattern whose name started with "a" (in other words, the
1525        length check was missing). Both these problems are fixed. "Subpattern name
1526        expected" is now given for (?&) (a zero-length name), and this patch also
1527        makes it give the same error for \k'' (previously it complained that that
1528        was a reference to a non-existent subpattern).
1529    
1530    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1531        this is right because (?- can be followed by option settings as well as by
1532        digits. I have, however, made the messages clearer.
1533    
1534    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1535        than the number used in the conditional) now cause a compile-time error.
1536        This is actually not compatible with Perl, which accepts such patterns, but
1537        treats the conditional as always being FALSE (as PCRE used to), but it
1538        seems to me that giving a diagnostic is better.
1539    
1540    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1541        and messages.
1542    
1543    9.  Fix two occurrences of "backslash" in comments that should have been
1544        "backspace".
1545    
1546    10. Remove two redundant lines of code that can never be obeyed (their function
1547        was moved elsewhere).
1548    
1549    11. The program that makes PCRE's Unicode character property table had a bug
1550        which caused it to generate incorrect table entries for sequences of
1551        characters that have the same character type, but are in different scripts.
1552        It amalgamated them into a single range, with the script of the first of
1553        them. In other words, some characters were in the wrong script. There were
1554        thirteen such cases, affecting characters in the following ranges:
1555    
1556          U+002b0 - U+002c1
1557          U+0060c - U+0060d
1558          U+0061e - U+00612
1559          U+0064b - U+0065e
1560          U+0074d - U+0076d
1561          U+01800 - U+01805
1562          U+01d00 - U+01d77
1563          U+01d9b - U+01dbf
1564          U+0200b - U+0200f
1565          U+030fc - U+030fe
1566          U+03260 - U+0327f
1567          U+0fb46 - U+0fbb1
1568          U+10450 - U+1049d
1569    
1570    12. The -o option (show only the matching part of a line) for pcregrep was not
1571        compatible with GNU grep in that, if there was more than one match in a
1572        line, it showed only the first of them. It now behaves in the same way as
1573        GNU grep.
1574    
1575    13. If the -o and -v options were combined for pcregrep, it printed a blank
1576        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1577        does the same. The return code can be used to tell if there were any
1578        non-matching lines.
1579    
1580    14. Added --file-offsets and --line-offsets to pcregrep.
1581    
1582    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1583        infinitely looping recursion. The bug was that positive lookaheads were not
1584        being skipped when checking for a possible empty match (negative lookaheads
1585        and both kinds of lookbehind were skipped).
1586    
1587    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1588        inclusion of <windows.h> to before rather than after the definition of
1589        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1590    
1591    17. Specifying a possessive quantifier with a specific limit for a Unicode
1592        character property caused pcre_compile() to compile bad code, which led at
1593        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1594        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1595        caused the error; without that there was no problem.
1596    
1597    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1598    
1599    19. Added --enable-pcretest-libreadline.
1600    
1601    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1602        RE::GlobalReplace(). As a result, the number of replacements returned was
1603        double what it should be. I removed one of the increments, but Craig sent a
1604        later patch that removed the other one (the right fix) and added unit tests
1605        that check the return values (which was not done before).
1606    
1607    21. Several CMake things:
1608    
1609        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1610            the names libpcre and libpcreposix, not just pcre and pcreposix.
1611    
1612        (2) The above change means that pcretest and pcregrep are now correctly
1613            linked with the newly-built libraries, not previously installed ones.
1614    
1615        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1616    
1617    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1618        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1619        UTF-8 newline character). The key issue is that the pattern starts .*;
1620        this means that the match must be either at the beginning, or after a
1621        newline. The bug was in the code for advancing after a failed match and
1622        checking that the new position followed a newline. It was not taking
1623        account of UTF-8 characters correctly.
1624    
1625    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1626        character classes. PCRE was not treating the sequence [:...:] as a
1627        character class unless the ... were all letters. Perl, however, seems to
1628        allow any characters between [: and :], though of course it rejects as
1629        unknown any "names" that contain non-letters, because all the known class
1630        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1631        for example, whereas PCRE did not - it did not recognize a POSIX character
1632        class. This seemed a bit dangerous, so the code has been changed to be
1633        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1634        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1635        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1636        Perl does, and where it didn't before.
1637    
1638    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1639        Windows environments %n is disabled by default.
1640    
1641    
1642    Version 7.4 21-Sep-07
1643    ---------------------
1644    
1645    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1646        means that a class such as [\s] counted as "explicit reference to CR or
1647        LF". That isn't really right - the whole point of the change was to try to
1648        help when there was an actual mention of one of the two characters. So now
1649        the change happens only if \r or \n (or a literal CR or LF) character is
1650        encountered.
1651    
1652    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1653        of both had grown to the point where there were only 3 bits left.
1654        Fortunately, there was spare space in the data structure, and so I have
1655        moved the internal flags into a new 16-bit field to free up more option
1656        bits.
1657    
1658    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1659        but did not set the internal JCHANGED flag - either of these is enough to
1660        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1661        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1662        start sets both bits.
1663    
1664    4.  Added options (at build time, compile time, exec time) to change \R from
1665        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1666    
1667    5.  doc/pcresyntax.html was missing from the distribution.
1668    
1669    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1670        compatibility, even though it is no longer used.
1671    
1672    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1673        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1674        windows.h file is present (where different names are used). [This was
1675        reversed later after testing - see 16 below.]
1676    
1677    8.  Changed all #include <config.h> to #include "config.h". There were also
1678        some further <pcre.h> cases that I changed to "pcre.h".
1679    
1680    9.  When pcregrep was used with the --colour option, it missed the line ending
1681        sequence off the lines that it output.
1682    
1683    10. It was pointed out to me that arrays of string pointers cause lots of
1684        relocations when a shared library is dynamically loaded. A technique of
1685        using a single long string with a table of offsets can drastically reduce
1686        these. I have refactored PCRE in four places to do this. The result is
1687        dramatic:
1688    
1689          Originally:                          290
1690          After changing UCP table:            187
1691          After changing error message table:   43
1692          After changing table of "verbs"       36
1693          After changing table of Posix names   22
1694    
1695        Thanks to the folks working on Gregex for glib for this insight.
1696    
1697    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1698        unicode-properties was also set.
1699    
1700    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1701    
1702    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1703        checked only for CRLF.
1704    
1705    14. Added casts to pcretest.c to avoid compiler warnings.
1706    
1707    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1708    
1709    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1710        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1711        entirely. This removes changes made in 7 above.
1712    
1713    17. The CMake files have been updated, and there is now more information about
1714        building with CMake in the NON-UNIX-USE document.
1715    
1716    
1717    Version 7.3 28-Aug-07
1718    ---------------------
1719    
1720     1. In the rejigging of the build system that eventually resulted in 7.1, the
1721        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1722        brackets there is not right, since it causes compilers to look for an
1723        installed pcre.h, not the version that is in the source that is being
1724        compiled (which of course may be different). I have changed it back to:
1725    
1726          #include "pcre.h"
1727    
1728        I have a vague recollection that the change was concerned with compiling in
1729        different directories, but in the new build system, that is taken care of
1730        by the VPATH setting the Makefile.
1731    
1732     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1733        when the subject happened to end in the byte 0x85 (e.g. if the last
1734        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1735        characters but of course it shouldn't be taken as a newline when it is part
1736        of another character. The bug was that, for an unlimited repeat of . in
1737        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1738        characters when looking for a newline.
1739    
1740     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1741    
1742     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1743        in debug output.
1744    
1745     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1746        long printing in the pcrecpp unittest when running under MinGW.
1747    
1748     6. ESC_K was left out of the EBCDIC table.
1749    
1750     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1751        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1752        limit also applies to "virtual nesting" when a pattern is recursive, and in
1753        this case 1000 isn't so big. I have been able to remove this limit at the
1754        expense of backing off one optimization in certain circumstances. Normally,
1755        when pcre_exec() would call its internal match() function recursively and
1756        immediately return the result unconditionally, it uses a "tail recursion"
1757        feature to save stack. However, when a subpattern that can match an empty
1758        string has an unlimited repetition quantifier, it no longer makes this
1759        optimization. That gives it a stack frame in which to save the data for
1760        checking that an empty string has been matched. Previously this was taken
1761        from the 1000-entry workspace that had been reserved. So now there is no
1762        explicit limit, but more stack is used.
1763    
1764     8. Applied Daniel's patches to solve problems with the import/export magic
1765        syntax that is required for Windows, and which was going wrong for the
1766        pcreposix and pcrecpp parts of the library. These were overlooked when this
1767        problem was solved for the main library.
1768    
1769     9. There were some crude static tests to avoid integer overflow when computing
1770        the size of patterns that contain repeated groups with explicit upper
1771        limits. As the maximum quantifier is 65535, the maximum group length was
1772        set at 30,000 so that the product of these two numbers did not overflow a
1773        32-bit integer. However, it turns out that people want to use groups that
1774        are longer than 30,000 bytes (though not repeat them that many times).
1775        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1776        made it possible to implement the integer overflow checks in a much more
1777        dynamic way, which I have now done. The artificial limitation on group
1778        length has been removed - we now have only the limit on the total length of
1779        the compiled pattern, which depends on the LINK_SIZE setting.
1780    
1781    10. Fixed a bug in the documentation for get/copy named substring when
1782        duplicate names are permitted. If none of the named substrings are set, the
1783        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1784        empty string.
1785    
1786    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1787        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1788        because the ] is interpreted as the first data character and the
1789        terminating ] is not found. PCRE has been made compatible with Perl in this
1790        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1791        cause memory overwriting.
1792    
1793    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1794        string has been matched (to stop an infinite loop). It was not recognizing
1795        a conditional subpattern that could match an empty string if that
1796        subpattern was within another subpattern. For example, it looped when
1797        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1798        condition was not nested. This bug has been fixed.
1799    
1800    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1801        past the start of the subject in the presence of bytes with the top bit
1802        set, for example "\x8aBCD".
1803    
1804    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1805        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1806    
1807    14. Optimized (?!) to (*FAIL).
1808    
1809    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1810        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1811        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1812        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1813        does: it's just the validity check that is more restrictive.
1814    
1815    16. Inserted checks for integer overflows during escape sequence (backslash)
1816        processing, and also fixed erroneous offset values for syntax errors during
1817        backslash processing.
1818    
1819    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1820        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1821    
1822    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1823        caused an overrun.
1824    
1825    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1826        something other than just ASCII characters) inside a group that had an
1827        unlimited repeat caused a loop at compile time (while checking to see
1828        whether the group could match an empty string).
1829    
1830    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1831        [\P{Any}] did so. (Error in the code for printing property names.)
1832    
1833    21. An orphan \E inside a character class could cause a crash.
1834    
1835    22. A repeated capturing bracket such as (A)? could cause a wild memory
1836        reference during compilation.
1837    
1838    23. There are several functions in pcre_compile() that scan along a compiled
1839        expression for various reasons (e.g. to see if it's fixed length for look
1840        behind). There were bugs in these functions when a repeated \p or \P was
1841        present in the pattern. These operators have additional parameters compared
1842        with \d, etc, and these were not being taken into account when moving along
1843        the compiled data. Specifically:
1844    
1845        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1846            length.
1847    
1848        (b) An item such as \pL+ within a repeated group could cause crashes or
1849            loops.
1850    
1851        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1852            "reference to non-existent subpattern" error.
1853    
1854        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1855    
1856    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1857        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1858    
1859    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1860    
1861    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1862        character were causing crashes (broken optimization).
1863    
1864    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1865        \p or \P) caused a compile-time loop.
1866    
1867    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1868        break. For example, the unstudied pattern [\r\n]A does not match the string
1869        "\r\nA" because change 7.0/46 below moves the current point on by two
1870        characters after failing to match at the start. However, the pattern \nA
1871        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1872        the same is true. There doesn't seem any very clean way out of this, but
1873        what I have chosen to do makes the common cases work: PCRE now takes note
1874        of whether there can be an explicit match for \r or \n anywhere in the
1875        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1876        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1877        pattern has explicit CR or LF references.
1878    
1879    29. Added (*CR) etc for changing newline setting at start of pattern.
1880    
1881    
1882    Version 7.2 19-Jun-07
1883    ---------------------
1884    
1885     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1886        which is apparently normally available under Windows.
1887    
1888     2. Re-jig the pcregrep tests with different newline settings in an attempt
1889        to make them independent of the local environment's newline setting.
1890    
1891     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1892    
1893     4. Some of the "internals" tests were previously cut out when the link size
1894        was not 2, because the output contained actual offsets. The recent new
1895        "Z" feature of pcretest means that these can be cut out, making the tests
1896        usable with all link sizes.
1897    
1898     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1899        stack recursion. This gives a massive performance boost under BSD, but just
1900        a small improvement under Linux. However, it saves one field in the frame
1901        in all cases.
1902    
1903     6. Added more features from the forthcoming Perl 5.10:
1904    
1905        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1906            recursion call. It refers to the nth most recently opened parentheses.
1907    
1908        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1909            to be opened parentheses.
1910    
1911        (c) Conditions that refer to capturing parentheses can be specified
1912            relatively, for example, (?(-2)... or (?(+3)...
1913    
1914        (d) \K resets the start of the current match so that everything before
1915            is not part of it.
1916    
1917        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1918    
1919        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1920            reference syntax.
1921    
1922        (g) (?| introduces a group in which the numbering of parentheses in each
1923            alternative starts with the same number.
1924    
1925        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1926    
1927     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1928        PCRE_INFO_JCHANGED.
1929    
1930     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1931        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1932        for detecting groups that can match an empty string.
1933    
1934     9. A pattern with a very large number of alternatives (more than several
1935        hundred) was running out of internal workspace during the pre-compile
1936        phase, where pcre_compile() figures out how much memory will be needed. A
1937        bit of new cunning has reduced the workspace needed for groups with
1938        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1939        workspace instead of running out of the 4096 that are available.
1940    
1941    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1942    
1943    11. Applied patch from Google to remove an optimization that didn't quite work.
1944        The report of the bug said:
1945    
1946          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1947          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1948          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1949    
1950    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1951        it matched the wrong number of bytes.
1952    
1953    
1954    Version 7.1 24-Apr-07
1955    ---------------------
1956    
1957     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1958        that is more "standard", making use of automake and other Autotools. There
1959        is some re-arrangement of the files and adjustment of comments consequent
1960        on this.
1961    
1962     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1963        for recursive directory scanning broke on some systems because the files
1964        are not scanned in any specific order and on different systems the order
1965        was different. A call to "sort" has been inserted into RunGrepTest for the
1966        approprate test as a short-term fix. In the longer term there may be an
1967        alternative.
1968    
1969     3. I had an email from Eric Raymond about problems translating some of PCRE's
1970        man pages to HTML (despite the fact that I distribute HTML pages, some
1971        people do their own conversions for various reasons). The problems
1972        concerned the use of low-level troff macros .br and .in. I have therefore
1973        removed all such uses from the man pages (some were redundant, some could
1974        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1975        HTML has been updated to handle .nf/.fi and to complain if it encounters
1976        .br or .in.
1977    
1978     4. Updated comments in configure.ac that get placed in config.h.in and also
1979        arranged for config.h to be included in the distribution, with the name
1980        config.h.generic, for the benefit of those who have to compile without
1981        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1982    
1983     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1984        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1985        makevp.bat for the current PCRE, using the additional files
1986        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1987    
1988     6. A Windows user reported a minor discrepancy with test 2, which turned out
1989        to be caused by a trailing space on an input line that had got lost in his
1990        copy. The trailing space was an accident, so I've just removed it.
1991    
1992     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1993        that is needed.
1994    
1995     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1996        as "const" (a) because they are and (b) because it helps the PHP
1997        maintainers who have recently made a script to detect big data structures
1998        in the php code that should be moved to the .rodata section. I remembered
1999        to update Builducptable as well, so it won't revert if ucptable.h is ever
2000        re-created.
2001    
2002     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2003        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2004        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2005        support is not required. This saves 1.5-2K of code, which is important in
2006        some applications.
2007    
2008        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2009        so as not to refer to the tables, even though these functions will never be
2010        called when UTF-8 support is disabled. Otherwise there are problems with a
2011        shared library.
2012    
2013    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2014    
2015        (a) It was defining its arguments as char * instead of void *.
2016    
2017        (b) It was assuming that all moves were upwards in memory; this was true
2018            a long time ago when I wrote it, but is no longer the case.
2019    
2020        The emulated memove() is provided for those environments that have neither
2021        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2022        is clearly not the case, as these two bugs were recently reported.
2023    
2024    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2025        and Detrail to create the HTML documentation, the .txt form of the man
2026        pages, and it removes trailing spaces from listed files. It also creates
2027        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2028        case, it wraps all the #defines with #ifndefs. This script should be run
2029        before "make dist".
2030    
2031    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2032        with Unicode property support.
2033    
2034        (a) For a maximizing quantifier, if the two different cases of the
2035            character were of different lengths in their UTF-8 codings (there are
2036            some cases like this - I found 11), and the matching function had to
2037            back up over a mixture of the two cases, it incorrectly assumed they
2038            were both the same length.
2039    
2040        (b) When PCRE was configured to use the heap rather than the stack for
2041            recursion during matching, it was not correctly preserving the data for
2042            the other case of a UTF-8 character when checking ahead for a match
2043            while processing a minimizing repeat. If the check also involved
2044            matching a wide character, but failed, corruption could cause an
2045            erroneous result when trying to check for a repeat of the original
2046            character.
2047    
2048    13. Some tidying changes to the testing mechanism:
2049    
2050        (a) The RunTest script now detects the internal link size and whether there
2051            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2052            values substituted by "configure". (The RunGrepTest script already did
2053            this for UTF-8.) The configure.ac script no longer substitutes the
2054            relevant variables.
2055    
2056        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2057            with length and offset values. This means that the output is different
2058            for different internal link sizes. Test 2 is skipped for link sizes
2059            other than 2 because of this, bypassing the problem. Unfortunately,
2060            there was also a test in test 3 (the locale tests) that used /B and
2061            failed for link sizes other than 2. Rather than cut the whole test out,
2062            I have added a new /Z option to pcretest that replaces the length and
2063            offset values with spaces. This is now used to make test 3 independent
2064            of link size. (Test 2 will be tidied up later.)
2065    
2066    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2067        segmentation fault instead of returning the appropriate error message.
2068    
2069    15. In multiline mode when the newline sequence was set to "any", the pattern
2070        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2071        This doesn't seem right; it now treats the CRLF combination as the line
2072        ending, and so does not match in that case. It's only a pattern such as ^$
2073        that would hit this one: something like ^ABC$ would have failed after \r
2074        and then tried again after \r\n.
2075    
2076    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2077        in an attempt to make files that differ only in their line terminators
2078        compare equal. This works on Linux.
2079    
2080    17. Under certain error circumstances pcregrep might try to free random memory
2081        as it exited. This is now fixed, thanks to valgrind.
2082    
2083    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2084        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2085        was because its rules for how to advance for /g after matching an empty
2086        string at the end of a line did not allow for this case. They now check for
2087        it specially.
2088    
2089    20. pcretest is supposed to handle patterns and data of any length, by
2090        extending its buffers when necessary. It was getting this wrong when the
2091        buffer for a data line had to be extended.
2092    
2093    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2094        CRLF as a newline sequence.
2095    
2096    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2097        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2098        I have nevertheless tidied it up.
2099    
2100    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2101    
2102    24. Added a man page for pcre-config.
2103    
2104    
2105    Version 7.0 19-Dec-06
2106    ---------------------
2107    
2108     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2109        moving to gcc 4.1.1.
2110    
2111     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2112        sys/time.h, which is documented as needed for this function. It doesn't
2113        seem to matter on Linux, but it showed up on some releases of OS X.
2114    
2115     3. It seems that there are systems where bytes whose values are greater than
2116        127 match isprint() in the "C" locale. The "C" locale should be the
2117        default when a C program starts up. In most systems, only ASCII printing
2118        characters match isprint(). This difference caused the output from pcretest
2119        to vary, making some of the tests fail. I have changed pcretest so that:
2120    
2121        (a) When it is outputting text in the compiled version of a pattern, bytes
2122            other than 32-126 are always shown as hex escapes.
2123    
2124        (b) When it is outputting text that is a matched part of a subject string,
2125            it does the same, unless a different locale has been set for the match
2126            (using the /L modifier). In this case, it uses isprint() to decide.
2127    
2128     4. Fixed a major bug that caused incorrect computation of the amount of memory
2129        required for a compiled pattern when options that changed within the
2130        pattern affected the logic of the preliminary scan that determines the
2131        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2132        that the computed length was too small. The symptoms of this bug were
2133        either the PCRE error "internal error: code overflow" from pcre_compile(),
2134        or a glibc crash with a message such as "pcretest: free(): invalid next
2135        size (fast)". Examples of patterns that provoked this bug (shown in
2136        pcretest format) are:
2137    
2138          /(?-x: )/x
2139          /(?x)(?-x: \s*#\s*)/
2140          /((?i)[\x{c0}])/8
2141          /(?i:[\x{c0}])/8
2142    
2143        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2144        is now done differently.
2145    
2146     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2147        wrapper classes; (b) implement a new function in the C++ scanner that is
2148        more efficient than the old way of doing things because it avoids levels of
2149        recursion in the regex matching; (c) add a paragraph to the documentation
2150        for the FullMatch() function.
2151    
2152     6. The escape sequence \n was being treated as whatever was defined as
2153        "newline". Not only was this contrary to the documentation, which states
2154        that \n is character 10 (hex 0A), but it also went horribly wrong when
2155        "newline" was defined as CRLF. This has been fixed.
2156    
2157     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2158        was being set to -1 for the "end of line" case (supposedly a value that no
2159        character can have). Though this value is never used (the check for end of
2160        line is "zero bytes in current character"), it caused compiler complaints.
2161        I've changed it to 0xffffffff.
2162    
2163     8. In pcre_version.c, the version string was being built by a sequence of
2164        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2165        string (as it is for production releases) called a macro with an empty
2166        argument. The C standard says the result of this is undefined. The gcc
2167        compiler treats it as an empty string (which was what was wanted) but it is
2168        reported that Visual C gives an error. The source has been hacked around to
2169        avoid this problem.
2170    
2171     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2172        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2173        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2174        of them did).
2175    
2176    10. Originally, pcretest opened its input and output without "b"; then I was
2177        told that "b" was needed in some environments, so it was added for release
2178        5.0 to both the input and output. (It makes no difference on Unix-like
2179        systems.) Later I was told that it is wrong for the input on Windows. I've
2180        now abstracted the modes into two macros, to make it easier to fiddle with
2181        them, and removed "b" from the input mode under Windows.
2182    
2183    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2184    
2185    12. Added -help and --help to pcretest as an official way of being reminded
2186        of the options.
2187    
2188    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2189        and pcrecpp.cc because they annoy compilers at high warning levels.
2190    
2191    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2192    
2193    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2194        scripts are not C programs :-) and which was not noticed because it works
2195        on Linux.
2196    
2197    16. pcretest is supposed to handle any length of pattern and data line (as one
2198        line or as a continued sequence of lines) by extending its input buffer if
2199        necessary. This feature was broken for very long pattern lines, leading to
2200        a string of junk being passed to pcre_compile() if the pattern was longer
2201        than about 50K.
2202    
2203    17. I have done a major re-factoring of the way pcre_compile() computes the
2204        amount of memory needed for a compiled pattern. Previously, there was code
2205        that made a preliminary scan of the pattern in order to do this. That was
2206        OK when PCRE was new, but as the facilities have expanded, it has become
2207        harder and harder to keep it in step with the real compile phase, and there
2208        have been a number of bugs (see for example, 4 above). I have now found a
2209        cunning way of running the real compile function in a "fake" mode that
2210        enables it to compute how much memory it would need, while actually only
2211        ever using a few hundred bytes of working memory and without too many
2212        tests of the mode. This should make future maintenance and development
2213        easier. A side effect of this work is that the limit of 200 on the nesting
2214        depth of parentheses has been removed (though this was never a serious
2215        limitation, I suspect). However, there is a downside: pcre_compile() now
2216        runs more slowly than before (30% or more, depending on the pattern). I
2217        hope this isn't a big issue. There is no effect on runtime performance.
2218    
2219    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2220        newline (only possible for the last line of a file) and it was a
2221        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2222    
2223    19. Added additional timing features to pcretest. (1) The -tm option now times
2224        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2225        separate command line item, by a number that specifies the number of
2226        repeats to use when timing. The default is 50000; this gives better
2227        precision, but takes uncomfortably long for very large patterns.
2228    
2229    20. Extended pcre_study() to be more clever in cases where a branch of a
2230        subpattern has no definite first character. For example, (a*|b*)[cd] would
2231        previously give no result from pcre_study(). Now it recognizes that the
2232        first character must be a, b, c, or d.
2233    
2234    21. There was an incorrect error "recursive call could loop indefinitely" if
2235        a subpattern (or the entire pattern) that was being tested for matching an
2236        empty string contained only one non-empty item after a nested subpattern.
2237        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2238        incorrectly, because the \d was being skipped in the check.
2239    
2240    22. The pcretest program now has a new pattern option /B and a command line
2241        option -b, which is equivalent to adding /B to every pattern. This causes
2242        it to show the compiled bytecode, without the additional information that
2243        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2244        is the same as /B/I).
2245    
2246    23. A new optimization is now able automatically to treat some sequences such
2247        as a*b as a*+b. More specifically, if something simple (such as a character
2248        or a simple class like \d) has an unlimited quantifier, and is followed by
2249        something that cannot possibly match the quantified thing, the quantifier
2250        is automatically "possessified".
2251    
2252    24. A recursive reference to a subpattern whose number was greater than 39
2253        went wrong under certain circumstances in UTF-8 mode. This bug could also
2254        have affected the operation of pcre_study().
2255    
2256    25. Realized that a little bit of performance could be had by replacing
2257        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2258    
2259    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2260    
2261    27. Possessive quantifiers such as a++ were previously implemented by turning
2262        them into atomic groups such as ($>a+). Now they have their own opcodes,
2263        which improves performance. This includes the automatically created ones
2264        from 23 above.
2265    
2266    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2267        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2268        the first matched character to be a colon. This applied both to named and
2269        numbered groups.
2270    
2271    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2272    
2273    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2274        building PCRE on Windows easier, so I have included it in the distribution.
2275    
2276    31. There is now a check in pcretest against a ridiculously large number being
2277        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2278        loop, the loop is abandoned.
2279    
2280    32. Forward references to subpatterns in conditions such as (?(2)...) where
2281        subpattern 2 is defined later cause pcre_compile() to search forwards in
2282        the pattern for the relevant set of parentheses. This search went wrong
2283        when there were unescaped parentheses in a character class, parentheses
2284        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2285    
2286    33. "Subroutine" calls and backreferences were previously restricted to
2287        referencing subpatterns earlier in the regex. This restriction has now
2288        been removed.
2289    
2290    34. Added a number of extra features that are going to be in Perl 5.10. On the
2291        whole, these are just syntactic alternatives for features that PCRE had
2292        previously implemented using the Python syntax or my own invention. The
2293        other formats are all retained for compatibility.
2294    
2295        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2296            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2297            also .NET compatible.
2298    
2299        (b) A recursion or subroutine call to a named group can now be defined as
2300            (?&name) as well as (?P>name).
2301    
2302        (c) A backreference to a named group can now be defined as \k<name> or
2303            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2304            5.10, are also .NET compatible.
2305    
2306        (d) A conditional reference to a named group can now use the syntax
2307            (?(<name>) or (?('name') as well as (?(name).
2308    
2309        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2310            groups (named and numbered) that are never evaluated inline, but can be
2311            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2312            is always false. There may be only one alternative in such a group.
2313    
2314        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2315            as the simple (?(R). The condition is true only if the most recent
2316            recursion is that of the given number or name. It does not search out
2317            through the entire recursion stack.
2318    
2319        (g) The escape \gN or \g{N} has been added, where N is a positive or
2320            negative number, specifying an absolute or relative reference.
2321    
2322    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2323        some "unreachable code" warnings.
2324    
2325    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2326        things, this adds five new scripts.
2327    
2328    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2329        There were also incompatibilities regarding the handling of \Q..\E inside
2330        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2331        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2332    
2333    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2334        matches an empty string, and forcibly breaks the loop. There were bugs in
2335        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2336        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2337        separate and independent bugs (that affected different cases) have been
2338        fixed.
2339    
2340    39. Refactored the code to abolish the use of different opcodes for small
2341        capturing bracket numbers. This is a tidy that I avoided doing when I
2342        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2343        The new approach is not only tidier, it makes it possible to reduce the
2344        memory needed to fix the previous bug (38).
2345    
2346    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2347        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2348        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2349        mode.
2350    
2351    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2352        report.
2353    
2354    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2355        copy construction and assignment in the C++ wrapper.
2356    
2357    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2358        couple of bugs that could have given wrong results in the "--newline=crlf"
2359        case.
2360    
2361    44. Added a number of casts and did some reorganization of signed/unsigned int
2362        variables following suggestions from Dair Grant. Also renamed the variable
2363        "this" as "item" because it is a C++ keyword.
2364    
2365    45. Arranged for dftables to add
2366    
2367          #include "pcre_internal.h"
2368    
2369        to pcre_chartables.c because without it, gcc 4.x may remove the array
2370        definition from the final binary if PCRE is built into a static library and
2371        dead code stripping is activated.
2372    
2373    46. For an unanchored pattern, if a match attempt fails at the start of a
2374        newline sequence, and the newline setting is CRLF or ANY, and the next two
2375        characters are CRLF, advance by two characters instead of one.
2376    
2377    
2378    Version 6.7 04-Jul-06
2379    ---------------------
2380    
2381     1. In order to handle tests when input lines are enormously long, pcretest has
2382        been re-factored so that it automatically extends its buffers when
2383        necessary. The code is crude, but this _is_ just a test program. The
2384        default size has been increased from 32K to 50K.
2385    
2386     2. The code in pcre_study() was using the value of the re argument before
2387        testing it for NULL. (Of course, in any sensible call of the function, it
2388        won't be NULL.)
2389    
2390     3. The memmove() emulation function in pcre_internal.h, which is used on
2391        systems that lack both memmove() and bcopy() - that is, hardly ever -
2392        was missing a "static" storage class specifier.
2393    
2394     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2395        containing an extended class (one that cannot be represented by a bitmap
2396        because it contains high-valued characters or Unicode property items, e.g.
2397        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2398        pattern, but PCRE should not loop if you do not (it no longer does).
2399        [Detail: two cases were found: (a) a repeated subpattern containing an
2400        extended class; (b) a recursive reference to a subpattern that followed a
2401        previous extended class. It wasn't skipping over the extended class
2402        correctly when UTF-8 mode was not set.]
2403    
2404     5. A negated single-character class was not being recognized as fixed-length
2405        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2406        compile error "lookbehind assertion is not fixed length".
2407    
2408     6. The RunPerlTest auxiliary script was showing an unexpected difference
2409        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2410        write a Perl script that can interpret lines of an input file either as
2411        byte characters or as UTF-8, which is what "perltest" was being required to
2412        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2413        can't do is switch easily at run time between having the "use utf8;" pragma
2414        or not. In the end, I fudged it by using the RunPerlTest script to insert
2415        "use utf8;" explicitly for the UTF-8 tests.
2416    
2417     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2418        the end of the subject string, contrary to the documentation and to what
2419        Perl does. This was true of both matching functions. Now it matches only at
2420        the start of the subject and immediately after *internal* newlines.
2421    
2422     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2423        a pointer to an int instead of a pointer to an unsigned long int. This
2424        caused problems on 64-bit systems.
2425    
2426     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2427        instance of the 'standard' template library not being so standard".
2428    
2429    10. There was no check on the number of named subpatterns nor the maximum
2430        length of a subpattern name. The product of these values is used to compute
2431        the size of the memory block for a compiled pattern. By supplying a very
2432        long subpattern name and a large number of named subpatterns, the size
2433        computation could be caused to overflow. This is now prevented by limiting
2434        the length of names to 32 characters, and the number of named subpatterns
2435        to 10,000.
2436    
2437    11. Subpatterns that are repeated with specific counts have to be replicated in
2438        the compiled pattern. The size of memory for this was computed from the
2439        length of the subpattern and the repeat count. The latter is limited to
2440        65535, but there was no limit on the former, meaning that integer overflow
2441        could in principle occur. The compiled length of a repeated subpattern is
2442        now limited to 30,000 bytes in order to prevent this.
2443    
2444    12. Added the optional facility to have named substrings with the same name.
2445    
2446    13. Added the ability to use a named substring as a condition, using the
2447        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2448        are numbers (not recommended). Forward references are permitted.
2449    
2450    14. Added forward references in named backreferences (if you see what I mean).
2451    
2452    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2453        pattern could run off the end of the subject. For example, the pattern
2454        "(?s)(.{1,5})"8 did this with the subject "ab".
2455    
2456    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2457        PCRE_CASELESS was set when matching characters that were quantified with ?
2458        or *.
2459    
2460    17. A character class other than a single negated character that had a minimum
2461        but no maximum quantifier - for example [ab]{6,} - was not handled
2462        correctly by pce_dfa_exec(). It would match only one character.
2463    
2464    18. A valid (though odd) pattern that looked like a POSIX character
2465        class but used an invalid character after [ (for example [[,abc,]]) caused
2466        pcre_compile() to give the error "Failed: internal error: code overflow" or
2467        in some cases to crash with a glibc free() error. This could even happen if
2468        the pattern terminated after [[ but there just happened to be a sequence of
2469        letters, a binary zero, and a closing ] in the memory that followed.
2470    
2471    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2472        over the years. Originally (before any Unicode support), just the bottom 8
2473        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2474        output from "man perlunicode" includes this:
2475    
2476          The regular expression compiler produces polymorphic opcodes.  That
2477          is, the pattern adapts to the data and automatically switches to
2478          the Unicode character scheme when presented with Unicode data--or
2479          instead uses a traditional byte scheme when presented with byte
2480          data.
2481    
2482        Sadly, a wide octal escape does not cause a switch, and in a string with
2483        no other multibyte characters, these octal escapes are treated as before.
2484        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2485        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2486        Unicode string.
2487    
2488        I have not perpetrated such confusion in PCRE. Up till now, it took just
2489        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2490        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2491        translate to the appropriate multibyte character.
2492    
2493    29. Applied some refactoring to reduce the number of warnings from Microsoft
2494        and Borland compilers. This has included removing the fudge introduced
2495        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2496        a warning about an unused variable.
2497    
2498    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2499        characters since release 4.0, because Perl (from release 5.004) does not.
2500        [Or at least, is documented not to: some releases seem to be in conflict
2501        with the documentation.] However, when a pattern was studied with
2502        pcre_study() and all its branches started with \s, PCRE still included VT
2503        as a possible starting character. Of course, this did no harm; it just
2504        caused an unnecessary match attempt.
2505    
2506    22. Removed a now-redundant internal flag bit that recorded the fact that case
2507        dependency changed within the pattern. This was once needed for "required
2508        byte" processing, but is no longer used. This recovers a now-scarce options
2509        bit. Also moved the least significant internal flag bit to the most-
2510        significant bit of the word, which was not previously used (hangover from
2511        the days when it was an int rather than a uint) to free up another bit for
2512        the future.
2513    
2514    23. Added support for CRLF line endings as well as CR and LF. As well as the
2515        default being selectable at build time, it can now be changed at runtime
2516        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2517        specify that it is scanning data with non-default line endings.
2518    
2519    24. Changed the definition of CXXLINK to make it agree with the definition of
2520        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2521    
2522    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2523        recursions. This makes a big different to stack usage for some patterns.
2524    
2525    26. If a subpattern containing a named recursion or subroutine reference such
2526        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2527        the space required for the compiled pattern went wrong and gave too small a
2528        value. Depending on the environment, this could lead to "Failed: internal
2529        error: code overflow at offset 49" or "glibc detected double free or
2530        corruption" errors.
2531    
2532    27. Applied patches from Google (a) to support the new newline modes and (b) to
2533        advance over multibyte UTF-8 characters in GlobalReplace.
2534    
2535    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2536        difference for some implementation of PCRE in some Windows version.
2537    
2538    29. Added some extra testing facilities to pcretest:
2539    
2540        \q<number>   in a data line sets the "match limit" value
2541        \Q<number>   in a data line sets the "match recursion limt" value
2542        -S <number>  sets the stack size, where <number> is in megabytes
2543    
2544        The -S option isn't available for Windows.
2545    
2546    
2547    Version 6.6 06-Feb-06
2548    ---------------------
2549    
2550     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2551        in pcreposix.h. I have copied the definition from pcre.h.
2552    
2553     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2554        because pcre.h is no longer a built file.
2555    
2556     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2557        not normally included in the compiled code.
2558    
2559    
2560    Version 6.5 01-Feb-06
2561    ---------------------
2562    
2563     1. When using the partial match feature with pcre_dfa_exec(), it was not
2564        anchoring the second and subsequent partial matches at the new starting
2565        point. This could lead to incorrect results. For example, with the pattern
2566        /1234/, partially matching against "123" and then "a4" gave a match.
2567    
2568     2. Changes to pcregrep:
2569    
2570        (a) All non-match returns from pcre_exec() were being treated as failures
2571            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2572            error message is output. Some extra information is given for the
2573            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2574            probably the only errors that are likely to be caused by users (by
2575            specifying a regex that has nested indefinite repeats, for instance).
2576            If there are more than 20 of these errors, pcregrep is abandoned.
2577    
2578        (b) A binary zero was treated as data while matching, but terminated the
2579            output line if it was written out. This has been fixed: binary zeroes
2580            are now no different to any other data bytes.
2581    
2582        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2583            used to set a locale for matching. The --locale=xxxx long option has
2584            been added (no short equivalent) to specify a locale explicitly on the
2585            pcregrep command, overriding the environment variables.
2586    
2587        (d) When -B was used with -n, some line numbers in the output were one less
2588            than they should have been.
2589    
2590        (e) Added the -o (--only-matching) option.
2591    
2592        (f) If -A or -C was used with -c (count only), some lines of context were
2593            accidentally printed for the final match.
2594    
2595        (g) Added the -H (--with-filename) option.
2596    
2597        (h) The combination of options -rh failed to suppress file names for files
2598            that were found from directory arguments.
2599    
2600        (i) Added the -D (--devices) and -d (--directories) options.
2601    
2602        (j) Added the -F (--fixed-strings) option.
2603    
2604        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2605    
2606        (l) Added the --colo(u)r option.
2607    
2608        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2609            is not present by default.
2610    
2611     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2612        items such as (?R) or (?1), when the recursion could match a number of
2613        alternatives. If it matched one of the alternatives, but subsequently,
2614        outside the recursion, there was a failure, the code tried to back up into
2615        the recursion. However, because of the way PCRE is implemented, this is not
2616        possible, and the result was an incorrect result from the match.
2617    
2618        In order to prevent this happening, the specification of recursion has
2619        been changed so that all such subpatterns are automatically treated as
2620        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2621    
2622     4. I had overlooked the fact that, in some locales, there are characters for
2623        which isalpha() is true but neither isupper() nor islower() are true. In
2624        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2625        and ordfeminine) are like this. This affected the treatment of \w and \W
2626        when they appeared in character classes, but not when they appeared outside
2627        a character class. The bit map for "word" characters is now created
2628        separately from the results of isalnum() instead of just taking it from the
2629        upper, lower, and digit maps. (Plus the underscore character, of course.)
2630    
2631     5. The above bug also affected the handling of POSIX character classes such as
2632        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2633        permanent tables. Instead, the bit maps for such a class were previously
2634        created as the appropriate unions of the upper, lower, and digit bitmaps.
2635        Now they are created by subtraction from the [[:word:]] class, which has
2636        its own bitmap.
2637    
2638     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2639        It is created by subtracting the vertical space characters (\x09, \x0a,
2640        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2641        subtraction was done in the overall bitmap for a character class, meaning
2642        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2643        be recognized. This bug has been fixed.
2644    
2645     7. Patches from the folks at Google:
2646    
2647          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2648          real life, but is still worth protecting against".
2649    
2650          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2651          regular expressions".
2652    
2653          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2654          have it.
2655    
2656          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2657          "configure" and the latter not, in order to fix a problem somebody had
2658          with compiling the Arg class on HP-UX.
2659    
2660          (e) Improve the error-handling of the C++ wrapper a little bit.
2661    
2662          (f) New tests for checking recursion limiting.
2663    
2664     8. The pcre_memmove() function, which is used only if the environment does not
2665        have a standard memmove() function (and is therefore rarely compiled),
2666        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2667        returning a result (though PCRE never actually uses the result).
2668    
2669     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2670        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2671        returned instead of calling malloc() with an overflowing number that would
2672        most likely cause subsequent chaos.
2673    
2674    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2675    
2676    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2677        with this option is matched, the nmatch and pmatch options of regexec() are
2678        ignored.
2679    
2680    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2681        provided in case anyone wants to the the POSIX interface with UTF-8
2682        strings.
2683    
2684    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2685        C++ linking (needed for some HP-UX environments).
2686    
2687    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2688        (unused parameter) and in the pcre_printint() function (omitted "default"
2689        switch label when the default is to do nothing).
2690    
2691    15. Added some code to make it possible, when PCRE is compiled as a C++
2692        library, to replace subject pointers for pcre_exec() with a smart pointer
2693        class, thus making it possible to process discontinuous strings.
2694    
2695    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2696        much the same function. They were added by different people who were trying
2697        to make PCRE easy to compile on non-Unix systems. It has been suggested
2698        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2699        for compiling on Windows systems. I have therefore replaced it with
2700        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2701        defaults to "extern" for C or "extern C" for C++, which works fine on
2702        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2703        SCOPE with something explicit in config.h. In addition:
2704    
2705        (a) pcreposix.h still had just "extern" instead of either of these macros;
2706            I have replaced it with PCRE_DATA_SCOPE.
2707    
2708        (b) Functions such as _pcre_xclass(), which are internal to the library,
2709            but external in the C sense, all had PCRE_EXPORT in their definitions.
2710            This is apparently wrong for the Windows case, so I have removed it.
2711            (It makes no difference on Unix-like systems.)
2712    
2713    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2714        of recursive calls to match(). This is different to MATCH_LIMIT because
2715        that limits the total number of calls to match(), not all of which increase
2716        the depth of recursion. Limiting the recursion depth limits the amount of
2717        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2718        when PCRE is compiled, and changed at run time. A patch from Google adds
2719        this functionality to the C++ interface.
2720    
2721    18. Changes to the handling of Unicode character properties:
2722    
2723        (a) Updated the table to Unicode 4.1.0.
2724    
2725        (b) Recognize characters that are not in the table as "Cn" (undefined).
2726    
2727        (c) I revised the way the table is implemented to a much improved format
2728            which includes recognition of ranges. It now supports the ranges that
2729            are defined in UnicodeData.txt, and it also amalgamates other
2730            characters into ranges. This has reduced the number of entries in the
2731            table from around 16,000 to around 3,000, thus reducing its size
2732            considerably. I realized I did not need to use a tree structure after
2733            all - a binary chop search is just as efficient. Having reduced the
2734            number of entries, I extended their size from 6 bytes to 8 bytes to
2735            allow for more data.
2736    
2737        (d) Added support for Unicode script names via properties such as \p{Han}.
2738    
2739    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2740        matching that character.
2741    
2742    20. When matching a repeated Unicode property with a minimum greater than zero,
2743        (for example \pL{2,}), PCRE could look past the end of the subject if it
2744        reached it while seeking the minimum number of characters. This could
2745        happen only if some of the characters were more than one byte long, because
2746        there is a check for at least the minimum number of bytes.
2747    
2748    21. Refactored the implementation of \p and \P so as to be more general, to
2749        allow for more different types of property in future. This has changed the
2750        compiled form incompatibly. Anybody with saved compiled patterns that use
2751        \p or \P will have to recompile them.
2752    
2753    22. Added "Any" and "L&" to the supported property types.
2754    
2755    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2756        but give a compile time error if the value is greater than 0xff.
2757    
2758    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2759        accidentally not being installed or uninstalled.
2760    
2761    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2762        made were to insert the current release number. This seemed silly, because
2763        it made things harder for people building PCRE on systems that don't run
2764        "configure". I have turned pcre.h into a distributed file, no longer built
2765        by "configure", with the version identification directly included. There is
2766        no longer a pcre.h.in file.
2767    
2768        However, this change necessitated a change to the pcre-config script as
2769        well. It is built from pcre-config.in, and one of the substitutions was the
2770        release number. I have updated configure.ac so that ./configure now finds
2771        the release number by grepping pcre.h.
2772    
2773    26. Added the ability to run the tests under valgrind.
2774    
2775    
2776    Version 6.4 05-Sep-05
2777    ---------------------
2778    
2779     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2780        "--" to be printed when multiple files were scanned, even when none of the
2781        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2782        consider it to be a bug, and have restored the previous behaviour.
2783    
2784     2. A couple of code tidies to get rid of compiler warnings.
2785    
2786     3. The pcretest program used to cheat by referring to symbols in the library
2787        whose names begin with _pcre_. These are internal symbols that are not
2788        really supposed to be visible externally, and in some environments it is
2789        possible to suppress them. The cheating is now confined to including
2790        certain files from the library's source, which is a bit cleaner.
2791    
2792     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2793        file's purpose clearer.
2794    
2795     5. Reorganized pcre_ucp_findchar().
2796    
2797    
2798    Version 6.3 15-Aug-05
2799    ---------------------
2800    
2801     1. The file libpcre.pc.in did not have general read permission in the tarball.
2802    
2803     2. There were some problems when building without C++ support:
2804    
2805        (a) If C++ support was not built, "make install" and "make test" still
2806            tried to test it.
2807    
2808        (b) There were problems when the value of CXX was explicitly set. Some
2809            changes have been made to try to fix these, and ...
2810    
2811        (c) --disable-cpp can now be used to explicitly disable C++ support.
2812    
2813        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2814            backslash in a target when C++ was disabled. This confuses some
2815            versions of "make", apparently. Using an intermediate variable solves
2816            this. (Same for CPP_LOBJ.)
2817    
2818     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2819        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2820        necessary on certain architectures.
2821    
2822     4. Added a setting of -export-symbols-regex to the link command to remove
2823        those symbols that are exported in the C sense, but actually are local
2824        within the library, and not documented. Their names all begin with
2825        "_pcre_". This is not a perfect job, because (a) we have to except some
2826        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2827        available (and never for static libraries). I have made a note to try to
2828        find a way round (a) in the future.
2829    
2830    
2831    Version 6.2 01-Aug-05
2832    ---------------------
2833    
2834     1. There was no test for integer overflow of quantifier values. A construction
2835        such as {1111111111111111} would give undefined results. What is worse, if
2836        a minimum quantifier for a parenthesized subpattern overflowed and became
2837        negative, the calculation of the memory size went wrong. This could have
2838        led to memory overwriting.
2839    
2840     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2841    
2842     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2843        operating environments where this matters.
2844    
2845     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2846        PCRE options from within the C++ wrapper.
2847    
2848     5. Named capturing subpatterns were not being correctly counted when a pattern
2849        was compiled. This caused two problems: (a) If there were more than 100
2850        such subpatterns, the calculation of the memory needed for the whole
2851        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2852        back references of the form \12, where the number was greater than 9, were
2853        not recognized as back references, even though there were sufficient
2854        previous subpatterns.
2855    
2856     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2857        versions of gcc, e.g. 2.95.4.
2858    
2859    
2860    Version 6.1 21-Jun-05
2861    ---------------------
2862    
2863     1. There was one reference to the variable "posix" in pcretest.c that was not
2864        surrounded by "#if !defined NOPOSIX".
2865    
2866     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2867        the cross-check on the old pcre_info() function, for the benefit of the
2868        cut-down version of PCRE that is currently imported into Exim.
2869    
2870     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2871        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2872        patterns that start (?i) but I don't think that matters. The use of (?i) is
2873        just an example; this all applies to the other options as well.
2874    
2875     4. Since libtool seems to echo the compile commands it is issuing, the output
2876        from "make" can be reduced a bit by putting "@" in front of each libtool
2877        compile command.
2878    
2879     5. Patch from the folks at Google for configure.in to be a bit more thorough
2880        in checking for a suitable C++ installation before trying to compile the
2881        C++ stuff. This should fix a reported problem when a compiler was present,
2882        but no suitable headers.
2883    
2884     6. The man pages all had just "PCRE" as their title. I have changed them to
2885        be the relevant file name. I have also arranged that these names are
2886        retained in the file doc/pcre.txt, which is a concatenation in text format
2887        of all the man pages except the little individual ones for each function.
2888    
2889     7. The NON-UNIX-USE file had not been updated for the different set of source
2890        files that come with release 6. I also added a few comments about the C++
2891        wrapper.
2892    
2893    
2894    Version 6.0 07-Jun-05
2895    ---------------------
2896    
2897     1. Some minor internal re-organization to help with my DFA experiments.
2898    
2899     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2900        didn't matter for the library itself when fully configured, but did matter
2901        when compiling without UCP support, or within Exim, where the ucp files are
2902        not imported.
2903    
2904     3. Refactoring of the library code to split up the various functions into
2905        different source modules. The addition of the new DFA matching code (see
2906        below) to a single monolithic source would have made it really too
2907        unwieldy, quite apart from causing all the code to be include in a
2908        statically linked application, when only some functions are used. This is
2909        relevant even without the DFA addition now that patterns can be compiled in
2910        one application and matched in another.
2911    
2912        The downside of splitting up is that there have to be some external
2913        functions and data tables that are used internally in different modules of
2914        the library but which are not part of the API. These have all had their
2915        names changed to start with "_pcre_" so that they are unlikely to clash
2916        with other external names.
2917    
2918     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2919        a different (DFA) algorithm. Although it is slower than the original
2920        function, it does have some advantages for certain types of matching
2921        problem.
2922    
2923     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2924        including restarting after a partial match.
2925    
2926     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2927        defined when compiling for Windows was sent to me. I have put it into the
2928        code, though I have no means of testing or verifying it.
2929    
2930     7. Added the pcre_refcount() auxiliary function.
2931    
2932     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2933        match before or at the first newline in the subject string. In pcretest,
2934        the /f option on a pattern can be used to set this.
2935    
2936     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2937        would behave wrongly. This has been present in PCRE since release 4.0.
2938    
2939    10. A number of changes to the pcregrep command:
2940    
2941        (a) Refactored how -x works; insert ^(...)$ instead of setting
2942            PCRE_ANCHORED and checking the length, in preparation for adding
2943            something similar for -w.
2944    
2945        (b) Added the -w (match as a word) option.
2946    
2947        (c) Refactored the way lines are read and buffered so as to have more
2948            than one at a time available.
2949    
2950        (d) Implemented a pcregrep test script.
2951    
2952        (e) Added the -M (multiline match) option. This allows patterns to match
2953            over several lines of the subject. The buffering ensures that at least
2954            8K, or the rest of the document (whichever is the shorter) is available
2955            for matching (and similarly the previous 8K for lookbehind assertions).
2956    
2957        (f) Changed the --help output so that it now says
2958    
2959              -w, --word-regex(p)
2960    
2961            instead of two lines, one with "regex" and the other with "regexp"
2962            because that confused at least one person since the short forms are the
2963            same. (This required a bit of code, as the output is generated
2964            automatically from a table. It wasn't just a text change.)
2965    
2966        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2967            option but starts with a hyphen. Could be a pattern or a path name
2968            starting with a hyphen, for instance.
2969    
2970        (h) "-" can be given as a file name to represent stdin.
2971    
2972        (i) When file names are being printed, "(standard input)" is used for
2973            the standard input, for compatibility with GNU grep. Previously
2974            "<stdin>" was used.
2975    
2976        (j) The option --label=xxx can be used to supply a name to be used for
2977            stdin when file names are being printed. There is no short form.
2978    
2979        (k) Re-factored the options decoding logic because we are going to add
2980            two more options that take data. Such options can now be given in four
2981            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2982    
2983        (l) Added the -A, -B, and -C options for requesting that lines of context
2984            around matches be printed.
2985    
2986        (m) Added the -L option to print the names of files that do not contain
2987            any matching lines, that is, the complement of -l.
2988    
2989        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2990            continue to scan other files.
2991    
2992        (o) The -s option was incorrectly implemented. For compatibility with other
2993            greps, it now suppresses the error message for a non-existent or non-
2994            accessible file (but not the return code). There is a new option called
2995            -q that suppresses the output of matching lines, which was what -s was
2996            previously doing.
2997    
2998        (p) Added --include and --exclude options to specify files for inclusion
2999            and exclusion when recursing.
3000    
3001    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3002        Hopefully, it now does.
3003    
3004    12. Missing cast in pcre_study().
3005    
3006    13. Added an "uninstall" target to the makefile.
3007    
3008    14. Replaced "extern" in the function prototypes in Makefile.in with
3009        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3010        world, but is set differently for Windows.
3011    
3012    15. Added a second compiling function called pcre_compile2(). The only
3013        difference is that it has an extra argument, which is a pointer to an
3014        integer error code. When there is a compile-time failure, this is set
3015        non-zero, in addition to the error test pointer being set to point to an
3016        error message. The new argument may be NULL if no error number is required
3017        (but then you may as well call pcre_compile(), which is now just a
3018        wrapper). This facility is provided because some applications need a
3019        numeric error indication, but it has also enabled me to tidy up the way
3020        compile-time errors are handled in the POSIX wrapper.
3021    
3022    16. Added VPATH=.libs to the makefile; this should help when building with one
3023        prefix path and installing with another. (Or so I'm told by someone who
3024        knows more about this stuff than I do.)
3025    
3026    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3027        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3028        match everything, including newlines. This is not POSIX-compatible, but
3029        somebody wanted the feature. From pcretest it can be activated by using
3030        both the P and the s flags.
3031    
3032    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3033    
3034    19. libpcre.pc was being incorrectly installed as executable.
3035    
3036    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3037        it now also looks for '\r' so that it will work unmodified on Windows.
3038    
3039    21. Added Google's contributed C++ wrapper to the distribution.
3040    
3041    22. Added some untidy missing memory free() calls in pcretest, to keep
3042        Electric Fence happy when testing.
3043    
3044    
3045    
3046    Version 5.0 13-Sep-04
3047    ---------------------
3048    
3049     1. Internal change: literal characters are no longer packed up into items
3050        containing multiple characters in a single byte-string. Each character
3051        is now matched using a separate opcode. However, there may be more than one
3052        byte in the character in UTF-8 mode.
3053    
3054     2. The pcre_callout_block structure has two new fields: pattern_position and
3055        next_item_length. These contain the offset in the pattern to the next match
3056        item, and its length, respectively.
3057    
3058     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3059        insertion of callouts before each pattern item. Added the /C option to
3060        pcretest to make use of this.
3061    
3062     4. On the advice of a Windows user, the lines
3063    
3064          #if defined(_WIN32) || defined(WIN32)
3065          _setmode( _fileno( stdout ), 0x8000 );
3066          #endif  /* defined(_WIN32) || defined(WIN32) */
3067    
3068        have been added to the source of pcretest. This apparently does useful
3069        magic in relation to line terminators.
3070    
3071     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3072        for the benefit of those environments where the "b" makes a difference.
3073    
3074     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3075        to know about it. I have put a hack into configure.in that adds in code
3076        to set GCC=yes if CC=icc. This seems to end up at a point in the
3077        generated configure script that is early enough to affect the setting of
3078        compiler options, which is what is needed, but I have no means of testing
3079        whether it really works. (The user who reported this had patched the
3080        generated configure script, which of course I cannot do.)
3081    
3082        LATER: After change 22 below (new libtool files), the configure script
3083        seems to know about icc (and also ecc). Therefore, I have commented out
3084        this hack in configure.in.
3085    
3086     7. Added support for pkg-config (2 patches were sent in).
3087    
3088     8. Negated POSIX character classes that used a combination of internal tables
3089        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3090        [[:^ascii]]. Typically, they would match almost any characters. The other
3091        POSIX classes were not broken in this way.
3092    
3093     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3094        to find the match, as PCRE was deluded into thinking that the match had to
3095        start at the start point or following a newline. The same bug applied to
3096        patterns with negative forward assertions or any backward assertions
3097        preceding ".*" at the start, unless the pattern required a fixed first
3098        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3099    
3100    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3101        starting at the last subject character, bytes beyond the end of the subject
3102        string were read.
3103    
3104    11. Renamed the variable "class" as "classbits" to make life easier for C++
3105        users. (Previously there was a macro definition, but it apparently wasn't
3106        enough.)
3107    
3108    12. Added the new field "tables" to the extra data so that tables can be passed
3109        in at exec time, or the internal tables can be re-selected. This allows
3110        a compiled regex to be saved and re-used at a later time by a different
3111        program that might have everything at different addresses.
3112    
3113    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3114        -R library as well as a -L library.
3115    
3116    14. The debugging options of pcretest (-d on the command line or D on a
3117        pattern) showed incorrect output for anything following an extended class
3118        that contained multibyte characters and which was followed by a quantifier.
3119    
3120    15. Added optional support for general category Unicode character properties
3121        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3122        support. It adds about 90K to the size of the library. The meanings of the
3123        inbuilt class escapes such as \d and \s have NOT been changed.
3124    
3125    16. Updated pcredemo.c to include calls to free() to release the memory for the
3126        compiled pattern.
3127    
3128    17. The generated file chartables.c was being created in the source directory
3129        instead of in the building directory. This caused the build to fail if the
3130        source directory was different from the building directory, and was
3131        read-only.
3132    
3133    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3134        file. No doubt somebody will tell me if they don't make sense... Also added
3135        Dan Mooney's comments about building on OpenVMS.
3136    
3137    19. Added support for partial matching via the PCRE_PARTIAL option for
3138        pcre_exec() and the \P data escape in pcretest.
3139    
3140    20. Extended pcretest with 3 new pattern features:
3141    
3142        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3143              write the compiled pattern to the file whose name is "rest-of-line".
3144              This is a straight binary dump of the data, with the saved pointer to
3145              the character tables forced to be NULL. The study data, if any, is
3146              written too. After writing, pcretest reads a new pattern.
3147    
3148        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3149              compiled pattern from the given file. There must not be any
3150              occurrences of "<" in the file name (pretty unlikely); if there are,
3151              pcretest will instead treat the initial "<" as a pattern delimiter.
3152              After reading in the pattern, pcretest goes on to read data lines as
3153              usual.
3154    
3155        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3156              and 16-bit fields in a compiled pattern, to simulate a pattern that
3157              was compiled on a host of opposite endianness.
3158    
3159    21. The pcre-exec() function can now cope with patterns that were compiled on
3160        hosts of opposite endianness, with this restriction:
3161    
3162          As for any compiled expression that is saved and used later, the tables
3163          pointer field cannot be preserved; the extra_data field in the arguments
3164          to pcre_exec() should be used to pass in a tables address if a value
3165          other than the default internal tables were used at compile time.
3166    
3167    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3168        now diagnosed as an error. Previously, most of the time, a negative number
3169        would have been treated as zero, but if in addition "ovector" was passed as
3170        NULL, a crash could occur.
3171    
3172    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3173        new versions from the libtool 1.5 distribution (the last one is a copy of
3174        a file called libtool.m4). This seems to have fixed the need to patch
3175        "configure" to support Darwin 1.3 (which I used to do). However, I still
3176        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3177        workstation).
3178    
3179    24. Changed the PCRE licence to be the more standard "BSD" licence.
3180    
3181    
3182    Version 4.5 01-Dec-03
3183    ---------------------
3184    
3185     1. There has been some re-arrangement of the code for the match() function so
3186        that it can be compiled in a version that does not call itself recursively.
3187        Instead, it keeps those local variables that need separate instances for
3188        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3189        needs to "recurse". Keeping track of where control must go is done by means
3190        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3191        hide most of the details from the main code, and operates only if
3192        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3193        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3194        operating.
3195    
3196        To make it easier for callers to provide specially tailored get/free
3197        functions for this usage, two new functions, pcre_stack_malloc, and
3198        pcre_stack_free, are used. They are always called in strict stacking order,
3199        and the size of block requested is always the same.
3200    
3201        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3202        PCRE has been compiled to use the stack or the heap for recursion. The
3203        -C option of pcretest uses this to show which version is compiled.
3204    
3205        A new data escape \S, is added to pcretest; it causes the amounts of store
3206        obtained and freed by both kinds of malloc/free at match time to be added
3207        to the output.
3208    
3209     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3210        what's available on my current Linux desktop machine.
3211    
3212     3. When matching a UTF-8 string, the test for a valid string at the start has
3213        been extended. If start_offset is not zero, PCRE now checks that it points
3214        to a byte that is the start of a UTF-8 character. If not, it returns
3215        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3216        this is necessary because there may be backward assertions in the pattern.
3217        When matching the same subject several times, it may save resources to use
3218        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3219    
3220     4. The code for checking the validity of UTF-8 strings has been tightened so
3221        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3222        containing "overlong sequences".
3223    
3224     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3225        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3226        should have been "&", but it just so happened that all the cases this let
3227        through by mistake were picked up later in the function.
3228    
3229     6. I had used a variable called "isblank" - this is a C99 function, causing
3230        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3231    
3232     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3233        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3234        so the version is shown; (c) stop "make test" echoing "./RunTest".
3235    
3236     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3237    
3238     9. The prototype for memmove() for systems that don't have it was using
3239        size_t, but the inclusion of the header that defines size_t was later. I've
3240        moved the #includes for the C headers earlier to avoid this.
3241    
3242    10. Added some adjustments to the code to make it easier to compiler on certain
3243        special systems:
3244    
3245          (a) Some "const" qualifiers were missing.
3246          (b) Added the macro EXPORT before all exported functions; by default this
3247              is defined to be empty.
3248          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3249              that it reads its output file name as an argument instead of writing
3250              to the standard output and assuming this can be redirected.
3251    
3252    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3253        class containing characters with values greater than 255, PCRE compilation
3254        went into a loop.
3255    
3256    12. A recursive reference to a subpattern that was within another subpattern
3257        that had a minimum quantifier of zero caused PCRE to crash. For example,
3258        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3259        recursion. If the recursively-called subpattern itself had a zero repeat,
3260        that was OK.
3261    
3262    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3263        buffer into which it was copied (for escape processing) was still set at
3264        1024, so long lines caused crashes.
3265    
3266    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3267        "internal error: code overflow...". This applied to any character class
3268        that was followed by a possessive quantifier.
3269    
3270    15. Modified the Makefile to add libpcre.la as a prerequisite for
3271        libpcreposix.la because I was told this is needed for a parallel build to
3272        work.
3273    
3274    16. If a pattern that contained .* following optional items at the start was
3275        studied, the wrong optimizing data was generated, leading to matching
3276        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3277        matching string must start with a or b or c. The correct conclusion for
3278        this pattern is that a match can start with any character.
3279    
3280    
3281    Version 4.4 13-Aug-03
3282    ---------------------
3283    
3284     1. In UTF-8 mode, a character class containing characters with values between
3285        127 and 255 was not handled correctly if the compiled pattern was studied.
3286        In fixing this, I have also improved the studying algorithm for such
3287        classes (slightly).
3288    
3289     2. Three internal functions had redundant arguments passed to them. Removal
3290        might give a very teeny performance improvement.
3291    
3292     3. Documentation bug: the value of the capture_top field in a callout is *one
3293        more than* the number of the hightest numbered captured substring.
3294    
3295     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3296        in incorrectly linking with a previously installed version. They now link
3297        explicitly with libpcre.la.
3298    
3299     5. configure.in no longer needs to recognize Cygwin specially.
3300    
3301     6. A problem in pcre.in for Windows platforms is fixed.
3302    
3303     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3304        pcretest, it used to include the size of the study block as part of its
3305        output. Unfortunately, the structure contains a field that has a different
3306        size on different hardware architectures. This meant that the tests that
3307        showed this size failed. As the block is currently always of a fixed size,
3308        this information isn't actually particularly useful in pcretest output, so
3309        I have just removed it.
3310    
3311     8. Three pre-processor statements accidentally did not start in column 1.
3312        Sadly, there are *still* compilers around that complain, even though
3313        standard C has not required this for well over a decade. Sigh.
3314    
3315     9. In pcretest, the code for checking callouts passed small integers in the
3316        callout_data field, which is a void * field. However, some picky compilers
3317        complained about the casts involved for this on 64-bit systems. Now
3318        pcretest passes the address of the small integer instead, which should get
3319        rid of the warnings.
3320    
3321    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3322        both compile and run time, and gives an error if an invalid UTF-8 sequence
3323        is found. There is a option for disabling this check in cases where the
3324        string is known to be correct and/or the maximum performance is wanted.
3325    
3326    11. In response to a bug report, I changed one line in Makefile.in from
3327    
3328            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3329        to
3330            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3331    
3332        to look similar to other lines, but I have no way of telling whether this
3333        is the right thing to do, as I do not use Windows. No doubt I'll get told
3334        if it's wrong...
3335    
3336    
3337    Version 4.3 21-May-03
3338    ---------------------
3339    
3340    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3341       Makefile.
3342    
3343    2. Some refactoring to improve the quality of the code:
3344    
3345       (i)   The utf8_table... variables are now declared "const".
3346    
3347       (ii)  The code for \cx, which used the "case flipping" table to upper case
3348             lower case letters, now just substracts 32. This is ASCII-specific,
3349             but the whole concept of \cx is ASCII-specific, so it seems
3350             reasonable.
3351    
3352       (iii) PCRE was using its character types table to recognize decimal and
3353             hexadecimal digits in the pattern. This is silly, because it handles
3354             only 0-9, a-f, and A-F, but the character types table is locale-
3355             specific, which means strange things might happen. A private
3356             table is now used for this - though it costs 256 bytes, a table is
3357             much faster than multiple explicit tests. Of course, the standard
3358             character types table is still used for matching digits in subject
3359             strings against \d.
3360    
3361       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3362             ending in _t are). So I've renamed it as ESC_tee.
3363    
3364    3. The first argument for regexec() in the POSIX wrapper should have been
3365       defined as "const".
3366    
3367    4. Changed pcretest to use malloc() for its buffers so that they can be
3368       Electric Fenced for debugging.
3369    
3370    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3371       to read one or more bytes before the start of the subject string. Often this
3372       had no effect on PCRE's behaviour, but in some circumstances it could
3373       provoke a segmentation fault.
3374    
3375    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3376       to try to read one or more bytes before the start of the subject string.
3377    
3378    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3379       UTF-8 support could misbehave in various ways if the subject string
3380       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3381       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3382       back over UTF-8 characters.)
3383    
3384    
3385    Version 4.2 14-Apr-03
3386    ---------------------
3387    
3388    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3389    
3390    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3391         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3392         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3393         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3394         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3395           and BUILD_EXEEXT
3396         Note: automatic setting of the BUILD variables is not yet working
3397         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3398           compile-time but not at link-time
3399         [LINK]: use for linking executables only
3400         make different versions for Windows and non-Windows
3401         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3402           libraries
3403         [LINK_FOR_BUILD]: new variable
3404         [OBJEXT]: use throughout
3405         [EXEEXT]: use throughout
3406         <winshared>: new target
3407         <wininstall>: new target
3408         <dftables.o>: use native compiler
3409         <dftables>: use native linker
3410         <install>: handle Windows platform correctly
3411         <clean>: ditto
3412         <check>: ditto
3413         copy DLL to top builddir before testing
3414    
3415       As part of these changes, -no-undefined was removed again. This was reported
3416       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3417       in any case.
3418    
3419    3. Some tidies to get rid of compiler warnings:
3420    
3421       . In the match_data structure, match_limit was an unsigned long int, whereas
3422         match_call_count was an int. I've made them both unsigned long ints.
3423    
3424       . In pcretest the fact that a const uschar * doesn't automatically cast to
3425         a void * provoked a warning.
3426    
3427       . Turning on some more compiler warnings threw up some "shadow" variables
3428         and a few more missing casts.
3429    
3430    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3431       option, a class that contained a single character with a value between 128
3432       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3433    
3434    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3435       option, a class that contained several characters, but with at least one
3436       whose value was between 128 and 255 caused PCRE to crash.
3437    
3438    
3439    Version 4.1 12-Mar-03
3440    ---------------------
3441    
3442    1. Compiling with gcc -pedantic found a couple of places where casts were
3443    needed, and a string in dftables.c that was longer than standard compilers are
3444    required to support.
3445    
3446    2. Compiling with Sun's compiler found a few more places where the code could
3447    be tidied up in order to avoid warnings.
3448    
3449    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3450    first of these names is deprecated in the latest Autoconf in favour of the name
3451    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3452    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3453    analogy I have changed it to CFLAGS_FOR_BUILD.
3454    
3455    4. Added -no-undefined to the linking command in the Makefile, because this is
3456    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3457    linking step for the pcreposix library.
3458    
3459    5. PCRE was failing to diagnose the case of two named groups with the same
3460    name.
3461    
3462    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3463    literal character that is needed in the subject for a match, and scans along to
3464    ensure that it is present before embarking on the full matching process. This
3465    saves time in cases of nested unlimited repeats that are never going to match.
3466    Problem: the scan can take a lot of time if the subject is very long (e.g.
3467    megabytes), thus penalizing straightforward matches. It is now done only if the
3468    amount of subject to be scanned is less than 1000 bytes.
3469    
3470    7. A lesser problem with the same optimization is that it was recording the
3471    first character of an anchored pattern as "needed", thus provoking a search
3472    right along the subject, even when the first match of the pattern was going to
3473    fail. The "needed" character is now not set for anchored patterns, unless it
3474    follows something in the pattern that is of non-fixed length. Thus, it still
3475    fulfils its original purpose of finding quick non-matches in cases of nested
3476    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3477    
3478    
3479    Version 4.0 17-Feb-03
3480    ---------------------
3481    
3482    1. If a comment in an extended regex that started immediately after a meta-item
3483    extended to the end of string, PCRE compiled incorrect data. This could lead to
3484    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3485    
3486    2. Moved to autoconf 2.53 and libtool 1.4.2.
3487    
3488    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3489    the special perltest8 script is no longer needed - all the tests can be run
3490    from a single perltest script.
3491    
3492    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3493    by \s. It has now been removed in PCRE. This means it isn't recognized as
3494    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3495    class [:space:] *does* include VT, thereby creating a mess.
3496    
3497    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3498    space and tab.
3499    
3500    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3501    its new features into the main test script, reducing the number of scripts.
3502    
3503    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3504    were backward compatible, and made the (?i) apply to the whole pattern, as if
3505    /i were given. Now it behaves more logically, and applies the option setting
3506    only to what follows. PCRE has been changed to follow suit. However, if it
3507    finds options settings right at the start of the pattern, it extracts them into
3508    the global options, as before. Thus, they show up in the info data.
3509    
3510    8. Added support for the \Q...\E escape sequence. Characters in between are
3511    treated as literals. This is slightly different from Perl in that $ and @ are
3512    also handled as literals inside the quotes. In Perl, they will cause variable
3513    interpolation. Note the following examples:
3514    
3515        Pattern            PCRE matches      Perl matches
3516    
3517        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3518        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3519        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3520    
3521    For compatibility with Perl, \Q...\E sequences are recognized inside character
3522    classes as well as outside them.
3523    
3524    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3525    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3526    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3527    signed/unsigned warnings.
3528    
3529    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3530    option for pcretest, so I've replaced it by a simple function that does just
3531    that job.
3532    
3533    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3534    "pcregrep -".
3535    
3536    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3537    Java package. This provides some syntactic sugar for simple cases of what my
3538    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3539    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3540    item, you can use this simplified notation. Note that only makes sense with
3541    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3542    greediness, whatever the setting of the PCRE_UNGREEDY option.
3543    
3544    13. A change of greediness default within a pattern was not taking effect at
3545    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3546    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3547    was abstracted outside.
3548    
3549    14. PCRE now supports the \G assertion. It is true when the current matching
3550    position is at the start point of the match. This differs from \A when the
3551    starting offset is non-zero. Used with the /g option of pcretest (or similar
3552    code), it works in the same way as it does for Perl's /g option. If all
3553    alternatives of a regex begin with \G, the expression is anchored to the start
3554    match position, and the "anchored" flag is set in the compiled expression.
3555    
3556    15. Some bugs concerning the handling of certain option changes within patterns
3557    have been fixed. These applied to options other than (?ims). For example,
3558    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3559    been the other way round. Some of this was related to change 7 above.
3560    
3561    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3562    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3563    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3564    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3565    
3566    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3567    mode. Unlike ".", it always matches newline, whatever the setting of
3568    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3569    assertions. Perl allows it, but it doesn't (in general) work because it can't
3570    calculate the length of the lookbehind. At least, that's the case for Perl
3571    5.8.0 - I've been told they are going to document that it doesn't work in
3572    future.
3573    
3574    18. Added an error diagnosis for escapes that PCRE does not support: these are
3575    \L, \l, \N, \P, \p, \U, \u, and \X.
3576    
3577    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3578    reading past the end of the pattern in cases such as /[abcd/.
3579    
3580    20. PCRE was getting more memory than necessary for patterns with classes that
3581    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3582    
3583    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3584    compiling PCRE for use with Virtual Pascal.
3585    
3586    22. Small fix to the Makefile to make it work properly if the build is done
3587    outside the source tree.
3588    
3589    23. Added a new extension: a condition to go with recursion. If a conditional
3590    subpattern starts with (?(R) the "true" branch is used if recursion has
3591    happened, whereas the "false" branch is used only at the top level.
3592    
3593    24. When there was a very long string of literal characters (over 255 bytes
3594    without UTF support, over 250 bytes with UTF support), the computation of how
3595    much memory was required could be incorrect, leading to segfaults or other
3596    strange effects.
3597    
3598    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3599    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3600    there was a subsequent back reference to those brackets. This meant that, for
3601    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3602    possible to check for precisely this case. All we can do is abandon the
3603    optimization if .* occurs inside capturing brackets when there are any back
3604    references whatsoever. (See below for a better fix that came later.)
3605    
3606    26. The handling of the optimization for finding the first character of a
3607    non-anchored pattern, and for finding a character that is required later in the
3608    match were failing in some cases. This didn't break the matching; it just
3609    failed to optimize when it could. The way this is done has been re-implemented.
3610    
3611    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3612    
3613    28. Added a new feature that provides some of the functionality that Perl
3614    provides with (?{...}). The facility is termed a "callout". The way it is done
3615    in PCRE is for the caller to provide an optional function, by setting
3616    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3617    global variable. By default it is unset, which disables all calling out. To get
3618    the function called, the regex must include (?C) at appropriate points. This
3619    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3620    This provides a means of identifying different callout points. When PCRE
3621    reaches such a point in the regex, if pcre_callout has been set, the external
3622    function is called. It is provided with data in a structure called
3623    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3624    matching continues; if it returns a non-zero value, the match at the current
3625    point fails. However, backtracking will occur if possible. [This was changed
3626    later and other features added - see item 49 below.]
3627    
3628    29. pcretest is upgraded to test the callout functionality. It provides a
3629    callout function that displays information. By default, it shows the start of
3630    the match and the current position in the text. There are some new data escapes
3631    to vary what happens:
3632    
3633        \C+         in addition, show current contents of captured substrings
3634        \C-         do not supply a callout function
3635        \C!n        return 1 when callout number n is reached
3636        \C!n!m      return 1 when callout number n is reached for the mth time
3637    
3638    30. If pcregrep was called with the -l option and just a single file name, it
3639    output "<stdin>" if a match was found, instead of the file name.
3640    
3641    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3642    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3643    pcre_exec(). This saves a malloc/free per call. The default value of
3644    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3645    when configuring.
3646    
3647    32. The default maximum size of a compiled pattern is 64K. There have been a
3648    few cases of people hitting this limit. The code now uses macros to handle the
3649    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3650    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3651    configuring. Tests 2 and 5 work only with 2-byte links because they output
3652    debugging information about compiled patterns.
3653    
3654    33. Internal code re-arrangements:
3655    
3656    (a) Moved the debugging function for printing out a compiled regex into
3657        its own source file (printint.c) and used #include to pull it into
3658        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3659        separate copies.
3660    
3661    (b) Defined the list of op-code names for debugging as a macro in
3662        internal.h so that it is next to the definition of the opcodes.
3663    
3664    (c) Defined a table of op-code lengths for simpler skipping along compiled
3665        code. This is again a macro in internal.h so that it is next to the
3666        definition of the opcodes.
3667    
3668    34. Added support for recursive calls to individual subpatterns, along the
3669    lines of Robin Houston's patch (but implemented somewhat differently).
3670    
3671    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3672    allow it to read and process whole directories in Win32. This code was
3673    contributed by Lionel Fourquaux; it has not been tested by me.
3674    
3675    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3676    used to name a group. Names consist of alphanumerics and underscores, and must
3677    be unique. Back references use the syntax (?P=name) and recursive calls use
3678    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3679    numbers. The function pcre_fullinfo() can be used after compilation to extract
3680    a name/number map. There are three relevant calls:
3681    
3682      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3683      PCRE_INFO_NAMECOUNT            yields the number of entries
3684      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3685    
3686    The map is a vector of fixed-size entries. The size of each entry depends on
3687    the length of the longest name used. The first two bytes of each entry are the
3688    group number, most significant byte first. There follows the corresponding
3689    name, zero terminated. The names are in alphabetical order.
3690    
3691    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3692    case instead of 255. Making it the same both with and without UTF-8 support
3693    means that the same test output works with both.
3694    
3695    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3696    calling malloc() with a zero argument.
3697    
3698    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3699    optimization. I've improved things by keeping a bitmap of backreferences with
3700    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3701    fact referenced, the optimization can be applied. It is unlikely that a
3702    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3703    the match to follow \n) will appear inside brackets with a number greater than
3704    31, but if it does, any back reference > 31 suppresses the optimization.
3705