ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 41 by nigel, Sat Feb 24 21:39:17 2007 UTC revision 836 by ph10, Wed Dec 28 17:16:11 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.30
5    ------------
7  Version 2.09 14-Sep-99  1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
8        name is defined in ctype.h.
10    2.  Fixed a bug in the code for calculating the fixed length of lookbehind
11        assertions.
13    3.  Removed the function pcre_info(), which has been obsolete and deprecated
14        since it was replaced by pcre_fullinfo() in February 2000.
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.
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.
26    6.  Add support for 16-bit character strings (a large amount of work involving
27        many changes and refactorings).
30    Version 8.21 12-Dec-2011
31    ------------------------
33    1.  Updating the JIT compiler.
35    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
36        are added as well.
38    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
39        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
40        calling _pcre_jit_exec. Some extra comments are added.
42    4.  (*MARK) settings inside atomic groups that do not contain any capturing
43        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
44        was introduced by change 18 for 8.20.
46    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
47        ECMA-262 standard.
49    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
50        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
51        This bug was probably introduced by change 9 of 8.13.
53    7.  While fixing 6 above, I noticed that a number of other items were being
54        incorrectly rejected as "not fixed length". This arose partly because newer
55        opcodes had not been added to the fixed-length checking code. I have (a)
56        corrected the bug and added tests for these items, and (b) arranged for an
57        error to occur if an unknown opcode is encountered while checking for fixed
58        length instead of just assuming "not fixed length". The items that were
59        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
60        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
61        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
63    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
64        being incorrectly compiled and would have given unpredicatble results.
66    9.  A possessively repeated subpattern with minimum repeat count greater than
67        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
68        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
69        the first (A) could occur when it should not.
71    10. Add a cast and remove a redundant test from the code.
73    11. JIT should use pcre_malloc/pcre_free for allocation.
75    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
76        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
77        is anything other than /usr, -L is still shown).
79    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
81    14. Perl does not support \N without a following name in a [] class; PCRE now
82        also gives an error.
84    15. If a forward reference was repeated with an upper limit of around 2000,
85        it caused the error "internal error: overran compiling workspace". The
86        maximum number of forward references (including repeats) was limited by the
87        internal workspace, and dependent on the LINK_SIZE. The code has been
88        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
89        the default depends on LINK_SIZE. There is a new upper limit (for safety)
90        of around 200,000 forward references. While doing this, I also speeded up
91        the filling in of repeated forward references.
93    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
94        incorrectly expecting the subject to contain another "a" after the start.
96    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
97        in the match, the SKIP should be ignored. This was not happening; instead
98        the SKIP was being treated as NOMATCH. For patterns such as
99        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
100        tested.
102    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
103        now much more compatible with Perl, in particular in cases where the result
104        is a non-match for a non-anchored pattern. For example, if
105        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
106        "m", where previously it did not return a name. A side effect of this
107        change is that for partial matches, the last encountered mark name is
108        returned, as for non matches. A number of tests that were previously not
109        Perl-compatible have been moved into the Perl-compatible test files. The
110        refactoring has had the pleasing side effect of removing one argument from
111        the match() function, thus reducing its stack requirements.
113    19. If the /S+ option was used in pcretest to study a pattern using JIT,
114        subsequent uses of /S (without +) incorrectly behaved like /S+.
116    21. Retrieve executable code size support for the JIT compiler and fixing
117        some warnings.
119    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
120        not work when the shorter character appeared right at the end of the
121        subject string.
123    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
124        systems.
126    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
127        output it when the /M option is used in pcretest.
129    25. The CheckMan script was not being included in the distribution. Also, added
130        an explicit "perl" to run Perl scripts from the PrepareRelease script
131        because this is reportedly needed in Windows.
133    26. If study data was being save in a file and studying had not found a set of
134        "starts with" bytes for the pattern, the data written to the file (though
135        never used) was taken from uninitialized memory and so caused valgrind to
136        complain.
138    27. Updated RunTest.bat as provided by Sheri Pierce.
140    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
142    29. Computation of memory usage for the table of capturing group names was
143        giving an unnecessarily large value.
146    Version 8.20 21-Oct-2011
147    ------------------------
149    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
150        a POSIX class. After further experiments with Perl, which convinced me that
151        Perl has bugs and confusions, a closing square bracket is no longer allowed
152        in a POSIX name. This bug also affected patterns with classes that started
153        with full stops.
155    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
156        captured substring, but while checking the failing first alternative,
157        substring 1 is temporarily captured. If the output vector supplied to
158        pcre_exec() was not big enough for this capture, the yield of the function
159        was still zero ("insufficient space for captured substrings"). This cannot
160        be totally fixed without adding another stack variable, which seems a lot
161        of expense for a edge case. However, I have improved the situation in cases
162        such as /(a)(b)x|abc/ matched against "abc", where the return code
163        indicates that fewer than the maximum number of slots in the ovector have
164        been set.
166    3.  Related to (2) above: when there are more back references in a pattern than
167        slots in the output vector, pcre_exec() uses temporary memory during
168        matching, and copies in the captures as far as possible afterwards. It was
169        using the entire output vector, but this conflicts with the specification
170        that only 2/3 is used for passing back captured substrings. Now it uses
171        only the first 2/3, for compatibility. This is, of course, another edge
172        case.
174    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
175        main code base, and can be used by building with --enable-jit. When this is
176        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
177        runtime --no-jit option is given.
179    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
180        ovector, the return from the function was zero, implying that there were
181        other matches that did not fit. The correct "exactly full" value is now
182        returned.
184    6.  If a subpattern that was called recursively or as a subroutine contained
185        (*PRUNE) or any other control that caused it to give a non-standard return,
186        invalid errors such as "Error -26 (nested recursion at the same subject
187        position)" or even infinite loops could occur.
189    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
190        computing the minimum length on reaching *ACCEPT, and so ended up with the
191        wrong value of 1 rather than 0. Further investigation indicates that
192        computing a minimum subject length in the presence of *ACCEPT is difficult
193        (think back references, subroutine calls), and so I have changed the code
194        so that no minimum is registered for a pattern that contains *ACCEPT.
196    8.  If (*THEN) was present in the first (true) branch of a conditional group,
197        it was not handled as intended. [But see 16 below.]
199    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
200        Sheri Pierce.
202    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
203        the first byte in a match must be "a".
205    11. Change 17 for 8.13 increased the recursion depth for patterns like
206        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
207        pattern contains any instances of (*THEN). If it does not, the old
208        optimizations are restored. It would be nice to do this on a per-group
209        basis, but at the moment that is not feasible.
211    12. In some environments, the output of pcretest -C is CRLF terminated. This
212        broke RunTest's code that checks for the link size. A single white space
213        character after the value is now allowed for.
215    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
216        For "fr", it uses the Windows-specific input and output files.
218    14. If (*THEN) appeared in a group that was called recursively or as a
219        subroutine, it did not work as intended. [But see next item.]
221    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
222        pattern fragments (but not containing any | characters). If A and B are
223        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
224        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
225        D. In other words, Perl considers parentheses that do not contain any |
226        characters to be part of a surrounding alternative, whereas PCRE was
227        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
228        differently. PCRE now behaves in the same way as Perl, except in the case
229        of subroutine/recursion calls such as (?1) which have in any case always
230        been different (but PCRE had them first :-).
232    16. Related to 15 above: Perl does not treat the | in a conditional group as
233        creating alternatives. Such a group is treated in the same way as an
234        ordinary group without any | characters when processing (*THEN). PCRE has
235        been changed to match Perl's behaviour.
237    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
238        RunGrepTest script failed.
240    18. Change 22 for version 13 caused atomic groups to use more stack. This is
241        inevitable for groups that contain captures, but it can lead to a lot of
242        stack use in large patterns. The old behaviour has been restored for atomic
243        groups that do not contain any capturing parentheses.
245    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
246        suppress the check for a minimum subject length at run time. (If it was
247        given to pcre_exec() or pcre_dfa_exec() it did work.)
249    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
250        fail to work when decoding hex characters in data strings in EBCDIC
251        environments.
253    21. It appears that in at least one Mac OS environment, the isxdigit() function
254        is implemented as a macro that evaluates to its argument more than once,
255        contravening the C 90 Standard (I haven't checked a later standard). There
256        was an instance in pcretest which caused it to go wrong when processing
257        \x{...} escapes in subject strings. The has been rewritten to avoid using
258        things like p++ in the argument of isxdigit().
261    Version 8.13 16-Aug-2011
262    ------------------------
264    1.  The Unicode data tables have been updated to Unicode 6.0.0.
266    2.  Two minor typos in pcre_internal.h have been fixed.
268    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
269        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
270        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
272    4.  There were a number of related bugs in the code for matching backrefences
273        caselessly in UTF-8 mode when codes for the characters concerned were
274        different numbers of bytes. For example, U+023A and U+2C65 are an upper
275        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
276        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
277        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
278        2-byte code at the end of the subject (it thought there wasn't enough data
279        left).
281    5.  Comprehensive information about what went wrong is now returned by
282        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
283        as the output vector has at least 2 elements. The offset of the start of
284        the failing character and a reason code are placed in the vector.
286    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
287        now returned is for the first byte of the failing character, instead of the
288        last byte inspected. This is an incompatible change, but I hope it is small
289        enough not to be a problem. It makes the returned offset consistent with
290        pcre_exec() and pcre_dfa_exec().
292    7.  pcretest now gives a text phrase as well as the error number when
293        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
294        failure, the offset and reason code are output.
296    8.  When \R was used with a maximizing quantifier it failed to skip backwards
297        over a \r\n pair if the subsequent match failed. Instead, it just skipped
298        back over a single character (\n). This seems wrong (because it treated the
299        two characters as a single entity when going forwards), conflicts with the
300        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
301        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
302        has been changed.
304    9.  Some internal refactoring has changed the processing so that the handling
305        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
306        time (the PCRE_DOTALL option was changed this way some time ago: version
307        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
308        which was always a bit of a fudge. It also means that there is one less
309        argument for the match() function, which reduces its stack requirements
310        slightly. This change also fixes an incompatibility with Perl: the pattern
311        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
313    10. More internal refactoring has drastically reduced the number of recursive
314        calls to match() for possessively repeated groups such as (abc)++ when
315        using pcre_exec().
317    11. While implementing 10, a number of bugs in the handling of groups were
318        discovered and fixed:
320        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
321        (a|)*(?1) gave a compile-time internal error.
322        ((a|)+)+  did not notice that the outer group could match an empty string.
323        (^a|^)+   was not marked as anchored.
324        (.*a|.*)+ was not marked as matching at start or after a newline.
326    12. Yet more internal refactoring has removed another argument from the match()
327        function. Special calls to this function are now indicated by setting a
328        value in a variable in the "match data" data block.
330    13. Be more explicit in pcre_study() instead of relying on "default" for
331        opcodes that mean there is no starting character; this means that when new
332        ones are added and accidentally left out of pcre_study(), testing should
333        pick them up.
335    14. The -s option of pcretest has been documented for ages as being an old
336        synonym of -m (show memory usage). I have changed it to mean "force study
337        for every regex", that is, assume /S for every regex. This is similar to -i
338        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
339        using it. It makes it easier to run collections of tests with and without
340        study enabled, and thereby test pcre_study() more easily. All the standard
341        tests are now run with and without -s (but some patterns can be marked as
342        "never study" - see 20 below).
344    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
345        restoration of the capturing data to the outer values was not happening
346        correctly.
348    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
349        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
350        pattern had matched an empty string, and so incorrectly returned a no
351        match.
353    17. There was optimizing code for the last branch of non-capturing parentheses,
354        and also for the obeyed branch of a conditional subexpression, which used
355        tail recursion to cut down on stack usage. Unfortunately, now that there is
356        the possibility of (*THEN) occurring in these branches, tail recursion is
357        no longer possible because the return has to be checked for (*THEN). These
358        two optimizations have therefore been removed. [But see 8.20/11 above.]
360    18. If a pattern containing \R was studied, it was assumed that \R always
361        matched two bytes, thus causing the minimum subject length to be
362        incorrectly computed because \R can also match just one byte.
364    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
365        was incorrectly computed.
367    20. If /S is present twice on a test pattern in pcretest input, it now
368        *disables* studying, thereby overriding the use of -s on the command line
369        (see 14 above). This is necessary for one or two tests to keep the output
370        identical in both cases.
372    21. When (*ACCEPT) was used in an assertion that matched an empty string and
373        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
375    22. When an atomic group that contained a capturing parenthesis was
376        successfully matched, but the branch in which it appeared failed, the
377        capturing was not being forgotten if a higher numbered group was later
378        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
379        group 1 to "a", when in fact it should be unset. This applied to multi-
380        branched capturing and non-capturing groups, repeated or not, and also to
381        positive assertions (capturing in negative assertions does not happen
382        in PCRE) and also to nested atomic groups.
384    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
385        subject after a captured substring, to make it easier to tell which of a
386        number of identical substrings has been captured.
388    24. The way atomic groups are processed by pcre_exec() has been changed so that
389        if they are repeated, backtracking one repetition now resets captured
390        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
391        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
392        "aaa". Previously, it would have been "a". As part of this code
393        refactoring, the way recursive calls are handled has also been changed.
395    25. If an assertion condition captured any substrings, they were not passed
396        back unless some other capturing happened later. For example, if
397        (?(?=(a))a) was matched against "a", no capturing was returned.
399    26. When studying a pattern that contained subroutine calls or assertions,
400        the code for finding the minimum length of a possible match was handling
401        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
402        group 1 called group 2 while simultaneously a separate group 2 called group
403        1). A stack overflow occurred in this case. I have fixed this by limiting
404        the recursion depth to 10.
406    27. Updated RunTest.bat in the distribution to the version supplied by Tom
407        Fortmann. This supports explicit test numbers on the command line, and has
408        argument validation and error reporting.
410    28. An instance of \X with an unlimited repeat could fail if at any point the
411        first character it looked at was a mark character.
413    29. Some minor code refactoring concerning Unicode properties and scripts
414        should reduce the stack requirement of match() slightly.
416    30. Added the '=' option to pcretest to check the setting of unused capturing
417        slots at the end of the pattern, which are documented as being -1, but are
418        not included in the return count.
420    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
421        compiled something random. Now it gives a compile-time error (as does
422        Perl).
424    32. A *MARK encountered during the processing of a positive assertion is now
425        recorded and passed back (compatible with Perl).
427    33. If --only-matching or --colour was set on a pcregrep call whose pattern
428        had alternative anchored branches, the search for a second match in a line
429        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
430        matched the line "0102" twice. The same bug affected patterns that started
431        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
432        twice.
434    34. Previously, PCRE did not allow quantification of assertions. However, Perl
435        does, and because of capturing effects, quantifying parenthesized
436        assertions may at times be useful. Quantifiers are now allowed for
437        parenthesized assertions.
439    35. A minor code tidy in pcre_compile() when checking options for \R usage.
441    36. \g was being checked for fancy things in a character class, when it should
442        just be a literal "g".
444    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
445        appearance of a nested POSIX class supersedes an apparent external class.
446        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
447        unescaped square brackets may also appear as part of class names. For
448        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
449        more like Perl. (But see 8.20/1 above.)
451    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
452        was because it thought it was \N{name}, which is not supported).
454    39. Add minix to OS list not supporting the -S option in pcretest.
456    40. PCRE tries to detect cases of infinite recursion at compile time, but it
457        cannot analyze patterns in sufficient detail to catch mutual recursions
458        such as ((?1))((?2)). There is now a runtime test that gives an error if a
459        subgroup is called recursively as a subpattern for a second time at the
460        same position in the subject string. In previous releases this might have
461        been caught by the recursion limit, or it might have run out of stack.
463    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
464        happen only once. PCRE was, however incorrectly giving a compile time error
465        "recursive call could loop indefinitely" because it cannot analyze the
466        pattern in sufficient detail. The compile time test no longer happens when
467        PCRE is compiling a conditional subpattern, but actual runaway loops are
468        now caught at runtime (see 40 above).
470    42. It seems that Perl allows any characters other than a closing parenthesis
471        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
472        has been changed to be the same.
474    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
475        as not to get warnings when autogen.sh is called. Also changed
476        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
478    44. To help people who use pcregrep to scan files containing exceedingly long
479        lines, the following changes have been made:
481        (a) The default value of the buffer size parameter has been increased from
482            8K to 20K. (The actual buffer used is three times this size.)
484        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
485            PCRE is built.
487        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
488            to be set at run time.
490        (d) Numerical values in pcregrep options can be followed by K or M, for
491            example --buffer-size=50K.
493        (e) If a line being scanned overflows pcregrep's buffer, an error is now
494            given and the return code is set to 2.
496    45. Add a pointer to the latest mark to the callout data block.
498    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
499        partial match of an empty string instead of no match. This was specific to
500        the use of ".".
502    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
503        complete match instead of a partial match. This bug was dependent on both
504        the PCRE_UTF8 and PCRE_DOTALL options being set.
506    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
507        starting byte set, because \b was not being ignored.
510    Version 8.12 15-Jan-2011
511    ------------------------
513    1.  Fixed some typos in the markup of the man pages, and wrote a script that
514        checks for such things as part of the documentation building process.
516    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
517        --match-limit and --recursion-limit options (added for 8.11). In
518        particular, this made one of the standard tests fail. (The integer value
519        went into the wrong half of a long int.)
521    3.  If the --colour option was given to pcregrep with -v (invert match), it
522        did strange things, either producing crazy output, or crashing. It should,
523        of course, ignore a request for colour when reporting lines that do not
524        match.
526    4.  Another pcregrep bug caused similar problems if --colour was specified with
527        -M (multiline) and the pattern match finished with a line ending.
529    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
530        matched in multiline mode, the following line was shown as part of the
531        match. This seems wrong, so I have changed it.
533    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
534        the check for further matches in the same line (so they could be coloured)
535        to overrun the end of the current line. If another match was found, it was
536        incorrectly shown (and then shown again when found in the next line).
538    7.  If pcregrep was compiled under Windows, there was a reference to the
539        function pcregrep_exit() before it was defined. I am assuming this was
540        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
541        reported by a user. I've moved the definition above the reference.
544    Version 8.11 10-Dec-2010
545    ------------------------
547    1.  (*THEN) was not working properly if there were untried alternatives prior
548        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
549        backtracked to try for "b" instead of moving to the next alternative branch
550        at the same level (in this case, to look for "c"). The Perl documentation
551        is clear that when (*THEN) is backtracked onto, it goes to the "next
552        alternative in the innermost enclosing group".
554    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
555        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
556        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
557        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
558        (*THEN).
560    3.  If \s appeared in a character class, it removed the VT character from
561        the class, even if it had been included by some previous item, for example
562        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
563        of \s, but is part of the POSIX "space" class.)
565    4.  A partial match never returns an empty string (because you can always
566        match an empty string at the end of the subject); however the checking for
567        an empty string was starting at the "start of match" point. This has been
568        changed to the "earliest inspected character" point, because the returned
569        data for a partial match starts at this character. This means that, for
570        example, /(?<=abc)def/ gives a partial match for the subject "abc"
571        (previously it gave "no match").
573    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
574        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
575        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
576        has an implication that the given string is incomplete (because a partial
577        match is preferred over a full match). For this reason, these items now
578        give a partial match in this situation. [Aside: previously, the one case
579        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
580        match rather than a full match, which was wrong by the old rules, but is
581        now correct.]
583    6.  There was a bug in the handling of #-introduced comments, recognized when
584        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
585        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
586        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
587        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
588        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
589        places in pcre_compile().
591    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
592        comments when looking ahead for named forward references to subpatterns,
593        the only newline sequence it recognized was NL. It now handles newlines
594        according to the set newline convention.
596    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
597        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
598        cater for a lack of strerror(). These oversights have been fixed.
600    9.  Added --match-limit and --recursion-limit to pcregrep.
602    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
604    11. When the -o option was used, pcregrep was setting a return code of 1, even
605        when matches were found, and --line-buffered was not being honoured.
607    12. Added an optional parentheses number to the -o and --only-matching options
608        of pcregrep.
610    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
611        can match an empty string. The code to do it in pcretest and pcredemo
612        needed fixing:
614        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
615            only one byte after an empty string match just before CRLF (this case
616            just got forgotten; "any" and "anycrlf" were OK).
618        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
619            mode when an empty string match preceded an ASCII character followed by
620            a non-ASCII character. (The code for advancing by one character rather
621            than one byte was nonsense.)
623        (c) The pcredemo.c sample program did not have any code at all to handle
624            the cases when CRLF is a valid newline sequence.
626    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
627        as a starting offset was within the subject string. There is now a new
628        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
629        negative or greater than the length of the string. In order to test this,
630        pcretest is extended to allow the setting of negative starting offsets.
632    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
633        starting offset points to the beginning of a UTF-8 character was
634        unnecessarily clumsy. I tidied it up.
636    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
637        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
639    17. Nobody had reported that the --include_dir option, which was added in
640        release 7.7 should have been called --include-dir (hyphen, not underscore)
641        for compatibility with GNU grep. I have changed it to --include-dir, but
642        left --include_dir as an undocumented synonym, and the same for
643        --exclude-dir, though that is not available in GNU grep, at least as of
644        release 2.5.4.
646    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
647        characters from a string of bytes) have been redefined so as not to use
648        loops, in order to improve performance in some environments. At the same
649        time, I abstracted some of the common code into auxiliary macros to save
650        repetition (this should not affect the compiled code).
652    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
653        compile-time error is now given if \c is not followed by an ASCII
654        character, that is, a byte less than 128. (In EBCDIC mode, the code is
655        different, and any byte value is allowed.)
657    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
658        START_OPTIMIZE option, which is now allowed at compile time - but just
659        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
660        to pcregrep and other applications that have no direct access to PCRE
661        options. The new /Y option in pcretest sets this option when calling
662        pcre_compile().
664    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
665        back references. Groups containing recursive back references were forced to
666        be atomic by that change, but in the case of named groups, the amount of
667        memory required was incorrectly computed, leading to "Failed: internal
668        error: code overflow". This has been fixed.
670    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
671        pcretest.c, to avoid build problems in some Borland environments.
674    Version 8.10 25-Jun-2010
675    ------------------------
677    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
678        THEN.
680    2.  (*ACCEPT) was not working when inside an atomic group.
682    3.  Inside a character class, \B is treated as a literal by default, but
683        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
684        causes the error). The code is unchanged, but I tidied the documentation.
686    4.  Inside a character class, PCRE always treated \R and \X as literals,
687        whereas Perl faults them if its -w option is set. I have changed PCRE so
688        that it faults them when PCRE_EXTRA is set.
690    5.  Added support for \N, which always matches any character other than
691        newline. (It is the same as "." when PCRE_DOTALL is not set.)
693    6.  When compiling pcregrep with newer versions of gcc which may have
694        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
695        declared with attribute warn_unused_result" were given. Just casting the
696        result to (void) does not stop the warnings; a more elaborate fudge is
697        needed. I've used a macro to implement this.
699    7.  Minor change to pcretest.c to avoid a compiler warning.
701    8.  Added four artifical Unicode properties to help with an option to make
702        \s etc use properties (see next item). The new properties are: Xan
703        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
705    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
706        use Unicode properties. (*UCP) at the start of a pattern can be used to set
707        this option. Modified pcretest to add /W to test this facility. Added
708        REG_UCP to make it available via the POSIX interface.
710    10. Added --line-buffered to pcregrep.
712    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
713        studied, and the match started with a letter with a code point greater than
714        127 whose first byte was different to the first byte of the other case of
715        the letter, the other case of this starting letter was not recognized
716        (#976).
718    12. If a pattern that was studied started with a repeated Unicode property
719        test, for example, \p{Nd}+, there was the theoretical possibility of
720        setting up an incorrect bitmap of starting bytes, but fortunately it could
721        not have actually happened in practice until change 8 above was made (it
722        added property types that matched character-matching opcodes).
724    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
725        possible starting bytes for non-anchored patterns.
727    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
728        \R, and also a number of cases that involve Unicode properties, both
729        explicit and implicit when PCRE_UCP is set.
731    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
732        input, it could crash or give wrong results if characters with values
733        greater than 0xc0 were present in the subject string. (Detail: it assumed
734        UTF-8 input when processing these items.)
736    16. Added a lot of (int) casts to avoid compiler warnings in systems where
737        size_t is 64-bit (#991).
739    17. Added a check for running out of memory when PCRE is compiled with
740        --disable-stack-for-recursion (#990).
742    18. If the last data line in a file for pcretest does not have a newline on
743        the end, a newline was missing in the output.
745    19. The default pcre_chartables.c file recognizes only ASCII characters (values
746        less than 128) in its various bitmaps. However, there is a facility for
747        generating tables according to the current locale when PCRE is compiled. It
748        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
749        characters, are recognized by isspace() and therefore were getting set in
750        these tables, and indeed these tables seem to approximate to ISO 8859. This
751        caused a problem in UTF-8 mode when pcre_study() was used to create a list
752        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
753        which of course cannot start UTF-8 characters. I have changed the code so
754        that only real ASCII characters (less than 128) and the correct starting
755        bytes for UTF-8 encodings are set for characters greater than 127 when in
756        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
757        altogether.)
759    20. Added the /T option to pcretest so as to be able to run tests with non-
760        standard character tables, thus making it possible to include the tests
761        used for 19 above in the standard set of tests.
763    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
764        reference to a subpattern the other side of a comment that contains an
765        opening parenthesis caused either an internal compiling error, or a
766        reference to the wrong subpattern.
769    Version 8.02 19-Mar-2010
770    ------------------------
772    1.  The Unicode data tables have been updated to Unicode 5.2.0.
774    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
775        configured.
777    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
778        original author of that file, following a query about its status.
780    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
781        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
783    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
784        quantifier applied to a forward-referencing subroutine call, could compile
785        incorrect code or give the error "internal error: previously-checked
786        referenced subpattern not found".
788    6.  Both MS Visual Studio and Symbian OS have problems with initializing
789        variables to point to external functions. For these systems, therefore,
790        pcre_malloc etc. are now initialized to local functions that call the
791        relevant global functions.
793    7.  There were two entries missing in the vectors called coptable and poptable
794        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
795        I've fixed the data, and added a kludgy way of testing at compile time that
796        the lengths are correct (equal to the number of opcodes).
798    8.  Following on from 7, I added a similar kludge to check the length of the
799        eint vector in pcreposix.c.
801    9.  Error texts for pcre_compile() are held as one long string to avoid too
802        much relocation at load time. To find a text, the string is searched,
803        counting zeros. There was no check for running off the end of the string,
804        which could happen if a new error number was added without updating the
805        string.
807    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
809    11. \K was not working if it appeared in an atomic group or in a group that
810        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
811        \K is "not well defined" if used in an assertion. PCRE now accepts it if
812        the assertion is positive, but not if it is negative.
814    12. Change 11 fortuitously reduced the size of the stack frame used in the
815        "match()" function of pcre_exec.c by one pointer. Forthcoming
816        implementation of support for (*MARK) will need an extra pointer on the
817        stack; I have reserved it now, so that the stack frame size does not
818        decrease.
820    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
821        item in branch that calls a recursion is a subroutine call - as in the
822        second branch in the above example - was incorrectly given the compile-
823        time error "recursive call could loop indefinitely" because pcre_compile()
824        was not correctly checking the subroutine for matching a non-empty string.
826    14. The checks for overrunning compiling workspace could trigger after an
827        overrun had occurred. This is a "should never occur" error, but it can be
828        triggered by pathological patterns such as hundreds of nested parentheses.
829        The checks now trigger 100 bytes before the end of the workspace.
831    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
834    Version 8.01 19-Jan-2010
835    ------------------------
837    1.  If a pattern contained a conditional subpattern with only one branch (in
838        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
839        computed the wrong minimum data length (which is of course zero for such
840        subpatterns). This could cause incorrect "no match" results.
842    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
843        the pattern is reset in the first branch, pcre_compile() failed with
844        "internal error: code overflow at offset...". This happened only when
845        the reset was to the original external option setting. (An optimization
846        abstracts leading options settings into an external setting, which was the
847        cause of this.)
849    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
850        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
851        assertion pattern did not match (meaning that the assertion was true), it
852        was incorrectly treated as false if the SKIP had been reached during the
853        matching. This also applied to assertions used as conditions.
855    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
856        assertion subpattern, including such a pattern used as a condition,
857        unpredictable results occurred, instead of the error return
860    5.  The C++ GlobalReplace function was not working like Perl for the special
861        situation when an empty string is matched. It now does the fancy magic
862        stuff that is necessary.
864    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
865        removed. (These were left over from very, very early versions of PCRE.)
867    7.  Some cosmetic changes to the code to make life easier when compiling it
868        as part of something else:
870        (a) Change DEBUG to PCRE_DEBUG.
872        (b) In pcre_compile(), rename the member of the "branch_chain" structure
873            called "current" as "current_branch", to prevent a collision with the
874            Linux macro when compiled as a kernel module.
876        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
877            prevent a collision with the Linux macro when compiled as a kernel
878            module.
880    8.  In pcre_compile() there are some checks for integer overflows that used to
881        cast potentially large values to (double). This has been changed to that
882        when building, a check for int64_t is made, and if it is found, it is used
883        instead, thus avoiding the use of floating point arithmetic. (There is no
884        other use of FP in PCRE.) If int64_t is not found, the fallback is to
885        double.
887    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
888        2005 (difference between two addresses compared to an unsigned value).
890    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
891        custom one, because of the following reported problem in Windows:
893          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
894              under Win32.
895          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
896              therefore missing the function definition.
897          - The compiler thus generates a "C" signature for the test function.
898          - The linker fails to find the "C" function.
899          - PCRE fails to configure if asked to do so against libbz2.
901    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
902        messages were output:
904          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
905          rerunning libtoolize, to keep the correct libtool macros in-tree.
906          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
908        I have done both of these things.
910    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
911        most of the time, it *can* run out if it is given a pattern that contains a
912        runaway infinite recursion. I updated the discussion in the pcrestack man
913        page.
915    13. Now that we have gone to the x.xx style of version numbers, the minor
916        version may start with zero. Using 08 or 09 is a bad idea because users
917        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
918        interpreted as invalid octal numbers. I've updated the previous comment in
919        configure.ac, and also added a check that gives an error if 08 or 09 are
920        used.
922    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
923        causing partial matching to fail when the end of the subject matched \W
924        in a UTF-8 pattern where \W was quantified with a minimum of 3.
926    15. There were some discrepancies between the declarations in pcre_internal.h
927        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
928        their definitions. The declarations used "const uschar *" and the
929        definitions used USPTR. Even though USPTR is normally defined as "const
930        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
931        reported that: "This difference in casting confuses some C++ compilers, for
932        example, SunCC recognizes above declarations as different functions and
933        generates broken code for hbpcre." I have changed the declarations to use
934        USPTR.
936    16. GNU libtool is named differently on some systems. The autogen.sh script now
937        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
938        (FreeBSD).
940    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
941        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
942        comment: "Figure out how to create a longlong from a string: strtoll and
943        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
944        instance, but it only takes 2 args instead of 3!"
946    18. A subtle bug concerned with back references has been fixed by a change of
947        specification, with a corresponding code fix. A pattern such as
948        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
949        refers, was giving matches when it shouldn't. For example, xa=xaaa would
950        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
951        same bug. Such groups have to be quantified to be useful, or contained
952        inside another quantified group. (If there's no repetition, the reference
953        can never match.) The problem arises because, having left the group and
954        moved on to the rest of the pattern, a later failure that backtracks into
955        the group uses the captured value from the final iteration of the group
956        rather than the correct earlier one. I have fixed this in PCRE by forcing
957        any group that contains a reference to itself to be an atomic group; that
958        is, there cannot be any backtracking into it once it has completed. This is
959        similar to recursive and subroutine calls.
962    Version 8.00 19-Oct-09
963  ----------------------  ----------------------
965    1.  The table for translating pcre_compile() error codes into POSIX error codes
966        was out-of-date, and there was no check on the pcre_compile() error code
967        being within the table. This could lead to an OK return being given in
968        error.
970    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
971        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
972        in a Windows environment.
974    3.  The pcregrep --count option prints the count for each file even when it is
975        zero, as does GNU grep. However, pcregrep was also printing all files when
976        --files-with-matches was added. Now, when both options are given, it prints
977        counts only for those files that have at least one match. (GNU grep just
978        prints the file name in this circumstance, but including the count seems
979        more useful - otherwise, why use --count?) Also ensured that the
980        combination -clh just lists non-zero counts, with no names.
982    4.  The long form of the pcregrep -F option was incorrectly implemented as
983        --fixed_strings instead of --fixed-strings. This is an incompatible change,
984        but it seems right to fix it, and I didn't think it was worth preserving
985        the old behaviour.
987    5.  The command line items --regex=pattern and --regexp=pattern were not
988        recognized by pcregrep, which required --regex pattern or --regexp pattern
989        (with a space rather than an '='). The man page documented the '=' forms,
990        which are compatible with GNU grep; these now work.
992    6.  No libpcreposix.pc file was created for pkg-config; there was just
993        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
995    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
996        when UCP support is not needed, by modifying the Python script that
997        generates it from Unicode data files. This should not matter if the module
998        is correctly used as a library, but I received one complaint about 50K of
999        unwanted data. My guess is that the person linked everything into his
1000        program rather than using a library. Anyway, it does no harm.
1002    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1003        was a minimum greater than 1 for a wide character in a possessive
1004        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1005        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1006        character. Chaos in the form of incorrect output or a compiling loop could
1007        result.
1009    9.  The restrictions on what a pattern can contain when partial matching is
1010        requested for pcre_exec() have been removed. All patterns can now be
1011        partially matched by this function. In addition, if there are at least two
1012        slots in the offset vector, the offset of the earliest inspected character
1013        for the match and the offset of the end of the subject are set in them when
1014        PCRE_ERROR_PARTIAL is returned.
1016    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1017        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1018        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1019        and may be more useful for multi-segment matching.
1021    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1022        used to be given if ever the end of the subject was reached; now it is
1023        given only if matching could not proceed because another character was
1024        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1025        string "Z", which now yields "no match" instead of "partial match". In the
1026        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1027        final character ended with (*FAIL).
1029    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1030        if the pattern had a "must contain" character that was already found in the
1031        earlier partial match, unless partial matching was again requested. For
1032        example, with the pattern /dog.(body)?/, the "must contain" character is
1033        "g". If the first part-match was for the string "dog", restarting with
1034        "sbody" failed. This bug has been fixed.
1036    13. The string returned by pcre_dfa_exec() after a partial match has been
1037        changed so that it starts at the first inspected character rather than the
1038        first character of the match. This makes a difference only if the pattern
1039        starts with a lookbehind assertion or \b or \B (\K is not supported by
1040        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1041        matching functions compatible, and I think it's the right thing to do.
1043    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1044        so that the demonstration program is easily available in environments where
1045        PCRE has not been installed from source.
1047    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1048        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1049        library.
1051    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1052        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1053        is not the first non-POSIX option to be added. Clearly some people find
1054        these options useful.
1056    17. If a caller to the POSIX matching function regexec() passes a non-zero
1057        value for nmatch with a NULL value for pmatch, the value of
1058        nmatch is forced to zero.
1060    18. RunGrepTest did not have a test for the availability of the -u option of
1061        the diff command, as RunTest does. It now checks in the same way as
1062        RunTest, and also checks for the -b option.
1064    19. If an odd number of negated classes containing just a single character
1065        interposed, within parentheses, between a forward reference to a named
1066        subpattern and the definition of the subpattern, compilation crashed with
1067        an internal error, complaining that it could not find the referenced
1068        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1069        [The bug was that it was starting one character too far in when skipping
1070        over the character class, thus treating the ] as data rather than
1071        terminating the class. This meant it could skip too much.]
1073    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1074        /g option in pcretest when the pattern contains \K, which makes it possible
1075        to have an empty string match not at the start, even when the pattern is
1076        anchored. Updated pcretest and pcredemo to use this option.
1078    21. If the maximum number of capturing subpatterns in a recursion was greater
1079        than the maximum at the outer level, the higher number was returned, but
1080        with unset values at the outer level. The correct (outer level) value is
1081        now given.
1083    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1084        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1085        make it do so. The string so far is captured, making this feature
1086        compatible with Perl.
1088    23. The tests have been re-organized, adding tests 11 and 12, to make it
1089        possible to check the Perl 5.10 features against Perl 5.10.
1091    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1092        pattern matches a fixed length string. PCRE did not allow this; now it
1093        does. Neither allows recursion.
1095    25. I finally figured out how to implement a request to provide the minimum
1096        length of subject string that was needed in order to match a given pattern.
1097        (It was back references and recursion that I had previously got hung up
1098        on.) This code has now been added to pcre_study(); it finds a lower bound
1099        to the length of subject needed. It is not necessarily the greatest lower
1100        bound, but using it to avoid searching strings that are too short does give
1101        some useful speed-ups. The value is available to calling programs via
1102        pcre_fullinfo().
1104    26. While implementing 25, I discovered to my embarrassment that pcretest had
1105        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1106        study optimizations had never been tested with that matching function.
1107        Oops. What is worse, even when it was passed study data, there was a bug in
1108        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1109        were also very few tests of studied patterns with pcre_dfa_exec().
1111    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1112        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1113        on the other side of the coin, they are no longer allowed to have different
1114        names, because these cannot be distinguished in PCRE, and this has caused
1115        confusion. (This is a difference from Perl.)
1117    28. When duplicate subpattern names are present (necessarily with different
1118        numbers, as required by 27 above), and a test is made by name in a
1119        conditional pattern, either for a subpattern having been matched, or for
1120        recursion in such a pattern, all the associated numbered subpatterns are
1121        tested, and the overall condition is true if the condition is true for any
1122        one of them. This is the way Perl works, and is also more like the way
1123        testing by number works.
1126    Version 7.9 11-Apr-09
1127    ---------------------
1129    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1130        (pcretest), all targets were linked against these libraries. This included
1131        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1132        libraries. This caused unwanted dependencies to be created. This problem
1133        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1134        pcretest is linked with readline.
1136    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1137        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1138        moved outside it again, because FALSE and TRUE are already defined in AIX,
1139        but BOOL is not.
1141    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1142        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1144    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1145        hyphens) following file names and line numbers when outputting matching
1146        lines. This is not true; no spaces are inserted. I have also clarified the
1147        wording for the --colour (or --color) option.
1149    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1150        was not coloured; this is different to GNU grep, so I have changed it to be
1151        the same.
1153    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1154        each matching line was coloured. Now it goes on to look for further matches
1155        of any of the test patterns, which is the same behaviour as GNU grep.
1157    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1158        doesn't make sense to accept an empty string match in pcregrep, so I have
1159        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1160        seems to be how GNU grep behaves.
1162    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1163        start or after a newline", because the conditional assertion was not being
1164        correctly handled. The rule now is that both the assertion and what follows
1165        in the first alternative must satisfy the test.
1167    9.  If auto-callout was enabled in a pattern with a conditional group whose
1168        condition was an assertion, PCRE could crash during matching, both with
1169        pcre_exec() and pcre_dfa_exec().
1171    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1172        used for matching.
1174    11. Unicode property support in character classes was not working for
1175        characters (bytes) greater than 127 when not in UTF-8 mode.
1177    12. Added the -M command line option to pcretest.
1179    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1181    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1183    16. Added comments and documentation about mis-use of no_arg in the C++
1184        wrapper.
1186    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1187        from Martin Jerabek that uses macro names for all relevant character and
1188        string constants.
1190    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1191        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1192        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1193        these, but not everybody uses configure.
1195    19. A conditional group that had only one branch was not being correctly
1196        recognized as an item that could match an empty string. This meant that an
1197        enclosing group might also not be so recognized, causing infinite looping
1198        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1199        with the subject "ab", where knowledge that the repeated group can match
1200        nothing is needed in order to break the loop.
1202    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1203        exec(), but without supplying a callout function, matching went wrong.
1205    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1206        leak if the size of the offset vector was greater than 30. When the vector
1207        is smaller, the saved offsets during recursion go onto a local stack
1208        vector, but for larger vectors malloc() is used. It was failing to free
1209        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1210        error, in fact).
1212    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1213        heapframe that is used only when UTF-8 support is enabled. This caused no
1214        problem, but was untidy.
1216    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1217        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1218        included within another project.
1220    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1221        slightly modified by me:
1223          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1224              not building pcregrep.
1226          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1227              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1229    25. Forward references, both numeric and by name, in patterns that made use of
1230        duplicate group numbers, could behave incorrectly or give incorrect errors,
1231        because when scanning forward to find the reference group, PCRE was not
1232        taking into account the duplicate group numbers. A pattern such as
1233        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1235    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1236        the feature of a custom pointer more persuasive (as requested by a user).
1238    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1239        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1240        pre-defined.
1242    28. Added support for (*UTF8) at the start of a pattern.
1244    29. Arrange for flags added by the "release type" setting in CMake to be shown
1245        in the configuration summary.
1248    Version 7.8 05-Sep-08
1249    ---------------------
1251    1.  Replaced UCP searching code with optimized version as implemented for Ad
1252        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1253        stage table and inline lookup instead of a function, giving speed ups of 2
1254        to 5 times on some simple patterns that I tested. Permission was given to
1255        distribute the MultiStage2.py script that generates the tables (it's not in
1256        the tarball, but is in the Subversion repository).
1258    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1259        scripts.
1261    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1262        a group with a zero qualifier. The result of the study could be incorrect,
1263        or the function might crash, depending on the pattern.
1265    4.  Caseless matching was not working for non-ASCII characters in back
1266        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1267        It now works when Unicode Property Support is available.
1269    5.  In pcretest, an escape such as \x{de} in the data was always generating
1270        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1271        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1272        truncation.
1274    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1276    7.  Added two (int) casts to pcregrep when printing the difference of two
1277        pointers, in case they are 64-bit values.
1279    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1280        test 2 if it fails.
1282    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1283        and a #define of that name to empty if it is not externally set. This is to
1284        allow users of MSVC to set it if necessary.
1286    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1287        the convenience functions in the pcre_get.c source file.
1289    11. An option change at the start of a pattern that had top-level alternatives
1290        could cause overwriting and/or a crash. This command provoked a crash in
1291        some environments:
1293          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1295        This potential security problem was recorded as CVE-2008-2371.
1297    12. For a pattern where the match had to start at the beginning or immediately
1298        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1299        pcre_dfa_exec() could read past the end of the passed subject if there was
1300        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1301        pcretest so that it places the subject at the end of its malloc-ed buffer.
1303    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1304        exec() might read past the end of the data buffer in UTF-8 mode.
1306    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1307        the data contained the byte 0x85 as part of a UTF-8 character within its
1308        first line. This applied both to normal and DFA matching.
1310    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1311        /^[^d]*?$/8 failed to match "abc".
1313    16. Added a missing copyright notice to pcrecpp_internal.h.
1315    17. Make it more clear in the documentation that values returned from
1316        pcre_exec() in ovector are byte offsets, not character counts.
1318    18. Tidied a few places to stop certain compilers from issuing warnings.
1320    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1321        supplied by Stefan Weber. I made a further small update for 7.8 because
1322        there is a change of source arrangements: the pcre_searchfuncs.c module is
1323        replaced by pcre_ucd.c.
1326    Version 7.7 07-May-08
1327    ---------------------
1329    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1330        a string to a long long, pretend we don't even have a long long." This is
1331        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1333    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1334        pre-7.6 versions, which defined a global no_arg variable instead of putting
1335        it in the RE class. (See also #8 below.)
1337    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1338        Lopes.
1340    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1342        (1) The include/exclude patterns were being applied to the whole pathnames
1343            of files, instead of just to the final components.
1345        (2) If there was more than one level of directory, the subdirectories were
1346            skipped unless they satisfied the include/exclude conditions. This is
1347            inconsistent with GNU grep (and could even be seen as contrary to the
1348            pcregrep specification - which I improved to make it absolutely clear).
1349            The action now is always to scan all levels of directory, and just
1350            apply the include/exclude patterns to regular files.
1352    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1353        --exclude_dir in the tests to avoid scanning .svn directories.
1355    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1356        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1357        doesn't support NULs in patterns.
1359    7.  Added some missing "const"s to declarations of static tables in
1360        pcre_compile.c and pcre_dfa_exec.c.
1362    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1363        caused by fix #2  above. (Subsequently also a second patch to fix the
1364        first patch. And a third patch - this was a messy problem.)
1366    9.  Applied Craig's patch to remove the use of push_back().
1368    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1369        matching function regexec().
1371    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1372        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1373        references. PCRE supports relative numbers with this syntax (I don't think
1374        Oniguruma does).
1376    12. Previously, a group with a zero repeat such as (...){0} was completely
1377        omitted from the compiled regex. However, this means that if the group
1378        was called as a subroutine from elsewhere in the pattern, things went wrong
1379        (an internal error was given). Such groups are now left in the compiled
1380        pattern, with a new opcode that causes them to be skipped at execution
1381        time.
1383    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1384        to the way PCRE behaves:
1386        (a) A lone ] character is dis-allowed (Perl treats it as data).
1388        (b) A back reference to an unmatched subpattern matches an empty string
1389            (Perl fails the current match path).
1391        (c) A data ] in a character class must be notated as \] because if the
1392            first data character in a class is ], it defines an empty class. (In
1393            Perl it is not possible to have an empty class.) The empty class []
1394            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1395            The negative empty class [^] matches any one character, independently
1396            of the DOTALL setting.
1398    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1399        non-existent subpattern following a character class starting with ']' and
1400        containing () gave an internal compiling error instead of "reference to
1401        non-existent subpattern". Fortunately, when the pattern did exist, the
1402        compiled code was correct. (When scanning forwards to check for the
1403        existencd of the subpattern, it was treating the data ']' as terminating
1404        the class, so got the count wrong. When actually compiling, the reference
1405        was subsequently set up correctly.)
1407    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1408        it was being rejected as not supported by pcre_dfa_exec(), even though
1409        other assertions are supported. I have made pcre_dfa_exec() support
1410        (*FAIL).
1412    16. The implementation of 13c above involved the invention of a new opcode,
1413        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1414        cannot be changed at match time, I realized I could make a small
1415        improvement to matching performance by compiling OP_ALLANY instead of
1416        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1417        on the OP_ANY path.
1419    17. Compiling pcretest on Windows with readline support failed without the
1420        following two fixes: (1) Make the unistd.h include conditional on
1421        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1423    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1424        ncurses library to be included for pcretest when ReadLine support is
1425        requested, but also to allow for it to be overridden. This patch came from
1426        Daniel Bergström.
1428    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1429        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1430        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1431        spotting this.
1434    Version 7.6 28-Jan-08
1435    ---------------------
1437    1.  A character class containing a very large number of characters with
1438        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1439        overflow.
1441    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1442        HAVE_LONG_LONG is not defined.
1444    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1445        bring them up to date and include new features. This patch includes:
1447        - Fixed PH's badly added libz and libbz2 support.
1448        - Fixed a problem with static linking.
1449        - Added pcredemo. [But later removed - see 7 below.]
1450        - Fixed dftables problem and added an option.
1451        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1452            HAVE_LONG_LONG.
1453        - Added readline support for pcretest.
1454        - Added an listing of the option settings after cmake has run.
1456    4.  A user submitted a patch to Makefile that makes it easy to create
1457        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1458        Makefile.am that cause it to include this special target, without
1459        affecting anything else. Note that the same mingw target plus all
1460        the other distribution libraries and programs are now supported
1461        when configuring with CMake (see 6 below) instead of with
1462        Configure/Make.
1464    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1465        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1466        exported in the Windows port". It has not yet been confirmed that the patch
1467        solves the problem, but it does no harm.
1469    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1470        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1471        with CMake, and also correct the comment about stack recursion.
1473    7.  Remove the automatic building of pcredemo from the ./configure system and
1474        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1475        of a program that users should build themselves after PCRE is installed, so
1476        building it automatically is not really right. What is more, it gave
1477        trouble in some build environments.
1479    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1482    Version 7.5 10-Jan-08
1483    ---------------------
1485    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1486        values in parens when parsing an RE using the C++ wrapper."
1488    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1489        Characters greater than 255 were excluded from the class instead of being
1490        included.
1492    3.  The same bug as (2) above applied to negated POSIX classes such as
1493        [:^space:].
1495    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1496        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1497        I have changed it.
1499    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1500        first named subpattern) and a construct such as (?&a) would reference the
1501        first named subpattern whose name started with "a" (in other words, the
1502        length check was missing). Both these problems are fixed. "Subpattern name
1503        expected" is now given for (?&) (a zero-length name), and this patch also
1504        makes it give the same error for \k'' (previously it complained that that
1505        was a reference to a non-existent subpattern).
1507    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1508        this is right because (?- can be followed by option settings as well as by
1509        digits. I have, however, made the messages clearer.
1511    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1512        than the number used in the conditional) now cause a compile-time error.
1513        This is actually not compatible with Perl, which accepts such patterns, but
1514        treats the conditional as always being FALSE (as PCRE used to), but it
1515        seems to me that giving a diagnostic is better.
1517    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1518        and messages.
1520    9.  Fix two occurrences of "backslash" in comments that should have been
1521        "backspace".
1523    10. Remove two redundant lines of code that can never be obeyed (their function
1524        was moved elsewhere).
1526    11. The program that makes PCRE's Unicode character property table had a bug
1527        which caused it to generate incorrect table entries for sequences of
1528        characters that have the same character type, but are in different scripts.
1529        It amalgamated them into a single range, with the script of the first of
1530        them. In other words, some characters were in the wrong script. There were
1531        thirteen such cases, affecting characters in the following ranges:
1533          U+002b0 - U+002c1
1534          U+0060c - U+0060d
1535          U+0061e - U+00612
1536          U+0064b - U+0065e
1537          U+0074d - U+0076d
1538          U+01800 - U+01805
1539          U+01d00 - U+01d77
1540          U+01d9b - U+01dbf
1541          U+0200b - U+0200f
1542          U+030fc - U+030fe
1543          U+03260 - U+0327f
1544          U+0fb46 - U+0fbb1
1545          U+10450 - U+1049d
1547    12. The -o option (show only the matching part of a line) for pcregrep was not
1548        compatible with GNU grep in that, if there was more than one match in a
1549        line, it showed only the first of them. It now behaves in the same way as
1550        GNU grep.
1552    13. If the -o and -v options were combined for pcregrep, it printed a blank
1553        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1554        does the same. The return code can be used to tell if there were any
1555        non-matching lines.
1557    14. Added --file-offsets and --line-offsets to pcregrep.
1559    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1560        infinitely looping recursion. The bug was that positive lookaheads were not
1561        being skipped when checking for a possible empty match (negative lookaheads
1562        and both kinds of lookbehind were skipped).
1564    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1565        inclusion of <windows.h> to before rather than after the definition of
1566        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1568    17. Specifying a possessive quantifier with a specific limit for a Unicode
1569        character property caused pcre_compile() to compile bad code, which led at
1570        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1571        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1572        caused the error; without that there was no problem.
1574    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1576    19. Added --enable-pcretest-libreadline.
1578    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1579        RE::GlobalReplace(). As a result, the number of replacements returned was
1580        double what it should be. I removed one of the increments, but Craig sent a
1581        later patch that removed the other one (the right fix) and added unit tests
1582        that check the return values (which was not done before).
1584    21. Several CMake things:
1586        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1587            the names libpcre and libpcreposix, not just pcre and pcreposix.
1589        (2) The above change means that pcretest and pcregrep are now correctly
1590            linked with the newly-built libraries, not previously installed ones.
1594    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1595        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1596        UTF-8 newline character). The key issue is that the pattern starts .*;
1597        this means that the match must be either at the beginning, or after a
1598        newline. The bug was in the code for advancing after a failed match and
1599        checking that the new position followed a newline. It was not taking
1600        account of UTF-8 characters correctly.
1602    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1603        character classes. PCRE was not treating the sequence [:...:] as a
1604        character class unless the ... were all letters. Perl, however, seems to
1605        allow any characters between [: and :], though of course it rejects as
1606        unknown any "names" that contain non-letters, because all the known class
1607        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1608        for example, whereas PCRE did not - it did not recognize a POSIX character
1609        class. This seemed a bit dangerous, so the code has been changed to be
1610        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1611        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1612        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1613        Perl does, and where it didn't before.
1615    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1616        Windows environments %n is disabled by default.
1619    Version 7.4 21-Sep-07
1620    ---------------------
1622    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1623        means that a class such as [\s] counted as "explicit reference to CR or
1624        LF". That isn't really right - the whole point of the change was to try to
1625        help when there was an actual mention of one of the two characters. So now
1626        the change happens only if \r or \n (or a literal CR or LF) character is
1627        encountered.
1629    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1630        of both had grown to the point where there were only 3 bits left.
1631        Fortunately, there was spare space in the data structure, and so I have
1632        moved the internal flags into a new 16-bit field to free up more option
1633        bits.
1635    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1636        but did not set the internal JCHANGED flag - either of these is enough to
1637        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1638        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1639        start sets both bits.
1641    4.  Added options (at build time, compile time, exec time) to change \R from
1642        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1644    5.  doc/pcresyntax.html was missing from the distribution.
1646    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1647        compatibility, even though it is no longer used.
1649    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1650        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1651        windows.h file is present (where different names are used). [This was
1652        reversed later after testing - see 16 below.]
1654    8.  Changed all #include <config.h> to #include "config.h". There were also
1655        some further <pcre.h> cases that I changed to "pcre.h".
1657    9.  When pcregrep was used with the --colour option, it missed the line ending
1658        sequence off the lines that it output.
1660    10. It was pointed out to me that arrays of string pointers cause lots of
1661        relocations when a shared library is dynamically loaded. A technique of
1662        using a single long string with a table of offsets can drastically reduce
1663        these. I have refactored PCRE in four places to do this. The result is
1664        dramatic:
1666          Originally:                          290
1667          After changing UCP table:            187
1668          After changing error message table:   43
1669          After changing table of "verbs"       36
1670          After changing table of Posix names   22
1672        Thanks to the folks working on Gregex for glib for this insight.
1674    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1675        unicode-properties was also set.
1677    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1679    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1680        checked only for CRLF.
1682    14. Added casts to pcretest.c to avoid compiler warnings.
1684    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1686    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1687        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1688        entirely. This removes changes made in 7 above.
1690    17. The CMake files have been updated, and there is now more information about
1691        building with CMake in the NON-UNIX-USE document.
1694    Version 7.3 28-Aug-07
1695    ---------------------
1697     1. In the rejigging of the build system that eventually resulted in 7.1, the
1698        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1699        brackets there is not right, since it causes compilers to look for an
1700        installed pcre.h, not the version that is in the source that is being
1701        compiled (which of course may be different). I have changed it back to:
1703          #include "pcre.h"
1705        I have a vague recollection that the change was concerned with compiling in
1706        different directories, but in the new build system, that is taken care of
1707        by the VPATH setting the Makefile.
1709     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1710        when the subject happened to end in the byte 0x85 (e.g. if the last
1711        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1712        characters but of course it shouldn't be taken as a newline when it is part
1713        of another character. The bug was that, for an unlimited repeat of . in
1714        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1715        characters when looking for a newline.
1717     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1719     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1720        in debug output.
1722     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1723        long printing in the pcrecpp unittest when running under MinGW.
1725     6. ESC_K was left out of the EBCDIC table.
1727     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1728        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1729        limit also applies to "virtual nesting" when a pattern is recursive, and in
1730        this case 1000 isn't so big. I have been able to remove this limit at the
1731        expense of backing off one optimization in certain circumstances. Normally,
1732        when pcre_exec() would call its internal match() function recursively and
1733        immediately return the result unconditionally, it uses a "tail recursion"
1734        feature to save stack. However, when a subpattern that can match an empty
1735        string has an unlimited repetition quantifier, it no longer makes this
1736        optimization. That gives it a stack frame in which to save the data for
1737        checking that an empty string has been matched. Previously this was taken
1738        from the 1000-entry workspace that had been reserved. So now there is no
1739        explicit limit, but more stack is used.
1741     8. Applied Daniel's patches to solve problems with the import/export magic
1742        syntax that is required for Windows, and which was going wrong for the
1743        pcreposix and pcrecpp parts of the library. These were overlooked when this
1744        problem was solved for the main library.
1746     9. There were some crude static tests to avoid integer overflow when computing
1747        the size of patterns that contain repeated groups with explicit upper
1748        limits. As the maximum quantifier is 65535, the maximum group length was
1749        set at 30,000 so that the product of these two numbers did not overflow a
1750        32-bit integer. However, it turns out that people want to use groups that
1751        are longer than 30,000 bytes (though not repeat them that many times).
1752        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1753        made it possible to implement the integer overflow checks in a much more
1754        dynamic way, which I have now done. The artificial limitation on group
1755        length has been removed - we now have only the limit on the total length of
1756        the compiled pattern, which depends on the LINK_SIZE setting.
1758    10. Fixed a bug in the documentation for get/copy named substring when
1759        duplicate names are permitted. If none of the named substrings are set, the
1760        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1761        empty string.
1763    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1764        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1765        because the ] is interpreted as the first data character and the
1766        terminating ] is not found. PCRE has been made compatible with Perl in this
1767        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1768        cause memory overwriting.
1770    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1771        string has been matched (to stop an infinite loop). It was not recognizing
1772        a conditional subpattern that could match an empty string if that
1773        subpattern was within another subpattern. For example, it looped when
1774        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1775        condition was not nested. This bug has been fixed.
1777    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1778        past the start of the subject in the presence of bytes with the top bit
1779        set, for example "\x8aBCD".
1781    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1782        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1784    14. Optimized (?!) to (*FAIL).
1786    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1787        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1788        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1789        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1790        does: it's just the validity check that is more restrictive.
1792    16. Inserted checks for integer overflows during escape sequence (backslash)
1793        processing, and also fixed erroneous offset values for syntax errors during
1794        backslash processing.
1796    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1797        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1799    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1800        caused an overrun.
1802    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1803        something other than just ASCII characters) inside a group that had an
1804        unlimited repeat caused a loop at compile time (while checking to see
1805        whether the group could match an empty string).
1807    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1808        [\P{Any}] did so. (Error in the code for printing property names.)
1810    21. An orphan \E inside a character class could cause a crash.
1812    22. A repeated capturing bracket such as (A)? could cause a wild memory
1813        reference during compilation.
1815    23. There are several functions in pcre_compile() that scan along a compiled
1816        expression for various reasons (e.g. to see if it's fixed length for look
1817        behind). There were bugs in these functions when a repeated \p or \P was
1818        present in the pattern. These operators have additional parameters compared
1819        with \d, etc, and these were not being taken into account when moving along
1820        the compiled data. Specifically:
1822        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1823            length.
1825        (b) An item such as \pL+ within a repeated group could cause crashes or
1826            loops.
1828        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1829            "reference to non-existent subpattern" error.
1831        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1833    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1834        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1836    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1838    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1839        character were causing crashes (broken optimization).
1841    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1842        \p or \P) caused a compile-time loop.
1844    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1845        break. For example, the unstudied pattern [\r\n]A does not match the string
1846        "\r\nA" because change 7.0/46 below moves the current point on by two
1847        characters after failing to match at the start. However, the pattern \nA
1848        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1849        the same is true. There doesn't seem any very clean way out of this, but
1850        what I have chosen to do makes the common cases work: PCRE now takes note
1851        of whether there can be an explicit match for \r or \n anywhere in the
1852        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1853        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1854        pattern has explicit CR or LF references.
1856    29. Added (*CR) etc for changing newline setting at start of pattern.
1859    Version 7.2 19-Jun-07
1860    ---------------------
1862     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1863        which is apparently normally available under Windows.
1865     2. Re-jig the pcregrep tests with different newline settings in an attempt
1866        to make them independent of the local environment's newline setting.
1868     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1870     4. Some of the "internals" tests were previously cut out when the link size
1871        was not 2, because the output contained actual offsets. The recent new
1872        "Z" feature of pcretest means that these can be cut out, making the tests
1873        usable with all link sizes.
1875     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1876        stack recursion. This gives a massive performance boost under BSD, but just
1877        a small improvement under Linux. However, it saves one field in the frame
1878        in all cases.
1880     6. Added more features from the forthcoming Perl 5.10:
1882        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1883            recursion call. It refers to the nth most recently opened parentheses.
1885        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1886            to be opened parentheses.
1888        (c) Conditions that refer to capturing parentheses can be specified
1889            relatively, for example, (?(-2)... or (?(+3)...
1891        (d) \K resets the start of the current match so that everything before
1892            is not part of it.
1894        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1896        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1897            reference syntax.
1899        (g) (?| introduces a group in which the numbering of parentheses in each
1900            alternative starts with the same number.
1902        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1904     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1907     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1908        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1909        for detecting groups that can match an empty string.
1911     9. A pattern with a very large number of alternatives (more than several
1912        hundred) was running out of internal workspace during the pre-compile
1913        phase, where pcre_compile() figures out how much memory will be needed. A
1914        bit of new cunning has reduced the workspace needed for groups with
1915        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1916        workspace instead of running out of the 4096 that are available.
1918    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1920    11. Applied patch from Google to remove an optimization that didn't quite work.
1921        The report of the bug said:
1923          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1924          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1925          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1927    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1928        it matched the wrong number of bytes.
1931    Version 7.1 24-Apr-07
1932    ---------------------
1934     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1935        that is more "standard", making use of automake and other Autotools. There
1936        is some re-arrangement of the files and adjustment of comments consequent
1937        on this.
1939     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1940        for recursive directory scanning broke on some systems because the files
1941        are not scanned in any specific order and on different systems the order
1942        was different. A call to "sort" has been inserted into RunGrepTest for the
1943        approprate test as a short-term fix. In the longer term there may be an
1944        alternative.
1946     3. I had an email from Eric Raymond about problems translating some of PCRE's
1947        man pages to HTML (despite the fact that I distribute HTML pages, some
1948        people do their own conversions for various reasons). The problems
1949        concerned the use of low-level troff macros .br and .in. I have therefore
1950        removed all such uses from the man pages (some were redundant, some could
1951        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1952        HTML has been updated to handle .nf/.fi and to complain if it encounters
1953        .br or .in.
1955     4. Updated comments in configure.ac that get placed in config.h.in and also
1956        arranged for config.h to be included in the distribution, with the name
1957        config.h.generic, for the benefit of those who have to compile without
1958        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1960     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1961        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1962        makevp.bat for the current PCRE, using the additional files
1963        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1965     6. A Windows user reported a minor discrepancy with test 2, which turned out
1966        to be caused by a trailing space on an input line that had got lost in his
1967        copy. The trailing space was an accident, so I've just removed it.
1969     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1970        that is needed.
1972     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1973        as "const" (a) because they are and (b) because it helps the PHP
1974        maintainers who have recently made a script to detect big data structures
1975        in the php code that should be moved to the .rodata section. I remembered
1976        to update Builducptable as well, so it won't revert if ucptable.h is ever
1977        re-created.
1979     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1980        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1981        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1982        support is not required. This saves 1.5-2K of code, which is important in
1983        some applications.
1985        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1986        so as not to refer to the tables, even though these functions will never be
1987        called when UTF-8 support is disabled. Otherwise there are problems with a
1988        shared library.
1990    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1992        (a) It was defining its arguments as char * instead of void *.
1994        (b) It was assuming that all moves were upwards in memory; this was true
1995            a long time ago when I wrote it, but is no longer the case.
1997        The emulated memove() is provided for those environments that have neither
1998        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1999        is clearly not the case, as these two bugs were recently reported.
2001    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2002        and Detrail to create the HTML documentation, the .txt form of the man
2003        pages, and it removes trailing spaces from listed files. It also creates
2004        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2005        case, it wraps all the #defines with #ifndefs. This script should be run
2006        before "make dist".
2008    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2009        with Unicode property support.
2011        (a) For a maximizing quantifier, if the two different cases of the
2012            character were of different lengths in their UTF-8 codings (there are
2013            some cases like this - I found 11), and the matching function had to
2014            back up over a mixture of the two cases, it incorrectly assumed they
2015            were both the same length.
2017        (b) When PCRE was configured to use the heap rather than the stack for
2018            recursion during matching, it was not correctly preserving the data for
2019            the other case of a UTF-8 character when checking ahead for a match
2020            while processing a minimizing repeat. If the check also involved
2021            matching a wide character, but failed, corruption could cause an
2022            erroneous result when trying to check for a repeat of the original
2023            character.
2025    13. Some tidying changes to the testing mechanism:
2027        (a) The RunTest script now detects the internal link size and whether there
2028            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2029            values substituted by "configure". (The RunGrepTest script already did
2030            this for UTF-8.) The configure.ac script no longer substitutes the
2031            relevant variables.
2033        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2034            with length and offset values. This means that the output is different
2035            for different internal link sizes. Test 2 is skipped for link sizes
2036            other than 2 because of this, bypassing the problem. Unfortunately,
2037            there was also a test in test 3 (the locale tests) that used /B and
2038            failed for link sizes other than 2. Rather than cut the whole test out,
2039            I have added a new /Z option to pcretest that replaces the length and
2040            offset values with spaces. This is now used to make test 3 independent
2041            of link size. (Test 2 will be tidied up later.)
2043    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2044        segmentation fault instead of returning the appropriate error message.
2046    15. In multiline mode when the newline sequence was set to "any", the pattern
2047        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2048        This doesn't seem right; it now treats the CRLF combination as the line
2049        ending, and so does not match in that case. It's only a pattern such as ^$
2050        that would hit this one: something like ^ABC$ would have failed after \r
2051        and then tried again after \r\n.
2053    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2054        in an attempt to make files that differ only in their line terminators
2055        compare equal. This works on Linux.
2057    17. Under certain error circumstances pcregrep might try to free random memory
2058        as it exited. This is now fixed, thanks to valgrind.
2060    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2061        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2062        was because its rules for how to advance for /g after matching an empty
2063        string at the end of a line did not allow for this case. They now check for
2064        it specially.
2066    20. pcretest is supposed to handle patterns and data of any length, by
2067        extending its buffers when necessary. It was getting this wrong when the
2068        buffer for a data line had to be extended.
2070    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2071        CRLF as a newline sequence.
2073    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2074        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2075        I have nevertheless tidied it up.
2077    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2079    24. Added a man page for pcre-config.
2082    Version 7.0 19-Dec-06
2083    ---------------------
2085     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2086        moving to gcc 4.1.1.
2088     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2089        sys/time.h, which is documented as needed for this function. It doesn't
2090        seem to matter on Linux, but it showed up on some releases of OS X.
2092     3. It seems that there are systems where bytes whose values are greater than
2093        127 match isprint() in the "C" locale. The "C" locale should be the
2094        default when a C program starts up. In most systems, only ASCII printing
2095        characters match isprint(). This difference caused the output from pcretest
2096        to vary, making some of the tests fail. I have changed pcretest so that:
2098        (a) When it is outputting text in the compiled version of a pattern, bytes
2099            other than 32-126 are always shown as hex escapes.
2101        (b) When it is outputting text that is a matched part of a subject string,
2102            it does the same, unless a different locale has been set for the match
2103            (using the /L modifier). In this case, it uses isprint() to decide.
2105     4. Fixed a major bug that caused incorrect computation of the amount of memory
2106        required for a compiled pattern when options that changed within the
2107        pattern affected the logic of the preliminary scan that determines the
2108        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2109        that the computed length was too small. The symptoms of this bug were
2110        either the PCRE error "internal error: code overflow" from pcre_compile(),
2111        or a glibc crash with a message such as "pcretest: free(): invalid next
2112        size (fast)". Examples of patterns that provoked this bug (shown in
2113        pcretest format) are:
2115          /(?-x: )/x
2116          /(?x)(?-x: \s*#\s*)/
2117          /((?i)[\x{c0}])/8
2118          /(?i:[\x{c0}])/8
2120        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2121        is now done differently.
2123     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2124        wrapper classes; (b) implement a new function in the C++ scanner that is
2125        more efficient than the old way of doing things because it avoids levels of
2126        recursion in the regex matching; (c) add a paragraph to the documentation
2127        for the FullMatch() function.
2129     6. The escape sequence \n was being treated as whatever was defined as
2130        "newline". Not only was this contrary to the documentation, which states
2131        that \n is character 10 (hex 0A), but it also went horribly wrong when
2132        "newline" was defined as CRLF. This has been fixed.
2134     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2135        was being set to -1 for the "end of line" case (supposedly a value that no
2136        character can have). Though this value is never used (the check for end of
2137        line is "zero bytes in current character"), it caused compiler complaints.
2138        I've changed it to 0xffffffff.
2140     8. In pcre_version.c, the version string was being built by a sequence of
2141        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2142        string (as it is for production releases) called a macro with an empty
2143        argument. The C standard says the result of this is undefined. The gcc
2144        compiler treats it as an empty string (which was what was wanted) but it is
2145        reported that Visual C gives an error. The source has been hacked around to
2146        avoid this problem.
2148     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2149        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2150        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2151        of them did).
2153    10. Originally, pcretest opened its input and output without "b"; then I was
2154        told that "b" was needed in some environments, so it was added for release
2155        5.0 to both the input and output. (It makes no difference on Unix-like
2156        systems.) Later I was told that it is wrong for the input on Windows. I've
2157        now abstracted the modes into two macros, to make it easier to fiddle with
2158        them, and removed "b" from the input mode under Windows.
2160    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2162    12. Added -help and --help to pcretest as an official way of being reminded
2163        of the options.
2165    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2166        and pcrecpp.cc because they annoy compilers at high warning levels.
2168    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2170    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2171        scripts are not C programs :-) and which was not noticed because it works
2172        on Linux.
2174    16. pcretest is supposed to handle any length of pattern and data line (as one
2175        line or as a continued sequence of lines) by extending its input buffer if
2176        necessary. This feature was broken for very long pattern lines, leading to
2177        a string of junk being passed to pcre_compile() if the pattern was longer
2178        than about 50K.
2180    17. I have done a major re-factoring of the way pcre_compile() computes the
2181        amount of memory needed for a compiled pattern. Previously, there was code
2182        that made a preliminary scan of the pattern in order to do this. That was
2183        OK when PCRE was new, but as the facilities have expanded, it has become
2184        harder and harder to keep it in step with the real compile phase, and there
2185        have been a number of bugs (see for example, 4 above). I have now found a
2186        cunning way of running the real compile function in a "fake" mode that
2187        enables it to compute how much memory it would need, while actually only
2188        ever using a few hundred bytes of working memory and without too many
2189        tests of the mode. This should make future maintenance and development
2190        easier. A side effect of this work is that the limit of 200 on the nesting
2191        depth of parentheses has been removed (though this was never a serious
2192        limitation, I suspect). However, there is a downside: pcre_compile() now
2193        runs more slowly than before (30% or more, depending on the pattern). I
2194        hope this isn't a big issue. There is no effect on runtime performance.
2196    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2197        newline (only possible for the last line of a file) and it was a
2198        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2200    19. Added additional timing features to pcretest. (1) The -tm option now times
2201        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2202        separate command line item, by a number that specifies the number of
2203        repeats to use when timing. The default is 50000; this gives better
2204        precision, but takes uncomfortably long for very large patterns.
2206    20. Extended pcre_study() to be more clever in cases where a branch of a
2207        subpattern has no definite first character. For example, (a*|b*)[cd] would
2208        previously give no result from pcre_study(). Now it recognizes that the
2209        first character must be a, b, c, or d.
2211    21. There was an incorrect error "recursive call could loop indefinitely" if
2212        a subpattern (or the entire pattern) that was being tested for matching an
2213        empty string contained only one non-empty item after a nested subpattern.
2214        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2215        incorrectly, because the \d was being skipped in the check.
2217    22. The pcretest program now has a new pattern option /B and a command line
2218        option -b, which is equivalent to adding /B to every pattern. This causes
2219        it to show the compiled bytecode, without the additional information that
2220        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2221        is the same as /B/I).
2223    23. A new optimization is now able automatically to treat some sequences such
2224        as a*b as a*+b. More specifically, if something simple (such as a character
2225        or a simple class like \d) has an unlimited quantifier, and is followed by
2226        something that cannot possibly match the quantified thing, the quantifier
2227        is automatically "possessified".
2229    24. A recursive reference to a subpattern whose number was greater than 39
2230        went wrong under certain circumstances in UTF-8 mode. This bug could also
2231        have affected the operation of pcre_study().
2233    25. Realized that a little bit of performance could be had by replacing
2234        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2236    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2238    27. Possessive quantifiers such as a++ were previously implemented by turning
2239        them into atomic groups such as ($>a+). Now they have their own opcodes,
2240        which improves performance. This includes the automatically created ones
2241        from 23 above.
2243    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2244        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2245        the first matched character to be a colon. This applied both to named and
2246        numbered groups.
2248    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2250    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2251        building PCRE on Windows easier, so I have included it in the distribution.
2253    31. There is now a check in pcretest against a ridiculously large number being
2254        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2255        loop, the loop is abandoned.
2257    32. Forward references to subpatterns in conditions such as (?(2)...) where
2258        subpattern 2 is defined later cause pcre_compile() to search forwards in
2259        the pattern for the relevant set of parentheses. This search went wrong
2260        when there were unescaped parentheses in a character class, parentheses
2261        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2263    33. "Subroutine" calls and backreferences were previously restricted to
2264        referencing subpatterns earlier in the regex. This restriction has now
2265        been removed.
2267    34. Added a number of extra features that are going to be in Perl 5.10. On the
2268        whole, these are just syntactic alternatives for features that PCRE had
2269        previously implemented using the Python syntax or my own invention. The
2270        other formats are all retained for compatibility.
2272        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2273            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2274            also .NET compatible.
2276        (b) A recursion or subroutine call to a named group can now be defined as
2277            (?&name) as well as (?P>name).
2279        (c) A backreference to a named group can now be defined as \k<name> or
2280            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2281            5.10, are also .NET compatible.
2283        (d) A conditional reference to a named group can now use the syntax
2284            (?(<name>) or (?('name') as well as (?(name).
2286        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2287            groups (named and numbered) that are never evaluated inline, but can be
2288            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2289            is always false. There may be only one alternative in such a group.
2291        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2292            as the simple (?(R). The condition is true only if the most recent
2293            recursion is that of the given number or name. It does not search out
2294            through the entire recursion stack.
2296        (g) The escape \gN or \g{N} has been added, where N is a positive or
2297            negative number, specifying an absolute or relative reference.
2299    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2300        some "unreachable code" warnings.
2302    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2303        things, this adds five new scripts.
2305    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2306        There were also incompatibilities regarding the handling of \Q..\E inside
2307        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2308        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2310    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2311        matches an empty string, and forcibly breaks the loop. There were bugs in
2312        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2313        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2314        separate and independent bugs (that affected different cases) have been
2315        fixed.
2317    39. Refactored the code to abolish the use of different opcodes for small
2318        capturing bracket numbers. This is a tidy that I avoided doing when I
2319        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2320        The new approach is not only tidier, it makes it possible to reduce the
2321        memory needed to fix the previous bug (38).
2323    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2324        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2325        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2326        mode.
2328    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2329        report.
2331    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2332        copy construction and assignment in the C++ wrapper.
2334    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2335        couple of bugs that could have given wrong results in the "--newline=crlf"
2336        case.
2338    44. Added a number of casts and did some reorganization of signed/unsigned int
2339        variables following suggestions from Dair Grant. Also renamed the variable
2340        "this" as "item" because it is a C++ keyword.
2342    45. Arranged for dftables to add
2344          #include "pcre_internal.h"
2346        to pcre_chartables.c because without it, gcc 4.x may remove the array
2347        definition from the final binary if PCRE is built into a static library and
2348        dead code stripping is activated.
2350    46. For an unanchored pattern, if a match attempt fails at the start of a
2351        newline sequence, and the newline setting is CRLF or ANY, and the next two
2352        characters are CRLF, advance by two characters instead of one.
2355    Version 6.7 04-Jul-06
2356    ---------------------
2358     1. In order to handle tests when input lines are enormously long, pcretest has
2359        been re-factored so that it automatically extends its buffers when
2360        necessary. The code is crude, but this _is_ just a test program. The
2361        default size has been increased from 32K to 50K.
2363     2. The code in pcre_study() was using the value of the re argument before
2364        testing it for NULL. (Of course, in any sensible call of the function, it
2365        won't be NULL.)
2367     3. The memmove() emulation function in pcre_internal.h, which is used on
2368        systems that lack both memmove() and bcopy() - that is, hardly ever -
2369        was missing a "static" storage class specifier.
2371     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2372        containing an extended class (one that cannot be represented by a bitmap
2373        because it contains high-valued characters or Unicode property items, e.g.
2374        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2375        pattern, but PCRE should not loop if you do not (it no longer does).
2376        [Detail: two cases were found: (a) a repeated subpattern containing an
2377        extended class; (b) a recursive reference to a subpattern that followed a
2378        previous extended class. It wasn't skipping over the extended class
2379        correctly when UTF-8 mode was not set.]
2381     5. A negated single-character class was not being recognized as fixed-length
2382        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2383        compile error "lookbehind assertion is not fixed length".
2385     6. The RunPerlTest auxiliary script was showing an unexpected difference
2386        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2387        write a Perl script that can interpret lines of an input file either as
2388        byte characters or as UTF-8, which is what "perltest" was being required to
2389        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2390        can't do is switch easily at run time between having the "use utf8;" pragma
2391        or not. In the end, I fudged it by using the RunPerlTest script to insert
2392        "use utf8;" explicitly for the UTF-8 tests.
2394     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2395        the end of the subject string, contrary to the documentation and to what
2396        Perl does. This was true of both matching functions. Now it matches only at
2397        the start of the subject and immediately after *internal* newlines.
2399     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2400        a pointer to an int instead of a pointer to an unsigned long int. This
2401        caused problems on 64-bit systems.
2403     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2404        instance of the 'standard' template library not being so standard".
2406    10. There was no check on the number of named subpatterns nor the maximum
2407        length of a subpattern name. The product of these values is used to compute
2408        the size of the memory block for a compiled pattern. By supplying a very
2409        long subpattern name and a large number of named subpatterns, the size
2410        computation could be caused to overflow. This is now prevented by limiting
2411        the length of names to 32 characters, and the number of named subpatterns
2412        to 10,000.
2414    11. Subpatterns that are repeated with specific counts have to be replicated in
2415        the compiled pattern. The size of memory for this was computed from the
2416        length of the subpattern and the repeat count. The latter is limited to
2417        65535, but there was no limit on the former, meaning that integer overflow
2418        could in principle occur. The compiled length of a repeated subpattern is
2419        now limited to 30,000 bytes in order to prevent this.
2421    12. Added the optional facility to have named substrings with the same name.
2423    13. Added the ability to use a named substring as a condition, using the
2424        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2425        are numbers (not recommended). Forward references are permitted.
2427    14. Added forward references in named backreferences (if you see what I mean).
2429    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2430        pattern could run off the end of the subject. For example, the pattern
2431        "(?s)(.{1,5})"8 did this with the subject "ab".
2433    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2434        PCRE_CASELESS was set when matching characters that were quantified with ?
2435        or *.
2437    17. A character class other than a single negated character that had a minimum
2438        but no maximum quantifier - for example [ab]{6,} - was not handled
2439        correctly by pce_dfa_exec(). It would match only one character.
2441    18. A valid (though odd) pattern that looked like a POSIX character
2442        class but used an invalid character after [ (for example [[,abc,]]) caused
2443        pcre_compile() to give the error "Failed: internal error: code overflow" or
2444        in some cases to crash with a glibc free() error. This could even happen if
2445        the pattern terminated after [[ but there just happened to be a sequence of
2446        letters, a binary zero, and a closing ] in the memory that followed.
2448    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2449        over the years. Originally (before any Unicode support), just the bottom 8
2450        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2451        output from "man perlunicode" includes this:
2453          The regular expression compiler produces polymorphic opcodes.  That
2454          is, the pattern adapts to the data and automatically switches to
2455          the Unicode character scheme when presented with Unicode data--or
2456          instead uses a traditional byte scheme when presented with byte
2457          data.
2459        Sadly, a wide octal escape does not cause a switch, and in a string with
2460        no other multibyte characters, these octal escapes are treated as before.
2461        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2462        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2463        Unicode string.
2465        I have not perpetrated such confusion in PCRE. Up till now, it took just
2466        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2467        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2468        translate to the appropriate multibyte character.
2470    29. Applied some refactoring to reduce the number of warnings from Microsoft
2471        and Borland compilers. This has included removing the fudge introduced
2472        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2473        a warning about an unused variable.
2475    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2476        characters since release 4.0, because Perl (from release 5.004) does not.
2477        [Or at least, is documented not to: some releases seem to be in conflict
2478        with the documentation.] However, when a pattern was studied with
2479        pcre_study() and all its branches started with \s, PCRE still included VT
2480        as a possible starting character. Of course, this did no harm; it just
2481        caused an unnecessary match attempt.
2483    22. Removed a now-redundant internal flag bit that recorded the fact that case
2484        dependency changed within the pattern. This was once needed for "required
2485        byte" processing, but is no longer used. This recovers a now-scarce options
2486        bit. Also moved the least significant internal flag bit to the most-
2487        significant bit of the word, which was not previously used (hangover from
2488        the days when it was an int rather than a uint) to free up another bit for
2489        the future.
2491    23. Added support for CRLF line endings as well as CR and LF. As well as the
2492        default being selectable at build time, it can now be changed at runtime
2493        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2494        specify that it is scanning data with non-default line endings.
2496    24. Changed the definition of CXXLINK to make it agree with the definition of
2497        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2499    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2500        recursions. This makes a big different to stack usage for some patterns.
2502    26. If a subpattern containing a named recursion or subroutine reference such
2503        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2504        the space required for the compiled pattern went wrong and gave too small a
2505        value. Depending on the environment, this could lead to "Failed: internal
2506        error: code overflow at offset 49" or "glibc detected double free or
2507        corruption" errors.
2509    27. Applied patches from Google (a) to support the new newline modes and (b) to
2510        advance over multibyte UTF-8 characters in GlobalReplace.
2512    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2513        difference for some implementation of PCRE in some Windows version.
2515    29. Added some extra testing facilities to pcretest:
2517        \q<number>   in a data line sets the "match limit" value
2518        \Q<number>   in a data line sets the "match recursion limt" value
2519        -S <number>  sets the stack size, where <number> is in megabytes
2521        The -S option isn't available for Windows.
2524    Version 6.6 06-Feb-06
2525    ---------------------
2527     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2528        in pcreposix.h. I have copied the definition from pcre.h.
2530     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2531        because pcre.h is no longer a built file.
2533     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2534        not normally included in the compiled code.
2537    Version 6.5 01-Feb-06
2538    ---------------------
2540     1. When using the partial match feature with pcre_dfa_exec(), it was not
2541        anchoring the second and subsequent partial matches at the new starting
2542        point. This could lead to incorrect results. For example, with the pattern
2543        /1234/, partially matching against "123" and then "a4" gave a match.
2545     2. Changes to pcregrep:
2547        (a) All non-match returns from pcre_exec() were being treated as failures
2548            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2549            error message is output. Some extra information is given for the
2550            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2551            probably the only errors that are likely to be caused by users (by
2552            specifying a regex that has nested indefinite repeats, for instance).
2553            If there are more than 20 of these errors, pcregrep is abandoned.
2555        (b) A binary zero was treated as data while matching, but terminated the
2556            output line if it was written out. This has been fixed: binary zeroes
2557            are now no different to any other data bytes.
2559        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2560            used to set a locale for matching. The --locale=xxxx long option has
2561            been added (no short equivalent) to specify a locale explicitly on the
2562            pcregrep command, overriding the environment variables.
2564        (d) When -B was used with -n, some line numbers in the output were one less
2565            than they should have been.
2567        (e) Added the -o (--only-matching) option.
2569        (f) If -A or -C was used with -c (count only), some lines of context were
2570            accidentally printed for the final match.
2572        (g) Added the -H (--with-filename) option.
2574        (h) The combination of options -rh failed to suppress file names for files
2575            that were found from directory arguments.
2577        (i) Added the -D (--devices) and -d (--directories) options.
2579        (j) Added the -F (--fixed-strings) option.
2581        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2583        (l) Added the --colo(u)r option.
2585        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2586            is not present by default.
2588     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2589        items such as (?R) or (?1), when the recursion could match a number of
2590        alternatives. If it matched one of the alternatives, but subsequently,
2591        outside the recursion, there was a failure, the code tried to back up into
2592        the recursion. However, because of the way PCRE is implemented, this is not
2593        possible, and the result was an incorrect result from the match.
2595        In order to prevent this happening, the specification of recursion has
2596        been changed so that all such subpatterns are automatically treated as
2597        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2599     4. I had overlooked the fact that, in some locales, there are characters for
2600        which isalpha() is true but neither isupper() nor islower() are true. In
2601        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2602        and ordfeminine) are like this. This affected the treatment of \w and \W
2603        when they appeared in character classes, but not when they appeared outside
2604        a character class. The bit map for "word" characters is now created
2605        separately from the results of isalnum() instead of just taking it from the
2606        upper, lower, and digit maps. (Plus the underscore character, of course.)
2608     5. The above bug also affected the handling of POSIX character classes such as
2609        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2610        permanent tables. Instead, the bit maps for such a class were previously
2611        created as the appropriate unions of the upper, lower, and digit bitmaps.
2612        Now they are created by subtraction from the [[:word:]] class, which has
2613        its own bitmap.
2615     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2616        It is created by subtracting the vertical space characters (\x09, \x0a,
2617        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2618        subtraction was done in the overall bitmap for a character class, meaning
2619        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2620        be recognized. This bug has been fixed.
2622     7. Patches from the folks at Google:
2624          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2625          real life, but is still worth protecting against".
2627          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2628          regular expressions".
2630          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2631          have it.
2633          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2634          "configure" and the latter not, in order to fix a problem somebody had
2635          with compiling the Arg class on HP-UX.
2637          (e) Improve the error-handling of the C++ wrapper a little bit.
2639          (f) New tests for checking recursion limiting.
2641     8. The pcre_memmove() function, which is used only if the environment does not
2642        have a standard memmove() function (and is therefore rarely compiled),
2643        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2644        returning a result (though PCRE never actually uses the result).
2646     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2647        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2648        returned instead of calling malloc() with an overflowing number that would
2649        most likely cause subsequent chaos.
2651    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2653    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2654        with this option is matched, the nmatch and pmatch options of regexec() are
2655        ignored.
2657    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2658        provided in case anyone wants to the the POSIX interface with UTF-8
2659        strings.
2661    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2662        C++ linking (needed for some HP-UX environments).
2664    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2665        (unused parameter) and in the pcre_printint() function (omitted "default"
2666        switch label when the default is to do nothing).
2668    15. Added some code to make it possible, when PCRE is compiled as a C++
2669        library, to replace subject pointers for pcre_exec() with a smart pointer
2670        class, thus making it possible to process discontinuous strings.
2672    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2673        much the same function. They were added by different people who were trying
2674        to make PCRE easy to compile on non-Unix systems. It has been suggested
2675        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2676        for compiling on Windows systems. I have therefore replaced it with
2677        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2678        defaults to "extern" for C or "extern C" for C++, which works fine on
2679        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2680        SCOPE with something explicit in config.h. In addition:
2682        (a) pcreposix.h still had just "extern" instead of either of these macros;
2683            I have replaced it with PCRE_DATA_SCOPE.
2685        (b) Functions such as _pcre_xclass(), which are internal to the library,
2686            but external in the C sense, all had PCRE_EXPORT in their definitions.
2687            This is apparently wrong for the Windows case, so I have removed it.
2688            (It makes no difference on Unix-like systems.)
2690    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2691        of recursive calls to match(). This is different to MATCH_LIMIT because
2692        that limits the total number of calls to match(), not all of which increase
2693        the depth of recursion. Limiting the recursion depth limits the amount of
2694        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2695        when PCRE is compiled, and changed at run time. A patch from Google adds
2696        this functionality to the C++ interface.
2698    18. Changes to the handling of Unicode character properties:
2700        (a) Updated the table to Unicode 4.1.0.
2702        (b) Recognize characters that are not in the table as "Cn" (undefined).
2704        (c) I revised the way the table is implemented to a much improved format
2705            which includes recognition of ranges. It now supports the ranges that
2706            are defined in UnicodeData.txt, and it also amalgamates other
2707            characters into ranges. This has reduced the number of entries in the
2708            table from around 16,000 to around 3,000, thus reducing its size
2709            considerably. I realized I did not need to use a tree structure after
2710            all - a binary chop search is just as efficient. Having reduced the
2711            number of entries, I extended their size from 6 bytes to 8 bytes to
2712            allow for more data.
2714        (d) Added support for Unicode script names via properties such as \p{Han}.
2716    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2717        matching that character.
2719    20. When matching a repeated Unicode property with a minimum greater than zero,
2720        (for example \pL{2,}), PCRE could look past the end of the subject if it
2721        reached it while seeking the minimum number of characters. This could
2722        happen only if some of the characters were more than one byte long, because
2723        there is a check for at least the minimum number of bytes.
2725    21. Refactored the implementation of \p and \P so as to be more general, to
2726        allow for more different types of property in future. This has changed the
2727        compiled form incompatibly. Anybody with saved compiled patterns that use
2728        \p or \P will have to recompile them.
2730    22. Added "Any" and "L&" to the supported property types.
2732    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2733        but give a compile time error if the value is greater than 0xff.
2735    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2736        accidentally not being installed or uninstalled.
2738    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2739        made were to insert the current release number. This seemed silly, because
2740        it made things harder for people building PCRE on systems that don't run
2741        "configure". I have turned pcre.h into a distributed file, no longer built
2742        by "configure", with the version identification directly included. There is
2743        no longer a pcre.h.in file.
2745        However, this change necessitated a change to the pcre-config script as
2746        well. It is built from pcre-config.in, and one of the substitutions was the
2747        release number. I have updated configure.ac so that ./configure now finds
2748        the release number by grepping pcre.h.
2750    26. Added the ability to run the tests under valgrind.
2753    Version 6.4 05-Sep-05
2754    ---------------------
2756     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2757        "--" to be printed when multiple files were scanned, even when none of the
2758        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2759        consider it to be a bug, and have restored the previous behaviour.
2761     2. A couple of code tidies to get rid of compiler warnings.
2763     3. The pcretest program used to cheat by referring to symbols in the library
2764        whose names begin with _pcre_. These are internal symbols that are not
2765        really supposed to be visible externally, and in some environments it is
2766        possible to suppress them. The cheating is now confined to including
2767        certain files from the library's source, which is a bit cleaner.
2769     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2770        file's purpose clearer.
2772     5. Reorganized pcre_ucp_findchar().
2775    Version 6.3 15-Aug-05
2776    ---------------------
2778     1. The file libpcre.pc.in did not have general read permission in the tarball.
2780     2. There were some problems when building without C++ support:
2782        (a) If C++ support was not built, "make install" and "make test" still
2783            tried to test it.
2785        (b) There were problems when the value of CXX was explicitly set. Some
2786            changes have been made to try to fix these, and ...
2788        (c) --disable-cpp can now be used to explicitly disable C++ support.
2790        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2791            backslash in a target when C++ was disabled. This confuses some
2792            versions of "make", apparently. Using an intermediate variable solves
2793            this. (Same for CPP_LOBJ.)
2795     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2796        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2797        necessary on certain architectures.
2799     4. Added a setting of -export-symbols-regex to the link command to remove
2800        those symbols that are exported in the C sense, but actually are local
2801        within the library, and not documented. Their names all begin with
2802        "_pcre_". This is not a perfect job, because (a) we have to except some
2803        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2804        available (and never for static libraries). I have made a note to try to
2805        find a way round (a) in the future.
2808    Version 6.2 01-Aug-05
2809    ---------------------
2811     1. There was no test for integer overflow of quantifier values. A construction
2812        such as {1111111111111111} would give undefined results. What is worse, if
2813        a minimum quantifier for a parenthesized subpattern overflowed and became
2814        negative, the calculation of the memory size went wrong. This could have
2815        led to memory overwriting.
2817     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2819     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2820        operating environments where this matters.
2822     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2823        PCRE options from within the C++ wrapper.
2825     5. Named capturing subpatterns were not being correctly counted when a pattern
2826        was compiled. This caused two problems: (a) If there were more than 100
2827        such subpatterns, the calculation of the memory needed for the whole
2828        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2829        back references of the form \12, where the number was greater than 9, were
2830        not recognized as back references, even though there were sufficient
2831        previous subpatterns.
2833     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2834        versions of gcc, e.g. 2.95.4.
2837    Version 6.1 21-Jun-05
2838    ---------------------
2840     1. There was one reference to the variable "posix" in pcretest.c that was not
2841        surrounded by "#if !defined NOPOSIX".
2843     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2844        the cross-check on the old pcre_info() function, for the benefit of the
2845        cut-down version of PCRE that is currently imported into Exim.
2847     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2848        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2849        patterns that start (?i) but I don't think that matters. The use of (?i) is
2850        just an example; this all applies to the other options as well.
2852     4. Since libtool seems to echo the compile commands it is issuing, the output
2853        from "make" can be reduced a bit by putting "@" in front of each libtool
2854        compile command.
2856     5. Patch from the folks at Google for configure.in to be a bit more thorough
2857        in checking for a suitable C++ installation before trying to compile the
2858        C++ stuff. This should fix a reported problem when a compiler was present,
2859        but no suitable headers.
2861     6. The man pages all had just "PCRE" as their title. I have changed them to
2862        be the relevant file name. I have also arranged that these names are
2863        retained in the file doc/pcre.txt, which is a concatenation in text format
2864        of all the man pages except the little individual ones for each function.
2866     7. The NON-UNIX-USE file had not been updated for the different set of source
2867        files that come with release 6. I also added a few comments about the C++
2868        wrapper.
2871    Version 6.0 07-Jun-05
2872    ---------------------
2874     1. Some minor internal re-organization to help with my DFA experiments.
2876     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2877        didn't matter for the library itself when fully configured, but did matter
2878        when compiling without UCP support, or within Exim, where the ucp files are
2879        not imported.
2881     3. Refactoring of the library code to split up the various functions into
2882        different source modules. The addition of the new DFA matching code (see
2883        below) to a single monolithic source would have made it really too
2884        unwieldy, quite apart from causing all the code to be include in a
2885        statically linked application, when only some functions are used. This is
2886        relevant even without the DFA addition now that patterns can be compiled in
2887        one application and matched in another.
2889        The downside of splitting up is that there have to be some external
2890        functions and data tables that are used internally in different modules of
2891        the library but which are not part of the API. These have all had their
2892        names changed to start with "_pcre_" so that they are unlikely to clash
2893        with other external names.
2895     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2896        a different (DFA) algorithm. Although it is slower than the original
2897        function, it does have some advantages for certain types of matching
2898        problem.
2900     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2901        including restarting after a partial match.
2903     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2904        defined when compiling for Windows was sent to me. I have put it into the
2905        code, though I have no means of testing or verifying it.
2907     7. Added the pcre_refcount() auxiliary function.
2909     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2910        match before or at the first newline in the subject string. In pcretest,
2911        the /f option on a pattern can be used to set this.
2913     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2914        would behave wrongly. This has been present in PCRE since release 4.0.
2916    10. A number of changes to the pcregrep command:
2918        (a) Refactored how -x works; insert ^(...)$ instead of setting
2919            PCRE_ANCHORED and checking the length, in preparation for adding
2920            something similar for -w.
2922        (b) Added the -w (match as a word) option.
2924        (c) Refactored the way lines are read and buffered so as to have more
2925            than one at a time available.
2927        (d) Implemented a pcregrep test script.
2929        (e) Added the -M (multiline match) option. This allows patterns to match
2930            over several lines of the subject. The buffering ensures that at least
2931            8K, or the rest of the document (whichever is the shorter) is available
2932            for matching (and similarly the previous 8K for lookbehind assertions).
2934        (f) Changed the --help output so that it now says
2936              -w, --word-regex(p)
2938            instead of two lines, one with "regex" and the other with "regexp"
2939            because that confused at least one person since the short forms are the
2940            same. (This required a bit of code, as the output is generated
2941            automatically from a table. It wasn't just a text change.)
2943        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2944            option but starts with a hyphen. Could be a pattern or a path name
2945            starting with a hyphen, for instance.
2947        (h) "-" can be given as a file name to represent stdin.
2949        (i) When file names are being printed, "(standard input)" is used for
2950            the standard input, for compatibility with GNU grep. Previously
2951            "<stdin>" was used.
2953        (j) The option --label=xxx can be used to supply a name to be used for
2954            stdin when file names are being printed. There is no short form.
2956        (k) Re-factored the options decoding logic because we are going to add
2957            two more options that take data. Such options can now be given in four
2958            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2960        (l) Added the -A, -B, and -C options for requesting that lines of context
2961            around matches be printed.
2963        (m) Added the -L option to print the names of files that do not contain
2964            any matching lines, that is, the complement of -l.
2966        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2967            continue to scan other files.
2969        (o) The -s option was incorrectly implemented. For compatibility with other
2970            greps, it now suppresses the error message for a non-existent or non-
2971            accessible file (but not the return code). There is a new option called
2972            -q that suppresses the output of matching lines, which was what -s was
2973            previously doing.
2975        (p) Added --include and --exclude options to specify files for inclusion
2976            and exclusion when recursing.
2978    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2979        Hopefully, it now does.
2981    12. Missing cast in pcre_study().
2983    13. Added an "uninstall" target to the makefile.
2985    14. Replaced "extern" in the function prototypes in Makefile.in with
2986        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2987        world, but is set differently for Windows.
2989    15. Added a second compiling function called pcre_compile2(). The only
2990        difference is that it has an extra argument, which is a pointer to an
2991        integer error code. When there is a compile-time failure, this is set
2992        non-zero, in addition to the error test pointer being set to point to an
2993        error message. The new argument may be NULL if no error number is required
2994        (but then you may as well call pcre_compile(), which is now just a
2995        wrapper). This facility is provided because some applications need a
2996        numeric error indication, but it has also enabled me to tidy up the way
2997        compile-time errors are handled in the POSIX wrapper.
2999    16. Added VPATH=.libs to the makefile; this should help when building with one
3000        prefix path and installing with another. (Or so I'm told by someone who
3001        knows more about this stuff than I do.)
3003    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3004        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3005        match everything, including newlines. This is not POSIX-compatible, but
3006        somebody wanted the feature. From pcretest it can be activated by using
3007        both the P and the s flags.
3009    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3011    19. libpcre.pc was being incorrectly installed as executable.
3013    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3014        it now also looks for '\r' so that it will work unmodified on Windows.
3016    21. Added Google's contributed C++ wrapper to the distribution.
3018    22. Added some untidy missing memory free() calls in pcretest, to keep
3019        Electric Fence happy when testing.
3023    Version 5.0 13-Sep-04
3024    ---------------------
3026     1. Internal change: literal characters are no longer packed up into items
3027        containing multiple characters in a single byte-string. Each character
3028        is now matched using a separate opcode. However, there may be more than one
3029        byte in the character in UTF-8 mode.
3031     2. The pcre_callout_block structure has two new fields: pattern_position and
3032        next_item_length. These contain the offset in the pattern to the next match
3033        item, and its length, respectively.
3035     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3036        insertion of callouts before each pattern item. Added the /C option to
3037        pcretest to make use of this.
3039     4. On the advice of a Windows user, the lines
3041          #if defined(_WIN32) || defined(WIN32)
3042          _setmode( _fileno( stdout ), 0x8000 );
3043          #endif  /* defined(_WIN32) || defined(WIN32) */
3045        have been added to the source of pcretest. This apparently does useful
3046        magic in relation to line terminators.
3048     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3049        for the benefit of those environments where the "b" makes a difference.
3051     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3052        to know about it. I have put a hack into configure.in that adds in code
3053        to set GCC=yes if CC=icc. This seems to end up at a point in the
3054        generated configure script that is early enough to affect the setting of
3055        compiler options, which is what is needed, but I have no means of testing
3056        whether it really works. (The user who reported this had patched the
3057        generated configure script, which of course I cannot do.)
3059        LATER: After change 22 below (new libtool files), the configure script
3060        seems to know about icc (and also ecc). Therefore, I have commented out
3061        this hack in configure.in.
3063     7. Added support for pkg-config (2 patches were sent in).
3065     8. Negated POSIX character classes that used a combination of internal tables
3066        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3067        [[:^ascii]]. Typically, they would match almost any characters. The other
3068        POSIX classes were not broken in this way.
3070     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3071        to find the match, as PCRE was deluded into thinking that the match had to
3072        start at the start point or following a newline. The same bug applied to
3073        patterns with negative forward assertions or any backward assertions
3074        preceding ".*" at the start, unless the pattern required a fixed first
3075        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3077    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3078        starting at the last subject character, bytes beyond the end of the subject
3079        string were read.
3081    11. Renamed the variable "class" as "classbits" to make life easier for C++
3082        users. (Previously there was a macro definition, but it apparently wasn't
3083        enough.)
3085    12. Added the new field "tables" to the extra data so that tables can be passed
3086        in at exec time, or the internal tables can be re-selected. This allows
3087        a compiled regex to be saved and re-used at a later time by a different
3088        program that might have everything at different addresses.
3090    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3091        -R library as well as a -L library.
3093    14. The debugging options of pcretest (-d on the command line or D on a
3094        pattern) showed incorrect output for anything following an extended class
3095        that contained multibyte characters and which was followed by a quantifier.
3097    15. Added optional support for general category Unicode character properties
3098        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3099        support. It adds about 90K to the size of the library. The meanings of the
3100        inbuilt class escapes such as \d and \s have NOT been changed.
3102    16. Updated pcredemo.c to include calls to free() to release the memory for the
3103        compiled pattern.
3105    17. The generated file chartables.c was being created in the source directory
3106        instead of in the building directory. This caused the build to fail if the
3107        source directory was different from the building directory, and was
3108        read-only.
3110    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3111        file. No doubt somebody will tell me if they don't make sense... Also added
3112        Dan Mooney's comments about building on OpenVMS.
3114    19. Added support for partial matching via the PCRE_PARTIAL option for
3115        pcre_exec() and the \P data escape in pcretest.
3117    20. Extended pcretest with 3 new pattern features:
3119        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3120              write the compiled pattern to the file whose name is "rest-of-line".
3121              This is a straight binary dump of the data, with the saved pointer to
3122              the character tables forced to be NULL. The study data, if any, is
3123              written too. After writing, pcretest reads a new pattern.
3125        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3126              compiled pattern from the given file. There must not be any
3127              occurrences of "<" in the file name (pretty unlikely); if there are,
3128              pcretest will instead treat the initial "<" as a pattern delimiter.
3129              After reading in the pattern, pcretest goes on to read data lines as
3130              usual.
3132        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3133              and 16-bit fields in a compiled pattern, to simulate a pattern that
3134              was compiled on a host of opposite endianness.
3136    21. The pcre-exec() function can now cope with patterns that were compiled on
3137        hosts of opposite endianness, with this restriction:
3139          As for any compiled expression that is saved and used later, the tables
3140          pointer field cannot be preserved; the extra_data field in the arguments
3141          to pcre_exec() should be used to pass in a tables address if a value
3142          other than the default internal tables were used at compile time.
3144    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3145        now diagnosed as an error. Previously, most of the time, a negative number
3146        would have been treated as zero, but if in addition "ovector" was passed as
3147        NULL, a crash could occur.
3149    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3150        new versions from the libtool 1.5 distribution (the last one is a copy of
3151        a file called libtool.m4). This seems to have fixed the need to patch
3152        "configure" to support Darwin 1.3 (which I used to do). However, I still
3153        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3154        workstation).
3156    24. Changed the PCRE licence to be the more standard "BSD" licence.
3159    Version 4.5 01-Dec-03
3160    ---------------------
3162     1. There has been some re-arrangement of the code for the match() function so
3163        that it can be compiled in a version that does not call itself recursively.
3164        Instead, it keeps those local variables that need separate instances for
3165        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3166        needs to "recurse". Keeping track of where control must go is done by means
3167        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3168        hide most of the details from the main code, and operates only if
3169        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3170        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3171        operating.
3173        To make it easier for callers to provide specially tailored get/free
3174        functions for this usage, two new functions, pcre_stack_malloc, and
3175        pcre_stack_free, are used. They are always called in strict stacking order,
3176        and the size of block requested is always the same.
3178        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3179        PCRE has been compiled to use the stack or the heap for recursion. The
3180        -C option of pcretest uses this to show which version is compiled.
3182        A new data escape \S, is added to pcretest; it causes the amounts of store
3183        obtained and freed by both kinds of malloc/free at match time to be added
3184        to the output.
3186     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3187        what's available on my current Linux desktop machine.
3189     3. When matching a UTF-8 string, the test for a valid string at the start has
3190        been extended. If start_offset is not zero, PCRE now checks that it points
3191        to a byte that is the start of a UTF-8 character. If not, it returns
3192        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3193        this is necessary because there may be backward assertions in the pattern.
3194        When matching the same subject several times, it may save resources to use
3195        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3197     4. The code for checking the validity of UTF-8 strings has been tightened so
3198        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3199        containing "overlong sequences".
3201     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3202        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3203        should have been "&", but it just so happened that all the cases this let
3204        through by mistake were picked up later in the function.
3206     6. I had used a variable called "isblank" - this is a C99 function, causing
3207        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3209     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3210        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3211        so the version is shown; (c) stop "make test" echoing "./RunTest".
3213     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3215     9. The prototype for memmove() for systems that don't have it was using
3216        size_t, but the inclusion of the header that defines size_t was later. I've
3217        moved the #includes for the C headers earlier to avoid this.
3219    10. Added some adjustments to the code to make it easier to compiler on certain
3220        special systems:
3222          (a) Some "const" qualifiers were missing.
3223          (b) Added the macro EXPORT before all exported functions; by default this
3224              is defined to be empty.
3225          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3226              that it reads its output file name as an argument instead of writing
3227              to the standard output and assuming this can be redirected.
3229    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3230        class containing characters with values greater than 255, PCRE compilation
3231        went into a loop.
3233    12. A recursive reference to a subpattern that was within another subpattern
3234        that had a minimum quantifier of zero caused PCRE to crash. For example,
3235        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3236        recursion. If the recursively-called subpattern itself had a zero repeat,
3237        that was OK.
3239    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3240        buffer into which it was copied (for escape processing) was still set at
3241        1024, so long lines caused crashes.
3243    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3244        "internal error: code overflow...". This applied to any character class
3245        that was followed by a possessive quantifier.
3247    15. Modified the Makefile to add libpcre.la as a prerequisite for
3248        libpcreposix.la because I was told this is needed for a parallel build to
3249        work.
3251    16. If a pattern that contained .* following optional items at the start was
3252        studied, the wrong optimizing data was generated, leading to matching
3253        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3254        matching string must start with a or b or c. The correct conclusion for
3255        this pattern is that a match can start with any character.
3258    Version 4.4 13-Aug-03
3259    ---------------------
3261     1. In UTF-8 mode, a character class containing characters with values between
3262        127 and 255 was not handled correctly if the compiled pattern was studied.
3263        In fixing this, I have also improved the studying algorithm for such
3264        classes (slightly).
3266     2. Three internal functions had redundant arguments passed to them. Removal
3267        might give a very teeny performance improvement.
3269     3. Documentation bug: the value of the capture_top field in a callout is *one
3270        more than* the number of the hightest numbered captured substring.
3272     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3273        in incorrectly linking with a previously installed version. They now link
3274        explicitly with libpcre.la.
3276     5. configure.in no longer needs to recognize Cygwin specially.
3278     6. A problem in pcre.in for Windows platforms is fixed.
3280     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3281        pcretest, it used to include the size of the study block as part of its
3282        output. Unfortunately, the structure contains a field that has a different
3283        size on different hardware architectures. This meant that the tests that
3284        showed this size failed. As the block is currently always of a fixed size,
3285        this information isn't actually particularly useful in pcretest output, so
3286        I have just removed it.
3288     8. Three pre-processor statements accidentally did not start in column 1.
3289        Sadly, there are *still* compilers around that complain, even though
3290        standard C has not required this for well over a decade. Sigh.
3292     9. In pcretest, the code for checking callouts passed small integers in the
3293        callout_data field, which is a void * field. However, some picky compilers
3294        complained about the casts involved for this on 64-bit systems. Now
3295        pcretest passes the address of the small integer instead, which should get
3296        rid of the warnings.
3298    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3299        both compile and run time, and gives an error if an invalid UTF-8 sequence
3300        is found. There is a option for disabling this check in cases where the
3301        string is known to be correct and/or the maximum performance is wanted.
3303    11. In response to a bug report, I changed one line in Makefile.in from
3305            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3306        to
3307            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3309        to look similar to other lines, but I have no way of telling whether this
3310        is the right thing to do, as I do not use Windows. No doubt I'll get told
3311        if it's wrong...
3314    Version 4.3 21-May-03
3315    ---------------------
3317    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3318       Makefile.
3320    2. Some refactoring to improve the quality of the code:
3322       (i)   The utf8_table... variables are now declared "const".
3324       (ii)  The code for \cx, which used the "case flipping" table to upper case
3325             lower case letters, now just substracts 32. This is ASCII-specific,
3326             but the whole concept of \cx is ASCII-specific, so it seems
3327             reasonable.
3329       (iii) PCRE was using its character types table to recognize decimal and
3330             hexadecimal digits in the pattern. This is silly, because it handles
3331             only 0-9, a-f, and A-F, but the character types table is locale-
3332             specific, which means strange things might happen. A private
3333             table is now used for this - though it costs 256 bytes, a table is
3334             much faster than multiple explicit tests. Of course, the standard
3335             character types table is still used for matching digits in subject
3336             strings against \d.
3338       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3339             ending in _t are). So I've renamed it as ESC_tee.
3341    3. The first argument for regexec() in the POSIX wrapper should have been
3342       defined as "const".
3344    4. Changed pcretest to use malloc() for its buffers so that they can be
3345       Electric Fenced for debugging.
3347    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3348       to read one or more bytes before the start of the subject string. Often this
3349       had no effect on PCRE's behaviour, but in some circumstances it could
3350       provoke a segmentation fault.
3352    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3353       to try to read one or more bytes before the start of the subject string.
3355    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3356       UTF-8 support could misbehave in various ways if the subject string
3357       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3358       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3359       back over UTF-8 characters.)
3362    Version 4.2 14-Apr-03
3363    ---------------------
3365    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3367    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3368         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3369         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3370         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3371         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3372           and BUILD_EXEEXT
3373         Note: automatic setting of the BUILD variables is not yet working
3374         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3375           compile-time but not at link-time
3376         [LINK]: use for linking executables only
3377         make different versions for Windows and non-Windows
3378         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3379           libraries
3380         [LINK_FOR_BUILD]: new variable
3381         [OBJEXT]: use throughout
3382         [EXEEXT]: use throughout
3383         <winshared>: new target
3384         <wininstall>: new target
3385         <dftables.o>: use native compiler
3386         <dftables>: use native linker
3387         <install>: handle Windows platform correctly
3388         <clean>: ditto
3389         <check>: ditto
3390         copy DLL to top builddir before testing
3392       As part of these changes, -no-undefined was removed again. This was reported
3393       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3394       in any case.
3396    3. Some tidies to get rid of compiler warnings:
3398       . In the match_data structure, match_limit was an unsigned long int, whereas
3399         match_call_count was an int. I've made them both unsigned long ints.
3401       . In pcretest the fact that a const uschar * doesn't automatically cast to
3402         a void * provoked a warning.
3404       . Turning on some more compiler warnings threw up some "shadow" variables
3405         and a few more missing casts.
3407    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3408       option, a class that contained a single character with a value between 128
3409       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3411    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3412       option, a class that contained several characters, but with at least one
3413       whose value was between 128 and 255 caused PCRE to crash.
3416    Version 4.1 12-Mar-03
3417    ---------------------
3419    1. Compiling with gcc -pedantic found a couple of places where casts were
3420    needed, and a string in dftables.c that was longer than standard compilers are
3421    required to support.
3423    2. Compiling with Sun's compiler found a few more places where the code could
3424    be tidied up in order to avoid warnings.
3426    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3427    first of these names is deprecated in the latest Autoconf in favour of the name
3428    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3429    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3430    analogy I have changed it to CFLAGS_FOR_BUILD.
3432    4. Added -no-undefined to the linking command in the Makefile, because this is
3433    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3434    linking step for the pcreposix library.
3436    5. PCRE was failing to diagnose the case of two named groups with the same
3437    name.
3439    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3440    literal character that is needed in the subject for a match, and scans along to
3441    ensure that it is present before embarking on the full matching process. This
3442    saves time in cases of nested unlimited repeats that are never going to match.
3443    Problem: the scan can take a lot of time if the subject is very long (e.g.
3444    megabytes), thus penalizing straightforward matches. It is now done only if the
3445    amount of subject to be scanned is less than 1000 bytes.
3447    7. A lesser problem with the same optimization is that it was recording the
3448    first character of an anchored pattern as "needed", thus provoking a search
3449    right along the subject, even when the first match of the pattern was going to
3450    fail. The "needed" character is now not set for anchored patterns, unless it
3451    follows something in the pattern that is of non-fixed length. Thus, it still
3452    fulfils its original purpose of finding quick non-matches in cases of nested
3453    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3456    Version 4.0 17-Feb-03
3457    ---------------------
3459    1. If a comment in an extended regex that started immediately after a meta-item
3460    extended to the end of string, PCRE compiled incorrect data. This could lead to
3461    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3463    2. Moved to autoconf 2.53 and libtool 1.4.2.
3465    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3466    the special perltest8 script is no longer needed - all the tests can be run
3467    from a single perltest script.
3469    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3470    by \s. It has now been removed in PCRE. This means it isn't recognized as
3471    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3472    class [:space:] *does* include VT, thereby creating a mess.
3474    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3475    space and tab.
3477    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3478    its new features into the main test script, reducing the number of scripts.
3480    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3481    were backward compatible, and made the (?i) apply to the whole pattern, as if
3482    /i were given. Now it behaves more logically, and applies the option setting
3483    only to what follows. PCRE has been changed to follow suit. However, if it
3484    finds options settings right at the start of the pattern, it extracts them into
3485    the global options, as before. Thus, they show up in the info data.
3487    8. Added support for the \Q...\E escape sequence. Characters in between are
3488    treated as literals. This is slightly different from Perl in that $ and @ are
3489    also handled as literals inside the quotes. In Perl, they will cause variable
3490    interpolation. Note the following examples:
3492        Pattern            PCRE matches      Perl matches
3494        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3495        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3496        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3498    For compatibility with Perl, \Q...\E sequences are recognized inside character
3499    classes as well as outside them.
3501    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3502    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3503    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3504    signed/unsigned warnings.
3506    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3507    option for pcretest, so I've replaced it by a simple function that does just
3508    that job.
3510    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3511    "pcregrep -".
3513    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3514    Java package. This provides some syntactic sugar for simple cases of what my
3515    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3516    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3517    item, you can use this simplified notation. Note that only makes sense with
3518    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3519    greediness, whatever the setting of the PCRE_UNGREEDY option.
3521    13. A change of greediness default within a pattern was not taking effect at
3522    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3523    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3524    was abstracted outside.
3526    14. PCRE now supports the \G assertion. It is true when the current matching
3527    position is at the start point of the match. This differs from \A when the
3528    starting offset is non-zero. Used with the /g option of pcretest (or similar
3529    code), it works in the same way as it does for Perl's /g option. If all
3530    alternatives of a regex begin with \G, the expression is anchored to the start
3531    match position, and the "anchored" flag is set in the compiled expression.