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

Diff of /code/trunk/ChangeLog

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

revision 7 by nigel, Sat Feb 24 21:38:09 2007 UTC revision 890 by ph10, Wed Jan 18 16:25:19 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.30
5    ------------
6    
7    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
8        name is defined in ctype.h.
9    
10    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
11        only in quite long subpatterns.
12    
13    3.  Removed the function pcre_info(), which has been obsolete and deprecated
14        since it was replaced by pcre_fullinfo() in February 2000.
15    
16    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
17        match a (*MARK), and the match failed at the start of the subject, a
18        reference to memory before the start of the subject could occur. This bug
19        was introduced by fix 17 of release 8.21.
20    
21    5.  A reference to an unset group with zero minimum repetition was giving
22        totally wrong answers (in non-JavaScript-compatibility mode). For example,
23        /(another)?(\1?)test/ matched against "hello world test". This bug was
24        introduced in release 8.13.
25    
26    6.  Add support for 16-bit character strings (a large amount of work involving
27        many changes and refactorings).
28    
29    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
30        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
31        from a file.
32    
33    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
34        rounding is not applied in this particular case).
35    
36    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
37        if they appear, or are escaped, in patterns.
38    
39    10. Get rid of a number of -Wunused-but-set-variable warnings.
40    
41    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
42        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
43        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
44        also returns the mark "x". This bug applied to capturing parentheses,
45        non-capturing parentheses, and atomic parentheses. It also applied to some
46        assertions.
47    
48    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
49        information out of configure.ac instead of relying on pcre.h.generic, which
50        is not stored in the repository.
51    
52    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
53        -lreadline.
54    
55    14. ZH added PCRE_CONFIG_JITTARGET; added it's output to pcretest -C.
56    
57    
58    Version 8.21 12-Dec-2011
59    ------------------------
60    
61    1.  Updating the JIT compiler.
62    
63    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
64        are added as well.
65    
66    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
67        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
68        calling _pcre_jit_exec. Some extra comments are added.
69    
70    4.  (*MARK) settings inside atomic groups that do not contain any capturing
71        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
72        was introduced by change 18 for 8.20.
73    
74    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
75        ECMA-262 standard.
76    
77    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
78        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
79        This bug was probably introduced by change 9 of 8.13.
80    
81    7.  While fixing 6 above, I noticed that a number of other items were being
82        incorrectly rejected as "not fixed length". This arose partly because newer
83        opcodes had not been added to the fixed-length checking code. I have (a)
84        corrected the bug and added tests for these items, and (b) arranged for an
85        error to occur if an unknown opcode is encountered while checking for fixed
86        length instead of just assuming "not fixed length". The items that were
87        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
88        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
89        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
90    
91    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
92        being incorrectly compiled and would have given unpredicatble results.
93    
94    9.  A possessively repeated subpattern with minimum repeat count greater than
95        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
96        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
97        the first (A) could occur when it should not.
98    
99    10. Add a cast and remove a redundant test from the code.
100    
101    11. JIT should use pcre_malloc/pcre_free for allocation.
102    
103    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
104        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
105        is anything other than /usr, -L is still shown).
106    
107    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
108    
109    14. Perl does not support \N without a following name in a [] class; PCRE now
110        also gives an error.
111    
112    15. If a forward reference was repeated with an upper limit of around 2000,
113        it caused the error "internal error: overran compiling workspace". The
114        maximum number of forward references (including repeats) was limited by the
115        internal workspace, and dependent on the LINK_SIZE. The code has been
116        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
117        the default depends on LINK_SIZE. There is a new upper limit (for safety)
118        of around 200,000 forward references. While doing this, I also speeded up
119        the filling in of repeated forward references.
120    
121    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
122        incorrectly expecting the subject to contain another "a" after the start.
123    
124    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
125        in the match, the SKIP should be ignored. This was not happening; instead
126        the SKIP was being treated as NOMATCH. For patterns such as
127        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
128        tested.
129    
130    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
131        now much more compatible with Perl, in particular in cases where the result
132        is a non-match for a non-anchored pattern. For example, if
133        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
134        "m", where previously it did not return a name. A side effect of this
135        change is that for partial matches, the last encountered mark name is
136        returned, as for non matches. A number of tests that were previously not
137        Perl-compatible have been moved into the Perl-compatible test files. The
138        refactoring has had the pleasing side effect of removing one argument from
139        the match() function, thus reducing its stack requirements.
140    
141    19. If the /S+ option was used in pcretest to study a pattern using JIT,
142        subsequent uses of /S (without +) incorrectly behaved like /S+.
143    
144    21. Retrieve executable code size support for the JIT compiler and fixing
145        some warnings.
146    
147    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
148        not work when the shorter character appeared right at the end of the
149        subject string.
150    
151    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
152        systems.
153    
154    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
155        output it when the /M option is used in pcretest.
156    
157    25. The CheckMan script was not being included in the distribution. Also, added
158        an explicit "perl" to run Perl scripts from the PrepareRelease script
159        because this is reportedly needed in Windows.
160    
161    26. If study data was being save in a file and studying had not found a set of
162        "starts with" bytes for the pattern, the data written to the file (though
163        never used) was taken from uninitialized memory and so caused valgrind to
164        complain.
165    
166    27. Updated RunTest.bat as provided by Sheri Pierce.
167    
168    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
169    
170    29. Computation of memory usage for the table of capturing group names was
171        giving an unnecessarily large value.
172    
173    
174    Version 8.20 21-Oct-2011
175    ------------------------
176    
177    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
178        a POSIX class. After further experiments with Perl, which convinced me that
179        Perl has bugs and confusions, a closing square bracket is no longer allowed
180        in a POSIX name. This bug also affected patterns with classes that started
181        with full stops.
182    
183    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
184        captured substring, but while checking the failing first alternative,
185        substring 1 is temporarily captured. If the output vector supplied to
186        pcre_exec() was not big enough for this capture, the yield of the function
187        was still zero ("insufficient space for captured substrings"). This cannot
188        be totally fixed without adding another stack variable, which seems a lot
189        of expense for a edge case. However, I have improved the situation in cases
190        such as /(a)(b)x|abc/ matched against "abc", where the return code
191        indicates that fewer than the maximum number of slots in the ovector have
192        been set.
193    
194    3.  Related to (2) above: when there are more back references in a pattern than
195        slots in the output vector, pcre_exec() uses temporary memory during
196        matching, and copies in the captures as far as possible afterwards. It was
197        using the entire output vector, but this conflicts with the specification
198        that only 2/3 is used for passing back captured substrings. Now it uses
199        only the first 2/3, for compatibility. This is, of course, another edge
200        case.
201    
202    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
203        main code base, and can be used by building with --enable-jit. When this is
204        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
205        runtime --no-jit option is given.
206    
207    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
208        ovector, the return from the function was zero, implying that there were
209        other matches that did not fit. The correct "exactly full" value is now
210        returned.
211    
212    6.  If a subpattern that was called recursively or as a subroutine contained
213        (*PRUNE) or any other control that caused it to give a non-standard return,
214        invalid errors such as "Error -26 (nested recursion at the same subject
215        position)" or even infinite loops could occur.
216    
217    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
218        computing the minimum length on reaching *ACCEPT, and so ended up with the
219        wrong value of 1 rather than 0. Further investigation indicates that
220        computing a minimum subject length in the presence of *ACCEPT is difficult
221        (think back references, subroutine calls), and so I have changed the code
222        so that no minimum is registered for a pattern that contains *ACCEPT.
223    
224    8.  If (*THEN) was present in the first (true) branch of a conditional group,
225        it was not handled as intended. [But see 16 below.]
226    
227    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
228        Sheri Pierce.
229    
230    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
231        the first byte in a match must be "a".
232    
233    11. Change 17 for 8.13 increased the recursion depth for patterns like
234        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
235        pattern contains any instances of (*THEN). If it does not, the old
236        optimizations are restored. It would be nice to do this on a per-group
237        basis, but at the moment that is not feasible.
238    
239    12. In some environments, the output of pcretest -C is CRLF terminated. This
240        broke RunTest's code that checks for the link size. A single white space
241        character after the value is now allowed for.
242    
243    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
244        For "fr", it uses the Windows-specific input and output files.
245    
246    14. If (*THEN) appeared in a group that was called recursively or as a
247        subroutine, it did not work as intended. [But see next item.]
248    
249    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
250        pattern fragments (but not containing any | characters). If A and B are
251        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
252        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
253        D. In other words, Perl considers parentheses that do not contain any |
254        characters to be part of a surrounding alternative, whereas PCRE was
255        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
256        differently. PCRE now behaves in the same way as Perl, except in the case
257        of subroutine/recursion calls such as (?1) which have in any case always
258        been different (but PCRE had them first :-).
259    
260    16. Related to 15 above: Perl does not treat the | in a conditional group as
261        creating alternatives. Such a group is treated in the same way as an
262        ordinary group without any | characters when processing (*THEN). PCRE has
263        been changed to match Perl's behaviour.
264    
265    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
266        RunGrepTest script failed.
267    
268    18. Change 22 for version 13 caused atomic groups to use more stack. This is
269        inevitable for groups that contain captures, but it can lead to a lot of
270        stack use in large patterns. The old behaviour has been restored for atomic
271        groups that do not contain any capturing parentheses.
272    
273    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
274        suppress the check for a minimum subject length at run time. (If it was
275        given to pcre_exec() or pcre_dfa_exec() it did work.)
276    
277    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
278        fail to work when decoding hex characters in data strings in EBCDIC
279        environments.
280    
281    21. It appears that in at least one Mac OS environment, the isxdigit() function
282        is implemented as a macro that evaluates to its argument more than once,
283        contravening the C 90 Standard (I haven't checked a later standard). There
284        was an instance in pcretest which caused it to go wrong when processing
285        \x{...} escapes in subject strings. The has been rewritten to avoid using
286        things like p++ in the argument of isxdigit().
287    
288    
289    Version 8.13 16-Aug-2011
290    ------------------------
291    
292    1.  The Unicode data tables have been updated to Unicode 6.0.0.
293    
294    2.  Two minor typos in pcre_internal.h have been fixed.
295    
296    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
297        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
298        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
299    
300    4.  There were a number of related bugs in the code for matching backrefences
301        caselessly in UTF-8 mode when codes for the characters concerned were
302        different numbers of bytes. For example, U+023A and U+2C65 are an upper
303        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
304        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
305        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
306        2-byte code at the end of the subject (it thought there wasn't enough data
307        left).
308    
309    5.  Comprehensive information about what went wrong is now returned by
310        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
311        as the output vector has at least 2 elements. The offset of the start of
312        the failing character and a reason code are placed in the vector.
313    
314    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
315        now returned is for the first byte of the failing character, instead of the
316        last byte inspected. This is an incompatible change, but I hope it is small
317        enough not to be a problem. It makes the returned offset consistent with
318        pcre_exec() and pcre_dfa_exec().
319    
320    7.  pcretest now gives a text phrase as well as the error number when
321        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
322        failure, the offset and reason code are output.
323    
324    8.  When \R was used with a maximizing quantifier it failed to skip backwards
325        over a \r\n pair if the subsequent match failed. Instead, it just skipped
326        back over a single character (\n). This seems wrong (because it treated the
327        two characters as a single entity when going forwards), conflicts with the
328        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
329        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
330        has been changed.
331    
332    9.  Some internal refactoring has changed the processing so that the handling
333        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
334        time (the PCRE_DOTALL option was changed this way some time ago: version
335        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
336        which was always a bit of a fudge. It also means that there is one less
337        argument for the match() function, which reduces its stack requirements
338        slightly. This change also fixes an incompatibility with Perl: the pattern
339        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
340    
341    10. More internal refactoring has drastically reduced the number of recursive
342        calls to match() for possessively repeated groups such as (abc)++ when
343        using pcre_exec().
344    
345    11. While implementing 10, a number of bugs in the handling of groups were
346        discovered and fixed:
347    
348        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
349        (a|)*(?1) gave a compile-time internal error.
350        ((a|)+)+  did not notice that the outer group could match an empty string.
351        (^a|^)+   was not marked as anchored.
352        (.*a|.*)+ was not marked as matching at start or after a newline.
353    
354    12. Yet more internal refactoring has removed another argument from the match()
355        function. Special calls to this function are now indicated by setting a
356        value in a variable in the "match data" data block.
357    
358    13. Be more explicit in pcre_study() instead of relying on "default" for
359        opcodes that mean there is no starting character; this means that when new
360        ones are added and accidentally left out of pcre_study(), testing should
361        pick them up.
362    
363    14. The -s option of pcretest has been documented for ages as being an old
364        synonym of -m (show memory usage). I have changed it to mean "force study
365        for every regex", that is, assume /S for every regex. This is similar to -i
366        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
367        using it. It makes it easier to run collections of tests with and without
368        study enabled, and thereby test pcre_study() more easily. All the standard
369        tests are now run with and without -s (but some patterns can be marked as
370        "never study" - see 20 below).
371    
372    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
373        restoration of the capturing data to the outer values was not happening
374        correctly.
375    
376    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
377        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
378        pattern had matched an empty string, and so incorrectly returned a no
379        match.
380    
381    17. There was optimizing code for the last branch of non-capturing parentheses,
382        and also for the obeyed branch of a conditional subexpression, which used
383        tail recursion to cut down on stack usage. Unfortunately, now that there is
384        the possibility of (*THEN) occurring in these branches, tail recursion is
385        no longer possible because the return has to be checked for (*THEN). These
386        two optimizations have therefore been removed. [But see 8.20/11 above.]
387    
388    18. If a pattern containing \R was studied, it was assumed that \R always
389        matched two bytes, thus causing the minimum subject length to be
390        incorrectly computed because \R can also match just one byte.
391    
392    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
393        was incorrectly computed.
394    
395    20. If /S is present twice on a test pattern in pcretest input, it now
396        *disables* studying, thereby overriding the use of -s on the command line
397        (see 14 above). This is necessary for one or two tests to keep the output
398        identical in both cases.
399    
400    21. When (*ACCEPT) was used in an assertion that matched an empty string and
401        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
402    
403    22. When an atomic group that contained a capturing parenthesis was
404        successfully matched, but the branch in which it appeared failed, the
405        capturing was not being forgotten if a higher numbered group was later
406        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
407        group 1 to "a", when in fact it should be unset. This applied to multi-
408        branched capturing and non-capturing groups, repeated or not, and also to
409        positive assertions (capturing in negative assertions does not happen
410        in PCRE) and also to nested atomic groups.
411    
412    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
413        subject after a captured substring, to make it easier to tell which of a
414        number of identical substrings has been captured.
415    
416    24. The way atomic groups are processed by pcre_exec() has been changed so that
417        if they are repeated, backtracking one repetition now resets captured
418        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
419        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
420        "aaa". Previously, it would have been "a". As part of this code
421        refactoring, the way recursive calls are handled has also been changed.
422    
423    25. If an assertion condition captured any substrings, they were not passed
424        back unless some other capturing happened later. For example, if
425        (?(?=(a))a) was matched against "a", no capturing was returned.
426    
427    26. When studying a pattern that contained subroutine calls or assertions,
428        the code for finding the minimum length of a possible match was handling
429        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
430        group 1 called group 2 while simultaneously a separate group 2 called group
431        1). A stack overflow occurred in this case. I have fixed this by limiting
432        the recursion depth to 10.
433    
434    27. Updated RunTest.bat in the distribution to the version supplied by Tom
435        Fortmann. This supports explicit test numbers on the command line, and has
436        argument validation and error reporting.
437    
438    28. An instance of \X with an unlimited repeat could fail if at any point the
439        first character it looked at was a mark character.
440    
441    29. Some minor code refactoring concerning Unicode properties and scripts
442        should reduce the stack requirement of match() slightly.
443    
444    30. Added the '=' option to pcretest to check the setting of unused capturing
445        slots at the end of the pattern, which are documented as being -1, but are
446        not included in the return count.
447    
448    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
449        compiled something random. Now it gives a compile-time error (as does
450        Perl).
451    
452    32. A *MARK encountered during the processing of a positive assertion is now
453        recorded and passed back (compatible with Perl).
454    
455    33. If --only-matching or --colour was set on a pcregrep call whose pattern
456        had alternative anchored branches, the search for a second match in a line
457        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
458        matched the line "0102" twice. The same bug affected patterns that started
459        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
460        twice.
461    
462    34. Previously, PCRE did not allow quantification of assertions. However, Perl
463        does, and because of capturing effects, quantifying parenthesized
464        assertions may at times be useful. Quantifiers are now allowed for
465        parenthesized assertions.
466    
467    35. A minor code tidy in pcre_compile() when checking options for \R usage.
468    
469    36. \g was being checked for fancy things in a character class, when it should
470        just be a literal "g".
471    
472    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
473        appearance of a nested POSIX class supersedes an apparent external class.
474        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
475        unescaped square brackets may also appear as part of class names. For
476        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
477        more like Perl. (But see 8.20/1 above.)
478    
479    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
480        was because it thought it was \N{name}, which is not supported).
481    
482    39. Add minix to OS list not supporting the -S option in pcretest.
483    
484    40. PCRE tries to detect cases of infinite recursion at compile time, but it
485        cannot analyze patterns in sufficient detail to catch mutual recursions
486        such as ((?1))((?2)). There is now a runtime test that gives an error if a
487        subgroup is called recursively as a subpattern for a second time at the
488        same position in the subject string. In previous releases this might have
489        been caught by the recursion limit, or it might have run out of stack.
490    
491    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
492        happen only once. PCRE was, however incorrectly giving a compile time error
493        "recursive call could loop indefinitely" because it cannot analyze the
494        pattern in sufficient detail. The compile time test no longer happens when
495        PCRE is compiling a conditional subpattern, but actual runaway loops are
496        now caught at runtime (see 40 above).
497    
498    42. It seems that Perl allows any characters other than a closing parenthesis
499        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
500        has been changed to be the same.
501    
502    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
503        as not to get warnings when autogen.sh is called. Also changed
504        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
505    
506    44. To help people who use pcregrep to scan files containing exceedingly long
507        lines, the following changes have been made:
508    
509        (a) The default value of the buffer size parameter has been increased from
510            8K to 20K. (The actual buffer used is three times this size.)
511    
512        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
513            PCRE is built.
514    
515        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
516            to be set at run time.
517    
518        (d) Numerical values in pcregrep options can be followed by K or M, for
519            example --buffer-size=50K.
520    
521        (e) If a line being scanned overflows pcregrep's buffer, an error is now
522            given and the return code is set to 2.
523    
524    45. Add a pointer to the latest mark to the callout data block.
525    
526    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
527        partial match of an empty string instead of no match. This was specific to
528        the use of ".".
529    
530    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
531        complete match instead of a partial match. This bug was dependent on both
532        the PCRE_UTF8 and PCRE_DOTALL options being set.
533    
534    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
535        starting byte set, because \b was not being ignored.
536    
537    
538    Version 8.12 15-Jan-2011
539    ------------------------
540    
541    1.  Fixed some typos in the markup of the man pages, and wrote a script that
542        checks for such things as part of the documentation building process.
543    
544    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
545        --match-limit and --recursion-limit options (added for 8.11). In
546        particular, this made one of the standard tests fail. (The integer value
547        went into the wrong half of a long int.)
548    
549    3.  If the --colour option was given to pcregrep with -v (invert match), it
550        did strange things, either producing crazy output, or crashing. It should,
551        of course, ignore a request for colour when reporting lines that do not
552        match.
553    
554    4.  Another pcregrep bug caused similar problems if --colour was specified with
555        -M (multiline) and the pattern match finished with a line ending.
556    
557    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
558        matched in multiline mode, the following line was shown as part of the
559        match. This seems wrong, so I have changed it.
560    
561    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
562        the check for further matches in the same line (so they could be coloured)
563        to overrun the end of the current line. If another match was found, it was
564        incorrectly shown (and then shown again when found in the next line).
565    
566    7.  If pcregrep was compiled under Windows, there was a reference to the
567        function pcregrep_exit() before it was defined. I am assuming this was
568        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
569        reported by a user. I've moved the definition above the reference.
570    
571    
572    Version 8.11 10-Dec-2010
573    ------------------------
574    
575    1.  (*THEN) was not working properly if there were untried alternatives prior
576        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
577        backtracked to try for "b" instead of moving to the next alternative branch
578        at the same level (in this case, to look for "c"). The Perl documentation
579        is clear that when (*THEN) is backtracked onto, it goes to the "next
580        alternative in the innermost enclosing group".
581    
582    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
583        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
584        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
585        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
586        (*THEN).
587    
588    3.  If \s appeared in a character class, it removed the VT character from
589        the class, even if it had been included by some previous item, for example
590        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
591        of \s, but is part of the POSIX "space" class.)
592    
593    4.  A partial match never returns an empty string (because you can always
594        match an empty string at the end of the subject); however the checking for
595        an empty string was starting at the "start of match" point. This has been
596        changed to the "earliest inspected character" point, because the returned
597        data for a partial match starts at this character. This means that, for
598        example, /(?<=abc)def/ gives a partial match for the subject "abc"
599        (previously it gave "no match").
600    
601    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
602        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
603        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
604        has an implication that the given string is incomplete (because a partial
605        match is preferred over a full match). For this reason, these items now
606        give a partial match in this situation. [Aside: previously, the one case
607        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
608        match rather than a full match, which was wrong by the old rules, but is
609        now correct.]
610    
611    6.  There was a bug in the handling of #-introduced comments, recognized when
612        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
613        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
614        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
615        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
616        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
617        places in pcre_compile().
618    
619    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
620        comments when looking ahead for named forward references to subpatterns,
621        the only newline sequence it recognized was NL. It now handles newlines
622        according to the set newline convention.
623    
624    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
625        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
626        cater for a lack of strerror(). These oversights have been fixed.
627    
628    9.  Added --match-limit and --recursion-limit to pcregrep.
629    
630    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
631    
632    11. When the -o option was used, pcregrep was setting a return code of 1, even
633        when matches were found, and --line-buffered was not being honoured.
634    
635    12. Added an optional parentheses number to the -o and --only-matching options
636        of pcregrep.
637    
638    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
639        can match an empty string. The code to do it in pcretest and pcredemo
640        needed fixing:
641    
642        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
643            only one byte after an empty string match just before CRLF (this case
644            just got forgotten; "any" and "anycrlf" were OK).
645    
646        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
647            mode when an empty string match preceded an ASCII character followed by
648            a non-ASCII character. (The code for advancing by one character rather
649            than one byte was nonsense.)
650    
651        (c) The pcredemo.c sample program did not have any code at all to handle
652            the cases when CRLF is a valid newline sequence.
653    
654    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
655        as a starting offset was within the subject string. There is now a new
656        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
657        negative or greater than the length of the string. In order to test this,
658        pcretest is extended to allow the setting of negative starting offsets.
659    
660    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
661        starting offset points to the beginning of a UTF-8 character was
662        unnecessarily clumsy. I tidied it up.
663    
664    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
665        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
666    
667    17. Nobody had reported that the --include_dir option, which was added in
668        release 7.7 should have been called --include-dir (hyphen, not underscore)
669        for compatibility with GNU grep. I have changed it to --include-dir, but
670        left --include_dir as an undocumented synonym, and the same for
671        --exclude-dir, though that is not available in GNU grep, at least as of
672        release 2.5.4.
673    
674    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
675        characters from a string of bytes) have been redefined so as not to use
676        loops, in order to improve performance in some environments. At the same
677        time, I abstracted some of the common code into auxiliary macros to save
678        repetition (this should not affect the compiled code).
679    
680    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
681        compile-time error is now given if \c is not followed by an ASCII
682        character, that is, a byte less than 128. (In EBCDIC mode, the code is
683        different, and any byte value is allowed.)
684    
685    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
686        START_OPTIMIZE option, which is now allowed at compile time - but just
687        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
688        to pcregrep and other applications that have no direct access to PCRE
689        options. The new /Y option in pcretest sets this option when calling
690        pcre_compile().
691    
692    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
693        back references. Groups containing recursive back references were forced to
694        be atomic by that change, but in the case of named groups, the amount of
695        memory required was incorrectly computed, leading to "Failed: internal
696        error: code overflow". This has been fixed.
697    
698    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
699        pcretest.c, to avoid build problems in some Borland environments.
700    
701    
702    Version 8.10 25-Jun-2010
703    ------------------------
704    
705    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
706        THEN.
707    
708    2.  (*ACCEPT) was not working when inside an atomic group.
709    
710    3.  Inside a character class, \B is treated as a literal by default, but
711        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
712        causes the error). The code is unchanged, but I tidied the documentation.
713    
714    4.  Inside a character class, PCRE always treated \R and \X as literals,
715        whereas Perl faults them if its -w option is set. I have changed PCRE so
716        that it faults them when PCRE_EXTRA is set.
717    
718    5.  Added support for \N, which always matches any character other than
719        newline. (It is the same as "." when PCRE_DOTALL is not set.)
720    
721    6.  When compiling pcregrep with newer versions of gcc which may have
722        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
723        declared with attribute warn_unused_result" were given. Just casting the
724        result to (void) does not stop the warnings; a more elaborate fudge is
725        needed. I've used a macro to implement this.
726    
727    7.  Minor change to pcretest.c to avoid a compiler warning.
728    
729    8.  Added four artifical Unicode properties to help with an option to make
730        \s etc use properties (see next item). The new properties are: Xan
731        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
732    
733    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
734        use Unicode properties. (*UCP) at the start of a pattern can be used to set
735        this option. Modified pcretest to add /W to test this facility. Added
736        REG_UCP to make it available via the POSIX interface.
737    
738    10. Added --line-buffered to pcregrep.
739    
740    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
741        studied, and the match started with a letter with a code point greater than
742        127 whose first byte was different to the first byte of the other case of
743        the letter, the other case of this starting letter was not recognized
744        (#976).
745    
746    12. If a pattern that was studied started with a repeated Unicode property
747        test, for example, \p{Nd}+, there was the theoretical possibility of
748        setting up an incorrect bitmap of starting bytes, but fortunately it could
749        not have actually happened in practice until change 8 above was made (it
750        added property types that matched character-matching opcodes).
751    
752    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
753        possible starting bytes for non-anchored patterns.
754    
755    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
756        \R, and also a number of cases that involve Unicode properties, both
757        explicit and implicit when PCRE_UCP is set.
758    
759    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
760        input, it could crash or give wrong results if characters with values
761        greater than 0xc0 were present in the subject string. (Detail: it assumed
762        UTF-8 input when processing these items.)
763    
764    16. Added a lot of (int) casts to avoid compiler warnings in systems where
765        size_t is 64-bit (#991).
766    
767    17. Added a check for running out of memory when PCRE is compiled with
768        --disable-stack-for-recursion (#990).
769    
770    18. If the last data line in a file for pcretest does not have a newline on
771        the end, a newline was missing in the output.
772    
773    19. The default pcre_chartables.c file recognizes only ASCII characters (values
774        less than 128) in its various bitmaps. However, there is a facility for
775        generating tables according to the current locale when PCRE is compiled. It
776        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
777        characters, are recognized by isspace() and therefore were getting set in
778        these tables, and indeed these tables seem to approximate to ISO 8859. This
779        caused a problem in UTF-8 mode when pcre_study() was used to create a list
780        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
781        which of course cannot start UTF-8 characters. I have changed the code so
782        that only real ASCII characters (less than 128) and the correct starting
783        bytes for UTF-8 encodings are set for characters greater than 127 when in
784        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
785        altogether.)
786    
787    20. Added the /T option to pcretest so as to be able to run tests with non-
788        standard character tables, thus making it possible to include the tests
789        used for 19 above in the standard set of tests.
790    
791    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
792        reference to a subpattern the other side of a comment that contains an
793        opening parenthesis caused either an internal compiling error, or a
794        reference to the wrong subpattern.
795    
796    
797    Version 8.02 19-Mar-2010
798    ------------------------
799    
800    1.  The Unicode data tables have been updated to Unicode 5.2.0.
801    
802    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
803        configured.
804    
805    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
806        original author of that file, following a query about its status.
807    
808    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
809        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
810    
811    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
812        quantifier applied to a forward-referencing subroutine call, could compile
813        incorrect code or give the error "internal error: previously-checked
814        referenced subpattern not found".
815    
816    6.  Both MS Visual Studio and Symbian OS have problems with initializing
817        variables to point to external functions. For these systems, therefore,
818        pcre_malloc etc. are now initialized to local functions that call the
819        relevant global functions.
820    
821    7.  There were two entries missing in the vectors called coptable and poptable
822        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
823        I've fixed the data, and added a kludgy way of testing at compile time that
824        the lengths are correct (equal to the number of opcodes).
825    
826    8.  Following on from 7, I added a similar kludge to check the length of the
827        eint vector in pcreposix.c.
828    
829    9.  Error texts for pcre_compile() are held as one long string to avoid too
830        much relocation at load time. To find a text, the string is searched,
831        counting zeros. There was no check for running off the end of the string,
832        which could happen if a new error number was added without updating the
833        string.
834    
835    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
836    
837    11. \K was not working if it appeared in an atomic group or in a group that
838        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
839        \K is "not well defined" if used in an assertion. PCRE now accepts it if
840        the assertion is positive, but not if it is negative.
841    
842    12. Change 11 fortuitously reduced the size of the stack frame used in the
843        "match()" function of pcre_exec.c by one pointer. Forthcoming
844        implementation of support for (*MARK) will need an extra pointer on the
845        stack; I have reserved it now, so that the stack frame size does not
846        decrease.
847    
848    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
849        item in branch that calls a recursion is a subroutine call - as in the
850        second branch in the above example - was incorrectly given the compile-
851        time error "recursive call could loop indefinitely" because pcre_compile()
852        was not correctly checking the subroutine for matching a non-empty string.
853    
854    14. The checks for overrunning compiling workspace could trigger after an
855        overrun had occurred. This is a "should never occur" error, but it can be
856        triggered by pathological patterns such as hundreds of nested parentheses.
857        The checks now trigger 100 bytes before the end of the workspace.
858    
859    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
860    
861    
862    Version 8.01 19-Jan-2010
863    ------------------------
864    
865    1.  If a pattern contained a conditional subpattern with only one branch (in
866        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
867        computed the wrong minimum data length (which is of course zero for such
868        subpatterns). This could cause incorrect "no match" results.
869    
870    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
871        the pattern is reset in the first branch, pcre_compile() failed with
872        "internal error: code overflow at offset...". This happened only when
873        the reset was to the original external option setting. (An optimization
874        abstracts leading options settings into an external setting, which was the
875        cause of this.)
876    
877    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
878        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
879        assertion pattern did not match (meaning that the assertion was true), it
880        was incorrectly treated as false if the SKIP had been reached during the
881        matching. This also applied to assertions used as conditions.
882    
883    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
884        assertion subpattern, including such a pattern used as a condition,
885        unpredictable results occurred, instead of the error return
886        PCRE_ERROR_DFA_UITEM.
887    
888    5.  The C++ GlobalReplace function was not working like Perl for the special
889        situation when an empty string is matched. It now does the fancy magic
890        stuff that is necessary.
891    
892    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
893        removed. (These were left over from very, very early versions of PCRE.)
894    
895    7.  Some cosmetic changes to the code to make life easier when compiling it
896        as part of something else:
897    
898        (a) Change DEBUG to PCRE_DEBUG.
899    
900        (b) In pcre_compile(), rename the member of the "branch_chain" structure
901            called "current" as "current_branch", to prevent a collision with the
902            Linux macro when compiled as a kernel module.
903    
904        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
905            prevent a collision with the Linux macro when compiled as a kernel
906            module.
907    
908    8.  In pcre_compile() there are some checks for integer overflows that used to
909        cast potentially large values to (double). This has been changed to that
910        when building, a check for int64_t is made, and if it is found, it is used
911        instead, thus avoiding the use of floating point arithmetic. (There is no
912        other use of FP in PCRE.) If int64_t is not found, the fallback is to
913        double.
914    
915    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
916        2005 (difference between two addresses compared to an unsigned value).
917    
918    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
919        custom one, because of the following reported problem in Windows:
920    
921          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
922              under Win32.
923          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
924              therefore missing the function definition.
925          - The compiler thus generates a "C" signature for the test function.
926          - The linker fails to find the "C" function.
927          - PCRE fails to configure if asked to do so against libbz2.
928    
929    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
930        messages were output:
931    
932          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
933          rerunning libtoolize, to keep the correct libtool macros in-tree.
934          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
935    
936        I have done both of these things.
937    
938    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
939        most of the time, it *can* run out if it is given a pattern that contains a
940        runaway infinite recursion. I updated the discussion in the pcrestack man
941        page.
942    
943    13. Now that we have gone to the x.xx style of version numbers, the minor
944        version may start with zero. Using 08 or 09 is a bad idea because users
945        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
946        interpreted as invalid octal numbers. I've updated the previous comment in
947        configure.ac, and also added a check that gives an error if 08 or 09 are
948        used.
949    
950    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
951        causing partial matching to fail when the end of the subject matched \W
952        in a UTF-8 pattern where \W was quantified with a minimum of 3.
953    
954    15. There were some discrepancies between the declarations in pcre_internal.h
955        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
956        their definitions. The declarations used "const uschar *" and the
957        definitions used USPTR. Even though USPTR is normally defined as "const
958        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
959        reported that: "This difference in casting confuses some C++ compilers, for
960        example, SunCC recognizes above declarations as different functions and
961        generates broken code for hbpcre." I have changed the declarations to use
962        USPTR.
963    
964    16. GNU libtool is named differently on some systems. The autogen.sh script now
965        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
966        (FreeBSD).
967    
968    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
969        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
970        comment: "Figure out how to create a longlong from a string: strtoll and
971        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
972        instance, but it only takes 2 args instead of 3!"
973    
974    18. A subtle bug concerned with back references has been fixed by a change of
975        specification, with a corresponding code fix. A pattern such as
976        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
977        refers, was giving matches when it shouldn't. For example, xa=xaaa would
978        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
979        same bug. Such groups have to be quantified to be useful, or contained
980        inside another quantified group. (If there's no repetition, the reference
981        can never match.) The problem arises because, having left the group and
982        moved on to the rest of the pattern, a later failure that backtracks into
983        the group uses the captured value from the final iteration of the group
984        rather than the correct earlier one. I have fixed this in PCRE by forcing
985        any group that contains a reference to itself to be an atomic group; that
986        is, there cannot be any backtracking into it once it has completed. This is
987        similar to recursive and subroutine calls.
988    
989    
990    Version 8.00 19-Oct-09
991    ----------------------
992    
993    1.  The table for translating pcre_compile() error codes into POSIX error codes
994        was out-of-date, and there was no check on the pcre_compile() error code
995        being within the table. This could lead to an OK return being given in
996        error.
997    
998    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
999        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1000        in a Windows environment.
1001    
1002    3.  The pcregrep --count option prints the count for each file even when it is
1003        zero, as does GNU grep. However, pcregrep was also printing all files when
1004        --files-with-matches was added. Now, when both options are given, it prints
1005        counts only for those files that have at least one match. (GNU grep just
1006        prints the file name in this circumstance, but including the count seems
1007        more useful - otherwise, why use --count?) Also ensured that the
1008        combination -clh just lists non-zero counts, with no names.
1009    
1010    4.  The long form of the pcregrep -F option was incorrectly implemented as
1011        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1012        but it seems right to fix it, and I didn't think it was worth preserving
1013        the old behaviour.
1014    
1015    5.  The command line items --regex=pattern and --regexp=pattern were not
1016        recognized by pcregrep, which required --regex pattern or --regexp pattern
1017        (with a space rather than an '='). The man page documented the '=' forms,
1018        which are compatible with GNU grep; these now work.
1019    
1020    6.  No libpcreposix.pc file was created for pkg-config; there was just
1021        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1022    
1023    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1024        when UCP support is not needed, by modifying the Python script that
1025        generates it from Unicode data files. This should not matter if the module
1026        is correctly used as a library, but I received one complaint about 50K of
1027        unwanted data. My guess is that the person linked everything into his
1028        program rather than using a library. Anyway, it does no harm.
1029    
1030    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1031        was a minimum greater than 1 for a wide character in a possessive
1032        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1033        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1034        character. Chaos in the form of incorrect output or a compiling loop could
1035        result.
1036    
1037    9.  The restrictions on what a pattern can contain when partial matching is
1038        requested for pcre_exec() have been removed. All patterns can now be
1039        partially matched by this function. In addition, if there are at least two
1040        slots in the offset vector, the offset of the earliest inspected character
1041        for the match and the offset of the end of the subject are set in them when
1042        PCRE_ERROR_PARTIAL is returned.
1043    
1044    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1045        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1046        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1047        and may be more useful for multi-segment matching.
1048    
1049    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1050        used to be given if ever the end of the subject was reached; now it is
1051        given only if matching could not proceed because another character was
1052        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1053        string "Z", which now yields "no match" instead of "partial match". In the
1054        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1055        final character ended with (*FAIL).
1056    
1057    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1058        if the pattern had a "must contain" character that was already found in the
1059        earlier partial match, unless partial matching was again requested. For
1060        example, with the pattern /dog.(body)?/, the "must contain" character is
1061        "g". If the first part-match was for the string "dog", restarting with
1062        "sbody" failed. This bug has been fixed.
1063    
1064    13. The string returned by pcre_dfa_exec() after a partial match has been
1065        changed so that it starts at the first inspected character rather than the
1066        first character of the match. This makes a difference only if the pattern
1067        starts with a lookbehind assertion or \b or \B (\K is not supported by
1068        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1069        matching functions compatible, and I think it's the right thing to do.
1070    
1071    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1072        so that the demonstration program is easily available in environments where
1073        PCRE has not been installed from source.
1074    
1075    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1076        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1077        library.
1078    
1079    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1080        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1081        is not the first non-POSIX option to be added. Clearly some people find
1082        these options useful.
1083    
1084    17. If a caller to the POSIX matching function regexec() passes a non-zero
1085        value for nmatch with a NULL value for pmatch, the value of
1086        nmatch is forced to zero.
1087    
1088    18. RunGrepTest did not have a test for the availability of the -u option of
1089        the diff command, as RunTest does. It now checks in the same way as
1090        RunTest, and also checks for the -b option.
1091    
1092    19. If an odd number of negated classes containing just a single character
1093        interposed, within parentheses, between a forward reference to a named
1094        subpattern and the definition of the subpattern, compilation crashed with
1095        an internal error, complaining that it could not find the referenced
1096        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1097        [The bug was that it was starting one character too far in when skipping
1098        over the character class, thus treating the ] as data rather than
1099        terminating the class. This meant it could skip too much.]
1100    
1101    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1102        /g option in pcretest when the pattern contains \K, which makes it possible
1103        to have an empty string match not at the start, even when the pattern is
1104        anchored. Updated pcretest and pcredemo to use this option.
1105    
1106    21. If the maximum number of capturing subpatterns in a recursion was greater
1107        than the maximum at the outer level, the higher number was returned, but
1108        with unset values at the outer level. The correct (outer level) value is
1109        now given.
1110    
1111    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1112        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1113        make it do so. The string so far is captured, making this feature
1114        compatible with Perl.
1115    
1116    23. The tests have been re-organized, adding tests 11 and 12, to make it
1117        possible to check the Perl 5.10 features against Perl 5.10.
1118    
1119    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1120        pattern matches a fixed length string. PCRE did not allow this; now it
1121        does. Neither allows recursion.
1122    
1123    25. I finally figured out how to implement a request to provide the minimum
1124        length of subject string that was needed in order to match a given pattern.
1125        (It was back references and recursion that I had previously got hung up
1126        on.) This code has now been added to pcre_study(); it finds a lower bound
1127        to the length of subject needed. It is not necessarily the greatest lower
1128        bound, but using it to avoid searching strings that are too short does give
1129        some useful speed-ups. The value is available to calling programs via
1130        pcre_fullinfo().
1131    
1132    26. While implementing 25, I discovered to my embarrassment that pcretest had
1133        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1134        study optimizations had never been tested with that matching function.
1135        Oops. What is worse, even when it was passed study data, there was a bug in
1136        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1137        were also very few tests of studied patterns with pcre_dfa_exec().
1138    
1139    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1140        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1141        on the other side of the coin, they are no longer allowed to have different
1142        names, because these cannot be distinguished in PCRE, and this has caused
1143        confusion. (This is a difference from Perl.)
1144    
1145    28. When duplicate subpattern names are present (necessarily with different
1146        numbers, as required by 27 above), and a test is made by name in a
1147        conditional pattern, either for a subpattern having been matched, or for
1148        recursion in such a pattern, all the associated numbered subpatterns are
1149        tested, and the overall condition is true if the condition is true for any
1150        one of them. This is the way Perl works, and is also more like the way
1151        testing by number works.
1152    
1153    
1154    Version 7.9 11-Apr-09
1155    ---------------------
1156    
1157    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1158        (pcretest), all targets were linked against these libraries. This included
1159        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1160        libraries. This caused unwanted dependencies to be created. This problem
1161        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1162        pcretest is linked with readline.
1163    
1164    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1165        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1166        moved outside it again, because FALSE and TRUE are already defined in AIX,
1167        but BOOL is not.
1168    
1169    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1170        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1171    
1172    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1173        hyphens) following file names and line numbers when outputting matching
1174        lines. This is not true; no spaces are inserted. I have also clarified the
1175        wording for the --colour (or --color) option.
1176    
1177    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1178        was not coloured; this is different to GNU grep, so I have changed it to be
1179        the same.
1180    
1181    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1182        each matching line was coloured. Now it goes on to look for further matches
1183        of any of the test patterns, which is the same behaviour as GNU grep.
1184    
1185    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1186        doesn't make sense to accept an empty string match in pcregrep, so I have
1187        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1188        seems to be how GNU grep behaves.
1189    
1190    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1191        start or after a newline", because the conditional assertion was not being
1192        correctly handled. The rule now is that both the assertion and what follows
1193        in the first alternative must satisfy the test.
1194    
1195    9.  If auto-callout was enabled in a pattern with a conditional group whose
1196        condition was an assertion, PCRE could crash during matching, both with
1197        pcre_exec() and pcre_dfa_exec().
1198    
1199    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1200        used for matching.
1201    
1202    11. Unicode property support in character classes was not working for
1203        characters (bytes) greater than 127 when not in UTF-8 mode.
1204    
1205    12. Added the -M command line option to pcretest.
1206    
1207    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1208    
1209    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1210    
1211    16. Added comments and documentation about mis-use of no_arg in the C++
1212        wrapper.
1213    
1214    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1215        from Martin Jerabek that uses macro names for all relevant character and
1216        string constants.
1217    
1218    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1219        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1220        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1221        these, but not everybody uses configure.
1222    
1223    19. A conditional group that had only one branch was not being correctly
1224        recognized as an item that could match an empty string. This meant that an
1225        enclosing group might also not be so recognized, causing infinite looping
1226        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1227        with the subject "ab", where knowledge that the repeated group can match
1228        nothing is needed in order to break the loop.
1229    
1230    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1231        exec(), but without supplying a callout function, matching went wrong.
1232    
1233    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1234        leak if the size of the offset vector was greater than 30. When the vector
1235        is smaller, the saved offsets during recursion go onto a local stack
1236        vector, but for larger vectors malloc() is used. It was failing to free
1237        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1238        error, in fact).
1239    
1240    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1241        heapframe that is used only when UTF-8 support is enabled. This caused no
1242        problem, but was untidy.
1243    
1244    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1245        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1246        included within another project.
1247    
1248    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1249        slightly modified by me:
1250    
1251          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1252              not building pcregrep.
1253    
1254          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1255              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1256    
1257    25. Forward references, both numeric and by name, in patterns that made use of
1258        duplicate group numbers, could behave incorrectly or give incorrect errors,
1259        because when scanning forward to find the reference group, PCRE was not
1260        taking into account the duplicate group numbers. A pattern such as
1261        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1262    
1263    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1264        the feature of a custom pointer more persuasive (as requested by a user).
1265    
1266    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1267        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1268        pre-defined.
1269    
1270    28. Added support for (*UTF8) at the start of a pattern.
1271    
1272    29. Arrange for flags added by the "release type" setting in CMake to be shown
1273        in the configuration summary.
1274    
1275    
1276    Version 7.8 05-Sep-08
1277    ---------------------
1278    
1279    1.  Replaced UCP searching code with optimized version as implemented for Ad
1280        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1281        stage table and inline lookup instead of a function, giving speed ups of 2
1282        to 5 times on some simple patterns that I tested. Permission was given to
1283        distribute the MultiStage2.py script that generates the tables (it's not in
1284        the tarball, but is in the Subversion repository).
1285    
1286    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1287        scripts.
1288    
1289    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1290        a group with a zero qualifier. The result of the study could be incorrect,
1291        or the function might crash, depending on the pattern.
1292    
1293    4.  Caseless matching was not working for non-ASCII characters in back
1294        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1295        It now works when Unicode Property Support is available.
1296    
1297    5.  In pcretest, an escape such as \x{de} in the data was always generating
1298        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1299        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1300        truncation.
1301    
1302    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1303    
1304    7.  Added two (int) casts to pcregrep when printing the difference of two
1305        pointers, in case they are 64-bit values.
1306    
1307    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1308        test 2 if it fails.
1309    
1310    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1311        and a #define of that name to empty if it is not externally set. This is to
1312        allow users of MSVC to set it if necessary.
1313    
1314    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1315        the convenience functions in the pcre_get.c source file.
1316    
1317    11. An option change at the start of a pattern that had top-level alternatives
1318        could cause overwriting and/or a crash. This command provoked a crash in
1319        some environments:
1320    
1321          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1322    
1323        This potential security problem was recorded as CVE-2008-2371.
1324    
1325    12. For a pattern where the match had to start at the beginning or immediately
1326        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1327        pcre_dfa_exec() could read past the end of the passed subject if there was
1328        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1329        pcretest so that it places the subject at the end of its malloc-ed buffer.
1330    
1331    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1332        exec() might read past the end of the data buffer in UTF-8 mode.
1333    
1334    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1335        the data contained the byte 0x85 as part of a UTF-8 character within its
1336        first line. This applied both to normal and DFA matching.
1337    
1338    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1339        /^[^d]*?$/8 failed to match "abc".
1340    
1341    16. Added a missing copyright notice to pcrecpp_internal.h.
1342    
1343    17. Make it more clear in the documentation that values returned from
1344        pcre_exec() in ovector are byte offsets, not character counts.
1345    
1346    18. Tidied a few places to stop certain compilers from issuing warnings.
1347    
1348    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1349        supplied by Stefan Weber. I made a further small update for 7.8 because
1350        there is a change of source arrangements: the pcre_searchfuncs.c module is
1351        replaced by pcre_ucd.c.
1352    
1353    
1354    Version 7.7 07-May-08
1355    ---------------------
1356    
1357    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1358        a string to a long long, pretend we don't even have a long long." This is
1359        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1360    
1361    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1362        pre-7.6 versions, which defined a global no_arg variable instead of putting
1363        it in the RE class. (See also #8 below.)
1364    
1365    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1366        Lopes.
1367    
1368    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1369    
1370        (1) The include/exclude patterns were being applied to the whole pathnames
1371            of files, instead of just to the final components.
1372    
1373        (2) If there was more than one level of directory, the subdirectories were
1374            skipped unless they satisfied the include/exclude conditions. This is
1375            inconsistent with GNU grep (and could even be seen as contrary to the
1376            pcregrep specification - which I improved to make it absolutely clear).
1377            The action now is always to scan all levels of directory, and just
1378            apply the include/exclude patterns to regular files.
1379    
1380    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1381        --exclude_dir in the tests to avoid scanning .svn directories.
1382    
1383    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1384        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1385        doesn't support NULs in patterns.
1386    
1387    7.  Added some missing "const"s to declarations of static tables in
1388        pcre_compile.c and pcre_dfa_exec.c.
1389    
1390    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1391        caused by fix #2  above. (Subsequently also a second patch to fix the
1392        first patch. And a third patch - this was a messy problem.)
1393    
1394    9.  Applied Craig's patch to remove the use of push_back().
1395    
1396    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1397        matching function regexec().
1398    
1399    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1400        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1401        references. PCRE supports relative numbers with this syntax (I don't think
1402        Oniguruma does).
1403    
1404    12. Previously, a group with a zero repeat such as (...){0} was completely
1405        omitted from the compiled regex. However, this means that if the group
1406        was called as a subroutine from elsewhere in the pattern, things went wrong
1407        (an internal error was given). Such groups are now left in the compiled
1408        pattern, with a new opcode that causes them to be skipped at execution
1409        time.
1410    
1411    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1412        to the way PCRE behaves:
1413    
1414        (a) A lone ] character is dis-allowed (Perl treats it as data).
1415    
1416        (b) A back reference to an unmatched subpattern matches an empty string
1417            (Perl fails the current match path).
1418    
1419        (c) A data ] in a character class must be notated as \] because if the
1420            first data character in a class is ], it defines an empty class. (In
1421            Perl it is not possible to have an empty class.) The empty class []
1422            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1423            The negative empty class [^] matches any one character, independently
1424            of the DOTALL setting.
1425    
1426    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1427        non-existent subpattern following a character class starting with ']' and
1428        containing () gave an internal compiling error instead of "reference to
1429        non-existent subpattern". Fortunately, when the pattern did exist, the
1430        compiled code was correct. (When scanning forwards to check for the
1431        existencd of the subpattern, it was treating the data ']' as terminating
1432        the class, so got the count wrong. When actually compiling, the reference
1433        was subsequently set up correctly.)
1434    
1435    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1436        it was being rejected as not supported by pcre_dfa_exec(), even though
1437        other assertions are supported. I have made pcre_dfa_exec() support
1438        (*FAIL).
1439    
1440    16. The implementation of 13c above involved the invention of a new opcode,
1441        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1442        cannot be changed at match time, I realized I could make a small
1443        improvement to matching performance by compiling OP_ALLANY instead of
1444        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1445        on the OP_ANY path.
1446    
1447    17. Compiling pcretest on Windows with readline support failed without the
1448        following two fixes: (1) Make the unistd.h include conditional on
1449        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1450    
1451    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1452        ncurses library to be included for pcretest when ReadLine support is
1453        requested, but also to allow for it to be overridden. This patch came from
1454        Daniel Bergström.
1455    
1456    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1457        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1458        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1459        spotting this.
1460    
1461    
1462    Version 7.6 28-Jan-08
1463    ---------------------
1464    
1465    1.  A character class containing a very large number of characters with
1466        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1467        overflow.
1468    
1469    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1470        HAVE_LONG_LONG is not defined.
1471    
1472    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1473        bring them up to date and include new features. This patch includes:
1474    
1475        - Fixed PH's badly added libz and libbz2 support.
1476        - Fixed a problem with static linking.
1477        - Added pcredemo. [But later removed - see 7 below.]
1478        - Fixed dftables problem and added an option.
1479        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1480            HAVE_LONG_LONG.
1481        - Added readline support for pcretest.
1482        - Added an listing of the option settings after cmake has run.
1483    
1484    4.  A user submitted a patch to Makefile that makes it easy to create
1485        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1486        Makefile.am that cause it to include this special target, without
1487        affecting anything else. Note that the same mingw target plus all
1488        the other distribution libraries and programs are now supported
1489        when configuring with CMake (see 6 below) instead of with
1490        Configure/Make.
1491    
1492    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1493        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1494        exported in the Windows port". It has not yet been confirmed that the patch
1495        solves the problem, but it does no harm.
1496    
1497    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1498        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1499        with CMake, and also correct the comment about stack recursion.
1500    
1501    7.  Remove the automatic building of pcredemo from the ./configure system and
1502        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1503        of a program that users should build themselves after PCRE is installed, so
1504        building it automatically is not really right. What is more, it gave
1505        trouble in some build environments.
1506    
1507    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1508    
1509    
1510    Version 7.5 10-Jan-08
1511    ---------------------
1512    
1513    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1514        values in parens when parsing an RE using the C++ wrapper."
1515    
1516    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1517        Characters greater than 255 were excluded from the class instead of being
1518        included.
1519    
1520    3.  The same bug as (2) above applied to negated POSIX classes such as
1521        [:^space:].
1522    
1523    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1524        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1525        I have changed it.
1526    
1527    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1528        first named subpattern) and a construct such as (?&a) would reference the
1529        first named subpattern whose name started with "a" (in other words, the
1530        length check was missing). Both these problems are fixed. "Subpattern name
1531        expected" is now given for (?&) (a zero-length name), and this patch also
1532        makes it give the same error for \k'' (previously it complained that that
1533        was a reference to a non-existent subpattern).
1534    
1535    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1536        this is right because (?- can be followed by option settings as well as by
1537        digits. I have, however, made the messages clearer.
1538    
1539    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1540        than the number used in the conditional) now cause a compile-time error.
1541        This is actually not compatible with Perl, which accepts such patterns, but
1542        treats the conditional as always being FALSE (as PCRE used to), but it
1543        seems to me that giving a diagnostic is better.
1544    
1545    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1546        and messages.
1547    
1548    9.  Fix two occurrences of "backslash" in comments that should have been
1549        "backspace".
1550    
1551    10. Remove two redundant lines of code that can never be obeyed (their function
1552        was moved elsewhere).
1553    
1554    11. The program that makes PCRE's Unicode character property table had a bug
1555        which caused it to generate incorrect table entries for sequences of
1556        characters that have the same character type, but are in different scripts.
1557        It amalgamated them into a single range, with the script of the first of
1558        them. In other words, some characters were in the wrong script. There were
1559        thirteen such cases, affecting characters in the following ranges:
1560    
1561          U+002b0 - U+002c1
1562          U+0060c - U+0060d
1563          U+0061e - U+00612
1564          U+0064b - U+0065e
1565          U+0074d - U+0076d
1566          U+01800 - U+01805
1567          U+01d00 - U+01d77
1568          U+01d9b - U+01dbf
1569          U+0200b - U+0200f
1570          U+030fc - U+030fe
1571          U+03260 - U+0327f
1572          U+0fb46 - U+0fbb1
1573          U+10450 - U+1049d
1574    
1575    12. The -o option (show only the matching part of a line) for pcregrep was not
1576        compatible with GNU grep in that, if there was more than one match in a
1577        line, it showed only the first of them. It now behaves in the same way as
1578        GNU grep.
1579    
1580    13. If the -o and -v options were combined for pcregrep, it printed a blank
1581        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1582        does the same. The return code can be used to tell if there were any
1583        non-matching lines.
1584    
1585    14. Added --file-offsets and --line-offsets to pcregrep.
1586    
1587    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1588        infinitely looping recursion. The bug was that positive lookaheads were not
1589        being skipped when checking for a possible empty match (negative lookaheads
1590        and both kinds of lookbehind were skipped).
1591    
1592    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1593        inclusion of <windows.h> to before rather than after the definition of
1594        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1595    
1596    17. Specifying a possessive quantifier with a specific limit for a Unicode
1597        character property caused pcre_compile() to compile bad code, which led at
1598        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1599        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1600        caused the error; without that there was no problem.
1601    
1602    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1603    
1604    19. Added --enable-pcretest-libreadline.
1605    
1606    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1607        RE::GlobalReplace(). As a result, the number of replacements returned was
1608        double what it should be. I removed one of the increments, but Craig sent a
1609        later patch that removed the other one (the right fix) and added unit tests
1610        that check the return values (which was not done before).
1611    
1612    21. Several CMake things:
1613    
1614        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1615            the names libpcre and libpcreposix, not just pcre and pcreposix.
1616    
1617        (2) The above change means that pcretest and pcregrep are now correctly
1618            linked with the newly-built libraries, not previously installed ones.
1619    
1620        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1621    
1622    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1623        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1624        UTF-8 newline character). The key issue is that the pattern starts .*;
1625        this means that the match must be either at the beginning, or after a
1626        newline. The bug was in the code for advancing after a failed match and
1627        checking that the new position followed a newline. It was not taking
1628        account of UTF-8 characters correctly.
1629    
1630    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1631        character classes. PCRE was not treating the sequence [:...:] as a
1632        character class unless the ... were all letters. Perl, however, seems to
1633        allow any characters between [: and :], though of course it rejects as
1634        unknown any "names" that contain non-letters, because all the known class
1635        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1636        for example, whereas PCRE did not - it did not recognize a POSIX character
1637        class. This seemed a bit dangerous, so the code has been changed to be
1638        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1639        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1640        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1641        Perl does, and where it didn't before.
1642    
1643    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1644        Windows environments %n is disabled by default.
1645    
1646    
1647    Version 7.4 21-Sep-07
1648    ---------------------
1649    
1650    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1651        means that a class such as [\s] counted as "explicit reference to CR or
1652        LF". That isn't really right - the whole point of the change was to try to
1653        help when there was an actual mention of one of the two characters. So now
1654        the change happens only if \r or \n (or a literal CR or LF) character is
1655        encountered.
1656    
1657    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1658        of both had grown to the point where there were only 3 bits left.
1659        Fortunately, there was spare space in the data structure, and so I have
1660        moved the internal flags into a new 16-bit field to free up more option
1661        bits.
1662    
1663    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1664        but did not set the internal JCHANGED flag - either of these is enough to
1665        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1666        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1667        start sets both bits.
1668    
1669    4.  Added options (at build time, compile time, exec time) to change \R from
1670        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1671    
1672    5.  doc/pcresyntax.html was missing from the distribution.
1673    
1674    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1675        compatibility, even though it is no longer used.
1676    
1677    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1678        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1679        windows.h file is present (where different names are used). [This was
1680        reversed later after testing - see 16 below.]
1681    
1682    8.  Changed all #include <config.h> to #include "config.h". There were also
1683        some further <pcre.h> cases that I changed to "pcre.h".
1684    
1685    9.  When pcregrep was used with the --colour option, it missed the line ending
1686        sequence off the lines that it output.
1687    
1688    10. It was pointed out to me that arrays of string pointers cause lots of
1689        relocations when a shared library is dynamically loaded. A technique of
1690        using a single long string with a table of offsets can drastically reduce
1691        these. I have refactored PCRE in four places to do this. The result is
1692        dramatic:
1693    
1694          Originally:                          290
1695          After changing UCP table:            187
1696          After changing error message table:   43
1697          After changing table of "verbs"       36
1698          After changing table of Posix names   22
1699    
1700        Thanks to the folks working on Gregex for glib for this insight.
1701    
1702    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1703        unicode-properties was also set.
1704    
1705    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1706    
1707    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1708        checked only for CRLF.
1709    
1710    14. Added casts to pcretest.c to avoid compiler warnings.
1711    
1712    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1713    
1714    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1715        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1716        entirely. This removes changes made in 7 above.
1717    
1718    17. The CMake files have been updated, and there is now more information about
1719        building with CMake in the NON-UNIX-USE document.
1720    
1721    
1722    Version 7.3 28-Aug-07
1723    ---------------------
1724    
1725     1. In the rejigging of the build system that eventually resulted in 7.1, the
1726        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1727        brackets there is not right, since it causes compilers to look for an
1728        installed pcre.h, not the version that is in the source that is being
1729        compiled (which of course may be different). I have changed it back to:
1730    
1731          #include "pcre.h"
1732    
1733        I have a vague recollection that the change was concerned with compiling in
1734        different directories, but in the new build system, that is taken care of
1735        by the VPATH setting the Makefile.
1736    
1737     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1738        when the subject happened to end in the byte 0x85 (e.g. if the last
1739        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1740        characters but of course it shouldn't be taken as a newline when it is part
1741        of another character. The bug was that, for an unlimited repeat of . in
1742        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1743        characters when looking for a newline.
1744    
1745     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1746    
1747     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1748        in debug output.
1749    
1750     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1751        long printing in the pcrecpp unittest when running under MinGW.
1752    
1753     6. ESC_K was left out of the EBCDIC table.
1754    
1755     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1756        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1757        limit also applies to "virtual nesting" when a pattern is recursive, and in
1758        this case 1000 isn't so big. I have been able to remove this limit at the
1759        expense of backing off one optimization in certain circumstances. Normally,
1760        when pcre_exec() would call its internal match() function recursively and
1761        immediately return the result unconditionally, it uses a "tail recursion"
1762        feature to save stack. However, when a subpattern that can match an empty
1763        string has an unlimited repetition quantifier, it no longer makes this
1764        optimization. That gives it a stack frame in which to save the data for
1765        checking that an empty string has been matched. Previously this was taken
1766        from the 1000-entry workspace that had been reserved. So now there is no
1767        explicit limit, but more stack is used.
1768    
1769     8. Applied Daniel's patches to solve problems with the import/export magic
1770        syntax that is required for Windows, and which was going wrong for the
1771        pcreposix and pcrecpp parts of the library. These were overlooked when this
1772        problem was solved for the main library.
1773    
1774     9. There were some crude static tests to avoid integer overflow when computing
1775        the size of patterns that contain repeated groups with explicit upper
1776        limits. As the maximum quantifier is 65535, the maximum group length was
1777        set at 30,000 so that the product of these two numbers did not overflow a
1778        32-bit integer. However, it turns out that people want to use groups that
1779        are longer than 30,000 bytes (though not repeat them that many times).
1780        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1781        made it possible to implement the integer overflow checks in a much more
1782        dynamic way, which I have now done. The artificial limitation on group
1783        length has been removed - we now have only the limit on the total length of
1784        the compiled pattern, which depends on the LINK_SIZE setting.
1785    
1786    10. Fixed a bug in the documentation for get/copy named substring when
1787        duplicate names are permitted. If none of the named substrings are set, the
1788        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1789        empty string.
1790    
1791    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1792        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1793        because the ] is interpreted as the first data character and the
1794        terminating ] is not found. PCRE has been made compatible with Perl in this
1795        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1796        cause memory overwriting.
1797    
1798    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1799        string has been matched (to stop an infinite loop). It was not recognizing
1800        a conditional subpattern that could match an empty string if that
1801        subpattern was within another subpattern. For example, it looped when
1802        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1803        condition was not nested. This bug has been fixed.
1804    
1805    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1806        past the start of the subject in the presence of bytes with the top bit
1807        set, for example "\x8aBCD".
1808    
1809    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1810        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1811    
1812    14. Optimized (?!) to (*FAIL).
1813    
1814    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1815        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1816        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1817        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1818        does: it's just the validity check that is more restrictive.
1819    
1820    16. Inserted checks for integer overflows during escape sequence (backslash)
1821        processing, and also fixed erroneous offset values for syntax errors during
1822        backslash processing.
1823    
1824    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1825        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1826    
1827    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1828        caused an overrun.
1829    
1830    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1831        something other than just ASCII characters) inside a group that had an
1832        unlimited repeat caused a loop at compile time (while checking to see
1833        whether the group could match an empty string).
1834    
1835    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1836        [\P{Any}] did so. (Error in the code for printing property names.)
1837    
1838    21. An orphan \E inside a character class could cause a crash.
1839    
1840    22. A repeated capturing bracket such as (A)? could cause a wild memory
1841        reference during compilation.
1842    
1843    23. There are several functions in pcre_compile() that scan along a compiled
1844        expression for various reasons (e.g. to see if it's fixed length for look
1845        behind). There were bugs in these functions when a repeated \p or \P was
1846        present in the pattern. These operators have additional parameters compared
1847        with \d, etc, and these were not being taken into account when moving along
1848        the compiled data. Specifically:
1849    
1850        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1851            length.
1852    
1853        (b) An item such as \pL+ within a repeated group could cause crashes or
1854            loops.
1855    
1856        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1857            "reference to non-existent subpattern" error.
1858    
1859        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1860    
1861    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1862        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1863    
1864    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1865    
1866    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1867        character were causing crashes (broken optimization).
1868    
1869    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1870        \p or \P) caused a compile-time loop.
1871    
1872    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1873        break. For example, the unstudied pattern [\r\n]A does not match the string
1874        "\r\nA" because change 7.0/46 below moves the current point on by two
1875        characters after failing to match at the start. However, the pattern \nA
1876        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1877        the same is true. There doesn't seem any very clean way out of this, but
1878        what I have chosen to do makes the common cases work: PCRE now takes note
1879        of whether there can be an explicit match for \r or \n anywhere in the
1880        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1881        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1882        pattern has explicit CR or LF references.
1883    
1884    29. Added (*CR) etc for changing newline setting at start of pattern.
1885    
1886    
1887    Version 7.2 19-Jun-07
1888    ---------------------
1889    
1890     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1891        which is apparently normally available under Windows.
1892    
1893     2. Re-jig the pcregrep tests with different newline settings in an attempt
1894        to make them independent of the local environment's newline setting.
1895    
1896     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1897    
1898     4. Some of the "internals" tests were previously cut out when the link size
1899        was not 2, because the output contained actual offsets. The recent new
1900        "Z" feature of pcretest means that these can be cut out, making the tests
1901        usable with all link sizes.
1902    
1903     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1904        stack recursion. This gives a massive performance boost under BSD, but just
1905        a small improvement under Linux. However, it saves one field in the frame
1906        in all cases.
1907    
1908     6. Added more features from the forthcoming Perl 5.10:
1909    
1910        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1911            recursion call. It refers to the nth most recently opened parentheses.
1912    
1913        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1914            to be opened parentheses.
1915    
1916        (c) Conditions that refer to capturing parentheses can be specified
1917            relatively, for example, (?(-2)... or (?(+3)...
1918    
1919        (d) \K resets the start of the current match so that everything before
1920            is not part of it.
1921    
1922        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1923    
1924        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1925            reference syntax.
1926    
1927        (g) (?| introduces a group in which the numbering of parentheses in each
1928            alternative starts with the same number.
1929    
1930        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1931    
1932     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1933        PCRE_INFO_JCHANGED.
1934    
1935     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1936        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1937        for detecting groups that can match an empty string.
1938    
1939     9. A pattern with a very large number of alternatives (more than several
1940        hundred) was running out of internal workspace during the pre-compile
1941        phase, where pcre_compile() figures out how much memory will be needed. A
1942        bit of new cunning has reduced the workspace needed for groups with
1943        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1944        workspace instead of running out of the 4096 that are available.
1945    
1946    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1947    
1948    11. Applied patch from Google to remove an optimization that didn't quite work.
1949        The report of the bug said:
1950    
1951          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1952          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1953          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1954    
1955    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1956        it matched the wrong number of bytes.
1957    
1958    
1959    Version 7.1 24-Apr-07
1960    ---------------------
1961    
1962     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1963        that is more "standard", making use of automake and other Autotools. There
1964        is some re-arrangement of the files and adjustment of comments consequent
1965        on this.
1966    
1967     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1968        for recursive directory scanning broke on some systems because the files
1969        are not scanned in any specific order and on different systems the order
1970        was different. A call to "sort" has been inserted into RunGrepTest for the
1971        approprate test as a short-term fix. In the longer term there may be an
1972        alternative.
1973    
1974     3. I had an email from Eric Raymond about problems translating some of PCRE's
1975        man pages to HTML (despite the fact that I distribute HTML pages, some
1976        people do their own conversions for various reasons). The problems
1977        concerned the use of low-level troff macros .br and .in. I have therefore
1978        removed all such uses from the man pages (some were redundant, some could
1979        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1980        HTML has been updated to handle .nf/.fi and to complain if it encounters
1981        .br or .in.
1982    
1983     4. Updated comments in configure.ac that get placed in config.h.in and also
1984        arranged for config.h to be included in the distribution, with the name
1985        config.h.generic, for the benefit of those who have to compile without
1986        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1987    
1988     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1989        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1990        makevp.bat for the current PCRE, using the additional files
1991        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1992    
1993     6. A Windows user reported a minor discrepancy with test 2, which turned out
1994        to be caused by a trailing space on an input line that had got lost in his
1995        copy. The trailing space was an accident, so I've just removed it.
1996    
1997     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1998        that is needed.
1999    
2000     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2001        as "const" (a) because they are and (b) because it helps the PHP
2002        maintainers who have recently made a script to detect big data structures
2003        in the php code that should be moved to the .rodata section. I remembered
2004        to update Builducptable as well, so it won't revert if ucptable.h is ever
2005        re-created.
2006    
2007     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2008        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2009        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2010        support is not required. This saves 1.5-2K of code, which is important in
2011        some applications.
2012    
2013        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2014        so as not to refer to the tables, even though these functions will never be
2015        called when UTF-8 support is disabled. Otherwise there are problems with a
2016        shared library.
2017    
2018    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2019    
2020        (a) It was defining its arguments as char * instead of void *.
2021    
2022        (b) It was assuming that all moves were upwards in memory; this was true
2023            a long time ago when I wrote it, but is no longer the case.
2024    
2025        The emulated memove() is provided for those environments that have neither
2026        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2027        is clearly not the case, as these two bugs were recently reported.
2028    
2029    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2030        and Detrail to create the HTML documentation, the .txt form of the man
2031        pages, and it removes trailing spaces from listed files. It also creates
2032        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2033        case, it wraps all the #defines with #ifndefs. This script should be run
2034        before "make dist".
2035    
2036    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2037        with Unicode property support.
2038    
2039        (a) For a maximizing quantifier, if the two different cases of the
2040            character were of different lengths in their UTF-8 codings (there are
2041            some cases like this - I found 11), and the matching function had to
2042            back up over a mixture of the two cases, it incorrectly assumed they
2043            were both the same length.
2044    
2045        (b) When PCRE was configured to use the heap rather than the stack for
2046            recursion during matching, it was not correctly preserving the data for
2047            the other case of a UTF-8 character when checking ahead for a match
2048            while processing a minimizing repeat. If the check also involved
2049            matching a wide character, but failed, corruption could cause an
2050            erroneous result when trying to check for a repeat of the original
2051            character.
2052    
2053    13. Some tidying changes to the testing mechanism:
2054    
2055        (a) The RunTest script now detects the internal link size and whether there
2056            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2057            values substituted by "configure". (The RunGrepTest script already did
2058            this for UTF-8.) The configure.ac script no longer substitutes the
2059            relevant variables.
2060    
2061        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2062            with length and offset values. This means that the output is different
2063            for different internal link sizes. Test 2 is skipped for link sizes
2064            other than 2 because of this, bypassing the problem. Unfortunately,
2065            there was also a test in test 3 (the locale tests) that used /B and
2066            failed for link sizes other than 2. Rather than cut the whole test out,
2067            I have added a new /Z option to pcretest that replaces the length and
2068            offset values with spaces. This is now used to make test 3 independent
2069            of link size. (Test 2 will be tidied up later.)
2070    
2071    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2072        segmentation fault instead of returning the appropriate error message.
2073    
2074    15. In multiline mode when the newline sequence was set to "any", the pattern
2075        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2076        This doesn't seem right; it now treats the CRLF combination as the line
2077        ending, and so does not match in that case. It's only a pattern such as ^$
2078        that would hit this one: something like ^ABC$ would have failed after \r
2079        and then tried again after \r\n.
2080    
2081    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2082        in an attempt to make files that differ only in their line terminators
2083        compare equal. This works on Linux.
2084    
2085    17. Under certain error circumstances pcregrep might try to free random memory
2086        as it exited. This is now fixed, thanks to valgrind.
2087    
2088    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2089        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2090        was because its rules for how to advance for /g after matching an empty
2091        string at the end of a line did not allow for this case. They now check for
2092        it specially.
2093    
2094    20. pcretest is supposed to handle patterns and data of any length, by
2095        extending its buffers when necessary. It was getting this wrong when the
2096        buffer for a data line had to be extended.
2097    
2098    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2099        CRLF as a newline sequence.
2100    
2101    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2102        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2103        I have nevertheless tidied it up.
2104    
2105    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2106    
2107    24. Added a man page for pcre-config.
2108    
2109    
2110    Version 7.0 19-Dec-06
2111    ---------------------
2112    
2113     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2114        moving to gcc 4.1.1.
2115    
2116     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2117        sys/time.h, which is documented as needed for this function. It doesn't
2118        seem to matter on Linux, but it showed up on some releases of OS X.
2119    
2120     3. It seems that there are systems where bytes whose values are greater than
2121        127 match isprint() in the "C" locale. The "C" locale should be the
2122        default when a C program starts up. In most systems, only ASCII printing
2123        characters match isprint(). This difference caused the output from pcretest
2124        to vary, making some of the tests fail. I have changed pcretest so that:
2125    
2126        (a) When it is outputting text in the compiled version of a pattern, bytes
2127            other than 32-126 are always shown as hex escapes.
2128    
2129        (b) When it is outputting text that is a matched part of a subject string,
2130            it does the same, unless a different locale has been set for the match
2131            (using the /L modifier). In this case, it uses isprint() to decide.
2132    
2133     4. Fixed a major bug that caused incorrect computation of the amount of memory
2134        required for a compiled pattern when options that changed within the
2135        pattern affected the logic of the preliminary scan that determines the
2136        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2137        that the computed length was too small. The symptoms of this bug were
2138        either the PCRE error "internal error: code overflow" from pcre_compile(),
2139        or a glibc crash with a message such as "pcretest: free(): invalid next
2140        size (fast)". Examples of patterns that provoked this bug (shown in
2141        pcretest format) are:
2142    
2143          /(?-x: )/x
2144          /(?x)(?-x: \s*#\s*)/
2145          /((?i)[\x{c0}])/8
2146          /(?i:[\x{c0}])/8
2147    
2148        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2149        is now done differently.
2150    
2151     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2152        wrapper classes; (b) implement a new function in the C++ scanner that is
2153        more efficient than the old way of doing things because it avoids levels of
2154        recursion in the regex matching; (c) add a paragraph to the documentation
2155        for the FullMatch() function.
2156    
2157     6. The escape sequence \n was being treated as whatever was defined as
2158        "newline". Not only was this contrary to the documentation, which states
2159        that \n is character 10 (hex 0A), but it also went horribly wrong when
2160        "newline" was defined as CRLF. This has been fixed.
2161    
2162     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2163        was being set to -1 for the "end of line" case (supposedly a value that no
2164        character can have). Though this value is never used (the check for end of
2165        line is "zero bytes in current character"), it caused compiler complaints.
2166        I've changed it to 0xffffffff.
2167    
2168     8. In pcre_version.c, the version string was being built by a sequence of
2169        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2170        string (as it is for production releases) called a macro with an empty
2171        argument. The C standard says the result of this is undefined. The gcc
2172        compiler treats it as an empty string (which was what was wanted) but it is
2173        reported that Visual C gives an error. The source has been hacked around to
2174        avoid this problem.
2175    
2176     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2177        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2178        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2179        of them did).
2180    
2181    10. Originally, pcretest opened its input and output without "b"; then I was
2182        told that "b" was needed in some environments, so it was added for release
2183        5.0 to both the input and output. (It makes no difference on Unix-like
2184        systems.) Later I was told that it is wrong for the input on Windows. I've
2185        now abstracted the modes into two macros, to make it easier to fiddle with
2186        them, and removed "b" from the input mode under Windows.
2187    
2188    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2189    
2190    12. Added -help and --help to pcretest as an official way of being reminded
2191        of the options.
2192    
2193    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2194        and pcrecpp.cc because they annoy compilers at high warning levels.
2195    
2196    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2197    
2198    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2199        scripts are not C programs :-) and which was not noticed because it works
2200        on Linux.
2201    
2202    16. pcretest is supposed to handle any length of pattern and data line (as one
2203        line or as a continued sequence of lines) by extending its input buffer if
2204        necessary. This feature was broken for very long pattern lines, leading to
2205        a string of junk being passed to pcre_compile() if the pattern was longer
2206        than about 50K.
2207    
2208    17. I have done a major re-factoring of the way pcre_compile() computes the
2209        amount of memory needed for a compiled pattern. Previously, there was code
2210        that made a preliminary scan of the pattern in order to do this. That was
2211        OK when PCRE was new, but as the facilities have expanded, it has become
2212        harder and harder to keep it in step with the real compile phase, and there
2213        have been a number of bugs (see for example, 4 above). I have now found a
2214        cunning way of running the real compile function in a "fake" mode that
2215        enables it to compute how much memory it would need, while actually only
2216        ever using a few hundred bytes of working memory and without too many
2217        tests of the mode. This should make future maintenance and development
2218        easier. A side effect of this work is that the limit of 200 on the nesting
2219        depth of parentheses has been removed (though this was never a serious
2220        limitation, I suspect). However, there is a downside: pcre_compile() now
2221        runs more slowly than before (30% or more, depending on the pattern). I
2222        hope this isn't a big issue. There is no effect on runtime performance.
2223    
2224    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2225        newline (only possible for the last line of a file) and it was a
2226        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2227    
2228    19. Added additional timing features to pcretest. (1) The -tm option now times
2229        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2230        separate command line item, by a number that specifies the number of
2231        repeats to use when timing. The default is 50000; this gives better
2232        precision, but takes uncomfortably long for very large patterns.
2233    
2234    20. Extended pcre_study() to be more clever in cases where a branch of a
2235        subpattern has no definite first character. For example, (a*|b*)[cd] would
2236        previously give no result from pcre_study(). Now it recognizes that the
2237        first character must be a, b, c, or d.
2238    
2239    21. There was an incorrect error "recursive call could loop indefinitely" if
2240        a subpattern (or the entire pattern) that was being tested for matching an
2241        empty string contained only one non-empty item after a nested subpattern.
2242        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2243        incorrectly, because the \d was being skipped in the check.
2244    
2245    22. The pcretest program now has a new pattern option /B and a command line
2246        option -b, which is equivalent to adding /B to every pattern. This causes
2247        it to show the compiled bytecode, without the additional information that
2248        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2249        is the same as /B/I).
2250    
2251    23. A new optimization is now able automatically to treat some sequences such
2252        as a*b as a*+b. More specifically, if something simple (such as a character
2253        or a simple class like \d) has an unlimited quantifier, and is followed by
2254        something that cannot possibly match the quantified thing, the quantifier
2255        is automatically "possessified".
2256    
2257    24. A recursive reference to a subpattern whose number was greater than 39
2258        went wrong under certain circumstances in UTF-8 mode. This bug could also
2259        have affected the operation of pcre_study().
2260    
2261    25. Realized that a little bit of performance could be had by replacing
2262        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2263    
2264    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2265    
2266    27. Possessive quantifiers such as a++ were previously implemented by turning
2267        them into atomic groups such as ($>a+). Now they have their own opcodes,
2268        which improves performance. This includes the automatically created ones
2269        from 23 above.
2270    
2271    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2272        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2273        the first matched character to be a colon. This applied both to named and
2274        numbered groups.
2275    
2276    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2277    
2278    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2279        building PCRE on Windows easier, so I have included it in the distribution.
2280    
2281    31. There is now a check in pcretest against a ridiculously large number being
2282        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2283        loop, the loop is abandoned.
2284    
2285    32. Forward references to subpatterns in conditions such as (?(2)...) where
2286        subpattern 2 is defined later cause pcre_compile() to search forwards in
2287        the pattern for the relevant set of parentheses. This search went wrong
2288        when there were unescaped parentheses in a character class, parentheses
2289        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2290    
2291    33. "Subroutine" calls and backreferences were previously restricted to
2292        referencing subpatterns earlier in the regex. This restriction has now
2293        been removed.
2294    
2295    34. Added a number of extra features that are going to be in Perl 5.10. On the
2296        whole, these are just syntactic alternatives for features that PCRE had
2297        previously implemented using the Python syntax or my own invention. The
2298        other formats are all retained for compatibility.
2299    
2300        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2301            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2302            also .NET compatible.
2303    
2304        (b) A recursion or subroutine call to a named group can now be defined as
2305            (?&name) as well as (?P>name).
2306    
2307        (c) A backreference to a named group can now be defined as \k<name> or
2308            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2309            5.10, are also .NET compatible.
2310    
2311        (d) A conditional reference to a named group can now use the syntax
2312            (?(<name>) or (?('name') as well as (?(name).
2313    
2314        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2315            groups (named and numbered) that are never evaluated inline, but can be
2316            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2317            is always false. There may be only one alternative in such a group.
2318    
2319        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2320            as the simple (?(R). The condition is true only if the most recent
2321            recursion is that of the given number or name. It does not search out
2322            through the entire recursion stack.
2323    
2324        (g) The escape \gN or \g{N} has been added, where N is a positive or
2325            negative number, specifying an absolute or relative reference.
2326    
2327    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2328        some "unreachable code" warnings.
2329    
2330    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2331        things, this adds five new scripts.
2332    
2333    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2334        There were also incompatibilities regarding the handling of \Q..\E inside
2335        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2336        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2337    
2338    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2339        matches an empty string, and forcibly breaks the loop. There were bugs in
2340        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2341        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2342        separate and independent bugs (that affected different cases) have been
2343        fixed.
2344    
2345    39. Refactored the code to abolish the use of different opcodes for small
2346        capturing bracket numbers. This is a tidy that I avoided doing when I
2347        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2348        The new approach is not only tidier, it makes it possible to reduce the
2349        memory needed to fix the previous bug (38).
2350    
2351    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2352        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2353        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2354        mode.
2355    
2356    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2357        report.
2358    
2359    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2360        copy construction and assignment in the C++ wrapper.
2361    
2362    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2363        couple of bugs that could have given wrong results in the "--newline=crlf"
2364        case.
2365    
2366    44. Added a number of casts and did some reorganization of signed/unsigned int
2367        variables following suggestions from Dair Grant. Also renamed the variable
2368        "this" as "item" because it is a C++ keyword.
2369    
2370    45. Arranged for dftables to add
2371    
2372          #include "pcre_internal.h"
2373    
2374        to pcre_chartables.c because without it, gcc 4.x may remove the array
2375        definition from the final binary if PCRE is built into a static library and
2376        dead code stripping is activated.
2377    
2378    46. For an unanchored pattern, if a match attempt fails at the start of a
2379        newline sequence, and the newline setting is CRLF or ANY, and the next two
2380        characters are CRLF, advance by two characters instead of one.
2381    
2382    
2383    Version 6.7 04-Jul-06
2384    ---------------------
2385    
2386     1. In order to handle tests when input lines are enormously long, pcretest has
2387        been re-factored so that it automatically extends its buffers when
2388        necessary. The code is crude, but this _is_ just a test program. The
2389        default size has been increased from 32K to 50K.
2390    
2391     2. The code in pcre_study() was using the value of the re argument before
2392        testing it for NULL. (Of course, in any sensible call of the function, it
2393        won't be NULL.)
2394    
2395     3. The memmove() emulation function in pcre_internal.h, which is used on
2396        systems that lack both memmove() and bcopy() - that is, hardly ever -
2397        was missing a "static" storage class specifier.
2398    
2399     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2400        containing an extended class (one that cannot be represented by a bitmap
2401        because it contains high-valued characters or Unicode property items, e.g.
2402        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2403        pattern, but PCRE should not loop if you do not (it no longer does).
2404        [Detail: two cases were found: (a) a repeated subpattern containing an
2405        extended class; (b) a recursive reference to a subpattern that followed a
2406        previous extended class. It wasn't skipping over the extended class
2407        correctly when UTF-8 mode was not set.]
2408    
2409     5. A negated single-character class was not being recognized as fixed-length
2410        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2411        compile error "lookbehind assertion is not fixed length".
2412    
2413     6. The RunPerlTest auxiliary script was showing an unexpected difference
2414        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2415        write a Perl script that can interpret lines of an input file either as
2416        byte characters or as UTF-8, which is what "perltest" was being required to
2417        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2418        can't do is switch easily at run time between having the "use utf8;" pragma
2419        or not. In the end, I fudged it by using the RunPerlTest script to insert
2420        "use utf8;" explicitly for the UTF-8 tests.
2421    
2422     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2423        the end of the subject string, contrary to the documentation and to what
2424        Perl does. This was true of both matching functions. Now it matches only at
2425        the start of the subject and immediately after *internal* newlines.
2426    
2427     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2428        a pointer to an int instead of a pointer to an unsigned long int. This
2429        caused problems on 64-bit systems.
2430    
2431     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2432        instance of the 'standard' template library not being so standard".
2433    
2434    10. There was no check on the number of named subpatterns nor the maximum
2435        length of a subpattern name. The product of these values is used to compute
2436        the size of the memory block for a compiled pattern. By supplying a very
2437        long subpattern name and a large number of named subpatterns, the size
2438        computation could be caused to overflow. This is now prevented by limiting
2439        the length of names to 32 characters, and the number of named subpatterns
2440        to 10,000.
2441    
2442    11. Subpatterns that are repeated with specific counts have to be replicated in
2443        the compiled pattern. The size of memory for this was computed from the
2444        length of the subpattern and the repeat count. The latter is limited to
2445        65535, but there was no limit on the former, meaning that integer overflow
2446        could in principle occur. The compiled length of a repeated subpattern is
2447        now limited to 30,000 bytes in order to prevent this.
2448    
2449    12. Added the optional facility to have named substrings with the same name.
2450    
2451    13. Added the ability to use a named substring as a condition, using the
2452        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2453        are numbers (not recommended). Forward references are permitted.
2454    
2455    14. Added forward references in named backreferences (if you see what I mean).
2456    
2457    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2458        pattern could run off the end of the subject. For example, the pattern
2459        "(?s)(.{1,5})"8 did this with the subject "ab".
2460    
2461    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2462        PCRE_CASELESS was set when matching characters that were quantified with ?
2463        or *.
2464    
2465    17. A character class other than a single negated character that had a minimum
2466        but no maximum quantifier - for example [ab]{6,} - was not handled
2467        correctly by pce_dfa_exec(). It would match only one character.
2468    
2469    18. A valid (though odd) pattern that looked like a POSIX character
2470        class but used an invalid character after [ (for example [[,abc,]]) caused
2471        pcre_compile() to give the error "Failed: internal error: code overflow" or
2472        in some cases to crash with a glibc free() error. This could even happen if
2473        the pattern terminated after [[ but there just happened to be a sequence of
2474        letters, a binary zero, and a closing ] in the memory that followed.
2475    
2476    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2477        over the years. Originally (before any Unicode support), just the bottom 8
2478        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2479        output from "man perlunicode" includes this:
2480    
2481          The regular expression compiler produces polymorphic opcodes.  That
2482          is, the pattern adapts to the data and automatically switches to
2483          the Unicode character scheme when presented with Unicode data--or
2484          instead uses a traditional byte scheme when presented with byte
2485          data.
2486    
2487        Sadly, a wide octal escape does not cause a switch, and in a string with
2488        no other multibyte characters, these octal escapes are treated as before.
2489        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2490        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2491        Unicode string.
2492    
2493        I have not perpetrated such confusion in PCRE. Up till now, it took just
2494        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2495        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2496        translate to the appropriate multibyte character.
2497    
2498    29. Applied some refactoring to reduce the number of warnings from Microsoft
2499        and Borland compilers. This has included removing the fudge introduced
2500        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2501        a warning about an unused variable.
2502    
2503    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2504        characters since release 4.0, because Perl (from release 5.004) does not.
2505        [Or at least, is documented not to: some releases seem to be in conflict
2506        with the documentation.] However, when a pattern was studied with
2507        pcre_study() and all its branches started with \s, PCRE still included VT
2508        as a possible starting character. Of course, this did no harm; it just
2509        caused an unnecessary match attempt.
2510    
2511    22. Removed a now-redundant internal flag bit that recorded the fact that case
2512        dependency changed within the pattern. This was once needed for "required
2513        byte" processing, but is no longer used. This recovers a now-scarce options
2514        bit. Also moved the least significant internal flag bit to the most-
2515        significant bit of the word, which was not previously used (hangover from
2516        the days when it was an int rather than a uint) to free up another bit for
2517        the future.
2518    
2519    23. Added support for CRLF line endings as well as CR and LF. As well as the
2520        default being selectable at build time, it can now be changed at runtime
2521        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2522        specify that it is scanning data with non-default line endings.
2523    
2524    24. Changed the definition of CXXLINK to make it agree with the definition of
2525        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2526    
2527    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2528        recursions. This makes a big different to stack usage for some patterns.
2529    
2530    26. If a subpattern containing a named recursion or subroutine reference such
2531        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2532        the space required for the compiled pattern went wrong and gave too small a
2533        value. Depending on the environment, this could lead to "Failed: internal
2534        error: code overflow at offset 49" or "glibc detected double free or
2535        corruption" errors.
2536    
2537    27. Applied patches from Google (a) to support the new newline modes and (b) to
2538        advance over multibyte UTF-8 characters in GlobalReplace.
2539    
2540    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2541        difference for some implementation of PCRE in some Windows version.
2542    
2543    29. Added some extra testing facilities to pcretest:
2544    
2545        \q<number>   in a data line sets the "match limit" value
2546        \Q<number>   in a data line sets the "match recursion limt" value
2547        -S <number>  sets the stack size, where <number> is in megabytes
2548    
2549        The -S option isn't available for Windows.
2550    
2551    
2552    Version 6.6 06-Feb-06
2553    ---------------------
2554    
2555     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2556        in pcreposix.h. I have copied the definition from pcre.h.
2557    
2558     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2559        because pcre.h is no longer a built file.
2560    
2561     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2562        not normally included in the compiled code.
2563    
2564    
2565    Version 6.5 01-Feb-06
2566    ---------------------
2567    
2568     1. When using the partial match feature with pcre_dfa_exec(), it was not
2569        anchoring the second and subsequent partial matches at the new starting
2570        point. This could lead to incorrect results. For example, with the pattern
2571        /1234/, partially matching against "123" and then "a4" gave a match.
2572    
2573     2. Changes to pcregrep:
2574    
2575        (a) All non-match returns from pcre_exec() were being treated as failures
2576            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2577            error message is output. Some extra information is given for the
2578            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2579            probably the only errors that are likely to be caused by users (by
2580            specifying a regex that has nested indefinite repeats, for instance).
2581            If there are more than 20 of these errors, pcregrep is abandoned.
2582    
2583        (b) A binary zero was treated as data while matching, but terminated the
2584            output line if it was written out. This has been fixed: binary zeroes
2585            are now no different to any other data bytes.
2586    
2587        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2588            used to set a locale for matching. The --locale=xxxx long option has
2589            been added (no short equivalent) to specify a locale explicitly on the
2590            pcregrep command, overriding the environment variables.
2591    
2592        (d) When -B was used with -n, some line numbers in the output were one less
2593            than they should have been.
2594    
2595        (e) Added the -o (--only-matching) option.
2596    
2597        (f) If -A or -C was used with -c (count only), some lines of context were
2598            accidentally printed for the final match.
2599    
2600        (g) Added the -H (--with-filename) option.
2601    
2602        (h) The combination of options -rh failed to suppress file names for files
2603            that were found from directory arguments.
2604    
2605        (i) Added the -D (--devices) and -d (--directories) options.
2606    
2607        (j) Added the -F (--fixed-strings) option.
2608    
2609        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2610    
2611        (l) Added the --colo(u)r option.
2612    
2613        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2614            is not present by default.
2615    
2616     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2617        items such as (?R) or (?1), when the recursion could match a number of
2618        alternatives. If it matched one of the alternatives, but subsequently,
2619        outside the recursion, there was a failure, the code tried to back up into
2620        the recursion. However, because of the way PCRE is implemented, this is not
2621        possible, and the result was an incorrect result from the match.
2622    
2623        In order to prevent this happening, the specification of recursion has
2624        been changed so that all such subpatterns are automatically treated as
2625        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2626    
2627     4. I had overlooked the fact that, in some locales, there are characters for
2628        which isalpha() is true but neither isupper() nor islower() are true. In
2629        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2630        and ordfeminine) are like this. This affected the treatment of \w and \W
2631        when they appeared in character classes, but not when they appeared outside
2632        a character class. The bit map for "word" characters is now created
2633        separately from the results of isalnum() instead of just taking it from the
2634        upper, lower, and digit maps. (Plus the underscore character, of course.)
2635    
2636     5. The above bug also affected the handling of POSIX character classes such as
2637        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2638        permanent tables. Instead, the bit maps for such a class were previously
2639        created as the appropriate unions of the upper, lower, and digit bitmaps.
2640        Now they are created by subtraction from the [[:word:]] class, which has
2641        its own bitmap.
2642    
2643     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2644        It is created by subtracting the vertical space characters (\x09, \x0a,
2645        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2646        subtraction was done in the overall bitmap for a character class, meaning
2647        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2648        be recognized. This bug has been fixed.
2649    
2650     7. Patches from the folks at Google:
2651    
2652          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2653          real life, but is still worth protecting against".
2654    
2655          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2656          regular expressions".
2657    
2658          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2659          have it.
2660    
2661          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2662          "configure" and the latter not, in order to fix a problem somebody had
2663          with compiling the Arg class on HP-UX.
2664    
2665          (e) Improve the error-handling of the C++ wrapper a little bit.
2666    
2667          (f) New tests for checking recursion limiting.
2668    
2669     8. The pcre_memmove() function, which is used only if the environment does not
2670        have a standard memmove() function (and is therefore rarely compiled),
2671        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2672        returning a result (though PCRE never actually uses the result).
2673    
2674     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2675        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2676        returned instead of calling malloc() with an overflowing number that would
2677        most likely cause subsequent chaos.
2678    
2679    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2680    
2681    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2682        with this option is matched, the nmatch and pmatch options of regexec() are
2683        ignored.
2684    
2685    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2686        provided in case anyone wants to the the POSIX interface with UTF-8
2687        strings.
2688    
2689    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2690        C++ linking (needed for some HP-UX environments).
2691    
2692    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2693        (unused parameter) and in the pcre_printint() function (omitted "default"
2694        switch label when the default is to do nothing).
2695    
2696    15. Added some code to make it possible, when PCRE is compiled as a C++
2697        library, to replace subject pointers for pcre_exec() with a smart pointer
2698        class, thus making it possible to process discontinuous strings.
2699    
2700    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2701        much the same function. They were added by different people who were trying
2702        to make PCRE easy to compile on non-Unix systems. It has been suggested
2703        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2704        for compiling on Windows systems. I have therefore replaced it with
2705        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2706        defaults to "extern" for C or "extern C" for C++, which works fine on
2707        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2708        SCOPE with something explicit in config.h. In addition:
2709    
2710        (a) pcreposix.h still had just "extern" instead of either of these macros;
2711            I have replaced it with PCRE_DATA_SCOPE.
2712    
2713        (b) Functions such as _pcre_xclass(), which are internal to the library,
2714            but external in the C sense, all had PCRE_EXPORT in their definitions.
2715            This is apparently wrong for the Windows case, so I have removed it.
2716            (It makes no difference on Unix-like systems.)
2717    
2718    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2719        of recursive calls to match(). This is different to MATCH_LIMIT because
2720        that limits the total number of calls to match(), not all of which increase
2721        the depth of recursion. Limiting the recursion depth limits the amount of
2722        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2723        when PCRE is compiled, and changed at run time. A patch from Google adds
2724        this functionality to the C++ interface.
2725    
2726    18. Changes to the handling of Unicode character properties:
2727    
2728        (a) Updated the table to Unicode 4.1.0.
2729    
2730        (b) Recognize characters that are not in the table as "Cn" (undefined).
2731    
2732        (c) I revised the way the table is implemented to a much improved format
2733            which includes recognition of ranges. It now supports the ranges that
2734            are defined in UnicodeData.txt, and it also amalgamates other
2735            characters into ranges. This has reduced the number of entries in the
2736            table from around 16,000 to around 3,000, thus reducing its size
2737            considerably. I realized I did not need to use a tree structure after
2738            all - a binary chop search is just as efficient. Having reduced the
2739            number of entries, I extended their size from 6 bytes to 8 bytes to
2740            allow for more data.
2741    
2742        (d) Added support for Unicode script names via properties such as \p{Han}.
2743    
2744    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2745        matching that character.
2746    
2747    20. When matching a repeated Unicode property with a minimum greater than zero,
2748        (for example \pL{2,}), PCRE could look past the end of the subject if it
2749        reached it while seeking the minimum number of characters. This could
2750        happen only if some of the characters were more than one byte long, because
2751        there is a check for at least the minimum number of bytes.
2752    
2753    21. Refactored the implementation of \p and \P so as to be more general, to
2754        allow for more different types of property in future. This has changed the
2755        compiled form incompatibly. Anybody with saved compiled patterns that use
2756        \p or \P will have to recompile them.
2757    
2758    22. Added "Any" and "L&" to the supported property types.
2759    
2760    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2761        but give a compile time error if the value is greater than 0xff.
2762    
2763    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2764        accidentally not being installed or uninstalled.
2765    
2766    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2767        made were to insert the current release number. This seemed silly, because
2768        it made things harder for people building PCRE on systems that don't run
2769        "configure". I have turned pcre.h into a distributed file, no longer built
2770        by "configure", with the version identification directly included. There is
2771        no longer a pcre.h.in file.
2772    
2773        However, this change necessitated a change to the pcre-config script as
2774        well. It is built from pcre-config.in, and one of the substitutions was the
2775        release number. I have updated configure.ac so that ./configure now finds
2776        the release number by grepping pcre.h.
2777    
2778    26. Added the ability to run the tests under valgrind.
2779    
2780    
2781    Version 6.4 05-Sep-05
2782    ---------------------
2783    
2784     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2785        "--" to be printed when multiple files were scanned, even when none of the
2786        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2787        consider it to be a bug, and have restored the previous behaviour.
2788    
2789     2. A couple of code tidies to get rid of compiler warnings.
2790    
2791     3. The pcretest program used to cheat by referring to symbols in the library
2792        whose names begin with _pcre_. These are internal symbols that are not
2793        really supposed to be visible externally, and in some environments it is
2794        possible to suppress them. The cheating is now confined to including
2795        certain files from the library's source, which is a bit cleaner.
2796    
2797     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2798        file's purpose clearer.
2799    
2800     5. Reorganized pcre_ucp_findchar().
2801    
2802    
2803    Version 6.3 15-Aug-05
2804    ---------------------
2805    
2806     1. The file libpcre.pc.in did not have general read permission in the tarball.
2807    
2808     2. There were some problems when building without C++ support:
2809    
2810        (a) If C++ support was not built, "make install" and "make test" still
2811            tried to test it.
2812    
2813        (b) There were problems when the value of CXX was explicitly set. Some
2814            changes have been made to try to fix these, and ...
2815    
2816        (c) --disable-cpp can now be used to explicitly disable C++ support.
2817    
2818        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2819            backslash in a target when C++ was disabled. This confuses some
2820            versions of "make", apparently. Using an intermediate variable solves
2821            this. (Same for CPP_LOBJ.)
2822    
2823     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2824        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2825        necessary on certain architectures.
2826    
2827     4. Added a setting of -export-symbols-regex to the link command to remove
2828        those symbols that are exported in the C sense, but actually are local
2829        within the library, and not documented. Their names all begin with
2830        "_pcre_". This is not a perfect job, because (a) we have to except some
2831        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2832        available (and never for static libraries). I have made a note to try to
2833        find a way round (a) in the future.
2834    
2835    
2836    Version 6.2 01-Aug-05
2837    ---------------------
2838    
2839     1. There was no test for integer overflow of quantifier values. A construction
2840        such as {1111111111111111} would give undefined results. What is worse, if
2841        a minimum quantifier for a parenthesized subpattern overflowed and became
2842        negative, the calculation of the memory size went wrong. This could have
2843        led to memory overwriting.
2844    
2845     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2846    
2847     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2848        operating environments where this matters.
2849    
2850     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2851        PCRE options from within the C++ wrapper.
2852    
2853     5. Named capturing subpatterns were not being correctly counted when a pattern
2854        was compiled. This caused two problems: (a) If there were more than 100
2855        such subpatterns, the calculation of the memory needed for the whole
2856        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2857        back references of the form \12, where the number was greater than 9, were
2858        not recognized as back references, even though there were sufficient
2859        previous subpatterns.
2860    
2861     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2862        versions of gcc, e.g. 2.95.4.
2863    
2864    
2865    Version 6.1 21-Jun-05
2866    ---------------------
2867    
2868     1. There was one reference to the variable "posix" in pcretest.c that was not
2869        surrounded by "#if !defined NOPOSIX".
2870    
2871     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2872        the cross-check on the old pcre_info() function, for the benefit of the
2873        cut-down version of PCRE that is currently imported into Exim.
2874    
2875     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2876        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2877        patterns that start (?i) but I don't think that matters. The use of (?i) is
2878        just an example; this all applies to the other options as well.
2879    
2880     4. Since libtool seems to echo the compile commands it is issuing, the output
2881        from "make" can be reduced a bit by putting "@" in front of each libtool
2882        compile command.
2883    
2884     5. Patch from the folks at Google for configure.in to be a bit more thorough
2885        in checking for a suitable C++ installation before trying to compile the
2886        C++ stuff. This should fix a reported problem when a compiler was present,
2887        but no suitable headers.
2888    
2889     6. The man pages all had just "PCRE" as their title. I have changed them to
2890        be the relevant file name. I have also arranged that these names are
2891        retained in the file doc/pcre.txt, which is a concatenation in text format
2892        of all the man pages except the little individual ones for each function.
2893    
2894     7. The NON-UNIX-USE file had not been updated for the different set of source
2895        files that come with release 6. I also added a few comments about the C++
2896        wrapper.
2897    
2898    
2899    Version 6.0 07-Jun-05
2900    ---------------------
2901    
2902     1. Some minor internal re-organization to help with my DFA experiments.
2903    
2904     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2905        didn't matter for the library itself when fully configured, but did matter
2906        when compiling without UCP support, or within Exim, where the ucp files are
2907        not imported.
2908    
2909     3. Refactoring of the library code to split up the various functions into
2910        different source modules. The addition of the new DFA matching code (see
2911        below) to a single monolithic source would have made it really too
2912        unwieldy, quite apart from causing all the code to be include in a
2913        statically linked application, when only some functions are used. This is
2914        relevant even without the DFA addition now that patterns can be compiled in
2915        one application and matched in another.
2916    
2917        The downside of splitting up is that there have to be some external
2918        functions and data tables that are used internally in different modules of
2919        the library but which are not part of the API. These have all had their
2920        names changed to start with "_pcre_" so that they are unlikely to clash
2921        with other external names.
2922    
2923     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2924        a different (DFA) algorithm. Although it is slower than the original
2925        function, it does have some advantages for certain types of matching
2926        problem.
2927    
2928     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2929        including restarting after a partial match.
2930    
2931     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2932        defined when compiling for Windows was sent to me. I have put it into the
2933        code, though I have no means of testing or verifying it.
2934    
2935     7. Added the pcre_refcount() auxiliary function.
2936    
2937     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2938        match before or at the first newline in the subject string. In pcretest,
2939        the /f option on a pattern can be used to set this.
2940    
2941     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2942        would behave wrongly. This has been present in PCRE since release 4.0.
2943    
2944    10. A number of changes to the pcregrep command:
2945    
2946        (a) Refactored how -x works; insert ^(...)$ instead of setting
2947            PCRE_ANCHORED and checking the length, in preparation for adding
2948            something similar for -w.
2949    
2950        (b) Added the -w (match as a word) option.
2951    
2952        (c) Refactored the way lines are read and buffered so as to have more
2953            than one at a time available.
2954    
2955        (d) Implemented a pcregrep test script.
2956    
2957        (e) Added the -M (multiline match) option. This allows patterns to match
2958            over several lines of the subject. The buffering ensures that at least
2959            8K, or the rest of the document (whichever is the shorter) is available
2960            for matching (and similarly the previous 8K for lookbehind assertions).
2961    
2962        (f) Changed the --help output so that it now says
2963    
2964              -w, --word-regex(p)
2965    
2966            instead of two lines, one with "regex" and the other with "regexp"
2967            because that confused at least one person since the short forms are the
2968            same. (This required a bit of code, as the output is generated
2969            automatically from a table. It wasn't just a text change.)
2970    
2971        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2972            option but starts with a hyphen. Could be a pattern or a path name
2973            starting with a hyphen, for instance.
2974    
2975        (h) "-" can be given as a file name to represent stdin.
2976    
2977        (i) When file names are being printed, "(standard input)" is used for
2978            the standard input, for compatibility with GNU grep. Previously
2979            "<stdin>" was used.
2980    
2981        (j) The option --label=xxx can be used to supply a name to be used for
2982            stdin when file names are being printed. There is no short form.
2983    
2984        (k) Re-factored the options decoding logic because we are going to add
2985            two more options that take data. Such options can now be given in four
2986            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2987    
2988        (l) Added the -A, -B, and -C options for requesting that lines of context
2989            around matches be printed.
2990    
2991        (m) Added the -L option to print the names of files that do not contain
2992            any matching lines, that is, the complement of -l.
2993    
2994        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2995            continue to scan other files.
2996    
2997        (o) The -s option was incorrectly implemented. For compatibility with other
2998            greps, it now suppresses the error message for a non-existent or non-
2999            accessible file (but not the return code). There is a new option called
3000            -q that suppresses the output of matching lines, which was what -s was
3001            previously doing.
3002    
3003        (p) Added --include and --exclude options to specify files for inclusion
3004            and exclusion when recursing.
3005    
3006    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3007        Hopefully, it now does.
3008    
3009    12. Missing cast in pcre_study().
3010    
3011    13. Added an "uninstall" target to the makefile.
3012    
3013    14. Replaced "extern" in the function prototypes in Makefile.in with
3014        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3015        world, but is set differently for Windows.
3016    
3017    15. Added a second compiling function called pcre_compile2(). The only
3018        difference is that it has an extra argument, which is a pointer to an
3019        integer error code. When there is a compile-time failure, this is set
3020        non-zero, in addition to the error test pointer being set to point to an
3021        error message. The new argument may be NULL if no error number is required
3022        (but then you may as well call pcre_compile(), which is now just a
3023        wrapper). This facility is provided because some applications need a
3024        numeric error indication, but it has also enabled me to tidy up the way
3025        compile-time errors are handled in the POSIX wrapper.
3026    
3027    16. Added VPATH=.libs to the makefile; this should help when building with one
3028        prefix path and installing with another. (Or so I'm told by someone who
3029        knows more about this stuff than I do.)
3030    
3031    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3032        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3033        match everything, including newlines. This is not POSIX-compatible, but
3034        somebody wanted the feature. From pcretest it can be activated by using
3035        both the P and the s flags.
3036    
3037    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3038    
3039    19. libpcre.pc was being incorrectly installed as executable.
3040    
3041    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3042        it now also looks for '\r' so that it will work unmodified on Windows.
3043    
3044    21. Added Google's contributed C++ wrapper to the distribution.
3045    
3046    22. Added some untidy missing memory free() calls in pcretest, to keep
3047        Electric Fence happy when testing.
3048    
3049    
3050    
3051    Version 5.0 13-Sep-04
3052    ---------------------
3053    
3054     1. Internal change: literal characters are no longer packed up into items
3055        containing multiple characters in a single byte-string. Each character
3056        is now matched using a separate opcode. However, there may be more than one
3057        byte in the character in UTF-8 mode.
3058    
3059     2. The pcre_callout_block structure has two new fields: pattern_position and
3060        next_item_length. These contain the offset in the pattern to the next match
3061        item, and its length, respectively.
3062    
3063     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3064        insertion of callouts before each pattern item. Added the /C option to
3065        pcretest to make use of this.
3066    
3067     4. On the advice of a Windows user, the lines
3068    
3069          #if defined(_WIN32) || defined(WIN32)
3070          _setmode( _fileno( stdout ), 0x8000 );
3071          #endif  /* defined(_WIN32) || defined(WIN32) */
3072    
3073        have been added to the source of pcretest. This apparently does useful
3074        magic in relation to line terminators.
3075    
3076     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3077        for the benefit of those environments where the "b" makes a difference.
3078    
3079     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3080        to know about it. I have put a hack into configure.in that adds in code
3081        to set GCC=yes if CC=icc. This seems to end up at a point in the
3082        generated configure script that is early enough to affect the setting of
3083        compiler options, which is what is needed, but I have no means of testing
3084        whether it really works. (The user who reported this had patched the
3085        generated configure script, which of course I cannot do.)
3086    
3087        LATER: After change 22 below (new libtool files), the configure script
3088        seems to know about icc (and also ecc). Therefore, I have commented out
3089        this hack in configure.in.
3090    
3091     7. Added support for pkg-config (2 patches were sent in).
3092    
3093     8. Negated POSIX character classes that used a combination of internal tables
3094        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3095        [[:^ascii]]. Typically, they would match almost any characters. The other
3096        POSIX classes were not broken in this way.
3097    
3098     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3099        to find the match, as PCRE was deluded into thinking that the match had to
3100        start at the start point or following a newline. The same bug applied to
3101        patterns with negative forward assertions or any backward assertions
3102        preceding ".*" at the start, unless the pattern required a fixed first
3103        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3104    
3105    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3106        starting at the last subject character, bytes beyond the end of the subject
3107        string were read.
3108    
3109    11. Renamed the variable "class" as "classbits" to make life easier for C++
3110        users. (Previously there was a macro definition, but it apparently wasn't
3111        enough.)
3112    
3113    12. Added the new field "tables" to the extra data so that tables can be passed
3114        in at exec time, or the internal tables can be re-selected. This allows
3115        a compiled regex to be saved and re-used at a later time by a different
3116        program that might have everything at different addresses.
3117    
3118    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3119        -R library as well as a -L library.
3120    
3121    14. The debugging options of pcretest (-d on the command line or D on a
3122        pattern) showed incorrect output for anything following an extended class
3123        that contained multibyte characters and which was followed by a quantifier.
3124    
3125    15. Added optional support for general category Unicode character properties
3126        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3127        support. It adds about 90K to the size of the library. The meanings of the
3128        inbuilt class escapes such as \d and \s have NOT been changed.
3129    
3130    16. Updated pcredemo.c to include calls to free() to release the memory for the
3131        compiled pattern.
3132    
3133    17. The generated file chartables.c was being created in the source directory
3134        instead of in the building directory. This caused the build to fail if the
3135        source directory was different from the building directory, and was
3136        read-only.
3137    
3138    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3139        file. No doubt somebody will tell me if they don't make sense... Also added
3140        Dan Mooney's comments about building on OpenVMS.
3141    
3142    19. Added support for partial matching via the PCRE_PARTIAL option for
3143        pcre_exec() and the \P data escape in pcretest.
3144    
3145    20. Extended pcretest with 3 new pattern features:
3146    
3147        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3148              write the compiled pattern to the file whose name is "rest-of-line".
3149              This is a straight binary dump of the data, with the saved pointer to
3150              the character tables forced to be NULL. The study data, if any, is
3151              written too. After writing, pcretest reads a new pattern.
3152    
3153        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3154              compiled pattern from the given file. There must not be any
3155              occurrences of "<" in the file name (pretty unlikely); if there are,
3156              pcretest will instead treat the initial "<" as a pattern delimiter.
3157              After reading in the pattern, pcretest goes on to read data lines as
3158              usual.
3159    
3160        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3161              and 16-bit fields in a compiled pattern, to simulate a pattern that
3162              was compiled on a host of opposite endianness.
3163    
3164    21. The pcre-exec() function can now cope with patterns that were compiled on
3165        hosts of opposite endianness, with this restriction:
3166    
3167          As for any compiled expression that is saved and used later, the tables
3168          pointer field cannot be preserved; the extra_data field in the arguments
3169          to pcre_exec() should be used to pass in a tables address if a value
3170          other than the default internal tables were used at compile time.
3171    
3172    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3173        now diagnosed as an error. Previously, most of the time, a negative number
3174        would have been treated as zero, but if in addition "ovector" was passed as
3175        NULL, a crash could occur.
3176    
3177    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3178        new versions from the libtool 1.5 distribution (the last one is a copy of
3179        a file called libtool.m4). This seems to have fixed the need to patch
3180        "configure" to support Darwin 1.3 (which I used to do). However, I still
3181        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3182        workstation).
3183    
3184    24. Changed the PCRE licence to be the more standard "BSD" licence.
3185    
3186    
3187    Version 4.5 01-Dec-03
3188    ---------------------
3189    
3190     1. There has been some re-arrangement of the code for the match() function so
3191        that it can be compiled in a version that does not call itself recursively.
3192        Instead, it keeps those local variables that need separate instances for
3193        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3194        needs to "recurse". Keeping track of where control must go is done by means
3195        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3196        hide most of the details from the main code, and operates only if
3197        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3198        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3199        operating.
3200    
3201        To make it easier for callers to provide specially tailored get/free
3202        functions for this usage, two new functions, pcre_stack_malloc, and
3203        pcre_stack_free, are used. They are always called in strict stacking order,
3204        and the size of block requested is always the same.
3205    
3206        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3207        PCRE has been compiled to use the stack or the heap for recursion. The
3208        -C option of pcretest uses this to show which version is compiled.
3209    
3210        A new data escape \S, is added to pcretest; it causes the amounts of store
3211        obtained and freed by both kinds of malloc/free at match time to be added
3212        to the output.
3213    
3214     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3215        what's available on my current Linux desktop machine.
3216    
3217     3. When matching a UTF-8 string, the test for a valid string at the start has
3218        been extended. If start_offset is not zero, PCRE now checks that it points
3219        to a byte that is the start of a UTF-8 character. If not, it returns
3220        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3221        this is necessary because there may be backward assertions in the pattern.
3222        When matching the same subject several times, it may save resources to use
3223        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3224    
3225     4. The code for checking the validity of UTF-8 strings has been tightened so
3226        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3227        containing "overlong sequences".
3228    
3229     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3230        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3231        should have been "&", but it just so happened that all the cases this let
3232        through by mistake were picked up later in the function.
3233    
3234     6. I had used a variable called "isblank" - this is a C99 function, causing
3235        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3236    
3237     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3238        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3239        so the version is shown; (c) stop "make test" echoing "./RunTest".
3240    
3241     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3242    
3243     9. The prototype for memmove() for systems that don't have it was using
3244        size_t, but the inclusion of the header that defines size_t was later. I've
3245        moved the #includes for the C headers earlier to avoid this.
3246    
3247    10. Added some adjustments to the code to make it easier to compiler on certain
3248        special systems:
3249    
3250          (a) Some "const" qualifiers were missing.
3251          (b) Added the macro EXPORT before all exported functions; by default this
3252              is defined to be empty.
3253          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3254              that it reads its output file name as an argument instead of writing
3255              to the standard output and assuming this can be redirected.
3256    
3257    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3258        class containing characters with values greater than 255, PCRE compilation
3259        went into a loop.
3260    
3261    12. A recursive reference to a subpattern that was within another subpattern
3262        that had a minimum quantifier of zero caused PCRE to crash. For example,
3263        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3264        recursion. If the recursively-called subpattern itself had a zero repeat,
3265        that was OK.
3266    
3267    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3268        buffer into which it was copied (for escape processing) was still set at
3269        1024, so long lines caused crashes.
3270    
3271    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3272        "internal error: code overflow...". This applied to any character class
3273        that was followed by a possessive quantifier.
3274    
3275    15. Modified the Makefile to add libpcre.la as a prerequisite for
3276        libpcreposix.la because I was told this is needed for a parallel build to
3277        work.
3278    
3279    16. If a pattern that contained .* following optional items at the start was
3280        studied, the wrong optimizing data was generated, leading to matching
3281        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3282        matching string must start with a or b or c. The correct conclusion for
3283        this pattern is that a match can start with any character.
3284    
3285    
3286    Version 4.4 13-Aug-03
3287    ---------------------
3288    
3289     1. In UTF-8 mode, a character class containing characters with values between
3290        127 and 255 was not handled correctly if the compiled pattern was studied.
3291        In fixing this, I have also improved the studying algorithm for such
3292        classes (slightly).
3293    
3294     2. Three internal functions had redundant arguments passed to them. Removal
3295        might give a very teeny performance improvement.
3296    
3297     3. Documentation bug: the value of the capture_top field in a callout is *one
3298        more than* the number of the hightest numbered captured substring.
3299    
3300     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3301        in incorrectly linking with a previously installed version. They now link
3302        explicitly with libpcre.la.
3303    
3304     5. configure.in no longer needs to recognize Cygwin specially.
3305    
3306     6. A problem in pcre.in for Windows platforms is fixed.
3307    
3308     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3309        pcretest, it used to include the size of the study block as part of its
3310        output. Unfortunately, the structure contains a field that has a different
3311        size on different hardware architectures. This meant that the tests that
3312        showed this size failed. As the block is currently always of a fixed size,
3313        this information isn't actually particularly useful in pcretest output, so
3314        I have just removed it.
3315    
3316     8. Three pre-processor statements accidentally did not start in column 1.
3317        Sadly, there are *still* compilers around that complain, even though
3318        standard C has not required this for well over a decade. Sigh.
3319    
3320     9. In pcretest, the code for checking callouts passed small integers in the
3321        callout_data field, which is a void * field. However, some picky compilers
3322        complained about the casts involved for this on 64-bit systems. Now
3323        pcretest passes the address of the small integer instead, which should get
3324        rid of the warnings.
3325    
3326    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3327        both compile and run time, and gives an error if an invalid UTF-8 sequence
3328        is found. There is a option for disabling this check in cases where the
3329        string is known to be correct and/or the maximum performance is wanted.
3330    
3331    11. In response to a bug report, I changed one line in Makefile.in from
3332    
3333            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3334        to
3335            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3336    
3337        to look similar to other lines, but I have no way of telling whether this
3338        is the right thing to do, as I do not use Windows. No doubt I'll get told
3339        if it's wrong...
3340    
3341    
3342    Version 4.3 21-May-03
3343    ---------------------
3344    
3345    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3346       Makefile.
3347    
3348    2. Some refactoring to improve the quality of the code:
3349    
3350       (i)   The utf8_table... variables are now declared "const".
3351    
3352       (ii)  The code for \cx, which used the "case flipping" table to upper case
3353             lower case letters, now just substracts 32. This is ASCII-specific,
3354             but the whole concept of \cx is ASCII-specific, so it seems
3355             reasonable.
3356    
3357       (iii) PCRE was using its character types table to recognize decimal and
3358             hexadecimal digits in the pattern. This is silly, because it handles
3359             only 0-9, a-f, and A-F, but the character types table is locale-
3360             specific, which means strange things might happen. A private
3361             table is now used for this - though it costs 256 bytes, a table is
3362             much faster than multiple explicit tests. Of course, the standard
3363             character types table is still used for matching digits in subject
3364             strings against \d.
3365    
3366       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3367             ending in _t are). So I've renamed it as ESC_tee.
3368    
3369    3. The first argument for regexec() in the POSIX wrapper should have been
3370       defined as "const".
3371    
3372    4. Changed pcretest to use malloc() for its buffers so that they can be
3373       Electric Fenced for debugging.
3374    
3375    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3376       to read one or more bytes before the start of the subject string. Often this
3377       had no effect on PCRE's behaviour, but in some circumstances it could
3378       provoke a segmentation fault.
3379    
3380    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3381       to try to read one or more bytes before the start of the subject string.
3382    
3383    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3384       UTF-8 support could misbehave in various ways if the subject string
3385       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3386       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3387       back over UTF-8 characters.)
3388    
3389    
3390    Version 4.2 14-Apr-03
3391    ---------------------
3392    
3393    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3394    
3395    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3396         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3397         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3398         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3399         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3400           and BUILD_EXEEXT
3401         Note: automatic setting of the BUILD variables is not yet working
3402         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3403           compile-time but not at link-time
3404         [LINK]: use for linking executables only
3405         make different versions for Windows and non-Windows
3406         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3407           libraries
3408         [LINK_FOR_BUILD]: new variable
3409         [OBJEXT]: use throughout
3410         [EXEEXT]: use throughout
3411         <winshared>: new target
3412         <wininstall>: new target
3413         <dftables.o>: use native compiler
3414         <dftables>: use native linker
3415         <install>: handle Windows platform correctly
3416         <clean>: ditto
3417         <check>: ditto
3418         copy DLL to top builddir before testing
3419    
3420       As part of these changes, -no-undefined was removed again. This was reported
3421       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3422       in any case.
3423    
3424    3. Some tidies to get rid of compiler warnings:
3425    
3426       . In the match_data structure, match_limit was an unsigned long int, whereas
3427         match_call_count was an int. I've made them both unsigned long ints.
3428    
3429       . In pcretest the fact that a const uschar * doesn't automatically cast to
3430         a void * provoked a warning.
3431    
3432       . Turning on some more compiler warnings threw up some "shadow" variables
3433         and a few more missing casts.
3434    
3435    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3436       option, a class that contained a single character with a value between 128
3437       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3438    
3439    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3440       option, a class that contained several characters, but with at least one
3441       whose value was between 128 and 255 caused PCRE to crash.
3442    
3443    
3444    Version 4.1 12-Mar-03
3445    ---------------------
3446    
3447    1. Compiling with gcc -pedantic found a couple of places where casts were
3448    needed, and a string in dftables.c that was longer than standard compilers are
3449    required to support.
3450    
3451    2. Compiling with Sun's compiler found a few more places where the code could
3452    be tidied up in order to avoid warnings.
3453    
3454    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3455    first of these names is deprecated in the latest Autoconf in favour of the name
3456    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3457    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3458    analogy I have changed it to CFLAGS_FOR_BUILD.
3459    
3460    4. Added -no-undefined to the linking command in the Makefile, because this is
3461    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3462    linking step for the pcreposix library.
3463    
3464    5. PCRE was failing to diagnose the case of two named groups with the same
3465    name.
3466    
3467    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3468    literal character that is needed in the subject for a match, and scans along to
3469    ensure that it is present before embarking on the full matching process. This
3470    saves time in cases of nested unlimited repeats that are never going to match.
3471    Problem: the scan can take a lot of time if the subject is very long (e.g.
3472    megabytes), thus penalizing straightforward matches. It is now done only if the
3473    amount of subject to be scanned is less than 1000 bytes.
3474    
3475    7. A lesser problem with the same optimization is that it was recording the
3476    first character of an anchored pattern as "needed", thus provoking a search
3477    right along the subject, even when the first match of the pattern was going to
3478    fail. The "needed" character is now not set for anchored patterns, unless it
3479    follows something in the pattern that is of non-fixed length. Thus, it still
3480    fulfils its original purpose of finding quick non-matches in cases of nested
3481    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3482    
3483    
3484    Version 4.0 17-Feb-03
3485    ---------------------
3486    
3487    1. If a comment in an extended regex that started immediately after a meta-item
3488    extended to the end of string, PCRE compiled incorrect data. This could lead to
3489    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3490    
3491    2. Moved to autoconf 2.53 and libtool 1.4.2.
3492    
3493    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3494    the special perltest8 script is no longer needed - all the tests can be run
3495    from a single perltest script.
3496    
3497    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3498    by \s. It has now been removed in PCRE. This means it isn't recognized as
3499    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3500    class [:space:] *does* include VT, thereby creating a mess.
3501    
3502    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3503    space and tab.
3504    
3505    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3506    its new features into the main test script, reducing the number of scripts.
3507    
3508    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3509    were backward compatible, and made the (?i) apply to the whole pattern, as if
3510    /i were given. Now it behaves more logically, and applies the option setting
3511    only to what follows. PCRE has been changed to follow suit. However, if it
3512    finds options settings right at the start of the pattern, it extracts them into
3513    the global options, as before. Thus, they show up in the info data.
3514    
3515    8. Added support for the \Q...\E escape sequence. Characters in between are
3516    treated as literals. This is slightly different from Perl in that $ and @ are
3517    also handled as literals inside the quotes. In Perl, they will cause variable
3518    interpolation. Note the following examples:
3519    
3520        Pattern            PCRE matches      Perl matches
3521    
3522        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3523        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3524        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3525    
3526    For compatibility with Perl, \Q...\E sequences are recognized inside character
3527    classes as well as outside them.
3528    
3529    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3530    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3531    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3532    signed/unsigned warnings.
3533    
3534    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3535    option for pcretest, so I've replaced it by a simple function that does just
3536    that job.
3537    
3538    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3539    "pcregrep -".
3540    
3541    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3542    Java package. This provides some syntactic sugar for simple cases of what my
3543    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3544    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3545    item, you can use this simplified notation. Note that only makes sense with
3546    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3547    greediness, whatever the setting of the PCRE_UNGREEDY option.
3548    
3549    13. A change of greediness default within a pattern was not taking effect at
3550    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3551    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3552    was abstracted outside.
3553    
3554    14. PCRE now supports the \G assertion. It is true when the current matching
3555    position is at the start point of the match. This differs from \A when the
3556    starting offset is non-zero. Used with the /g option of pcretest (or similar
3557    code), it works in the same way as it does for Perl's /g option. If all
3558    alternatives of a regex begin with \G, the expression is anchored to the start
3559    match position, and the "anchored" flag is set in the compiled expression.
3560    
3561    15. Some bugs concerning the handling of certain option changes within patterns
3562    have been fixed. These applied to options other than (?ims). For example,
3563    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3564    been the other way round. Some of this was related to change 7 above.
3565    
3566    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3567    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3568    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3569    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3570    
3571    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3572    mode. Unlike ".", it always matches newline, whatever the setting of
3573    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3574    assertions. Perl allows it, but it doesn't (in general) work because it can't
3575    calculate the length of the lookbehind. At least, that's the case for Perl
3576    5.8.0 - I've been told they are going to document that it doesn't work in
3577    future.
3578    
3579    18. Added an error diagnosis for escapes that PCRE does not support: these are
3580    \L, \l, \N, \P, \p, \U, \u, and \X.
3581    
3582    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3583    reading past the end of the pattern in cases such as /[abcd/.
3584    
3585    20. PCRE was getting more memory than necessary for patterns with classes that
3586    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3587    
3588    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3589    compiling PCRE for use with Virtual Pascal.
3590    
3591    22. Small fix to the Makefile to make it work properly if the build is done
3592    outside the source tree.
3593    
3594    23. Added a new extension: a condition to go with recursion. If a conditional
3595    subpattern starts with (?(R) the "true" branch is used if recursion has
3596    happened, whereas the "false" branch is used only at the top level.
3597    
3598    24. When there was a very long string of literal characters (over 255 bytes
3599    without UTF support, over 250 bytes with UTF support), the computation of how
3600    much memory was required could be incorrect, leading to segfaults or other
3601    strange effects.
3602    
3603    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3604    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3605    there was a subsequent back reference to those brackets. This meant that, for
3606    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3607    possible to check for precisely this case. All we can do is abandon the
3608    optimization if .* occurs inside capturing brackets when there are any back
3609    references whatsoever. (See below for a better fix that came later.)
3610    
3611    26. The handling of the optimization for finding the first character of a
3612    non-anchored pattern, and for finding a character that is required later in the
3613    match were failing in some cases. This didn't break the matching; it just
3614    failed to optimize when it could. The way this is done has been re-implemented.
3615    
3616    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3617    
3618    28. Added a new feature that provides some of the functionality that Perl
3619    provides with (?{...}). The facility is termed a "callout". The way it is done
3620    in PCRE is for the caller to provide an optional function, by setting
3621    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3622    global variable. By default it is unset, which disables all calling out. To get
3623    the function called, the regex must include (?C) at appropriate points. This
3624    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3625    This provides a means of identifying different callout points. When PCRE
3626    reaches such a point in the regex, if pcre_callout has been set, the external
3627    function is called. It is provided with data in a structure called
3628    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3629    matching continues; if it returns a non-zero value, the match at the current
3630    point fails. However, backtracking will occur if possible. [This was changed
3631    later and other features added - see item 49 below.]
3632    
3633    29. pcretest is upgraded to test the callout functionality. It provides a
3634    callout function that displays information. By default, it shows the start of
3635    the match and the current position in the text. There are some new data escapes
3636    to vary what happens:
3637    
3638        \C+         in addition, show current contents of captured substrings
3639        \C-         do not supply a callout function
3640        \C!n        return 1 when callout number n is reached
3641        \C!n!m      return 1 when callout number n is reached for the mth time
3642    
3643    30. If pcregrep was called with the -l option and just a single file name, it
3644    output "<stdin>" if a match was found, instead of the file name.
3645    
3646    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3647    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3648    pcre_exec(). This saves a malloc/free per call. The default value of
3649    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3650    when configuring.
3651    
3652    32. The default maximum size of a compiled pattern is 64K. There have been a
3653    few cases of people hitting this limit. The code now uses macros to handle the
3654    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3655    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3656    configuring. Tests 2 and 5 work only with 2-byte links because they output
3657    debugging information about compiled patterns.
3658    
3659    33. Internal code re-arrangements:
3660    
3661    (a) Moved the debugging function for printing out a compiled regex into
3662        its own source file (printint.c) and used #include to pull it into
3663        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3664        separate copies.
3665    
3666    (b) Defined the list of op-code names for debugging as a macro in
3667        internal.h so that it is next to the definition of the opcodes.
3668    
3669    (c) Defined a table of op-code lengths for simpler skipping along compiled
3670        code. This is again a macro in internal.h so that it is next to the
3671        definition of the opcodes.
3672    
3673    34. Added support for recursive calls to individual subpatterns, along the
3674    lines of Robin Houston's patch (but implemented somewhat differently).
3675    
3676    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3677    allow it to read and process whole directories in Win32. This code was
3678    contributed by Lionel Fourquaux; it has not been tested by me.
3679    
3680    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3681    used to name a group. Names consist of alphanumerics and underscores, and must
3682    be unique. Back references use the syntax (?P=name) and recursive calls use
3683    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3684    numbers. The function pcre_fullinfo() can be used after compilation to extract
3685    a name/number map. There are three relevant calls:
3686    
3687      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3688      PCRE_INFO_NAMECOUNT            yields the number of entries
3689      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3690    
3691    The map is a vector of fixed-size entries. The size of each entry depends on
3692    the length of the longest name used. The first two bytes of each entry are the
3693    group number, most significant byte first. There follows the corresponding
3694    name, zero terminated. The names are in alphabetical order.
3695    
3696    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3697    case instead of 255. Making it the same both with and without UTF-8 support
3698    means that the same test output works with both.
3699    
3700    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3701    calling malloc() with a zero argument.
3702    
3703    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3704    optimization. I've improved things by keeping a bitmap of backreferences with
3705    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3706    fact referenced, the optimization can be applied. It is unlikely that a
3707    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3708    the match to follow \n) will appear inside brackets with a number greater than
3709    31, but if it does, any back reference > 31 suppresses the optimization.
3710    
3711    40. Added a new compi