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

Diff of /code/trunk/ChangeLog

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

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