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

Diff of /code/trunk/ChangeLog

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

revision 33 by nigel, Sat Feb 24 21:39:01 2007 UTC revision 659 by ph10, Tue Aug 16 09:48:29 2011 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.13 16-Aug-2011
5    ------------------------
6    
7    1.  The Unicode data tables have been updated to Unicode 6.0.0.
8    
9    2.  Two minor typos in pcre_internal.h have been fixed.
10    
11    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
12        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
13        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
14    
15    4.  There were a number of related bugs in the code for matching backrefences
16        caselessly in UTF-8 mode when codes for the characters concerned were
17        different numbers of bytes. For example, U+023A and U+2C65 are an upper
18        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
19        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
20        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
21        2-byte code at the end of the subject (it thought there wasn't enough data
22        left).
23    
24    5.  Comprehensive information about what went wrong is now returned by
25        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
26        as the output vector has at least 2 elements. The offset of the start of
27        the failing character and a reason code are placed in the vector.
28    
29    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
30        now returned is for the first byte of the failing character, instead of the
31        last byte inspected. This is an incompatible change, but I hope it is small
32        enough not to be a problem. It makes the returned offset consistent with
33        pcre_exec() and pcre_dfa_exec().
34    
35    7.  pcretest now gives a text phrase as well as the error number when
36        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
37        failure, the offset and reason code are output.
38    
39    8.  When \R was used with a maximizing quantifier it failed to skip backwards
40        over a \r\n pair if the subsequent match failed. Instead, it just skipped
41        back over a single character (\n). This seems wrong (because it treated the
42        two characters as a single entity when going forwards), conflicts with the
43        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
44        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
45        has been changed.
46    
47    9.  Some internal refactoring has changed the processing so that the handling
48        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
49        time (the PCRE_DOTALL option was changed this way some time ago: version
50        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
51        which was always a bit of a fudge. It also means that there is one less
52        argument for the match() function, which reduces its stack requirements
53        slightly. This change also fixes an incompatibility with Perl: the pattern
54        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
55    
56    10. More internal refactoring has drastically reduced the number of recursive
57        calls to match() for possessively repeated groups such as (abc)++ when
58        using pcre_exec().
59    
60    11. While implementing 10, a number of bugs in the handling of groups were
61        discovered and fixed:
62    
63        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
64        (a|)*(?1) gave a compile-time internal error.
65        ((a|)+)+  did not notice that the outer group could match an empty string.
66        (^a|^)+   was not marked as anchored.
67        (.*a|.*)+ was not marked as matching at start or after a newline.
68    
69    12. Yet more internal refactoring has removed another argument from the match()
70        function. Special calls to this function are now indicated by setting a
71        value in a variable in the "match data" data block.
72    
73    13. Be more explicit in pcre_study() instead of relying on "default" for
74        opcodes that mean there is no starting character; this means that when new
75        ones are added and accidentally left out of pcre_study(), testing should
76        pick them up.
77    
78    14. The -s option of pcretest has been documented for ages as being an old
79        synonym of -m (show memory usage). I have changed it to mean "force study
80        for every regex", that is, assume /S for every regex. This is similar to -i
81        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
82        using it. It makes it easier to run collections of tests with and without
83        study enabled, and thereby test pcre_study() more easily. All the standard
84        tests are now run with and without -s (but some patterns can be marked as
85        "never study" - see 20 below).
86    
87    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
88        restoration of the capturing data to the outer values was not happening
89        correctly.
90    
91    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
92        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
93        pattern had matched an empty string, and so incorrectly returned a no
94        match.
95    
96    17. There was optimizing code for the last branch of non-capturing parentheses,
97        and also for the obeyed branch of a conditional subexpression, which used
98        tail recursion to cut down on stack usage. Unfortunately, now that there is
99        the possibility of (*THEN) occurring in these branches, tail recursion is
100        no longer possible because the return has to be checked for (*THEN). These
101        two optimizations have therefore been removed.
102    
103    18. If a pattern containing \R was studied, it was assumed that \R always
104        matched two bytes, thus causing the minimum subject length to be
105        incorrectly computed because \R can also match just one byte.
106    
107    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
108        was incorrectly computed.
109    
110    20. If /S is present twice on a test pattern in pcretest input, it now
111        *disables* studying, thereby overriding the use of -s on the command line
112        (see 14 above). This is necessary for one or two tests to keep the output
113        identical in both cases.
114    
115    21. When (*ACCEPT) was used in an assertion that matched an empty string and
116        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
117    
118    22. When an atomic group that contained a capturing parenthesis was
119        successfully matched, but the branch in which it appeared failed, the
120        capturing was not being forgotten if a higher numbered group was later
121        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
122        group 1 to "a", when in fact it should be unset. This applied to multi-
123        branched capturing and non-capturing groups, repeated or not, and also to
124        positive assertions (capturing in negative assertions does not happen
125        in PCRE) and also to nested atomic groups.
126    
127    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
128        subject after a captured substring, to make it easier to tell which of a
129        number of identical substrings has been captured.
130    
131    24. The way atomic groups are processed by pcre_exec() has been changed so that
132        if they are repeated, backtracking one repetition now resets captured
133        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
134        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
135        "aaa". Previously, it would have been "a". As part of this code
136        refactoring, the way recursive calls are handled has also been changed.
137    
138    25. If an assertion condition captured any substrings, they were not passed
139        back unless some other capturing happened later. For example, if
140        (?(?=(a))a) was matched against "a", no capturing was returned.
141    
142    26. When studying a pattern that contained subroutine calls or assertions,
143        the code for finding the minimum length of a possible match was handling
144        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
145        group 1 called group 2 while simultaneously a separate group 2 called group
146        1). A stack overflow occurred in this case. I have fixed this by limiting
147        the recursion depth to 10.
148    
149    27. Updated RunTest.bat in the distribution to the version supplied by Tom
150        Fortmann. This supports explicit test numbers on the command line, and has
151        argument validation and error reporting.
152    
153    28. An instance of \X with an unlimited repeat could fail if at any point the
154        first character it looked at was a mark character.
155    
156    29. Some minor code refactoring concerning Unicode properties and scripts
157        should reduce the stack requirement of match() slightly.
158    
159    30. Added the '=' option to pcretest to check the setting of unused capturing
160        slots at the end of the pattern, which are documented as being -1, but are
161        not included in the return count.
162    
163    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
164        compiled something random. Now it gives a compile-time error (as does
165        Perl).
166    
167    32. A *MARK encountered during the processing of a positive assertion is now
168        recorded and passed back (compatible with Perl).
169    
170    33. If --only-matching or --colour was set on a pcregrep call whose pattern
171        had alternative anchored branches, the search for a second match in a line
172        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
173        matched the line "0102" twice. The same bug affected patterns that started
174        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
175        twice.
176    
177    34. Previously, PCRE did not allow quantification of assertions. However, Perl
178        does, and because of capturing effects, quantifying parenthesized
179        assertions may at times be useful. Quantifiers are now allowed for
180        parenthesized assertions.
181    
182    35. A minor code tidy in pcre_compile() when checking options for \R usage.
183    
184    36. \g was being checked for fancy things in a character class, when it should
185        just be a literal "g".
186    
187    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
188        appearance of a nested POSIX class supersedes an apparent external class.
189        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
190        unescaped square brackets may also appear as part of class names. For
191        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
192        more like Perl.
193    
194    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
195        was because it thought it was \N{name}, which is not supported).
196    
197    39. Add minix to OS list not supporting the -S option in pcretest.
198    
199    40. PCRE tries to detect cases of infinite recursion at compile time, but it
200        cannot analyze patterns in sufficient detail to catch mutual recursions
201        such as ((?1))((?2)). There is now a runtime test that gives an error if a
202        subgroup is called recursively as a subpattern for a second time at the
203        same position in the subject string. In previous releases this might have
204        been caught by the recursion limit, or it might have run out of stack.
205    
206    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
207        happen only once. PCRE was, however incorrectly giving a compile time error
208        "recursive call could loop indefinitely" because it cannot analyze the
209        pattern in sufficient detail. The compile time test no longer happens when
210        PCRE is compiling a conditional subpattern, but actual runaway loops are
211        now caught at runtime (see 40 above).
212    
213    42. It seems that Perl allows any characters other than a closing parenthesis
214        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
215        has been changed to be the same.
216    
217    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
218        as not to get warnings when autogen.sh is called. Also changed
219        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
220    
221    44. To help people who use pcregrep to scan files containing exceedingly long
222        lines, the following changes have been made:
223    
224        (a) The default value of the buffer size parameter has been increased from
225            8K to 20K. (The actual buffer used is three times this size.)
226    
227        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
228            PCRE is built.
229    
230        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
231            to be set at run time.
232    
233        (d) Numerical values in pcregrep options can be followed by K or M, for
234            example --buffer-size=50K.
235    
236        (e) If a line being scanned overflows pcregrep's buffer, an error is now
237            given and the return code is set to 2.
238    
239    45. Add a pointer to the latest mark to the callout data block.
240    
241    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
242        partial match of an empty string instead of no match. This was specific to
243        the use of ".".
244    
245    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
246        complete match instead of a partial match. This bug was dependent on both
247        the PCRE_UTF8 and PCRE_DOTALL options being set.
248    
249    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
250        starting byte set, because \b was not being ignored.
251    
252    
253    Version 8.12 15-Jan-2011
254    ------------------------
255    
256    1.  Fixed some typos in the markup of the man pages, and wrote a script that
257        checks for such things as part of the documentation building process.
258    
259    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
260        --match-limit and --recursion-limit options (added for 8.11). In
261        particular, this made one of the standard tests fail. (The integer value
262        went into the wrong half of a long int.)
263    
264    3.  If the --colour option was given to pcregrep with -v (invert match), it
265        did strange things, either producing crazy output, or crashing. It should,
266        of course, ignore a request for colour when reporting lines that do not
267        match.
268    
269    4.  Another pcregrep bug caused similar problems if --colour was specified with
270        -M (multiline) and the pattern match finished with a line ending.
271    
272    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
273        matched in multiline mode, the following line was shown as part of the
274        match. This seems wrong, so I have changed it.
275    
276    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
277        the check for further matches in the same line (so they could be coloured)
278        to overrun the end of the current line. If another match was found, it was
279        incorrectly shown (and then shown again when found in the next line).
280    
281    7.  If pcregrep was compiled under Windows, there was a reference to the
282        function pcregrep_exit() before it was defined. I am assuming this was
283        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
284        reported by a user. I've moved the definition above the reference.
285    
286    
287    Version 8.11 10-Dec-2010
288    ------------------------
289    
290    1.  (*THEN) was not working properly if there were untried alternatives prior
291        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
292        backtracked to try for "b" instead of moving to the next alternative branch
293        at the same level (in this case, to look for "c"). The Perl documentation
294        is clear that when (*THEN) is backtracked onto, it goes to the "next
295        alternative in the innermost enclosing group".
296    
297    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
298        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
299        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
300        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
301        (*THEN).
302    
303    3.  If \s appeared in a character class, it removed the VT character from
304        the class, even if it had been included by some previous item, for example
305        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
306        of \s, but is part of the POSIX "space" class.)
307    
308    4.  A partial match never returns an empty string (because you can always
309        match an empty string at the end of the subject); however the checking for
310        an empty string was starting at the "start of match" point. This has been
311        changed to the "earliest inspected character" point, because the returned
312        data for a partial match starts at this character. This means that, for
313        example, /(?<=abc)def/ gives a partial match for the subject "abc"
314        (previously it gave "no match").
315    
316    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
317        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
318        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
319        has an implication that the given string is incomplete (because a partial
320        match is preferred over a full match). For this reason, these items now
321        give a partial match in this situation. [Aside: previously, the one case
322        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
323        match rather than a full match, which was wrong by the old rules, but is
324        now correct.]
325    
326    6.  There was a bug in the handling of #-introduced comments, recognized when
327        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
328        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
329        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
330        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
331        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
332        places in pcre_compile().
333    
334    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
335        comments when looking ahead for named forward references to subpatterns,
336        the only newline sequence it recognized was NL. It now handles newlines
337        according to the set newline convention.
338    
339    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
340        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
341        cater for a lack of strerror(). These oversights have been fixed.
342    
343    9.  Added --match-limit and --recursion-limit to pcregrep.
344    
345    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
346    
347    11. When the -o option was used, pcregrep was setting a return code of 1, even
348        when matches were found, and --line-buffered was not being honoured.
349    
350    12. Added an optional parentheses number to the -o and --only-matching options
351        of pcregrep.
352    
353    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
354        can match an empty string. The code to do it in pcretest and pcredemo
355        needed fixing:
356    
357        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
358            only one byte after an empty string match just before CRLF (this case
359            just got forgotten; "any" and "anycrlf" were OK).
360    
361        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
362            mode when an empty string match preceded an ASCII character followed by
363            a non-ASCII character. (The code for advancing by one character rather
364            than one byte was nonsense.)
365    
366        (c) The pcredemo.c sample program did not have any code at all to handle
367            the cases when CRLF is a valid newline sequence.
368    
369    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
370        as a starting offset was within the subject string. There is now a new
371        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
372        negative or greater than the length of the string. In order to test this,
373        pcretest is extended to allow the setting of negative starting offsets.
374    
375    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
376        starting offset points to the beginning of a UTF-8 character was
377        unnecessarily clumsy. I tidied it up.
378    
379    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
380        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
381    
382    17. Nobody had reported that the --include_dir option, which was added in
383        release 7.7 should have been called --include-dir (hyphen, not underscore)
384        for compatibility with GNU grep. I have changed it to --include-dir, but
385        left --include_dir as an undocumented synonym, and the same for
386        --exclude-dir, though that is not available in GNU grep, at least as of
387        release 2.5.4.
388    
389    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
390        characters from a string of bytes) have been redefined so as not to use
391        loops, in order to improve performance in some environments. At the same
392        time, I abstracted some of the common code into auxiliary macros to save
393        repetition (this should not affect the compiled code).
394    
395    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
396        compile-time error is now given if \c is not followed by an ASCII
397        character, that is, a byte less than 128. (In EBCDIC mode, the code is
398        different, and any byte value is allowed.)
399    
400    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
401        START_OPTIMIZE option, which is now allowed at compile time - but just
402        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
403        to pcregrep and other applications that have no direct access to PCRE
404        options. The new /Y option in pcretest sets this option when calling
405        pcre_compile().
406    
407    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
408        back references. Groups containing recursive back references were forced to
409        be atomic by that change, but in the case of named groups, the amount of
410        memory required was incorrectly computed, leading to "Failed: internal
411        error: code overflow". This has been fixed.
412    
413    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
414        pcretest.c, to avoid build problems in some Borland environments.
415    
416    
417    Version 8.10 25-Jun-2010
418    ------------------------
419    
420    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
421        THEN.
422    
423    2.  (*ACCEPT) was not working when inside an atomic group.
424    
425    3.  Inside a character class, \B is treated as a literal by default, but
426        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
427        causes the error). The code is unchanged, but I tidied the documentation.
428    
429    4.  Inside a character class, PCRE always treated \R and \X as literals,
430        whereas Perl faults them if its -w option is set. I have changed PCRE so
431        that it faults them when PCRE_EXTRA is set.
432    
433    5.  Added support for \N, which always matches any character other than
434        newline. (It is the same as "." when PCRE_DOTALL is not set.)
435    
436    6.  When compiling pcregrep with newer versions of gcc which may have
437        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
438        declared with attribute warn_unused_result" were given. Just casting the
439        result to (void) does not stop the warnings; a more elaborate fudge is
440        needed. I've used a macro to implement this.
441    
442    7.  Minor change to pcretest.c to avoid a compiler warning.
443    
444    8.  Added four artifical Unicode properties to help with an option to make
445        \s etc use properties (see next item). The new properties are: Xan
446        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
447    
448    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
449        use Unicode properties. (*UCP) at the start of a pattern can be used to set
450        this option. Modified pcretest to add /W to test this facility. Added
451        REG_UCP to make it available via the POSIX interface.
452    
453    10. Added --line-buffered to pcregrep.
454    
455    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
456        studied, and the match started with a letter with a code point greater than
457        127 whose first byte was different to the first byte of the other case of
458        the letter, the other case of this starting letter was not recognized
459        (#976).
460    
461    12. If a pattern that was studied started with a repeated Unicode property
462        test, for example, \p{Nd}+, there was the theoretical possibility of
463        setting up an incorrect bitmap of starting bytes, but fortunately it could
464        not have actually happened in practice until change 8 above was made (it
465        added property types that matched character-matching opcodes).
466    
467    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
468        possible starting bytes for non-anchored patterns.
469    
470    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
471        \R, and also a number of cases that involve Unicode properties, both
472        explicit and implicit when PCRE_UCP is set.
473    
474    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
475        input, it could crash or give wrong results if characters with values
476        greater than 0xc0 were present in the subject string. (Detail: it assumed
477        UTF-8 input when processing these items.)
478    
479    16. Added a lot of (int) casts to avoid compiler warnings in systems where
480        size_t is 64-bit (#991).
481    
482    17. Added a check for running out of memory when PCRE is compiled with
483        --disable-stack-for-recursion (#990).
484    
485    18. If the last data line in a file for pcretest does not have a newline on
486        the end, a newline was missing in the output.
487    
488    19. The default pcre_chartables.c file recognizes only ASCII characters (values
489        less than 128) in its various bitmaps. However, there is a facility for
490        generating tables according to the current locale when PCRE is compiled. It
491        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
492        characters, are recognized by isspace() and therefore were getting set in
493        these tables, and indeed these tables seem to approximate to ISO 8859. This
494        caused a problem in UTF-8 mode when pcre_study() was used to create a list
495        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
496        which of course cannot start UTF-8 characters. I have changed the code so
497        that only real ASCII characters (less than 128) and the correct starting
498        bytes for UTF-8 encodings are set for characters greater than 127 when in
499        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
500        altogether.)
501    
502    20. Added the /T option to pcretest so as to be able to run tests with non-
503        standard character tables, thus making it possible to include the tests
504        used for 19 above in the standard set of tests.
505    
506    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
507        reference to a subpattern the other side of a comment that contains an
508        opening parenthesis caused either an internal compiling error, or a
509        reference to the wrong subpattern.
510    
511    
512    Version 8.02 19-Mar-2010
513    ------------------------
514    
515    1.  The Unicode data tables have been updated to Unicode 5.2.0.
516    
517    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
518        configured.
519    
520    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
521        original author of that file, following a query about its status.
522    
523    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
524        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
525    
526    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
527        quantifier applied to a forward-referencing subroutine call, could compile
528        incorrect code or give the error "internal error: previously-checked
529        referenced subpattern not found".
530    
531    6.  Both MS Visual Studio and Symbian OS have problems with initializing
532        variables to point to external functions. For these systems, therefore,
533        pcre_malloc etc. are now initialized to local functions that call the
534        relevant global functions.
535    
536    7.  There were two entries missing in the vectors called coptable and poptable
537        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
538        I've fixed the data, and added a kludgy way of testing at compile time that
539        the lengths are correct (equal to the number of opcodes).
540    
541    8.  Following on from 7, I added a similar kludge to check the length of the
542        eint vector in pcreposix.c.
543    
544    9.  Error texts for pcre_compile() are held as one long string to avoid too
545        much relocation at load time. To find a text, the string is searched,
546        counting zeros. There was no check for running off the end of the string,
547        which could happen if a new error number was added without updating the
548        string.
549    
550    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
551    
552    11. \K was not working if it appeared in an atomic group or in a group that
553        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
554        \K is "not well defined" if used in an assertion. PCRE now accepts it if
555        the assertion is positive, but not if it is negative.
556    
557    12. Change 11 fortuitously reduced the size of the stack frame used in the
558        "match()" function of pcre_exec.c by one pointer. Forthcoming
559        implementation of support for (*MARK) will need an extra pointer on the
560        stack; I have reserved it now, so that the stack frame size does not
561        decrease.
562    
563    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
564        item in branch that calls a recursion is a subroutine call - as in the
565        second branch in the above example - was incorrectly given the compile-
566        time error "recursive call could loop indefinitely" because pcre_compile()
567        was not correctly checking the subroutine for matching a non-empty string.
568    
569    14. The checks for overrunning compiling workspace could trigger after an
570        overrun had occurred. This is a "should never occur" error, but it can be
571        triggered by pathological patterns such as hundreds of nested parentheses.
572        The checks now trigger 100 bytes before the end of the workspace.
573    
574    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
575    
576    
577    Version 8.01 19-Jan-2010
578    ------------------------
579    
580    1.  If a pattern contained a conditional subpattern with only one branch (in
581        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
582        computed the wrong minimum data length (which is of course zero for such
583        subpatterns). This could cause incorrect "no match" results.
584    
585    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
586        the pattern is reset in the first branch, pcre_compile() failed with
587        "internal error: code overflow at offset...". This happened only when
588        the reset was to the original external option setting. (An optimization
589        abstracts leading options settings into an external setting, which was the
590        cause of this.)
591    
592    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
593        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
594        assertion pattern did not match (meaning that the assertion was true), it
595        was incorrectly treated as false if the SKIP had been reached during the
596        matching. This also applied to assertions used as conditions.
597    
598    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
599        assertion subpattern, including such a pattern used as a condition,
600        unpredictable results occurred, instead of the error return
601        PCRE_ERROR_DFA_UITEM.
602    
603    5.  The C++ GlobalReplace function was not working like Perl for the special
604        situation when an empty string is matched. It now does the fancy magic
605        stuff that is necessary.
606    
607    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
608        removed. (These were left over from very, very early versions of PCRE.)
609    
610    7.  Some cosmetic changes to the code to make life easier when compiling it
611        as part of something else:
612    
613        (a) Change DEBUG to PCRE_DEBUG.
614    
615        (b) In pcre_compile(), rename the member of the "branch_chain" structure
616            called "current" as "current_branch", to prevent a collision with the
617            Linux macro when compiled as a kernel module.
618    
619        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
620            prevent a collision with the Linux macro when compiled as a kernel
621            module.
622    
623    8.  In pcre_compile() there are some checks for integer overflows that used to
624        cast potentially large values to (double). This has been changed to that
625        when building, a check for int64_t is made, and if it is found, it is used
626        instead, thus avoiding the use of floating point arithmetic. (There is no
627        other use of FP in PCRE.) If int64_t is not found, the fallback is to
628        double.
629    
630    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
631        2005 (difference between two addresses compared to an unsigned value).
632    
633    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
634        custom one, because of the following reported problem in Windows:
635    
636          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
637              under Win32.
638          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
639              therefore missing the function definition.
640          - The compiler thus generates a "C" signature for the test function.
641          - The linker fails to find the "C" function.
642          - PCRE fails to configure if asked to do so against libbz2.
643    
644    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
645        messages were output:
646    
647          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
648          rerunning libtoolize, to keep the correct libtool macros in-tree.
649          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
650    
651        I have done both of these things.
652    
653    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
654        most of the time, it *can* run out if it is given a pattern that contains a
655        runaway infinite recursion. I updated the discussion in the pcrestack man
656        page.
657    
658    13. Now that we have gone to the x.xx style of version numbers, the minor
659        version may start with zero. Using 08 or 09 is a bad idea because users
660        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
661        interpreted as invalid octal numbers. I've updated the previous comment in
662        configure.ac, and also added a check that gives an error if 08 or 09 are
663        used.
664    
665    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
666        causing partial matching to fail when the end of the subject matched \W
667        in a UTF-8 pattern where \W was quantified with a minimum of 3.
668    
669    15. There were some discrepancies between the declarations in pcre_internal.h
670        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
671        their definitions. The declarations used "const uschar *" and the
672        definitions used USPTR. Even though USPTR is normally defined as "const
673        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
674        reported that: "This difference in casting confuses some C++ compilers, for
675        example, SunCC recognizes above declarations as different functions and
676        generates broken code for hbpcre." I have changed the declarations to use
677        USPTR.
678    
679    16. GNU libtool is named differently on some systems. The autogen.sh script now
680        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
681        (FreeBSD).
682    
683    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
684        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
685        comment: "Figure out how to create a longlong from a string: strtoll and
686        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
687        instance, but it only takes 2 args instead of 3!"
688    
689    18. A subtle bug concerned with back references has been fixed by a change of
690        specification, with a corresponding code fix. A pattern such as
691        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
692        refers, was giving matches when it shouldn't. For example, xa=xaaa would
693        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
694        same bug. Such groups have to be quantified to be useful, or contained
695        inside another quantified group. (If there's no repetition, the reference
696        can never match.) The problem arises because, having left the group and
697        moved on to the rest of the pattern, a later failure that backtracks into
698        the group uses the captured value from the final iteration of the group
699        rather than the correct earlier one. I have fixed this in PCRE by forcing
700        any group that contains a reference to itself to be an atomic group; that
701        is, there cannot be any backtracking into it once it has completed. This is
702        similar to recursive and subroutine calls.
703    
704    
705    Version 8.00 19-Oct-09
706    ----------------------
707    
708    1.  The table for translating pcre_compile() error codes into POSIX error codes
709        was out-of-date, and there was no check on the pcre_compile() error code
710        being within the table. This could lead to an OK return being given in
711        error.
712    
713    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
714        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
715        in a Windows environment.
716    
717    3.  The pcregrep --count option prints the count for each file even when it is
718        zero, as does GNU grep. However, pcregrep was also printing all files when
719        --files-with-matches was added. Now, when both options are given, it prints
720        counts only for those files that have at least one match. (GNU grep just
721        prints the file name in this circumstance, but including the count seems
722        more useful - otherwise, why use --count?) Also ensured that the
723        combination -clh just lists non-zero counts, with no names.
724    
725    4.  The long form of the pcregrep -F option was incorrectly implemented as
726        --fixed_strings instead of --fixed-strings. This is an incompatible change,
727        but it seems right to fix it, and I didn't think it was worth preserving
728        the old behaviour.
729    
730    5.  The command line items --regex=pattern and --regexp=pattern were not
731        recognized by pcregrep, which required --regex pattern or --regexp pattern
732        (with a space rather than an '='). The man page documented the '=' forms,
733        which are compatible with GNU grep; these now work.
734    
735    6.  No libpcreposix.pc file was created for pkg-config; there was just
736        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
737    
738    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
739        when UCP support is not needed, by modifying the Python script that
740        generates it from Unicode data files. This should not matter if the module
741        is correctly used as a library, but I received one complaint about 50K of
742        unwanted data. My guess is that the person linked everything into his
743        program rather than using a library. Anyway, it does no harm.
744    
745    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
746        was a minimum greater than 1 for a wide character in a possessive
747        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
748        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
749        character. Chaos in the form of incorrect output or a compiling loop could
750        result.
751    
752    9.  The restrictions on what a pattern can contain when partial matching is
753        requested for pcre_exec() have been removed. All patterns can now be
754        partially matched by this function. In addition, if there are at least two
755        slots in the offset vector, the offset of the earliest inspected character
756        for the match and the offset of the end of the subject are set in them when
757        PCRE_ERROR_PARTIAL is returned.
758    
759    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
760        synonymous with PCRE_PARTIAL, for backwards compatibility, and
761        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
762        and may be more useful for multi-segment matching.
763    
764    11. Partial matching with pcre_exec() is now more intuitive. A partial match
765        used to be given if ever the end of the subject was reached; now it is
766        given only if matching could not proceed because another character was
767        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
768        string "Z", which now yields "no match" instead of "partial match". In the
769        case of pcre_dfa_exec(), "no match" is given if every matching path for the
770        final character ended with (*FAIL).
771    
772    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
773        if the pattern had a "must contain" character that was already found in the
774        earlier partial match, unless partial matching was again requested. For
775        example, with the pattern /dog.(body)?/, the "must contain" character is
776        "g". If the first part-match was for the string "dog", restarting with
777        "sbody" failed. This bug has been fixed.
778    
779    13. The string returned by pcre_dfa_exec() after a partial match has been
780        changed so that it starts at the first inspected character rather than the
781        first character of the match. This makes a difference only if the pattern
782        starts with a lookbehind assertion or \b or \B (\K is not supported by
783        pcre_dfa_exec()). It's an incompatible change, but it makes the two
784        matching functions compatible, and I think it's the right thing to do.
785    
786    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
787        so that the demonstration program is easily available in environments where
788        PCRE has not been installed from source.
789    
790    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
791        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
792        library.
793    
794    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
795        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
796        is not the first non-POSIX option to be added. Clearly some people find
797        these options useful.
798    
799    17. If a caller to the POSIX matching function regexec() passes a non-zero
800        value for nmatch with a NULL value for pmatch, the value of
801        nmatch is forced to zero.
802    
803    18. RunGrepTest did not have a test for the availability of the -u option of
804        the diff command, as RunTest does. It now checks in the same way as
805        RunTest, and also checks for the -b option.
806    
807    19. If an odd number of negated classes containing just a single character
808        interposed, within parentheses, between a forward reference to a named
809        subpattern and the definition of the subpattern, compilation crashed with
810        an internal error, complaining that it could not find the referenced
811        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
812        [The bug was that it was starting one character too far in when skipping
813        over the character class, thus treating the ] as data rather than
814        terminating the class. This meant it could skip too much.]
815    
816    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
817        /g option in pcretest when the pattern contains \K, which makes it possible
818        to have an empty string match not at the start, even when the pattern is
819        anchored. Updated pcretest and pcredemo to use this option.
820    
821    21. If the maximum number of capturing subpatterns in a recursion was greater
822        than the maximum at the outer level, the higher number was returned, but
823        with unset values at the outer level. The correct (outer level) value is
824        now given.
825    
826    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
827        PCRE did not set those parentheses (unlike Perl). I have now found a way to
828        make it do so. The string so far is captured, making this feature
829        compatible with Perl.
830    
831    23. The tests have been re-organized, adding tests 11 and 12, to make it
832        possible to check the Perl 5.10 features against Perl 5.10.
833    
834    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
835        pattern matches a fixed length string. PCRE did not allow this; now it
836        does. Neither allows recursion.
837    
838    25. I finally figured out how to implement a request to provide the minimum
839        length of subject string that was needed in order to match a given pattern.
840        (It was back references and recursion that I had previously got hung up
841        on.) This code has now been added to pcre_study(); it finds a lower bound
842        to the length of subject needed. It is not necessarily the greatest lower
843        bound, but using it to avoid searching strings that are too short does give
844        some useful speed-ups. The value is available to calling programs via
845        pcre_fullinfo().
846    
847    26. While implementing 25, I discovered to my embarrassment that pcretest had
848        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
849        study optimizations had never been tested with that matching function.
850        Oops. What is worse, even when it was passed study data, there was a bug in
851        pcre_dfa_exec() that meant it never actually used it. Double oops. There
852        were also very few tests of studied patterns with pcre_dfa_exec().
853    
854    27. If (?| is used to create subpatterns with duplicate numbers, they are now
855        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
856        on the other side of the coin, they are no longer allowed to have different
857        names, because these cannot be distinguished in PCRE, and this has caused
858        confusion. (This is a difference from Perl.)
859    
860    28. When duplicate subpattern names are present (necessarily with different
861        numbers, as required by 27 above), and a test is made by name in a
862        conditional pattern, either for a subpattern having been matched, or for
863        recursion in such a pattern, all the associated numbered subpatterns are
864        tested, and the overall condition is true if the condition is true for any
865        one of them. This is the way Perl works, and is also more like the way
866        testing by number works.
867    
868    
869    Version 7.9 11-Apr-09
870    ---------------------
871    
872    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
873        (pcretest), all targets were linked against these libraries. This included
874        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
875        libraries. This caused unwanted dependencies to be created. This problem
876        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
877        pcretest is linked with readline.
878    
879    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
880        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
881        moved outside it again, because FALSE and TRUE are already defined in AIX,
882        but BOOL is not.
883    
884    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
885        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
886    
887    4.  The pcregrep documentation said spaces were inserted as well as colons (or
888        hyphens) following file names and line numbers when outputting matching
889        lines. This is not true; no spaces are inserted. I have also clarified the
890        wording for the --colour (or --color) option.
891    
892    5.  In pcregrep, when --colour was used with -o, the list of matching strings
893        was not coloured; this is different to GNU grep, so I have changed it to be
894        the same.
895    
896    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
897        each matching line was coloured. Now it goes on to look for further matches
898        of any of the test patterns, which is the same behaviour as GNU grep.
899    
900    7.  A pattern that could match an empty string could cause pcregrep to loop; it
901        doesn't make sense to accept an empty string match in pcregrep, so I have
902        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
903        seems to be how GNU grep behaves.
904    
905    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
906        start or after a newline", because the conditional assertion was not being
907        correctly handled. The rule now is that both the assertion and what follows
908        in the first alternative must satisfy the test.
909    
910    9.  If auto-callout was enabled in a pattern with a conditional group whose
911        condition was an assertion, PCRE could crash during matching, both with
912        pcre_exec() and pcre_dfa_exec().
913    
914    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
915        used for matching.
916    
917    11. Unicode property support in character classes was not working for
918        characters (bytes) greater than 127 when not in UTF-8 mode.
919    
920    12. Added the -M command line option to pcretest.
921    
922    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
923    
924    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
925    
926    16. Added comments and documentation about mis-use of no_arg in the C++
927        wrapper.
928    
929    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
930        from Martin Jerabek that uses macro names for all relevant character and
931        string constants.
932    
933    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
934        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
935        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
936        these, but not everybody uses configure.
937    
938    19. A conditional group that had only one branch was not being correctly
939        recognized as an item that could match an empty string. This meant that an
940        enclosing group might also not be so recognized, causing infinite looping
941        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
942        with the subject "ab", where knowledge that the repeated group can match
943        nothing is needed in order to break the loop.
944    
945    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
946        exec(), but without supplying a callout function, matching went wrong.
947    
948    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
949        leak if the size of the offset vector was greater than 30. When the vector
950        is smaller, the saved offsets during recursion go onto a local stack
951        vector, but for larger vectors malloc() is used. It was failing to free
952        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
953        error, in fact).
954    
955    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
956        heapframe that is used only when UTF-8 support is enabled. This caused no
957        problem, but was untidy.
958    
959    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
960        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
961        included within another project.
962    
963    24. Steven Van Ingelgem's patches to add more options to the CMake support,
964        slightly modified by me:
965    
966          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
967              not building pcregrep.
968    
969          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
970              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
971    
972    25. Forward references, both numeric and by name, in patterns that made use of
973        duplicate group numbers, could behave incorrectly or give incorrect errors,
974        because when scanning forward to find the reference group, PCRE was not
975        taking into account the duplicate group numbers. A pattern such as
976        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
977    
978    26. Changed a few more instances of "const unsigned char *" to USPTR, making
979        the feature of a custom pointer more persuasive (as requested by a user).
980    
981    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
982        pcretest.c, inside #ifndefs, because it seems they are sometimes already
983        pre-defined.
984    
985    28. Added support for (*UTF8) at the start of a pattern.
986    
987    29. Arrange for flags added by the "release type" setting in CMake to be shown
988        in the configuration summary.
989    
990    
991    Version 7.8 05-Sep-08
992    ---------------------
993    
994    1.  Replaced UCP searching code with optimized version as implemented for Ad
995        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
996        stage table and inline lookup instead of a function, giving speed ups of 2
997        to 5 times on some simple patterns that I tested. Permission was given to
998        distribute the MultiStage2.py script that generates the tables (it's not in
999        the tarball, but is in the Subversion repository).
1000    
1001    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1002        scripts.
1003    
1004    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1005        a group with a zero qualifier. The result of the study could be incorrect,
1006        or the function might crash, depending on the pattern.
1007    
1008    4.  Caseless matching was not working for non-ASCII characters in back
1009        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1010        It now works when Unicode Property Support is available.
1011    
1012    5.  In pcretest, an escape such as \x{de} in the data was always generating
1013        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1014        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1015        truncation.
1016    
1017    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1018    
1019    7.  Added two (int) casts to pcregrep when printing the difference of two
1020        pointers, in case they are 64-bit values.
1021    
1022    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1023        test 2 if it fails.
1024    
1025    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1026        and a #define of that name to empty if it is not externally set. This is to
1027        allow users of MSVC to set it if necessary.
1028    
1029    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1030        the convenience functions in the pcre_get.c source file.
1031    
1032    11. An option change at the start of a pattern that had top-level alternatives
1033        could cause overwriting and/or a crash. This command provoked a crash in
1034        some environments:
1035    
1036          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1037    
1038        This potential security problem was recorded as CVE-2008-2371.
1039    
1040    12. For a pattern where the match had to start at the beginning or immediately
1041        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1042        pcre_dfa_exec() could read past the end of the passed subject if there was
1043        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1044        pcretest so that it places the subject at the end of its malloc-ed buffer.
1045    
1046    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1047        exec() might read past the end of the data buffer in UTF-8 mode.
1048    
1049    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1050        the data contained the byte 0x85 as part of a UTF-8 character within its
1051        first line. This applied both to normal and DFA matching.
1052    
1053    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1054        /^[^d]*?$/8 failed to match "abc".
1055    
1056    16. Added a missing copyright notice to pcrecpp_internal.h.
1057    
1058    17. Make it more clear in the documentation that values returned from
1059        pcre_exec() in ovector are byte offsets, not character counts.
1060    
1061    18. Tidied a few places to stop certain compilers from issuing warnings.
1062    
1063    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1064        supplied by Stefan Weber. I made a further small update for 7.8 because
1065        there is a change of source arrangements: the pcre_searchfuncs.c module is
1066        replaced by pcre_ucd.c.
1067    
1068    
1069    Version 7.7 07-May-08
1070    ---------------------
1071    
1072    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1073        a string to a long long, pretend we don't even have a long long." This is
1074        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1075    
1076    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1077        pre-7.6 versions, which defined a global no_arg variable instead of putting
1078        it in the RE class. (See also #8 below.)
1079    
1080    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1081        Lopes.
1082    
1083    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1084    
1085        (1) The include/exclude patterns were being applied to the whole pathnames
1086            of files, instead of just to the final components.
1087    
1088        (2) If there was more than one level of directory, the subdirectories were
1089            skipped unless they satisfied the include/exclude conditions. This is
1090            inconsistent with GNU grep (and could even be seen as contrary to the
1091            pcregrep specification - which I improved to make it absolutely clear).
1092            The action now is always to scan all levels of directory, and just
1093            apply the include/exclude patterns to regular files.
1094    
1095    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1096        --exclude_dir in the tests to avoid scanning .svn directories.
1097    
1098    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1099        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1100        doesn't support NULs in patterns.
1101    
1102    7.  Added some missing "const"s to declarations of static tables in
1103        pcre_compile.c and pcre_dfa_exec.c.
1104    
1105    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1106        caused by fix #2  above. (Subsequently also a second patch to fix the
1107        first patch. And a third patch - this was a messy problem.)
1108    
1109    9.  Applied Craig's patch to remove the use of push_back().
1110    
1111    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1112        matching function regexec().
1113    
1114    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1115        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1116        references. PCRE supports relative numbers with this syntax (I don't think
1117        Oniguruma does).
1118    
1119    12. Previously, a group with a zero repeat such as (...){0} was completely
1120        omitted from the compiled regex. However, this means that if the group
1121        was called as a subroutine from elsewhere in the pattern, things went wrong
1122        (an internal error was given). Such groups are now left in the compiled
1123        pattern, with a new opcode that causes them to be skipped at execution
1124        time.
1125    
1126    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1127        to the way PCRE behaves:
1128    
1129        (a) A lone ] character is dis-allowed (Perl treats it as data).
1130    
1131        (b) A back reference to an unmatched subpattern matches an empty string
1132            (Perl fails the current match path).
1133    
1134        (c) A data ] in a character class must be notated as \] because if the
1135            first data character in a class is ], it defines an empty class. (In
1136            Perl it is not possible to have an empty class.) The empty class []
1137            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1138            The negative empty class [^] matches any one character, independently
1139            of the DOTALL setting.
1140    
1141    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1142        non-existent subpattern following a character class starting with ']' and
1143        containing () gave an internal compiling error instead of "reference to
1144        non-existent subpattern". Fortunately, when the pattern did exist, the
1145        compiled code was correct. (When scanning forwards to check for the
1146        existencd of the subpattern, it was treating the data ']' as terminating
1147        the class, so got the count wrong. When actually compiling, the reference
1148        was subsequently set up correctly.)
1149    
1150    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1151        it was being rejected as not supported by pcre_dfa_exec(), even though
1152        other assertions are supported. I have made pcre_dfa_exec() support
1153        (*FAIL).
1154    
1155    16. The implementation of 13c above involved the invention of a new opcode,
1156        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1157        cannot be changed at match time, I realized I could make a small
1158        improvement to matching performance by compiling OP_ALLANY instead of
1159        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1160        on the OP_ANY path.
1161    
1162    17. Compiling pcretest on Windows with readline support failed without the
1163        following two fixes: (1) Make the unistd.h include conditional on
1164        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1165    
1166    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1167        ncurses library to be included for pcretest when ReadLine support is
1168        requested, but also to allow for it to be overridden. This patch came from
1169        Daniel Bergström.
1170    
1171    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1172        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1173        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1174        spotting this.
1175    
1176    
1177    Version 7.6 28-Jan-08
1178    ---------------------
1179    
1180    1.  A character class containing a very large number of characters with
1181        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1182        overflow.
1183    
1184    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1185        HAVE_LONG_LONG is not defined.
1186    
1187    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1188        bring them up to date and include new features. This patch includes:
1189    
1190        - Fixed PH's badly added libz and libbz2 support.
1191        - Fixed a problem with static linking.
1192        - Added pcredemo. [But later removed - see 7 below.]
1193        - Fixed dftables problem and added an option.
1194        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1195            HAVE_LONG_LONG.
1196        - Added readline support for pcretest.
1197        - Added an listing of the option settings after cmake has run.
1198    
1199    4.  A user submitted a patch to Makefile that makes it easy to create
1200        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1201        Makefile.am that cause it to include this special target, without
1202        affecting anything else. Note that the same mingw target plus all
1203        the other distribution libraries and programs are now supported
1204        when configuring with CMake (see 6 below) instead of with
1205        Configure/Make.
1206    
1207    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1208        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1209        exported in the Windows port". It has not yet been confirmed that the patch
1210        solves the problem, but it does no harm.
1211    
1212    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1213        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1214        with CMake, and also correct the comment about stack recursion.
1215    
1216    7.  Remove the automatic building of pcredemo from the ./configure system and
1217        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1218        of a program that users should build themselves after PCRE is installed, so
1219        building it automatically is not really right. What is more, it gave
1220        trouble in some build environments.
1221    
1222    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1223    
1224    
1225    Version 7.5 10-Jan-08
1226    ---------------------
1227    
1228    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1229        values in parens when parsing an RE using the C++ wrapper."
1230    
1231    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1232        Characters greater than 255 were excluded from the class instead of being
1233        included.
1234    
1235    3.  The same bug as (2) above applied to negated POSIX classes such as
1236        [:^space:].
1237    
1238    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1239        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1240        I have changed it.
1241    
1242    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1243        first named subpattern) and a construct such as (?&a) would reference the
1244        first named subpattern whose name started with "a" (in other words, the
1245        length check was missing). Both these problems are fixed. "Subpattern name
1246        expected" is now given for (?&) (a zero-length name), and this patch also
1247        makes it give the same error for \k'' (previously it complained that that
1248        was a reference to a non-existent subpattern).
1249    
1250    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1251        this is right because (?- can be followed by option settings as well as by
1252        digits. I have, however, made the messages clearer.
1253    
1254    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1255        than the number used in the conditional) now cause a compile-time error.
1256        This is actually not compatible with Perl, which accepts such patterns, but
1257        treats the conditional as always being FALSE (as PCRE used to), but it
1258        seems to me that giving a diagnostic is better.
1259    
1260    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1261        and messages.
1262    
1263    9.  Fix two occurrences of "backslash" in comments that should have been
1264        "backspace".
1265    
1266    10. Remove two redundant lines of code that can never be obeyed (their function
1267        was moved elsewhere).
1268    
1269    11. The program that makes PCRE's Unicode character property table had a bug
1270        which caused it to generate incorrect table entries for sequences of
1271        characters that have the same character type, but are in different scripts.
1272        It amalgamated them into a single range, with the script of the first of
1273        them. In other words, some characters were in the wrong script. There were
1274        thirteen such cases, affecting characters in the following ranges:
1275    
1276          U+002b0 - U+002c1
1277          U+0060c - U+0060d
1278          U+0061e - U+00612
1279          U+0064b - U+0065e
1280          U+0074d - U+0076d
1281          U+01800 - U+01805
1282          U+01d00 - U+01d77
1283          U+01d9b - U+01dbf
1284          U+0200b - U+0200f
1285          U+030fc - U+030fe
1286          U+03260 - U+0327f
1287          U+0fb46 - U+0fbb1
1288          U+10450 - U+1049d
1289    
1290    12. The -o option (show only the matching part of a line) for pcregrep was not
1291        compatible with GNU grep in that, if there was more than one match in a
1292        line, it showed only the first of them. It now behaves in the same way as
1293        GNU grep.
1294    
1295    13. If the -o and -v options were combined for pcregrep, it printed a blank
1296        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1297        does the same. The return code can be used to tell if there were any
1298        non-matching lines.
1299    
1300    14. Added --file-offsets and --line-offsets to pcregrep.
1301    
1302    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1303        infinitely looping recursion. The bug was that positive lookaheads were not
1304        being skipped when checking for a possible empty match (negative lookaheads
1305        and both kinds of lookbehind were skipped).
1306    
1307    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1308        inclusion of <windows.h> to before rather than after the definition of
1309        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1310    
1311    17. Specifying a possessive quantifier with a specific limit for a Unicode
1312        character property caused pcre_compile() to compile bad code, which led at
1313        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1314        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1315        caused the error; without that there was no problem.
1316    
1317    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1318    
1319    19. Added --enable-pcretest-libreadline.
1320    
1321    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1322        RE::GlobalReplace(). As a result, the number of replacements returned was
1323        double what it should be. I removed one of the increments, but Craig sent a
1324        later patch that removed the other one (the right fix) and added unit tests
1325        that check the return values (which was not done before).
1326    
1327    21. Several CMake things:
1328    
1329        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1330            the names libpcre and libpcreposix, not just pcre and pcreposix.
1331    
1332        (2) The above change means that pcretest and pcregrep are now correctly
1333            linked with the newly-built libraries, not previously installed ones.
1334    
1335        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
1336    
1337    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1338        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1339        UTF-8 newline character). The key issue is that the pattern starts .*;
1340        this means that the match must be either at the beginning, or after a
1341        newline. The bug was in the code for advancing after a failed match and
1342        checking that the new position followed a newline. It was not taking
1343        account of UTF-8 characters correctly.
1344    
1345    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1346        character classes. PCRE was not treating the sequence [:...:] as a
1347        character class unless the ... were all letters. Perl, however, seems to
1348        allow any characters between [: and :], though of course it rejects as
1349        unknown any "names" that contain non-letters, because all the known class
1350        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1351        for example, whereas PCRE did not - it did not recognize a POSIX character
1352        class. This seemed a bit dangerous, so the code has been changed to be
1353        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1354        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1355        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1356        Perl does, and where it didn't before.
1357    
1358    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1359        Windows environments %n is disabled by default.
1360    
1361    
1362    Version 7.4 21-Sep-07
1363    ---------------------
1364    
1365    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1366        means that a class such as [\s] counted as "explicit reference to CR or
1367        LF". That isn't really right - the whole point of the change was to try to
1368        help when there was an actual mention of one of the two characters. So now
1369        the change happens only if \r or \n (or a literal CR or LF) character is
1370        encountered.
1371    
1372    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1373        of both had grown to the point where there were only 3 bits left.
1374        Fortunately, there was spare space in the data structure, and so I have
1375        moved the internal flags into a new 16-bit field to free up more option
1376        bits.
1377    
1378    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1379        but did not set the internal JCHANGED flag - either of these is enough to
1380        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1381        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1382        start sets both bits.
1383    
1384    4.  Added options (at build time, compile time, exec time) to change \R from
1385        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1386    
1387    5.  doc/pcresyntax.html was missing from the distribution.
1388    
1389    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1390        compatibility, even though it is no longer used.
1391    
1392    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1393        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1394        windows.h file is present (where different names are used). [This was
1395        reversed later after testing - see 16 below.]
1396    
1397    8.  Changed all #include <config.h> to #include "config.h". There were also
1398        some further <pcre.h> cases that I changed to "pcre.h".
1399    
1400    9.  When pcregrep was used with the --colour option, it missed the line ending
1401        sequence off the lines that it output.
1402    
1403    10. It was pointed out to me that arrays of string pointers cause lots of
1404        relocations when a shared library is dynamically loaded. A technique of
1405        using a single long string with a table of offsets can drastically reduce
1406        these. I have refactored PCRE in four places to do this. The result is
1407        dramatic:
1408    
1409          Originally:                          290
1410          After changing UCP table:            187
1411          After changing error message table:   43
1412          After changing table of "verbs"       36
1413          After changing table of Posix names   22
1414    
1415        Thanks to the folks working on Gregex for glib for this insight.
1416    
1417    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1418        unicode-properties was also set.
1419    
1420    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1421    
1422    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1423        checked only for CRLF.
1424    
1425    14. Added casts to pcretest.c to avoid compiler warnings.
1426    
1427    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1428    
1429    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1430        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1431        entirely. This removes changes made in 7 above.
1432    
1433    17. The CMake files have been updated, and there is now more information about
1434        building with CMake in the NON-UNIX-USE document.
1435    
1436    
1437    Version 7.3 28-Aug-07
1438    ---------------------
1439    
1440     1. In the rejigging of the build system that eventually resulted in 7.1, the
1441        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1442        brackets there is not right, since it causes compilers to look for an
1443        installed pcre.h, not the version that is in the source that is being
1444        compiled (which of course may be different). I have changed it back to:
1445    
1446          #include "pcre.h"
1447    
1448        I have a vague recollection that the change was concerned with compiling in
1449        different directories, but in the new build system, that is taken care of
1450        by the VPATH setting the Makefile.
1451    
1452     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1453        when the subject happened to end in the byte 0x85 (e.g. if the last
1454        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1455        characters but of course it shouldn't be taken as a newline when it is part
1456        of another character. The bug was that, for an unlimited repeat of . in
1457        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1458        characters when looking for a newline.
1459    
1460     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1461    
1462     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1463        in debug output.
1464    
1465     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1466        long printing in the pcrecpp unittest when running under MinGW.
1467    
1468     6. ESC_K was left out of the EBCDIC table.
1469    
1470     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1471        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1472        limit also applies to "virtual nesting" when a pattern is recursive, and in
1473        this case 1000 isn't so big. I have been able to remove this limit at the
1474        expense of backing off one optimization in certain circumstances. Normally,
1475        when pcre_exec() would call its internal match() function recursively and
1476        immediately return the result unconditionally, it uses a "tail recursion"
1477        feature to save stack. However, when a subpattern that can match an empty
1478        string has an unlimited repetition quantifier, it no longer makes this
1479        optimization. That gives it a stack frame in which to save the data for
1480        checking that an empty string has been matched. Previously this was taken
1481        from the 1000-entry workspace that had been reserved. So now there is no
1482        explicit limit, but more stack is used.
1483    
1484     8. Applied Daniel's patches to solve problems with the import/export magic
1485        syntax that is required for Windows, and which was going wrong for the
1486        pcreposix and pcrecpp parts of the library. These were overlooked when this
1487        problem was solved for the main library.
1488    
1489     9. There were some crude static tests to avoid integer overflow when computing
1490        the size of patterns that contain repeated groups with explicit upper
1491        limits. As the maximum quantifier is 65535, the maximum group length was
1492        set at 30,000 so that the product of these two numbers did not overflow a
1493        32-bit integer. However, it turns out that people want to use groups that
1494        are longer than 30,000 bytes (though not repeat them that many times).
1495        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1496        made it possible to implement the integer overflow checks in a much more
1497        dynamic way, which I have now done. The artificial limitation on group
1498        length has been removed - we now have only the limit on the total length of
1499        the compiled pattern, which depends on the LINK_SIZE setting.
1500    
1501    10. Fixed a bug in the documentation for get/copy named substring when
1502        duplicate names are permitted. If none of the named substrings are set, the
1503        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1504        empty string.
1505    
1506    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1507        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1508        because the ] is interpreted as the first data character and the
1509        terminating ] is not found. PCRE has been made compatible with Perl in this
1510        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1511        cause memory overwriting.
1512    
1513    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1514        string has been matched (to stop an infinite loop). It was not recognizing
1515        a conditional subpattern that could match an empty string if that
1516        subpattern was within another subpattern. For example, it looped when
1517        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1518        condition was not nested. This bug has been fixed.
1519    
1520    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1521        past the start of the subject in the presence of bytes with the top bit
1522        set, for example "\x8aBCD".
1523    
1524    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1525        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1526    
1527    14. Optimized (?!) to (*FAIL).
1528    
1529    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1530        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1531        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1532        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1533        does: it's just the validity check that is more restrictive.
1534    
1535    16. Inserted checks for integer overflows during escape sequence (backslash)
1536        processing, and also fixed erroneous offset values for syntax errors during
1537        backslash processing.
1538    
1539    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1540        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1541    
1542    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1543        caused an overrun.
1544    
1545    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1546        something other than just ASCII characters) inside a group that had an
1547        unlimited repeat caused a loop at compile time (while checking to see
1548        whether the group could match an empty string).
1549    
1550    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1551        [\P{Any}] did so. (Error in the code for printing property names.)
1552    
1553    21. An orphan \E inside a character class could cause a crash.
1554    
1555    22. A repeated capturing bracket such as (A)? could cause a wild memory
1556        reference during compilation.
1557    
1558    23. There are several functions in pcre_compile() that scan along a compiled
1559        expression for various reasons (e.g. to see if it's fixed length for look
1560        behind). There were bugs in these functions when a repeated \p or \P was
1561        present in the pattern. These operators have additional parameters compared
1562        with \d, etc, and these were not being taken into account when moving along
1563        the compiled data. Specifically:
1564    
1565        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1566            length.
1567    
1568        (b) An item such as \pL+ within a repeated group could cause crashes or
1569            loops.
1570    
1571        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1572            "reference to non-existent subpattern" error.
1573    
1574        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1575    
1576    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1577        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1578    
1579    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1580    
1581    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1582        character were causing crashes (broken optimization).
1583    
1584    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1585        \p or \P) caused a compile-time loop.
1586    
1587    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1588        break. For example, the unstudied pattern [\r\n]A does not match the string
1589        "\r\nA" because change 7.0/46 below moves the current point on by two
1590        characters after failing to match at the start. However, the pattern \nA
1591        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1592        the same is true. There doesn't seem any very clean way out of this, but
1593        what I have chosen to do makes the common cases work: PCRE now takes note
1594        of whether there can be an explicit match for \r or \n anywhere in the
1595        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1596        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1597        pattern has explicit CR or LF references.
1598    
1599    29. Added (*CR) etc for changing newline setting at start of pattern.
1600    
1601    
1602    Version 7.2 19-Jun-07
1603    ---------------------
1604    
1605     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1606        which is apparently normally available under Windows.
1607    
1608     2. Re-jig the pcregrep tests with different newline settings in an attempt
1609        to make them independent of the local environment's newline setting.
1610    
1611     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1612    
1613     4. Some of the "internals" tests were previously cut out when the link size
1614        was not 2, because the output contained actual offsets. The recent new
1615        "Z" feature of pcretest means that these can be cut out, making the tests
1616        usable with all link sizes.
1617    
1618     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1619        stack recursion. This gives a massive performance boost under BSD, but just
1620        a small improvement under Linux. However, it saves one field in the frame
1621        in all cases.
1622    
1623     6. Added more features from the forthcoming Perl 5.10:
1624    
1625        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1626            recursion call. It refers to the nth most recently opened parentheses.
1627    
1628        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1629            to be opened parentheses.
1630    
1631        (c) Conditions that refer to capturing parentheses can be specified
1632            relatively, for example, (?(-2)... or (?(+3)...
1633    
1634        (d) \K resets the start of the current match so that everything before
1635            is not part of it.
1636    
1637        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1638    
1639        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1640            reference syntax.
1641    
1642        (g) (?| introduces a group in which the numbering of parentheses in each
1643            alternative starts with the same number.
1644    
1645        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1646    
1647     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1648        PCRE_INFO_JCHANGED.
1649    
1650     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1651        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1652        for detecting groups that can match an empty string.
1653    
1654     9. A pattern with a very large number of alternatives (more than several
1655        hundred) was running out of internal workspace during the pre-compile
1656        phase, where pcre_compile() figures out how much memory will be needed. A
1657        bit of new cunning has reduced the workspace needed for groups with
1658        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1659        workspace instead of running out of the 4096 that are available.
1660    
1661    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1662    
1663    11. Applied patch from Google to remove an optimization that didn't quite work.
1664        The report of the bug said:
1665    
1666          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1667          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1668          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1669    
1670    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1671        it matched the wrong number of bytes.
1672    
1673    
1674    Version 7.1 24-Apr-07
1675    ---------------------
1676    
1677     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1678        that is more "standard", making use of automake and other Autotools. There
1679        is some re-arrangement of the files and adjustment of comments consequent
1680        on this.
1681    
1682     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1683        for recursive directory scanning broke on some systems because the files
1684        are not scanned in any specific order and on different systems the order
1685        was different. A call to "sort" has been inserted into RunGrepTest for the
1686        approprate test as a short-term fix. In the longer term there may be an
1687        alternative.
1688    
1689     3. I had an email from Eric Raymond about problems translating some of PCRE's
1690        man pages to HTML (despite the fact that I distribute HTML pages, some
1691        people do their own conversions for various reasons). The problems
1692        concerned the use of low-level troff macros .br and .in. I have therefore
1693        removed all such uses from the man pages (some were redundant, some could
1694        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1695        HTML has been updated to handle .nf/.fi and to complain if it encounters
1696        .br or .in.
1697    
1698     4. Updated comments in configure.ac that get placed in config.h.in and also
1699        arranged for config.h to be included in the distribution, with the name
1700        config.h.generic, for the benefit of those who have to compile without
1701        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1702    
1703     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1704        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1705        makevp.bat for the current PCRE, using the additional files
1706        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1707    
1708     6. A Windows user reported a minor discrepancy with test 2, which turned out
1709        to be caused by a trailing space on an input line that had got lost in his
1710        copy. The trailing space was an accident, so I've just removed it.
1711    
1712     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1713        that is needed.
1714    
1715     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1716        as "const" (a) because they are and (b) because it helps the PHP
1717        maintainers who have recently made a script to detect big data structures
1718        in the php code that should be moved to the .rodata section. I remembered
1719        to update Builducptable as well, so it won't revert if ucptable.h is ever
1720        re-created.
1721    
1722     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1723        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1724        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1725        support is not required. This saves 1.5-2K of code, which is important in
1726        some applications.
1727    
1728        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1729        so as not to refer to the tables, even though these functions will never be
1730        called when UTF-8 support is disabled. Otherwise there are problems with a
1731        shared library.
1732    
1733    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1734    
1735        (a) It was defining its arguments as char * instead of void *.
1736    
1737        (b) It was assuming that all moves were upwards in memory; this was true
1738            a long time ago when I wrote it, but is no longer the case.
1739    
1740        The emulated memove() is provided for those environments that have neither
1741        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1742        is clearly not the case, as these two bugs were recently reported.
1743    
1744    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1745        and Detrail to create the HTML documentation, the .txt form of the man
1746        pages, and it removes trailing spaces from listed files. It also creates
1747        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1748        case, it wraps all the #defines with #ifndefs. This script should be run
1749        before "make dist".
1750    
1751    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1752        with Unicode property support.
1753    
1754        (a) For a maximizing quantifier, if the two different cases of the
1755            character were of different lengths in their UTF-8 codings (there are
1756            some cases like this - I found 11), and the matching function had to
1757            back up over a mixture of the two cases, it incorrectly assumed they
1758            were both the same length.
1759    
1760        (b) When PCRE was configured to use the heap rather than the stack for
1761            recursion during matching, it was not correctly preserving the data for
1762            the other case of a UTF-8 character when checking ahead for a match
1763            while processing a minimizing repeat. If the check also involved
1764            matching a wide character, but failed, corruption could cause an
1765            erroneous result when trying to check for a repeat of the original
1766            character.
1767    
1768    13. Some tidying changes to the testing mechanism:
1769    
1770        (a) The RunTest script now detects the internal link size and whether there
1771            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1772            values substituted by "configure". (The RunGrepTest script already did
1773            this for UTF-8.) The configure.ac script no longer substitutes the
1774            relevant variables.
1775    
1776        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1777            with length and offset values. This means that the output is different
1778            for different internal link sizes. Test 2 is skipped for link sizes
1779            other than 2 because of this, bypassing the problem. Unfortunately,
1780            there was also a test in test 3 (the locale tests) that used /B and
1781            failed for link sizes other than 2. Rather than cut the whole test out,
1782            I have added a new /Z option to pcretest that replaces the length and
1783            offset values with spaces. This is now used to make test 3 independent
1784            of link size. (Test 2 will be tidied up later.)
1785    
1786    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1787        segmentation fault instead of returning the appropriate error message.
1788    
1789    15. In multiline mode when the newline sequence was set to "any", the pattern
1790        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1791        This doesn't seem right; it now treats the CRLF combination as the line
1792        ending, and so does not match in that case. It's only a pattern such as ^$
1793        that would hit this one: something like ^ABC$ would have failed after \r
1794        and then tried again after \r\n.
1795    
1796    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1797        in an attempt to make files that differ only in their line terminators
1798        compare equal. This works on Linux.
1799    
1800    17. Under certain error circumstances pcregrep might try to free random memory
1801        as it exited. This is now fixed, thanks to valgrind.
1802    
1803    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1804        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1805        was because its rules for how to advance for /g after matching an empty
1806        string at the end of a line did not allow for this case. They now check for
1807        it specially.
1808    
1809    20. pcretest is supposed to handle patterns and data of any length, by
1810        extending its buffers when necessary. It was getting this wrong when the
1811        buffer for a data line had to be extended.
1812    
1813    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1814        CRLF as a newline sequence.
1815    
1816    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1817        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1818        I have nevertheless tidied it up.
1819    
1820    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1821    
1822    24. Added a man page for pcre-config.
1823    
1824    
1825    Version 7.0 19-Dec-06
1826    ---------------------
1827    
1828     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1829        moving to gcc 4.1.1.
1830    
1831     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1832        sys/time.h, which is documented as needed for this function. It doesn't
1833        seem to matter on Linux, but it showed up on some releases of OS X.
1834    
1835     3. It seems that there are systems where bytes whose values are greater than
1836        127 match isprint() in the "C" locale. The "C" locale should be the
1837        default when a C program starts up. In most systems, only ASCII printing
1838        characters match isprint(). This difference caused the output from pcretest
1839        to vary, making some of the tests fail. I have changed pcretest so that:
1840    
1841        (a) When it is outputting text in the compiled version of a pattern, bytes
1842            other than 32-126 are always shown as hex escapes.
1843    
1844        (b) When it is outputting text that is a matched part of a subject string,
1845            it does the same, unless a different locale has been set for the match
1846            (using the /L modifier). In this case, it uses isprint() to decide.
1847    
1848     4. Fixed a major bug that caused incorrect computation of the amount of memory
1849        required for a compiled pattern when options that changed within the
1850        pattern affected the logic of the preliminary scan that determines the
1851        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1852        that the computed length was too small. The symptoms of this bug were
1853        either the PCRE error "internal error: code overflow" from pcre_compile(),
1854        or a glibc crash with a message such as "pcretest: free(): invalid next
1855        size (fast)". Examples of patterns that provoked this bug (shown in
1856        pcretest format) are:
1857    
1858          /(?-x: )/x
1859          /(?x)(?-x: \s*#\s*)/
1860          /((?i)[\x{c0}])/8
1861          /(?i:[\x{c0}])/8
1862    
1863        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1864        is now done differently.
1865    
1866     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1867        wrapper classes; (b) implement a new function in the C++ scanner that is
1868        more efficient than the old way of doing things because it avoids levels of
1869        recursion in the regex matching; (c) add a paragraph to the documentation
1870        for the FullMatch() function.
1871    
1872     6. The escape sequence \n was being treated as whatever was defined as
1873        "newline". Not only was this contrary to the documentation, which states
1874        that \n is character 10 (hex 0A), but it also went horribly wrong when
1875        "newline" was defined as CRLF. This has been fixed.
1876    
1877     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1878        was being set to -1 for the "end of line" case (supposedly a value that no
1879        character can have). Though this value is never used (the check for end of
1880        line is "zero bytes in current character"), it caused compiler complaints.
1881        I've changed it to 0xffffffff.
1882    
1883     8. In pcre_version.c, the version string was being built by a sequence of
1884        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1885        string (as it is for production releases) called a macro with an empty
1886        argument. The C standard says the result of this is undefined. The gcc
1887        compiler treats it as an empty string (which was what was wanted) but it is
1888        reported that Visual C gives an error. The source has been hacked around to
1889        avoid this problem.
1890    
1891     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1892        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1893        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1894        of them did).
1895    
1896    10. Originally, pcretest opened its input and output without "b"; then I was
1897        told that "b" was needed in some environments, so it was added for release
1898        5.0 to both the input and output. (It makes no difference on Unix-like
1899        systems.) Later I was told that it is wrong for the input on Windows. I've
1900        now abstracted the modes into two macros, to make it easier to fiddle with
1901        them, and removed "b" from the input mode under Windows.
1902    
1903    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1904    
1905    12. Added -help and --help to pcretest as an official way of being reminded
1906        of the options.
1907    
1908    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1909        and pcrecpp.cc because they annoy compilers at high warning levels.
1910    
1911    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1912    
1913    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1914        scripts are not C programs :-) and which was not noticed because it works
1915        on Linux.
1916    
1917    16. pcretest is supposed to handle any length of pattern and data line (as one
1918        line or as a continued sequence of lines) by extending its input buffer if
1919        necessary. This feature was broken for very long pattern lines, leading to
1920        a string of junk being passed to pcre_compile() if the pattern was longer
1921        than about 50K.
1922    
1923    17. I have done a major re-factoring of the way pcre_compile() computes the
1924        amount of memory needed for a compiled pattern. Previously, there was code
1925        that made a preliminary scan of the pattern in order to do this. That was
1926        OK when PCRE was new, but as the facilities have expanded, it has become
1927        harder and harder to keep it in step with the real compile phase, and there
1928        have been a number of bugs (see for example, 4 above). I have now found a
1929        cunning way of running the real compile function in a "fake" mode that
1930        enables it to compute how much memory it would need, while actually only
1931        ever using a few hundred bytes of working memory and without too many
1932        tests of the mode. This should make future maintenance and development
1933        easier. A side effect of this work is that the limit of 200 on the nesting
1934        depth of parentheses has been removed (though this was never a serious
1935        limitation, I suspect). However, there is a downside: pcre_compile() now
1936        runs more slowly than before (30% or more, depending on the pattern). I
1937        hope this isn't a big issue. There is no effect on runtime performance.
1938    
1939    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1940        newline (only possible for the last line of a file) and it was a
1941        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1942    
1943    19. Added additional timing features to pcretest. (1) The -tm option now times
1944        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1945        separate command line item, by a number that specifies the number of
1946        repeats to use when timing. The default is 50000; this gives better
1947        precision, but takes uncomfortably long for very large patterns.
1948    
1949    20. Extended pcre_study() to be more clever in cases where a branch of a
1950        subpattern has no definite first character. For example, (a*|b*)[cd] would
1951        previously give no result from pcre_study(). Now it recognizes that the
1952        first character must be a, b, c, or d.
1953    
1954    21. There was an incorrect error "recursive call could loop indefinitely" if
1955        a subpattern (or the entire pattern) that was being tested for matching an
1956        empty string contained only one non-empty item after a nested subpattern.
1957        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1958        incorrectly, because the \d was being skipped in the check.
1959    
1960    22. The pcretest program now has a new pattern option /B and a command line
1961        option -b, which is equivalent to adding /B to every pattern. This causes
1962        it to show the compiled bytecode, without the additional information that
1963        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1964        is the same as /B/I).
1965    
1966    23. A new optimization is now able automatically to treat some sequences such
1967        as a*b as a*+b. More specifically, if something simple (such as a character
1968        or a simple class like \d) has an unlimited quantifier, and is followed by
1969        something that cannot possibly match the quantified thing, the quantifier
1970        is automatically "possessified".
1971    
1972    24. A recursive reference to a subpattern whose number was greater than 39
1973        went wrong under certain circumstances in UTF-8 mode. This bug could also
1974        have affected the operation of pcre_study().
1975    
1976    25. Realized that a little bit of performance could be had by replacing
1977        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1978    
1979    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1980    
1981    27. Possessive quantifiers such as a++ were previously implemented by turning
1982        them into atomic groups such as ($>a+). Now they have their own opcodes,
1983        which improves performance. This includes the automatically created ones
1984        from 23 above.
1985    
1986    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1987        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1988        the first matched character to be a colon. This applied both to named and
1989        numbered groups.
1990    
1991    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1992    
1993    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1994        building PCRE on Windows easier, so I have included it in the distribution.
1995    
1996    31. There is now a check in pcretest against a ridiculously large number being
1997        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1998        loop, the loop is abandoned.
1999    
2000    32. Forward references to subpatterns in conditions such as (?(2)...) where
2001        subpattern 2 is defined later cause pcre_compile() to search forwards in
2002        the pattern for the relevant set of parentheses. This search went wrong
2003        when there were unescaped parentheses in a character class, parentheses
2004        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2005    
2006    33. "Subroutine" calls and backreferences were previously restricted to
2007        referencing subpatterns earlier in the regex. This restriction has now
2008        been removed.
2009    
2010    34. Added a number of extra features that are going to be in Perl 5.10. On the
2011        whole, these are just syntactic alternatives for features that PCRE had
2012        previously implemented using the Python syntax or my own invention. The
2013        other formats are all retained for compatibility.
2014    
2015        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2016            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2017            also .NET compatible.
2018    
2019        (b) A recursion or subroutine call to a named group can now be defined as
2020            (?&name) as well as (?P>name).
2021    
2022        (c) A backreference to a named group can now be defined as \k<name> or
2023            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2024            5.10, are also .NET compatible.
2025    
2026        (d) A conditional reference to a named group can now use the syntax
2027            (?(<name>) or (?('name') as well as (?(name).
2028    
2029        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2030            groups (named and numbered) that are never evaluated inline, but can be
2031            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2032            is always false. There may be only one alternative in such a group.
2033    
2034        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2035            as the simple (?(R). The condition is true only if the most recent
2036            recursion is that of the given number or name. It does not search out
2037            through the entire recursion stack.
2038    
2039        (g) The escape \gN or \g{N} has been added, where N is a positive or
2040            negative number, specifying an absolute or relative reference.
2041    
2042    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2043        some "unreachable code" warnings.
2044    
2045    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2046        things, this adds five new scripts.
2047    
2048    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2049        There were also incompatibilities regarding the handling of \Q..\E inside
2050        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2051        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2052    
2053    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2054        matches an empty string, and forcibly breaks the loop. There were bugs in
2055        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2056        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2057        separate and independent bugs (that affected different cases) have been
2058        fixed.
2059    
2060    39. Refactored the code to abolish the use of different opcodes for small
2061        capturing bracket numbers. This is a tidy that I avoided doing when I
2062        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2063        The new approach is not only tidier, it makes it possible to reduce the
2064        memory needed to fix the previous bug (38).
2065    
2066    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2067        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2068        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2069        mode.
2070    
2071    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2072        report.
2073    
2074    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2075        copy construction and assignment in the C++ wrapper.
2076    
2077    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2078        couple of bugs that could have given wrong results in the "--newline=crlf"
2079        case.
2080    
2081    44. Added a number of casts and did some reorganization of signed/unsigned int
2082        variables following suggestions from Dair Grant. Also renamed the variable
2083        "this" as "item" because it is a C++ keyword.
2084    
2085    45. Arranged for dftables to add
2086    
2087          #include "pcre_internal.h"
2088    
2089        to pcre_chartables.c because without it, gcc 4.x may remove the array
2090        definition from the final binary if PCRE is built into a static library and
2091        dead code stripping is activated.
2092    
2093    46. For an unanchored pattern, if a match attempt fails at the start of a
2094        newline sequence, and the newline setting is CRLF or ANY, and the next two
2095        characters are CRLF, advance by two characters instead of one.
2096    
2097    
2098    Version 6.7 04-Jul-06
2099    ---------------------
2100    
2101     1. In order to handle tests when input lines are enormously long, pcretest has
2102        been re-factored so that it automatically extends its buffers when
2103        necessary. The code is crude, but this _is_ just a test program. The
2104        default size has been increased from 32K to 50K.
2105    
2106     2. The code in pcre_study() was using the value of the re argument before
2107        testing it for NULL. (Of course, in any sensible call of the function, it
2108        won't be NULL.)
2109    
2110     3. The memmove() emulation function in pcre_internal.h, which is used on
2111        systems that lack both memmove() and bcopy() - that is, hardly ever -
2112        was missing a "static" storage class specifier.
2113    
2114     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2115        containing an extended class (one that cannot be represented by a bitmap
2116        because it contains high-valued characters or Unicode property items, e.g.
2117        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2118        pattern, but PCRE should not loop if you do not (it no longer does).
2119        [Detail: two cases were found: (a) a repeated subpattern containing an
2120        extended class; (b) a recursive reference to a subpattern that followed a
2121        previous extended class. It wasn't skipping over the extended class
2122        correctly when UTF-8 mode was not set.]
2123    
2124     5. A negated single-character class was not being recognized as fixed-length
2125        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2126        compile error "lookbehind assertion is not fixed length".
2127    
2128     6. The RunPerlTest auxiliary script was showing an unexpected difference
2129        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2130        write a Perl script that can interpret lines of an input file either as
2131        byte characters or as UTF-8, which is what "perltest" was being required to
2132        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2133        can't do is switch easily at run time between having the "use utf8;" pragma
2134        or not. In the end, I fudged it by using the RunPerlTest script to insert
2135        "use utf8;" explicitly for the UTF-8 tests.
2136    
2137     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2138        the end of the subject string, contrary to the documentation and to what
2139        Perl does. This was true of both matching functions. Now it matches only at
2140        the start of the subject and immediately after *internal* newlines.
2141    
2142     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2143        a pointer to an int instead of a pointer to an unsigned long int. This
2144        caused problems on 64-bit systems.
2145    
2146     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2147        instance of the 'standard' template library not being so standard".
2148    
2149    10. There was no check on the number of named subpatterns nor the maximum
2150        length of a subpattern name. The product of these values is used to compute
2151        the size of the memory block for a compiled pattern. By supplying a very
2152        long subpattern name and a large number of named subpatterns, the size
2153        computation could be caused to overflow. This is now prevented by limiting
2154        the length of names to 32 characters, and the number of named subpatterns
2155        to 10,000.
2156    
2157    11. Subpatterns that are repeated with specific counts have to be replicated in
2158        the compiled pattern. The size of memory for this was computed from the
2159        length of the subpattern and the repeat count. The latter is limited to
2160        65535, but there was no limit on the former, meaning that integer overflow
2161        could in principle occur. The compiled length of a repeated subpattern is
2162        now limited to 30,000 bytes in order to prevent this.
2163    
2164    12. Added the optional facility to have named substrings with the same name.
2165    
2166    13. Added the ability to use a named substring as a condition, using the
2167        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2168        are numbers (not recommended). Forward references are permitted.
2169    
2170    14. Added forward references in named backreferences (if you see what I mean).
2171    
2172    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2173        pattern could run off the end of the subject. For example, the pattern
2174        "(?s)(.{1,5})"8 did this with the subject "ab".
2175    
2176    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2177        PCRE_CASELESS was set when matching characters that were quantified with ?
2178        or *.
2179    
2180    17. A character class other than a single negated character that had a minimum
2181        but no maximum quantifier - for example [ab]{6,} - was not handled
2182        correctly by pce_dfa_exec(). It would match only one character.
2183    
2184    18. A valid (though odd) pattern that looked like a POSIX character
2185        class but used an invalid character after [ (for example [[,abc,]]) caused
2186        pcre_compile() to give the error "Failed: internal error: code overflow" or
2187        in some cases to crash with a glibc free() error. This could even happen if
2188        the pattern terminated after [[ but there just happened to be a sequence of
2189        letters, a binary zero, and a closing ] in the memory that followed.
2190    
2191    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2192        over the years. Originally (before any Unicode support), just the bottom 8
2193        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2194        output from "man perlunicode" includes this:
2195    
2196          The regular expression compiler produces polymorphic opcodes.  That
2197          is, the pattern adapts to the data and automatically switches to
2198          the Unicode character scheme when presented with Unicode data--or
2199          instead uses a traditional byte scheme when presented with byte
2200          data.
2201    
2202        Sadly, a wide octal escape does not cause a switch, and in a string with
2203        no other multibyte characters, these octal escapes are treated as before.
2204        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2205        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2206        Unicode string.
2207    
2208        I have not perpetrated such confusion in PCRE. Up till now, it took just
2209        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2210        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2211        translate to the appropriate multibyte character.
2212    
2213    29. Applied some refactoring to reduce the number of warnings from Microsoft
2214        and Borland compilers. This has included removing the fudge introduced
2215        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2216        a warning about an unused variable.
2217    
2218    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2219        characters since release 4.0, because Perl (from release 5.004) does not.
2220        [Or at least, is documented not to: some releases seem to be in conflict
2221        with the documentation.] However, when a pattern was studied with
2222        pcre_study() and all its branches started with \s, PCRE still included VT
2223        as a possible starting character. Of course, this did no harm; it just
2224        caused an unnecessary match attempt.
2225    
2226    22. Removed a now-redundant internal flag bit that recorded the fact that case
2227        dependency changed within the pattern. This was once needed for "required
2228        byte" processing, but is no longer used. This recovers a now-scarce options
2229        bit. Also moved the least significant internal flag bit to the most-
2230        significant bit of the word, which was not previously used (hangover from
2231        the days when it was an int rather than a uint) to free up another bit for
2232        the future.
2233    
2234    23. Added support for CRLF line endings as well as CR and LF. As well as the
2235        default being selectable at build time, it can now be changed at runtime
2236        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2237        specify that it is scanning data with non-default line endings.
2238    
2239    24. Changed the definition of CXXLINK to make it agree with the definition of
2240        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2241    
2242    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2243        recursions. This makes a big different to stack usage for some patterns.
2244    
2245    26. If a subpattern containing a named recursion or subroutine reference such
2246        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2247        the space required for the compiled pattern went wrong and gave too small a
2248        value. Depending on the environment, this could lead to "Failed: internal
2249        error: code overflow at offset 49" or "glibc detected double free or
2250        corruption" errors.
2251    
2252    27. Applied patches from Google (a) to support the new newline modes and (b) to
2253        advance over multibyte UTF-8 characters in GlobalReplace.
2254    
2255    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2256        difference for some implementation of PCRE in some Windows version.
2257    
2258    29. Added some extra testing facilities to pcretest:
2259    
2260        \q<number>   in a data line sets the "match limit" value
2261        \Q<number>   in a data line sets the "match recursion limt" value
2262        -S <number>  sets the stack size, where <number> is in megabytes
2263    
2264        The -S option isn't available for Windows.
2265    
2266    
2267    Version 6.6 06-Feb-06
2268    ---------------------
2269    
2270     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2271        in pcreposix.h. I have copied the definition from pcre.h.
2272    
2273     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2274        because pcre.h is no longer a built file.
2275    
2276     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2277        not normally included in the compiled code.
2278    
2279    
2280    Version 6.5 01-Feb-06
2281    ---------------------
2282    
2283     1. When using the partial match feature with pcre_dfa_exec(), it was not
2284        anchoring the second and subsequent partial matches at the new starting
2285        point. This could lead to incorrect results. For example, with the pattern
2286        /1234/, partially matching against "123" and then "a4" gave a match.
2287    
2288     2. Changes to pcregrep:
2289    
2290        (a) All non-match returns from pcre_exec() were being treated as failures
2291            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2292            error message is output. Some extra information is given for the
2293            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2294            probably the only errors that are likely to be caused by users (by
2295            specifying a regex that has nested indefinite repeats, for instance).
2296            If there are more than 20 of these errors, pcregrep is abandoned.
2297    
2298        (b) A binary zero was treated as data while matching, but terminated the
2299            output line if it was written out. This has been fixed: binary zeroes
2300            are now no different to any other data bytes.
2301    
2302        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2303            used to set a locale for matching. The --locale=xxxx long option has
2304            been added (no short equivalent) to specify a locale explicitly on the
2305            pcregrep command, overriding the environment variables.
2306    
2307        (d) When -B was used with -n, some line numbers in the output were one less
2308            than they should have been.
2309    
2310        (e) Added the -o (--only-matching) option.
2311    
2312        (f) If -A or -C was used with -c (count only), some lines of context were
2313            accidentally printed for the final match.
2314    
2315        (g) Added the -H (--with-filename) option.
2316    
2317        (h) The combination of options -rh failed to suppress file names for files
2318            that were found from directory arguments.
2319    
2320        (i) Added the -D (--devices) and -d (--directories) options.
2321    
2322        (j) Added the -F (--fixed-strings) option.
2323    
2324        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2325    
2326        (l) Added the --colo(u)r option.
2327    
2328        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2329            is not present by default.
2330    
2331     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2332        items such as (?R) or (?1), when the recursion could match a number of
2333        alternatives. If it matched one of the alternatives, but subsequently,
2334        outside the recursion, there was a failure, the code tried to back up into
2335        the recursion. However, because of the way PCRE is implemented, this is not
2336        possible, and the result was an incorrect result from the match.
2337    
2338        In order to prevent this happening, the specification of recursion has
2339        been changed so that all such subpatterns are automatically treated as
2340        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2341    
2342     4. I had overlooked the fact that, in some locales, there are characters for
2343        which isalpha() is true but neither isupper() nor islower() are true. In
2344        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2345        and ordfeminine) are like this. This affected the treatment of \w and \W
2346        when they appeared in character classes, but not when they appeared outside
2347        a character class. The bit map for "word" characters is now created
2348        separately from the results of isalnum() instead of just taking it from the
2349        upper, lower, and digit maps. (Plus the underscore character, of course.)
2350    
2351     5. The above bug also affected the handling of POSIX character classes such as
2352        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2353        permanent tables. Instead, the bit maps for such a class were previously
2354        created as the appropriate unions of the upper, lower, and digit bitmaps.
2355        Now they are created by subtraction from the [[:word:]] class, which has
2356        its own bitmap.
2357    
2358     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2359        It is created by subtracting the vertical space characters (\x09, \x0a,
2360        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2361        subtraction was done in the overall bitmap for a character class, meaning
2362        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2363        be recognized. This bug has been fixed.
2364    
2365     7. Patches from the folks at Google:
2366    
2367          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2368          real life, but is still worth protecting against".
2369    
2370          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2371          regular expressions".
2372    
2373          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2374          have it.
2375    
2376          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2377          "configure" and the latter not, in order to fix a problem somebody had
2378          with compiling the Arg class on HP-UX.
2379    
2380          (e) Improve the error-handling of the C++ wrapper a little bit.
2381    
2382          (f) New tests for checking recursion limiting.
2383    
2384     8. The pcre_memmove() function, which is used only if the environment does not
2385        have a standard memmove() function (and is therefore rarely compiled),
2386        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2387        returning a result (though PCRE never actually uses the result).
2388    
2389     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2390        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2391        returned instead of calling malloc() with an overflowing number that would
2392        most likely cause subsequent chaos.
2393    
2394    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2395    
2396    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2397        with this option is matched, the nmatch and pmatch options of regexec() are
2398        ignored.
2399    
2400    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2401        provided in case anyone wants to the the POSIX interface with UTF-8
2402        strings.
2403    
2404    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2405        C++ linking (needed for some HP-UX environments).
2406    
2407    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2408        (unused parameter) and in the pcre_printint() function (omitted "default"
2409        switch label when the default is to do nothing).
2410    
2411    15. Added some code to make it possible, when PCRE is compiled as a C++
2412        library, to replace subject pointers for pcre_exec() with a smart pointer
2413        class, thus making it possible to process discontinuous strings.
2414    
2415    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2416        much the same function. They were added by different people who were trying
2417        to make PCRE easy to compile on non-Unix systems. It has been suggested
2418        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2419        for compiling on Windows systems. I have therefore replaced it with
2420        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2421        defaults to "extern" for C or "extern C" for C++, which works fine on
2422        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2423        SCOPE with something explicit in config.h. In addition:
2424    
2425        (a) pcreposix.h still had just "extern" instead of either of these macros;
2426            I have replaced it with PCRE_DATA_SCOPE.
2427    
2428        (b) Functions such as _pcre_xclass(), which are internal to the library,
2429            but external in the C sense, all had PCRE_EXPORT in their definitions.
2430            This is apparently wrong for the Windows case, so I have removed it.
2431            (It makes no difference on Unix-like systems.)
2432    
2433    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2434        of recursive calls to match(). This is different to MATCH_LIMIT because
2435        that limits the total number of calls to match(), not all of which increase
2436        the depth of recursion. Limiting the recursion depth limits the amount of
2437        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2438        when PCRE is compiled, and changed at run time. A patch from Google adds
2439        this functionality to the C++ interface.
2440    
2441    18. Changes to the handling of Unicode character properties:
2442    
2443        (a) Updated the table to Unicode 4.1.0.
2444    
2445        (b) Recognize characters that are not in the table as "Cn" (undefined).
2446    
2447        (c) I revised the way the table is implemented to a much improved format
2448            which includes recognition of ranges. It now supports the ranges that
2449            are defined in UnicodeData.txt, and it also amalgamates other
2450            characters into ranges. This has reduced the number of entries in the
2451            table from around 16,000 to around 3,000, thus reducing its size
2452            considerably. I realized I did not need to use a tree structure after
2453            all - a binary chop search is just as efficient. Having reduced the
2454            number of entries, I extended their size from 6 bytes to 8 bytes to
2455            allow for more data.
2456    
2457        (d) Added support for Unicode script names via properties such as \p{Han}.
2458    
2459    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2460        matching that character.
2461    
2462    20. When matching a repeated Unicode property with a minimum greater than zero,
2463        (for example \pL{2,}), PCRE could look past the end of the subject if it
2464        reached it while seeking the minimum number of characters. This could
2465        happen only if some of the characters were more than one byte long, because
2466        there is a check for at least the minimum number of bytes.
2467    
2468    21. Refactored the implementation of \p and \P so as to be more general, to
2469        allow for more different types of property in future. This has changed the
2470        compiled form incompatibly. Anybody with saved compiled patterns that use
2471        \p or \P will have to recompile them.
2472    
2473    22. Added "Any" and "L&" to the supported property types.
2474    
2475    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2476        but give a compile time error if the value is greater than 0xff.
2477    
2478    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2479        accidentally not being installed or uninstalled.
2480    
2481    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2482        made were to insert the current release number. This seemed silly, because
2483        it made things harder for people building PCRE on systems that don't run
2484        "configure". I have turned pcre.h into a distributed file, no longer built
2485        by "configure", with the version identification directly included. There is
2486        no longer a pcre.h.in file.
2487    
2488        However, this change necessitated a change to the pcre-config script as
2489        well. It is built from pcre-config.in, and one of the substitutions was the
2490        release number. I have updated configure.ac so that ./configure now finds
2491        the release number by grepping pcre.h.
2492    
2493    26. Added the ability to run the tests under valgrind.
2494    
2495    
2496    Version 6.4 05-Sep-05
2497    ---------------------
2498    
2499     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2500        "--" to be printed when multiple files were scanned, even when none of the
2501        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2502        consider it to be a bug, and have restored the previous behaviour.
2503    
2504     2. A couple of code tidies to get rid of compiler warnings.
2505    
2506     3. The pcretest program used to cheat by referring to symbols in the library
2507        whose names begin with _pcre_. These are internal symbols that are not
2508        really supposed to be visible externally, and in some environments it is
2509        possible to suppress them. The cheating is now confined to including
2510        certain files from the library's source, which is a bit cleaner.
2511    
2512     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2513        file's purpose clearer.
2514    
2515     5. Reorganized pcre_ucp_findchar().
2516    
2517    
2518    Version 6.3 15-Aug-05
2519    ---------------------
2520    
2521     1. The file libpcre.pc.in did not have general read permission in the tarball.
2522    
2523     2. There were some problems when building without C++ support:
2524    
2525        (a) If C++ support was not built, "make install" and "make test" still
2526            tried to test it.
2527    
2528        (b) There were problems when the value of CXX was explicitly set. Some
2529            changes have been made to try to fix these, and ...
2530    
2531        (c) --disable-cpp can now be used to explicitly disable C++ support.
2532    
2533        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2534            backslash in a target when C++ was disabled. This confuses some
2535            versions of "make", apparently. Using an intermediate variable solves
2536            this. (Same for CPP_LOBJ.)
2537    
2538     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2539        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2540        necessary on certain architectures.
2541    
2542     4. Added a setting of -export-symbols-regex to the link command to remove
2543        those symbols that are exported in the C sense, but actually are local
2544        within the library, and not documented. Their names all begin with
2545        "_pcre_". This is not a perfect job, because (a) we have to except some
2546        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2547        available (and never for static libraries). I have made a note to try to
2548        find a way round (a) in the future.
2549    
2550    
2551    Version 6.2 01-Aug-05
2552    ---------------------
2553    
2554     1. There was no test for integer overflow of quantifier values. A construction
2555        such as {1111111111111111} would give undefined results. What is worse, if
2556        a minimum quantifier for a parenthesized subpattern overflowed and became
2557        negative, the calculation of the memory size went wrong. This could have
2558        led to memory overwriting.
2559    
2560     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2561    
2562     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2563        operating environments where this matters.
2564    
2565     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2566        PCRE options from within the C++ wrapper.
2567    
2568     5. Named capturing subpatterns were not being correctly counted when a pattern
2569        was compiled. This caused two problems: (a) If there were more than 100
2570        such subpatterns, the calculation of the memory needed for the whole
2571        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2572        back references of the form \12, where the number was greater than 9, were
2573        not recognized as back references, even though there were sufficient
2574        previous subpatterns.
2575    
2576     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2577        versions of gcc, e.g. 2.95.4.
2578    
2579    
2580    Version 6.1 21-Jun-05
2581    ---------------------
2582    
2583     1. There was one reference to the variable "posix" in pcretest.c that was not
2584        surrounded by "#if !defined NOPOSIX".
2585    
2586     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2587        the cross-check on the old pcre_info() function, for the benefit of the
2588        cut-down version of PCRE that is currently imported into Exim.
2589    
2590     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2591        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2592        patterns that start (?i) but I don't think that matters. The use of (?i) is
2593        just an example; this all applies to the other options as well.
2594    
2595     4. Since libtool seems to echo the compile commands it is issuing, the output
2596        from "make" can be reduced a bit by putting "@" in front of each libtool
2597        compile command.
2598    
2599     5. Patch from the folks at Google for configure.in to be a bit more thorough
2600        in checking for a suitable C++ installation before trying to compile the
2601        C++ stuff. This should fix a reported problem when a compiler was present,
2602        but no suitable headers.
2603    
2604     6. The man pages all had just "PCRE" as their title. I have changed them to
2605        be the relevant file name. I have also arranged that these names are
2606        retained in the file doc/pcre.txt, which is a concatenation in text format
2607        of all the man pages except the little individual ones for each function.
2608    
2609     7. The NON-UNIX-USE file had not been updated for the different set of source
2610        files that come with release 6. I also added a few comments about the C++
2611        wrapper.
2612    
2613    
2614    Version 6.0 07-Jun-05
2615    ---------------------
2616    
2617     1. Some minor internal re-organization to help with my DFA experiments.
2618    
2619     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2620        didn't matter for the library itself when fully configured, but did matter
2621        when compiling without UCP support, or within Exim, where the ucp files are
2622        not imported.
2623    
2624     3. Refactoring of the library code to split up the various functions into
2625        different source modules. The addition of the new DFA matching code (see
2626        below) to a single monolithic source would have made it really too
2627        unwieldy, quite apart from causing all the code to be include in a
2628        statically linked application, when only some functions are used. This is
2629        relevant even without the DFA addition now that patterns can be compiled in
2630        one application and matched in another.
2631    
2632        The downside of splitting up is that there have to be some external
2633        functions and data tables that are used internally in different modules of
2634        the library but which are not part of the API. These have all had their
2635        names changed to start with "_pcre_" so that they are unlikely to clash
2636        with other external names.
2637    
2638     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2639        a different (DFA) algorithm. Although it is slower than the original
2640        function, it does have some advantages for certain types of matching
2641        problem.
2642    
2643     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2644        including restarting after a partial match.
2645    
2646     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2647        defined when compiling for Windows was sent to me. I have put it into the
2648        code, though I have no means of testing or verifying it.
2649    
2650     7. Added the pcre_refcount() auxiliary function.
2651    
2652     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2653        match before or at the first newline in the subject string. In pcretest,
2654        the /f option on a pattern can be used to set this.
2655    
2656     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2657        would behave wrongly. This has been present in PCRE since release 4.0.
2658    
2659    10. A number of changes to the pcregrep command:
2660    
2661        (a) Refactored how -x works; insert ^(...)$ instead of setting
2662            PCRE_ANCHORED and checking the length, in preparation for adding
2663            something similar for -w.
2664    
2665        (b) Added the -w (match as a word) option.
2666    
2667        (c) Refactored the way lines are read and buffered so as to have more
2668            than one at a time available.
2669    
2670        (d) Implemented a pcregrep test script.
2671    
2672        (e) Added the -M (multiline match) option. This allows patterns to match
2673            over several lines of the subject. The buffering ensures that at least
2674            8K, or the rest of the document (whichever is the shorter) is available
2675            for matching (and similarly the previous 8K for lookbehind assertions).
2676    
2677        (f) Changed the --help output so that it now says
2678    
2679              -w, --word-regex(p)
2680    
2681            instead of two lines, one with "regex" and the other with "regexp"
2682            because that confused at least one person since the short forms are the
2683            same. (This required a bit of code, as the output is generated
2684            automatically from a table. It wasn't just a text change.)
2685    
2686        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2687            option but starts with a hyphen. Could be a pattern or a path name
2688            starting with a hyphen, for instance.
2689    
2690        (h) "-" can be given as a file name to represent stdin.
2691    
2692        (i) When file names are being printed, "(standard input)" is used for
2693            the standard input, for compatibility with GNU grep. Previously
2694            "<stdin>" was used.
2695    
2696        (j) The option --label=xxx can be used to supply a name to be used for
2697            stdin when file names are being printed. There is no short form.
2698    
2699        (k) Re-factored the options decoding logic because we are going to add
2700            two more options that take data. Such options can now be given in four
2701            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2702    
2703        (l) Added the -A, -B, and -C options for requesting that lines of context
2704            around matches be printed.
2705    
2706        (m) Added the -L option to print the names of files that do not contain
2707            any matching lines, that is, the complement of -l.
2708    
2709        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2710            continue to scan other files.
2711    
2712        (o) The -s option was incorrectly implemented. For compatibility with other
2713            greps, it now suppresses the error message for a non-existent or non-
2714            accessible file (but not the return code). There is a new option called
2715            -q that suppresses the output of matching lines, which was what -s was
2716            previously doing.
2717    
2718        (p) Added --include and --exclude options to specify files for inclusion
2719            and exclusion when recursing.
2720    
2721    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2722        Hopefully, it now does.
2723    
2724    12. Missing cast in pcre_study().
2725    
2726    13. Added an "uninstall" target to the makefile.
2727    
2728    14. Replaced "extern" in the function prototypes in Makefile.in with
2729        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2730        world, but is set differently for Windows.
2731    
2732    15. Added a second compiling function called pcre_compile2(). The only
2733        difference is that it has an extra argument, which is a pointer to an
2734        integer error code. When there is a compile-time failure, this is set
2735        non-zero, in addition to the error test pointer being set to point to an
2736        error message. The new argument may be NULL if no error number is required
2737        (but then you may as well call pcre_compile(), which is now just a
2738        wrapper). This facility is provided because some applications need a
2739        numeric error indication, but it has also enabled me to tidy up the way
2740        compile-time errors are handled in the POSIX wrapper.
2741    
2742    16. Added VPATH=.libs to the makefile; this should help when building with one
2743        prefix path and installing with another. (Or so I'm told by someone who
2744        knows more about this stuff than I do.)
2745    
2746    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2747        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2748        match everything, including newlines. This is not POSIX-compatible, but
2749        somebody wanted the feature. From pcretest it can be activated by using
2750        both the P and the s flags.
2751    
2752    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2753    
2754    19. libpcre.pc was being incorrectly installed as executable.
2755    
2756    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2757        it now also looks for '\r' so that it will work unmodified on Windows.
2758    
2759    21. Added Google's contributed C++ wrapper to the distribution.
2760    
2761    22. Added some untidy missing memory free() calls in pcretest, to keep
2762        Electric Fence happy when testing.
2763    
2764    
2765    
2766    Version 5.0 13-Sep-04
2767    ---------------------
2768    
2769     1. Internal change: literal characters are no longer packed up into items
2770        containing multiple characters in a single byte-string. Each character
2771        is now matched using a separate opcode. However, there may be more than one
2772        byte in the character in UTF-8 mode.
2773    
2774     2. The pcre_callout_block structure has two new fields: pattern_position and
2775        next_item_length. These contain the offset in the pattern to the next match
2776        item, and its length, respectively.
2777    
2778     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2779        insertion of callouts before each pattern item. Added the /C option to
2780        pcretest to make use of this.
2781    
2782     4. On the advice of a Windows user, the lines
2783    
2784          #if defined(_WIN32) || defined(WIN32)
2785          _setmode( _fileno( stdout ), 0x8000 );
2786          #endif  /* defined(_WIN32) || defined(WIN32) */
2787    
2788        have been added to the source of pcretest. This apparently does useful
2789        magic in relation to line terminators.
2790    
2791     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2792        for the benefit of those environments where the "b" makes a difference.
2793    
2794     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2795        to know about it. I have put a hack into configure.in that adds in code
2796        to set GCC=yes if CC=icc. This seems to end up at a point in the
2797        generated configure script that is early enough to affect the setting of
2798        compiler options, which is what is needed, but I have no means of testing
2799        whether it really works. (The user who reported this had patched the
2800        generated configure script, which of course I cannot do.)
2801    
2802        LATER: After change 22 below (new libtool files), the configure script
2803        seems to know about icc (and also ecc). Therefore, I have commented out
2804        this hack in configure.in.
2805    
2806     7. Added support for pkg-config (2 patches were sent in).
2807    
2808     8. Negated POSIX character classes that used a combination of internal tables
2809        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2810        [[:^ascii]]. Typically, they would match almost any characters. The other
2811        POSIX classes were not broken in this way.
2812    
2813     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2814        to find the match, as PCRE was deluded into thinking that the match had to
2815        start at the start point or following a newline. The same bug applied to
2816        patterns with negative forward assertions or any backward assertions
2817        preceding ".*" at the start, unless the pattern required a fixed first
2818        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2819    
2820    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2821        starting at the last subject character, bytes beyond the end of the subject
2822        string were read.
2823    
2824    11. Renamed the variable "class" as "classbits" to make life easier for C++
2825        users. (Previously there was a macro definition, but it apparently wasn't
2826        enough.)
2827    
2828    12. Added the new field "tables" to the extra data so that tables can be passed
2829        in at exec time, or the internal tables can be re-selected. This allows
2830        a compiled regex to be saved and re-used at a later time by a different
2831        program that might have everything at different addresses.
2832    
2833    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2834        -R library as well as a -L library.
2835    
2836    14. The debugging options of pcretest (-d on the command line or D on a
2837        pattern) showed incorrect output for anything following an extended class
2838        that contained multibyte characters and which was followed by a quantifier.
2839    
2840    15. Added optional support for general category Unicode character properties
2841        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2842        support. It adds about 90K to the size of the library. The meanings of the
2843        inbuilt class escapes such as \d and \s have NOT been changed.
2844    
2845    16. Updated pcredemo.c to include calls to free() to release the memory for the
2846        compiled pattern.
2847    
2848    17. The generated file chartables.c was being created in the source directory
2849        instead of in the building directory. This caused the build to fail if the
2850        source directory was different from the building directory, and was
2851        read-only.
2852    
2853    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2854        file. No doubt somebody will tell me if they don't make sense... Also added
2855        Dan Mooney's comments about building on OpenVMS.
2856    
2857    19. Added support for partial matching via the PCRE_PARTIAL option for
2858        pcre_exec() and the \P data escape in pcretest.
2859    
2860    20. Extended pcretest with 3 new pattern features:
2861    
2862        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2863              write the compiled pattern to the file whose name is "rest-of-line".
2864              This is a straight binary dump of the data, with the saved pointer to
2865              the character tables forced to be NULL. The study data, if any, is
2866              written too. After writing, pcretest reads a new pattern.
2867    
2868        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2869              compiled pattern from the given file. There must not be any
2870              occurrences of "<" in the file name (pretty unlikely); if there are,
2871              pcretest will instead treat the initial "<" as a pattern delimiter.
2872              After reading in the pattern, pcretest goes on to read data lines as
2873              usual.
2874    
2875        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2876              and 16-bit fields in a compiled pattern, to simulate a pattern that
2877              was compiled on a host of opposite endianness.
2878    
2879    21. The pcre-exec() function can now cope with patterns that were compiled on
2880        hosts of opposite endianness, with this restriction:
2881    
2882          As for any compiled expression that is saved and used later, the tables
2883          pointer field cannot be preserved; the extra_data field in the arguments
2884          to pcre_exec() should be used to pass in a tables address if a value
2885          other than the default internal tables were used at compile time.
2886    
2887    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2888        now diagnosed as an error. Previously, most of the time, a negative number
2889        would have been treated as zero, but if in addition "ovector" was passed as
2890        NULL, a crash could occur.
2891    
2892    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2893        new versions from the libtool 1.5 distribution (the last one is a copy of
2894        a file called libtool.m4). This seems to have fixed the need to patch
2895        "configure" to support Darwin 1.3 (which I used to do). However, I still
2896        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2897        workstation).
2898    
2899    24. Changed the PCRE licence to be the more standard "BSD" licence.
2900    
2901    
2902    Version 4.5 01-Dec-03
2903    ---------------------
2904    
2905     1. There has been some re-arrangement of the code for the match() function so
2906        that it can be compiled in a version that does not call itself recursively.
2907        Instead, it keeps those local variables that need separate instances for
2908        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2909        needs to "recurse". Keeping track of where control must go is done by means
2910        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2911        hide most of the details from the main code, and operates only if
2912        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2913        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2914        operating.
2915    
2916        To make it easier for callers to provide specially tailored get/free
2917        functions for this usage, two new functions, pcre_stack_malloc, and
2918        pcre_stack_free, are used. They are always called in strict stacking order,
2919        and the size of block requested is always the same.
2920    
2921        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2922        PCRE has been compiled to use the stack or the heap for recursion. The
2923        -C option of pcretest uses this to show which version is compiled.
2924    
2925        A new data escape \S, is added to pcretest; it causes the amounts of store
2926        obtained and freed by both kinds of malloc/free at match time to be added
2927        to the output.
2928    
2929     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2930        what's available on my current Linux desktop machine.
2931    
2932     3. When matching a UTF-8 string, the test for a valid string at the start has
2933        been extended. If start_offset is not zero, PCRE now checks that it points
2934        to a byte that is the start of a UTF-8 character. If not, it returns
2935        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2936        this is necessary because there may be backward assertions in the pattern.
2937        When matching the same subject several times, it may save resources to use
2938        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2939    
2940     4. The code for checking the validity of UTF-8 strings has been tightened so
2941        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2942        containing "overlong sequences".
2943    
2944     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2945        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2946        should have been "&", but it just so happened that all the cases this let
2947        through by mistake were picked up later in the function.
2948    
2949     6. I had used a variable called "isblank" - this is a C99 function, causing
2950        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2951    
2952     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2953        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2954        so the version is shown; (c) stop "make test" echoing "./RunTest".
2955    
2956     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2957    
2958     9. The prototype for memmove() for systems that don't have it was using
2959        size_t, but the inclusion of the header that defines size_t was later. I've
2960        moved the #includes for the C headers earlier to avoid this.
2961    
2962    10. Added some adjustments to the code to make it easier to compiler on certain
2963        special systems:
2964    
2965          (a) Some "const" qualifiers were missing.
2966          (b) Added the macro EXPORT before all exported functions; by default this
2967              is defined to be empty.
2968          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2969              that it reads its output file name as an argument instead of writing
2970              to the standard output and assuming this can be redirected.
2971    
2972    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2973        class containing characters with values greater than 255, PCRE compilation
2974        went into a loop.
2975    
2976    12. A recursive reference to a subpattern that was within another subpattern
2977        that had a minimum quantifier of zero caused PCRE to crash. For example,
2978        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2979        recursion. If the recursively-called subpattern itself had a zero repeat,
2980        that was OK.
2981    
2982    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2983        buffer into which it was copied (for escape processing) was still set at
2984        1024, so long lines caused crashes.
2985    
2986    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2987        "internal error: code overflow...". This applied to any character class
2988        that was followed by a possessive quantifier.
2989    
2990    15. Modified the Makefile to add libpcre.la as a prerequisite for
2991        libpcreposix.la because I was told this is needed for a parallel build to
2992        work.
2993    
2994    16. If a pattern that contained .* following optional items at the start was
2995        studied, the wrong optimizing data was generated, leading to matching
2996        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2997        matching string must start with a or b or c. The correct conclusion for
2998        this pattern is that a match can start with any character.
2999    
3000    
3001    Version 4.4 13-Aug-03
3002    ---------------------
3003    
3004     1. In UTF-8 mode, a character class containing characters with values between
3005        127 and 255 was not handled correctly if the compiled pattern was studied.
3006        In fixing this, I have also improved the studying algorithm for such
3007        classes (slightly).
3008    
3009     2. Three internal functions had redundant arguments passed to them. Removal
3010        might give a very teeny performance improvement.
3011    
3012     3. Documentation bug: the value of the capture_top field in a callout is *one
3013        more than* the number of the hightest numbered captured substring.
3014    
3015     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3016        in incorrectly linking with a previously installed version. They now link
3017        explicitly with libpcre.la.
3018    
3019     5. configure.in no longer needs to recognize Cygwin specially.
3020    
3021     6. A problem in pcre.in for Windows platforms is fixed.
3022    
3023     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3024        pcretest, it used to include the size of the study block as part of its
3025        output. Unfortunately, the structure contains a field that has a different
3026        size on different hardware architectures. This meant that the tests that
3027        showed this size failed. As the block is currently always of a fixed size,
3028        this information isn't actually particularly useful in pcretest output, so
3029        I have just removed it.
3030    
3031     8. Three pre-processor statements accidentally did not start in column 1.
3032        Sadly, there are *still* compilers around that complain, even though
3033        standard C has not required this for well over a decade. Sigh.
3034    
3035     9. In pcretest, the code for checking callouts passed small integers in the
3036        callout_data field, which is a void * field. However, some picky compilers
3037        complained about the casts involved for this on 64-bit systems. Now
3038        pcretest passes the address of the small integer instead, which should get
3039        rid of the warnings.
3040    
3041    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3042        both compile and run time, and gives an error if an invalid UTF-8 sequence
3043        is found. There is a option for disabling this check in cases where the
3044        string is known to be correct and/or the maximum performance is wanted.
3045    
3046    11. In response to a bug report, I changed one line in Makefile.in from
3047    
3048            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3049        to
3050            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3051    
3052        to look similar to other lines, but I have no way of telling whether this
3053        is the right thing to do, as I do not use Windows. No doubt I'll get told
3054        if it's wrong...
3055    
3056    
3057    Version 4.3 21-May-03
3058    ---------------------
3059    
3060    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3061       Makefile.
3062    
3063    2. Some refactoring to improve the quality of the code:
3064    
3065       (i)   The utf8_table... variables are now declared "const".
3066    
3067       (ii)  The code for \cx, which used the "case flipping" table to upper case
3068             lower case letters, now just substracts 32. This is ASCII-specific,
3069             but the whole concept of \cx is ASCII-specific, so it seems
3070             reasonable.
3071    
3072       (iii) PCRE was using its character types table to recognize decimal and
3073             hexadecimal digits in the pattern. This is silly, because it handles
3074             only 0-9, a-f, and A-F, but the character types table is locale-
3075             specific, which means strange things might happen. A private
3076             table is now used for this - though it costs 256 bytes, a table is
3077             much faster than multiple explicit tests. Of course, the standard
3078             character types table is still used for matching digits in subject
3079             strings against \d.
3080    
3081       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3082             ending in _t are). So I've renamed it as ESC_tee.
3083    
3084    3. The first argument for regexec() in the POSIX wrapper should have been
3085       defined as "const".
3086    
3087    4. Changed pcretest to use malloc() for its buffers so that they can be
3088       Electric Fenced for debugging.
3089    
3090    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3091       to read one or more bytes before the start of the subject string. Often this
3092       had no effect on PCRE's behaviour, but in some circumstances it could
3093       provoke a segmentation fault.
3094    
3095    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3096       to try to read one or more bytes before the start of the subject string.
3097    
3098    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3099       UTF-8 support could misbehave in various ways if the subject string
3100       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3101       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3102       back over UTF-8 characters.)
3103    
3104    
3105    Version 4.2 14-Apr-03
3106    ---------------------
3107    
3108    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3109    
3110    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3111         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3112         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3113         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3114         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3115           and BUILD_EXEEXT
3116         Note: automatic setting of the BUILD variables is not yet working
3117         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3118           compile-time but not at link-time
3119         [LINK]: use for linking executables only
3120         make different versions for Windows and non-Windows
3121         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3122           libraries
3123         [LINK_FOR_BUILD]: new variable
3124         [OBJEXT]: use throughout
3125         [EXEEXT]: use throughout
3126         <winshared>: new target
3127         <wininstall>: new target
3128         <dftables.o>: use native compiler
3129         <dftables>: use native linker
3130         <install>: handle Windows platform correctly
3131         <clean>: ditto
3132         <check>: ditto
3133         copy DLL to top builddir before testing
3134    
3135       As part of these changes, -no-undefined was removed again. This was reported
3136       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3137       in any case.
3138    
3139    3. Some tidies to get rid of compiler warnings:
3140    
3141       . In the match_data structure, match_limit was an unsigned long int, whereas
3142         match_call_count was an int. I've made them both unsigned long ints.
3143    
3144       . In pcretest the fact that a const uschar * doesn't automatically cast to
3145         a void * provoked a warning.
3146    
3147       . Turning on some more compiler warnings threw up some "shadow" variables
3148         and a few more missing casts.
3149    
3150    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3151       option, a class that contained a single character with a value between 128
3152       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3153    
3154    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3155       option, a class that contained several characters, but with at least one
3156       whose value was between 128 and 255 caused PCRE to crash.
3157    
3158    
3159    Version 4.1 12-Mar-03
3160    ---------------------
3161    
3162    1. Compiling with gcc -pedantic found a couple of places where casts were
3163    needed, and a string in dftables.c that was longer than standard compilers are
3164    required to support.
3165    
3166    2. Compiling with Sun's compiler found a few more places where the code could
3167    be tidied up in order to avoid warnings.
3168    
3169    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3170    first of these names is deprecated in the latest Autoconf in favour of the name
3171    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3172    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3173    analogy I have changed it to CFLAGS_FOR_BUILD.
3174    
3175    4. Added -no-undefined to the linking command in the Makefile, because this is
3176    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3177    linking step for the pcreposix library.
3178    
3179    5. PCRE was failing to diagnose the case of two named groups with the same
3180    name.
3181    
3182    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3183    literal character that is needed in the subject for a match, and scans along to
3184    ensure that it is present before embarking on the full matching process. This
3185    saves time in cases of nested unlimited repeats that are never going to match.
3186    Problem: the scan can take a lot of time if the subject is very long (e.g.
3187    megabytes), thus penalizing straightforward matches. It is now done only if the
3188    amount of subject to be scanned is less than 1000 bytes.
3189    
3190    7. A lesser problem with the same optimization is that it was recording the
3191    first character of an anchored pattern as "needed", thus provoking a search
3192    right along the subject, even when the first match of the pattern was going to
3193    fail. The "needed" character is now not set for anchored patterns, unless it
3194    follows something in the pattern that is of non-fixed length. Thus, it still
3195    fulfils its original purpose of finding quick non-matches in cases of nested
3196    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3197    
3198    
3199    Version 4.0 17-Feb-03
3200    ---------------------
3201    
3202    1. If a comment in an extended regex that started immediately after a meta-item
3203    extended to the end of string, PCRE compiled incorrect data. This could lead to
3204    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3205    
3206    2. Moved to autoconf 2.53 and libtool 1.4.2.
3207    
3208    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3209    the special perltest8 script is no longer needed - all the tests can be run
3210    from a single perltest script.
3211    
3212    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3213    by \s. It has now been removed in PCRE. This means it isn't recognized as
3214    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3215    class [:space:] *does* include VT, thereby creating a mess.
3216    
3217    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3218    space and tab.
3219    
3220    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3221    its new features into the main test script, reducing the number of scripts.
3222    
3223    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3224    were backward compatible, and made the (?i) apply to the whole pattern, as if
3225    /i were given. Now it behaves more logically, and applies the option setting
3226    only to what follows. PCRE has been changed to follow suit. However, if it
3227    finds options settings right at the start of the pattern, it extracts them into
3228    the global options, as before. Thus, they show up in the info data.
3229    
3230    8. Added support for the \Q...\E escape sequence. Characters in between are
3231    treated as literals. This is slightly different from Perl in that $ and @ are
3232    also handled as literals inside the quotes. In Perl, they will cause variable
3233    interpolation. Note the following examples:
3234    
3235        Pattern            PCRE matches      Perl matches
3236    
3237        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3238        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3239        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3240    
3241    For compatibility with Perl, \Q...\E sequences are recognized inside character
3242    classes as well as outside them.
3243    
3244    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3245    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3246    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3247    signed/unsigned warnings.
3248    
3249    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3250    option for pcretest, so I've replaced it by a simple function that does just
3251    that job.
3252    
3253    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3254    "pcregrep -".
3255    
3256    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3257    Java package. This provides some syntactic sugar for simple cases of what my
3258    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3259    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3260    item, you can use this simplified notation. Note that only makes sense with
3261    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3262    greediness, whatever the setting of the PCRE_UNGREEDY option.
3263    
3264    13. A change of greediness default within a pattern was not taking effect at
3265    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3266    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3267    was abstracted outside.
3268    
3269    14. PCRE now supports the \G assertion. It is true when the current matching
3270    position is at the start point of the match. This differs from \A when the
3271    starting offset is non-zero. Used with the /g option of pcretest (or similar
3272    code), it works in the same way as it does for Perl's /g option. If all
3273    alternatives of a regex begin with \G, the expression is anchored to the start
3274    match position, and the "anchored" flag is set in the compiled expression.
3275    
3276    15. Some bugs concerning the handling of certain option changes within patterns
3277    have been fixed. These applied to options other than (?ims). For example,
3278    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3279    been the other way round. Some of this was related to change 7 above.
3280    
3281    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3282    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3283    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3284    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3285    
3286    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3287    mode. Unlike ".", it always matches newline, whatever the setting of
3288    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3289    assertions. Perl allows it, but it doesn't (in general) work because it can't
3290    calculate the length of the lookbehind. At least, that's the case for Perl
3291    5.8.0 - I've been told they are going to document that it doesn't work in
3292    future.
3293    
3294    18. Added an error diagnosis for escapes that PCRE does not support: these are
3295    \L, \l, \N, \P, \p, \U, \u, and \X.
3296    
3297    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
3298    reading past the end of the pattern in cases such as /[abcd/.
3299    
3300    20. PCRE was getting more memory than necessary for patterns with classes that
3301    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
3302    
3303    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
3304    compiling PCRE for use with Virtual Pascal.
3305    
3306    22. Small fix to the Makefile to make it work properly if the build is done
3307    outside the source tree.
3308    
3309    23. Added a new extension: a condition to go with recursion. If a conditional
3310    subpattern starts with (?(R) the "true" branch is used if recursion has
3311    happened, whereas the "false" branch is used only at the top level.
3312    
3313    24. When there was a very long string of literal characters (over 255 bytes
3314    without UTF support, over 250 bytes with UTF support), the computation of how
3315    much memory was required could be incorrect, leading to segfaults or other
3316    strange effects.
3317    
3318    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
3319    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
3320    there was a subsequent back reference to those brackets. This meant that, for
3321    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
3322    possible to check for precisely this case. All we can do is abandon the
3323    optimization if .* occurs inside capturing brackets when there are any back
3324    references whatsoever. (See below for a better fix that came later.)
3325    
3326    26. The handling of the optimization for finding the first character of a
3327    non-anchored pattern, and for finding a character that is required later in the
3328    match were failing in some cases. This didn't break the matching; it just
3329    failed to optimize when it could. The way this is done has been re-implemented.
3330    
3331    27. Fixed typo in error message for invalid (?R item (it said "(?p").
3332    
3333    28. Added a new feature that provides some of the functionality that Perl
3334    provides with (?{...}). The facility is termed a "callout". The way it is done
3335    in PCRE is for the caller to provide an optional function, by setting
3336    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
3337    global variable. By default it is unset, which disables all calling out. To get
3338    the function called, the regex must include (?C) at appropriate points. This
3339    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
3340    This provides a means of identifying different callout points. When PCRE
3341    reaches such a point in the regex, if pcre_callout has been set, the external
3342    function is called. It is provided with data in a structure called
3343    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
3344    matching continues; if it returns a non-zero value, the match at the current
3345    point fails. However, backtracking will occur if possible. [This was changed
3346    later and other features added - see item 49 below.]
3347    
3348    29. pcretest is upgraded to test the callout functionality. It provides a
3349    callout function that displays information. By default, it shows the start of
3350    the match and the current position in the text. There are some new data escapes
3351    to vary what happens:
3352    
3353        \C+         in addition, show current contents of captured substrings
3354        \C-         do not supply a callout function
3355        \C!n        return 1 when callout number n is reached
3356        \C!n!m      return 1 when callout number n is reached for the mth time
3357    
3358    30. If pcregrep was called with the -l option and just a single file name, it
3359    output "<stdin>" if a match was found, instead of the file name.
3360    
3361    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
3362    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
3363    pcre_exec(). This saves a malloc/free per call. The default value of
3364    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
3365    when configuring.
3366    
3367    32. The default maximum size of a compiled pattern is 64K. There have been a
3368    few cases of people hitting this limit. The code now uses macros to handle the
3369    storing of links as offsets within the compiled pattern. It defaults to 2-byte
3370    links, but this can be changed to 3 or 4 bytes by --with-link-size when
3371    configuring. Tests 2 and 5 work only with 2-byte links because they output
3372    debugging information about compiled patterns.
3373    
3374    33. Internal code re-arrangements:
3375    
3376    (a) Moved the debugging function for printing out a compiled regex into
3377        its own source file (printint.c) and used #include to pull it into
3378        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
3379        separate copies.
3380    
3381    (b) Defined the list of op-code names for debugging as a macro in
3382        internal.h so that it is next to the definition of the opcodes.
3383    
3384    (c) Defined a table of op-code lengths for simpler skipping along compiled
3385        code. This is again a macro in internal.h so that it is next to the
3386        definition of the opcodes.
3387    
3388    34. Added support for recursive calls to individual subpatterns, along the
3389    lines of Robin Houston's patch (but implemented somewhat differently).
3390    
3391    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
3392    allow it to read and process whole directories in Win32. This code was
3393    contributed by Lionel Fourquaux; it has not been tested by me.
3394    
3395    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
3396    used to name a group. Names consist of alphanumerics and underscores, and must
3397    be unique. Back references use the syntax (?P=name) and recursive calls use
3398    (?P>name) which is a PCRE extension to the Python extension. Groups still have
3399    numbers. The function pcre_fullinfo() can be used after compilation to extract
3400    a name/number map. There are three relevant calls:
3401    
3402      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
3403      PCRE_INFO_NAMECOUNT            yields the number of entries
3404      PCRE_INFO_NAMETABLE            yields a pointer to the map.
3405    
3406    The map is a vector of fixed-size entries. The size of each entry depends on
3407    the length of the longest name used. The first two bytes of each entry are the
3408    group number, most significant byte first. There follows the corresponding
3409    name, zero terminated. The names are in alphabetical order.
3410    
3411    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
3412    case instead of 255. Making it the same both with and without UTF-8 support
3413    means that the same test output works with both.
3414    
3415    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
3416    calling malloc() with a zero argument.
3417    
3418    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
3419    optimization. I've improved things by keeping a bitmap of backreferences with
3420    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
3421    fact referenced, the optimization can be applied. It is unlikely that a
3422    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
3423    the match to follow \n) will appear inside brackets with a number greater than
3424    31, but if it does, any back reference > 31 suppresses the optimization.
3425    
3426    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
3427    of disabling numbered capturing parentheses. Any opening parenthesis that is
3428    not followed by ? behaves as if it were followed by ?: but named parentheses
3429    can still be used for capturing (and they will acquire numbers in the usual
3430    way).
3431    
3432    41. Redesigned the return codes from the match() function into yes/no/error so
3433    that errors can be passed back from deep inside the nested calls. A malloc
3434    failure while inside a recursive subpattern call now causes the
3435    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
3436    
3437    42. It is now possible to set a limit on the number of times the match()
3438    function is called in a call to pcre_exec(). This facility makes it possible to
3439    limit the amount of recursion and backtracking, though not in a directly
3440    obvious way, because the match() function is used in a number of different
3441    circumstances. The count starts from zero for each position in the subject
3442    string (for non-anchored patterns). The default limit is, for compatibility, a
3443    large number, namely 10 000 000. You can change this in two ways:
3444    
3445    (a) When configuring PCRE before making, you can use --with-match-limit=n
3446        to set a default value for the compiled library.
3447    
3448    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
3449        a different value is set. See 45 below.
3450    
3451    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
3452    
3453    43. Added a new function pcre_config(int, void *) to enable run-time extraction
3454    of things that can be changed at compile time. The first argument specifies
3455    what is wanted and the second points to where the information is to be placed.
3456    The current list of available information is:
3457    
3458      PCRE_CONFIG_UTF8
3459    
3460    The output is an integer that is set to one if UTF-8 support is available;
3461    otherwise it is set to zero.
3462    
3463      PCRE_CONFIG_NEWLINE
3464    
3465    The output is an integer that it set to the value of the code that is used for
3466    newline. It is either LF (10) or CR (13).
3467    
3468      PCRE_CONFIG_LINK_SIZE
3469    
3470    The output is an integer that contains the number of bytes used for internal
3471    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
3472    
3473      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
3474    
3475    The output is an integer that contains the threshold above which the POSIX
3476    interface uses malloc() for output vectors. See item 31 above.
3477    
3478      PCRE_CONFIG_MATCH_LIMIT
3479    
3480    The output is an unsigned integer that contains the default limit of the number
3481    of match() calls in a pcre_exec() execution. See 42 above.
3482    
3483    44. pcretest has been upgraded by the addition of the -C option. This causes it
3484    to extract all the available output from the new pcre_config() function, and to
3485    output it. The program then exits immediately.
3486    
3487    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3488    order to support additional features. One way would have been to define
3489    pcre_exec2() (for example) with extra arguments, but this would not have been
3490    extensible, and would also have required all calls to the original function to
3491    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3492    is used for passing in "extra" data from pcre_study().
3493    
3494    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3495    contains the following fields:
3496    
3497      flags         a bitmap indicating which of the following fields are set
3498      study_data    opaque data from pcre_study()
3499      match_limit   a way of specifying a limit on match() calls for a specific
3500                      call to pcre_exec()
3501      callout_data  data for callouts (see 49 below)
3502    
3503    The flag bits are also defined in pcre.h, and are
3504    
3505      PCRE_EXTRA_STUDY_DATA
3506      PCRE_EXTRA_MATCH_LIMIT
3507      PCRE_EXTRA_CALLOUT_DATA
3508    
3509    The pcre_study() function now returns one of these new pcre_extra blocks, with
3510    the actual study data pointed to by the study_data field, and the
3511    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3512    before. That is, this change is entirely upwards-compatible and requires no
3513    change to existing code.
3514    
3515    If you want to pass in additional data to pcre_exec(), you can either place it
3516    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3517    block.
3518    
3519    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3520    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3521    times with different match limits, until it finds the minimum value needed for
3522    pcre_exec() to complete. The value is then output. This can be instructive; for
3523    most simple matches the number is quite small, but for pathological cases it
3524    gets very large very quickly.
3525    
3526    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3527    returns the size of the data block pointed to by the study_data field in a
3528    pcre_extra block, that is, the value that was passed as the argument to
3529    pcre_malloc() when PCRE was getting memory in which to place the information
3530    created by pcre_study(). The fourth argument should point to a size_t variable.
3531    pcretest has been extended so that this information is shown after a successful
3532    pcre_study() call when information about the compiled regex is being displayed.
3533    
3534    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3535    because what follows is always an absolute path. (Later: it turns out that this
3536    is more than cosmetic for MinGW, because it doesn't like empty path
3537    components.)
3538    
3539    49. Some changes have been made to the callout feature (see 28 above):
3540    
3541    (i)  A callout function now has three choices for what it returns:
3542    
3543           0  =>  success, carry on matching
3544         > 0  =>  failure at this point, but backtrack if possible
3545         < 0  =>  serious error, return this value from pcre_exec()
3546    
3547         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3548         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3549         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3550         use by callout functions. It will never be used by PCRE itself.
3551    
3552    (ii) The pcre_extra structure (see 45 above) has a void * field called
3553         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3554         pcre_callout_block structure has a field of the same name. The contents of
3555         the field passed in the pcre_extra structure are passed to the callout
3556         function in the corresponding field in the callout block. This makes it
3557         easier to use the same callout-containing regex from multiple threads. For
3558         testing, the pcretest program has a new data escape
3559    
3560           \C*n        pass the number n (may be negative) as callout_data
3561    
3562         If the callout function in pcretest receives a non-zero value as
3563         callout_data, it returns that value.
3564    
3565    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3566    there were some redundant $(CFLAGS) in commands that are now specified as
3567    $(LINK), which already includes $(CFLAGS).
3568    
3569    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3570    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3571    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3572    one-byte characters throughout. Note that case-insensitive matching applies
3573    only to characters whose values are less than 256. PCRE doesn't support the
3574    notion of cases for higher-valued characters.
3575    
3576    (i)   A character class whose characters are all within 0-255 is handled as
3577          a bit map, and the map is inverted for negative classes. Previously, a
3578          character > 255 always failed to match such a class; however it should
3579          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3580    
3581    (ii)  A negated character class with a single character < 255 is coded as
3582          "not this character" (OP_NOT). This wasn't working properly when the test
3583          character was multibyte, either singly or repeated.
3584    
3585    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3586          mode, for example: \x{100}{2,3}.
3587    
3588    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3589          singly or repeated) now correctly test multibyte characters. However,
3590          PCRE doesn't recognize any characters with values greater than 255 as
3591          digits, spaces, or word characters. Such characters always match \D, \S,
3592          and \W, and never match \d, \s, or \w.
3593    
3594    (v)   Classes may now contain characters and character ranges with values
3595          greater than 255. For example: [ab\x{100}-\x{400}].
3596    
3597    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3598          PCRE in UTF-8 mode.
3599    
3600    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3601    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3602    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3603    value.)
3604    
3605    53. The single man page has become too large. I have therefore split it up into
3606    a number of separate man pages. These also give rise to individual HTML pages;
3607    these are now put in a separate directory, and there is an index.html page that
3608    lists them all. Some hyperlinking between the pages has been installed.
3609    
3610    54. Added convenience functions for handling named capturing parentheses.
3611    
3612    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3613    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3614    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3615    are faulted.
3616    
3617    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3618    calling configure. These values are used when compiling the dftables.c program
3619    which is run to generate the source of the default character tables. They
3620    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3621    you will need to set these values.
3622    
3623    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3624    
3625    
3626    Version 3.9 02-Jan-02
3627    ---------------------
3628    
3629    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3630    
3631    2. If --disable-static was given, the building process failed when trying to
3632    build pcretest and pcregrep. (For some reason it was using libtool to compile
3633    them, which is not right, as they aren't part of the library.)
3634    
3635    
3636    Version 3.8 18-Dec-01
3637    ---------------------
3638    
3639    1. The experimental UTF-8 code was completely screwed up. It was packing the
3640    bytes in the wrong order. How dumb can you get?
3641    
3642    
3643    Version 3.7 29-Oct-01
3644    ---------------------
3645    
3646    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3647    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3648    this didn't happen under Solaris 8, where I normally test things.
3649    
3650    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3651    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3652    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3653    this fix an hour or so after the initial 3.7 release.)
3654    
3655    
3656    Version 3.6 23-Oct-01
3657    ---------------------
3658    
3659    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3660    offsets passed as NULL with zero offset count.
3661    
3662    2. The config.guess and config.sub files had not been updated when I moved to
3663    the latest autoconf.
3664    
3665    
3666    Version 3.5 15-Aug-01
3667    ---------------------
3668    
3669    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3670    had been forgotten.
3671    
3672    2. By using declared but undefined structures, we can avoid using "void"
3673    definitions in pcre.h while keeping the internal definitions of the structures
3674    private.
3675    
3676    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3677    user point of view, this means that both static and shared libraries are built
3678    by default, but this can be individually controlled. More of the work of
3679    handling this static/shared cases is now inside libtool instead of PCRE's make
3680    file.
3681    
3682    4. The pcretest utility is now installed along with pcregrep because it is
3683    useful for users (to test regexs) and by doing this, it automatically gets
3684    relinked by libtool. The documentation has been turned into a man page, so
3685    there are now .1, .txt, and .html versions in /doc.
3686    
3687    5. Upgrades to pcregrep:
3688       (i)   Added long-form option names like gnu grep.
3689       (ii)  Added --help to list all options with an explanatory phrase.
3690       (iii) Added -r, --recursive to recurse into sub-directories.
3691       (iv)  Added -f, --file to read patterns from a file.
3692    
3693    6. pcre_exec() was referring to its "code" argument before testing that
3694    argument for NULL (and giving an error if it was NULL).
3695    
3696    7. Upgraded Makefile.in to allow for compiling in a different directory from
3697    the source directory.
3698    
3699    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3700    options bits, the pointer it was passed was to an int instead of to an unsigned
3701    long int. This mattered only on 64-bit systems.
3702    
3703    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3704    generated) instead of pcre.in, which it its source. Also made the same change
3705    in several of the .c files.
3706    
3707    10. A new release of gcc defines printf() as a macro, which broke pcretest
3708    because it had an ifdef in the middle of a string argument for printf(). Fixed
3709    by using separate calls to printf().
3710    
3711    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3712    script, to force use of CR or LF instead of \n in the source. On non-Unix
3713