ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 13 by nigel, Sat Feb 24 21:38:21 2007 UTC revision 914 by zherczeg, Mon Feb 13 06:04:50 2012 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.31
5    -----------------------------
7    1.  Fixing a wrong JIT test case and some compiler warnings.
9    2.  Removed a bashism from the RunTest script.
11    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
12        to unsigned type, result still unsigned" that was given by an MS compiler
13        on encountering the code "-sizeof(xxx)".
15    4.  Partial matching support is added to the JIT compiler.
18    Version 8.30 04-February-2012
19    -----------------------------
21    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
22        name is defined in ctype.h.
24    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
25        only in quite long subpatterns.
27    3.  Removed the function pcre_info(), which has been obsolete and deprecated
28        since it was replaced by pcre_fullinfo() in February 2000.
30    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
31        match a (*MARK), and the match failed at the start of the subject, a
32        reference to memory before the start of the subject could occur. This bug
33        was introduced by fix 17 of release 8.21.
35    5.  A reference to an unset group with zero minimum repetition was giving
36        totally wrong answers (in non-JavaScript-compatibility mode). For example,
37        /(another)?(\1?)test/ matched against "hello world test". This bug was
38        introduced in release 8.13.
40    6.  Add support for 16-bit character strings (a large amount of work involving
41        many changes and refactorings).
43    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
44        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
45        from a file.
47    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
48        rounding is not applied in this particular case).
50    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
51        if they appear, or are escaped, in patterns.
53    10. Get rid of a number of -Wunused-but-set-variable warnings.
55    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
56        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
57        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
58        also returns the mark "x". This bug applied to capturing parentheses,
59        non-capturing parentheses, and atomic parentheses. It also applied to some
60        assertions.
62    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
63        information out of configure.ac instead of relying on pcre.h.generic, which
64        is not stored in the repository.
66    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
67        -lreadline.
69    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
71    15. Applied Graycode's patch to put the top-level frame on the stack rather
72        than the heap when not using the stack for recursion. This gives a
73        performance improvement in many cases when recursion is not deep.
75    16. Experimental code added to "pcretest -C" to output the stack frame size.
78    Version 8.21 12-Dec-2011
79    ------------------------
81    1.  Updating the JIT compiler.
83    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
84        are added as well.
86    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
87        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
88        calling _pcre_jit_exec. Some extra comments are added.
90    4.  (*MARK) settings inside atomic groups that do not contain any capturing
91        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
92        was introduced by change 18 for 8.20.
94    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
95        ECMA-262 standard.
97    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
98        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
99        This bug was probably introduced by change 9 of 8.13.
101    7.  While fixing 6 above, I noticed that a number of other items were being
102        incorrectly rejected as "not fixed length". This arose partly because newer
103        opcodes had not been added to the fixed-length checking code. I have (a)
104        corrected the bug and added tests for these items, and (b) arranged for an
105        error to occur if an unknown opcode is encountered while checking for fixed
106        length instead of just assuming "not fixed length". The items that were
107        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
108        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
109        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
111    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
112        being incorrectly compiled and would have given unpredicatble results.
114    9.  A possessively repeated subpattern with minimum repeat count greater than
115        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
116        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
117        the first (A) could occur when it should not.
119    10. Add a cast and remove a redundant test from the code.
121    11. JIT should use pcre_malloc/pcre_free for allocation.
123    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
124        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
125        is anything other than /usr, -L is still shown).
127    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
129    14. Perl does not support \N without a following name in a [] class; PCRE now
130        also gives an error.
132    15. If a forward reference was repeated with an upper limit of around 2000,
133        it caused the error "internal error: overran compiling workspace". The
134        maximum number of forward references (including repeats) was limited by the
135        internal workspace, and dependent on the LINK_SIZE. The code has been
136        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
137        the default depends on LINK_SIZE. There is a new upper limit (for safety)
138        of around 200,000 forward references. While doing this, I also speeded up
139        the filling in of repeated forward references.
141    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
142        incorrectly expecting the subject to contain another "a" after the start.
144    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
145        in the match, the SKIP should be ignored. This was not happening; instead
146        the SKIP was being treated as NOMATCH. For patterns such as
147        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
148        tested.
150    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
151        now much more compatible with Perl, in particular in cases where the result
152        is a non-match for a non-anchored pattern. For example, if
153        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
154        "m", where previously it did not return a name. A side effect of this
155        change is that for partial matches, the last encountered mark name is
156        returned, as for non matches. A number of tests that were previously not
157        Perl-compatible have been moved into the Perl-compatible test files. The
158        refactoring has had the pleasing side effect of removing one argument from
159        the match() function, thus reducing its stack requirements.
161    19. If the /S+ option was used in pcretest to study a pattern using JIT,
162        subsequent uses of /S (without +) incorrectly behaved like /S+.
164    21. Retrieve executable code size support for the JIT compiler and fixing
165        some warnings.
167    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
168        not work when the shorter character appeared right at the end of the
169        subject string.
171    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
172        systems.
174    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
175        output it when the /M option is used in pcretest.
177    25. The CheckMan script was not being included in the distribution. Also, added
178        an explicit "perl" to run Perl scripts from the PrepareRelease script
179        because this is reportedly needed in Windows.
181    26. If study data was being save in a file and studying had not found a set of
182        "starts with" bytes for the pattern, the data written to the file (though
183        never used) was taken from uninitialized memory and so caused valgrind to
184        complain.
186    27. Updated RunTest.bat as provided by Sheri Pierce.
188    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
190    29. Computation of memory usage for the table of capturing group names was
191        giving an unnecessarily large value.
194    Version 8.20 21-Oct-2011
195    ------------------------
197    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
198        a POSIX class. After further experiments with Perl, which convinced me that
199        Perl has bugs and confusions, a closing square bracket is no longer allowed
200        in a POSIX name. This bug also affected patterns with classes that started
201        with full stops.
203    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
204        captured substring, but while checking the failing first alternative,
205        substring 1 is temporarily captured. If the output vector supplied to
206        pcre_exec() was not big enough for this capture, the yield of the function
207        was still zero ("insufficient space for captured substrings"). This cannot
208        be totally fixed without adding another stack variable, which seems a lot
209        of expense for a edge case. However, I have improved the situation in cases
210        such as /(a)(b)x|abc/ matched against "abc", where the return code
211        indicates that fewer than the maximum number of slots in the ovector have
212        been set.
214    3.  Related to (2) above: when there are more back references in a pattern than
215        slots in the output vector, pcre_exec() uses temporary memory during
216        matching, and copies in the captures as far as possible afterwards. It was
217        using the entire output vector, but this conflicts with the specification
218        that only 2/3 is used for passing back captured substrings. Now it uses
219        only the first 2/3, for compatibility. This is, of course, another edge
220        case.
222    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
223        main code base, and can be used by building with --enable-jit. When this is
224        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
225        runtime --no-jit option is given.
227    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
228        ovector, the return from the function was zero, implying that there were
229        other matches that did not fit. The correct "exactly full" value is now
230        returned.
232    6.  If a subpattern that was called recursively or as a subroutine contained
233        (*PRUNE) or any other control that caused it to give a non-standard return,
234        invalid errors such as "Error -26 (nested recursion at the same subject
235        position)" or even infinite loops could occur.
237    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
238        computing the minimum length on reaching *ACCEPT, and so ended up with the
239        wrong value of 1 rather than 0. Further investigation indicates that
240        computing a minimum subject length in the presence of *ACCEPT is difficult
241        (think back references, subroutine calls), and so I have changed the code
242        so that no minimum is registered for a pattern that contains *ACCEPT.
244    8.  If (*THEN) was present in the first (true) branch of a conditional group,
245        it was not handled as intended. [But see 16 below.]
247    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
248        Sheri Pierce.
250    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
251        the first byte in a match must be "a".
253    11. Change 17 for 8.13 increased the recursion depth for patterns like
254        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
255        pattern contains any instances of (*THEN). If it does not, the old
256        optimizations are restored. It would be nice to do this on a per-group
257        basis, but at the moment that is not feasible.
259    12. In some environments, the output of pcretest -C is CRLF terminated. This
260        broke RunTest's code that checks for the link size. A single white space
261        character after the value is now allowed for.
263    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
264        For "fr", it uses the Windows-specific input and output files.
266    14. If (*THEN) appeared in a group that was called recursively or as a
267        subroutine, it did not work as intended. [But see next item.]
269    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
270        pattern fragments (but not containing any | characters). If A and B are
271        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
272        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
273        D. In other words, Perl considers parentheses that do not contain any |
274        characters to be part of a surrounding alternative, whereas PCRE was
275        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
276        differently. PCRE now behaves in the same way as Perl, except in the case
277        of subroutine/recursion calls such as (?1) which have in any case always
278        been different (but PCRE had them first :-).
280    16. Related to 15 above: Perl does not treat the | in a conditional group as
281        creating alternatives. Such a group is treated in the same way as an
282        ordinary group without any | characters when processing (*THEN). PCRE has
283        been changed to match Perl's behaviour.
285    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
286        RunGrepTest script failed.
288    18. Change 22 for version 13 caused atomic groups to use more stack. This is
289        inevitable for groups that contain captures, but it can lead to a lot of
290        stack use in large patterns. The old behaviour has been restored for atomic
291        groups that do not contain any capturing parentheses.
293    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
294        suppress the check for a minimum subject length at run time. (If it was
295        given to pcre_exec() or pcre_dfa_exec() it did work.)
297    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
298        fail to work when decoding hex characters in data strings in EBCDIC
299        environments.
301    21. It appears that in at least one Mac OS environment, the isxdigit() function
302        is implemented as a macro that evaluates to its argument more than once,
303        contravening the C 90 Standard (I haven't checked a later standard). There
304        was an instance in pcretest which caused it to go wrong when processing
305        \x{...} escapes in subject strings. The has been rewritten to avoid using
306        things like p++ in the argument of isxdigit().
309    Version 8.13 16-Aug-2011
310    ------------------------
312    1.  The Unicode data tables have been updated to Unicode 6.0.0.
314    2.  Two minor typos in pcre_internal.h have been fixed.
316    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
317        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
318        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
320    4.  There were a number of related bugs in the code for matching backrefences
321        caselessly in UTF-8 mode when codes for the characters concerned were
322        different numbers of bytes. For example, U+023A and U+2C65 are an upper
323        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
324        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
325        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
326        2-byte code at the end of the subject (it thought there wasn't enough data
327        left).
329    5.  Comprehensive information about what went wrong is now returned by
330        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
331        as the output vector has at least 2 elements. The offset of the start of
332        the failing character and a reason code are placed in the vector.
334    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
335        now returned is for the first byte of the failing character, instead of the
336        last byte inspected. This is an incompatible change, but I hope it is small
337        enough not to be a problem. It makes the returned offset consistent with
338        pcre_exec() and pcre_dfa_exec().
340    7.  pcretest now gives a text phrase as well as the error number when
341        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
342        failure, the offset and reason code are output.
344    8.  When \R was used with a maximizing quantifier it failed to skip backwards
345        over a \r\n pair if the subsequent match failed. Instead, it just skipped
346        back over a single character (\n). This seems wrong (because it treated the
347        two characters as a single entity when going forwards), conflicts with the
348        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
349        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
350        has been changed.
352    9.  Some internal refactoring has changed the processing so that the handling
353        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
354        time (the PCRE_DOTALL option was changed this way some time ago: version
355        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
356        which was always a bit of a fudge. It also means that there is one less
357        argument for the match() function, which reduces its stack requirements
358        slightly. This change also fixes an incompatibility with Perl: the pattern
359        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
361    10. More internal refactoring has drastically reduced the number of recursive
362        calls to match() for possessively repeated groups such as (abc)++ when
363        using pcre_exec().
365    11. While implementing 10, a number of bugs in the handling of groups were
366        discovered and fixed:
368        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
369        (a|)*(?1) gave a compile-time internal error.
370        ((a|)+)+  did not notice that the outer group could match an empty string.
371        (^a|^)+   was not marked as anchored.
372        (.*a|.*)+ was not marked as matching at start or after a newline.
374    12. Yet more internal refactoring has removed another argument from the match()
375        function. Special calls to this function are now indicated by setting a
376        value in a variable in the "match data" data block.
378    13. Be more explicit in pcre_study() instead of relying on "default" for
379        opcodes that mean there is no starting character; this means that when new
380        ones are added and accidentally left out of pcre_study(), testing should
381        pick them up.
383    14. The -s option of pcretest has been documented for ages as being an old
384        synonym of -m (show memory usage). I have changed it to mean "force study
385        for every regex", that is, assume /S for every regex. This is similar to -i
386        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
387        using it. It makes it easier to run collections of tests with and without
388        study enabled, and thereby test pcre_study() more easily. All the standard
389        tests are now run with and without -s (but some patterns can be marked as
390        "never study" - see 20 below).
392    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
393        restoration of the capturing data to the outer values was not happening
394        correctly.
396    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
397        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
398        pattern had matched an empty string, and so incorrectly returned a no
399        match.
401    17. There was optimizing code for the last branch of non-capturing parentheses,
402        and also for the obeyed branch of a conditional subexpression, which used
403        tail recursion to cut down on stack usage. Unfortunately, now that there is
404        the possibility of (*THEN) occurring in these branches, tail recursion is
405        no longer possible because the return has to be checked for (*THEN). These
406        two optimizations have therefore been removed. [But see 8.20/11 above.]
408    18. If a pattern containing \R was studied, it was assumed that \R always
409        matched two bytes, thus causing the minimum subject length to be
410        incorrectly computed because \R can also match just one byte.
412    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
413        was incorrectly computed.
415    20. If /S is present twice on a test pattern in pcretest input, it now
416        *disables* studying, thereby overriding the use of -s on the command line
417        (see 14 above). This is necessary for one or two tests to keep the output
418        identical in both cases.
420    21. When (*ACCEPT) was used in an assertion that matched an empty string and
421        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
423    22. When an atomic group that contained a capturing parenthesis was
424        successfully matched, but the branch in which it appeared failed, the
425        capturing was not being forgotten if a higher numbered group was later
426        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
427        group 1 to "a", when in fact it should be unset. This applied to multi-
428        branched capturing and non-capturing groups, repeated or not, and also to
429        positive assertions (capturing in negative assertions does not happen
430        in PCRE) and also to nested atomic groups.
432    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
433        subject after a captured substring, to make it easier to tell which of a
434        number of identical substrings has been captured.
436    24. The way atomic groups are processed by pcre_exec() has been changed so that
437        if they are repeated, backtracking one repetition now resets captured
438        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
439        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
440        "aaa". Previously, it would have been "a". As part of this code
441        refactoring, the way recursive calls are handled has also been changed.
443    25. If an assertion condition captured any substrings, they were not passed
444        back unless some other capturing happened later. For example, if
445        (?(?=(a))a) was matched against "a", no capturing was returned.
447    26. When studying a pattern that contained subroutine calls or assertions,
448        the code for finding the minimum length of a possible match was handling
449        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
450        group 1 called group 2 while simultaneously a separate group 2 called group
451        1). A stack overflow occurred in this case. I have fixed this by limiting
452        the recursion depth to 10.
454    27. Updated RunTest.bat in the distribution to the version supplied by Tom
455        Fortmann. This supports explicit test numbers on the command line, and has
456        argument validation and error reporting.
458    28. An instance of \X with an unlimited repeat could fail if at any point the
459        first character it looked at was a mark character.
461    29. Some minor code refactoring concerning Unicode properties and scripts
462        should reduce the stack requirement of match() slightly.
464    30. Added the '=' option to pcretest to check the setting of unused capturing
465        slots at the end of the pattern, which are documented as being -1, but are
466        not included in the return count.
468    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
469        compiled something random. Now it gives a compile-time error (as does
470        Perl).
472    32. A *MARK encountered during the processing of a positive assertion is now
473        recorded and passed back (compatible with Perl).
475    33. If --only-matching or --colour was set on a pcregrep call whose pattern
476        had alternative anchored branches, the search for a second match in a line
477        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
478        matched the line "0102" twice. The same bug affected patterns that started
479        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
480        twice.
482    34. Previously, PCRE did not allow quantification of assertions. However, Perl
483        does, and because of capturing effects, quantifying parenthesized
484        assertions may at times be useful. Quantifiers are now allowed for
485        parenthesized assertions.
487    35. A minor code tidy in pcre_compile() when checking options for \R usage.
489    36. \g was being checked for fancy things in a character class, when it should
490        just be a literal "g".
492    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
493        appearance of a nested POSIX class supersedes an apparent external class.
494        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
495        unescaped square brackets may also appear as part of class names. For
496        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
497        more like Perl. (But see 8.20/1 above.)
499    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
500        was because it thought it was \N{name}, which is not supported).
502    39. Add minix to OS list not supporting the -S option in pcretest.
504    40. PCRE tries to detect cases of infinite recursion at compile time, but it
505        cannot analyze patterns in sufficient detail to catch mutual recursions
506        such as ((?1))((?2)). There is now a runtime test that gives an error if a
507        subgroup is called recursively as a subpattern for a second time at the
508        same position in the subject string. In previous releases this might have
509        been caught by the recursion limit, or it might have run out of stack.
511    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
512        happen only once. PCRE was, however incorrectly giving a compile time error
513        "recursive call could loop indefinitely" because it cannot analyze the
514        pattern in sufficient detail. The compile time test no longer happens when
515        PCRE is compiling a conditional subpattern, but actual runaway loops are
516        now caught at runtime (see 40 above).
518    42. It seems that Perl allows any characters other than a closing parenthesis
519        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
520        has been changed to be the same.
522    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
523        as not to get warnings when autogen.sh is called. Also changed
524        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
526    44. To help people who use pcregrep to scan files containing exceedingly long
527        lines, the following changes have been made:
529        (a) The default value of the buffer size parameter has been increased from
530            8K to 20K. (The actual buffer used is three times this size.)
532        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
533            PCRE is built.
535        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
536            to be set at run time.
538        (d) Numerical values in pcregrep options can be followed by K or M, for
539            example --buffer-size=50K.
541        (e) If a line being scanned overflows pcregrep's buffer, an error is now
542            given and the return code is set to 2.
544    45. Add a pointer to the latest mark to the callout data block.
546    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
547        partial match of an empty string instead of no match. This was specific to
548        the use of ".".
550    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
551        complete match instead of a partial match. This bug was dependent on both
552        the PCRE_UTF8 and PCRE_DOTALL options being set.
554    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
555        starting byte set, because \b was not being ignored.
558    Version 8.12 15-Jan-2011
559    ------------------------
561    1.  Fixed some typos in the markup of the man pages, and wrote a script that
562        checks for such things as part of the documentation building process.
564    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
565        --match-limit and --recursion-limit options (added for 8.11). In
566        particular, this made one of the standard tests fail. (The integer value
567        went into the wrong half of a long int.)
569    3.  If the --colour option was given to pcregrep with -v (invert match), it
570        did strange things, either producing crazy output, or crashing. It should,
571        of course, ignore a request for colour when reporting lines that do not
572        match.
574    4.  Another pcregrep bug caused similar problems if --colour was specified with
575        -M (multiline) and the pattern match finished with a line ending.
577    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
578        matched in multiline mode, the following line was shown as part of the
579        match. This seems wrong, so I have changed it.
581    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
582        the check for further matches in the same line (so they could be coloured)
583        to overrun the end of the current line. If another match was found, it was
584        incorrectly shown (and then shown again when found in the next line).
586    7.  If pcregrep was compiled under Windows, there was a reference to the
587        function pcregrep_exit() before it was defined. I am assuming this was
588        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
589        reported by a user. I've moved the definition above the reference.
592    Version 8.11 10-Dec-2010
593    ------------------------
595    1.  (*THEN) was not working properly if there were untried alternatives prior
596        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
597        backtracked to try for "b" instead of moving to the next alternative branch
598        at the same level (in this case, to look for "c"). The Perl documentation
599        is clear that when (*THEN) is backtracked onto, it goes to the "next
600        alternative in the innermost enclosing group".
602    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
603        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
604        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
605        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
606        (*THEN).
608    3.  If \s appeared in a character class, it removed the VT character from
609        the class, even if it had been included by some previous item, for example
610        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
611        of \s, but is part of the POSIX "space" class.)
613    4.  A partial match never returns an empty string (because you can always
614        match an empty string at the end of the subject); however the checking for
615        an empty string was starting at the "start of match" point. This has been
616        changed to the "earliest inspected character" point, because the returned
617        data for a partial match starts at this character. This means that, for
618        example, /(?<=abc)def/ gives a partial match for the subject "abc"
619        (previously it gave "no match").
621    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
622        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
623        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
624        has an implication that the given string is incomplete (because a partial
625        match is preferred over a full match). For this reason, these items now
626        give a partial match in this situation. [Aside: previously, the one case
627        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
628        match rather than a full match, which was wrong by the old rules, but is
629        now correct.]
631    6.  There was a bug in the handling of #-introduced comments, recognized when
632        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
633        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
634        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
635        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
636        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
637        places in pcre_compile().
639    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
640        comments when looking ahead for named forward references to subpatterns,
641        the only newline sequence it recognized was NL. It now handles newlines
642        according to the set newline convention.
644    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
645        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
646        cater for a lack of strerror(). These oversights have been fixed.
648    9.  Added --match-limit and --recursion-limit to pcregrep.
650    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
652    11. When the -o option was used, pcregrep was setting a return code of 1, even
653        when matches were found, and --line-buffered was not being honoured.
655    12. Added an optional parentheses number to the -o and --only-matching options
656        of pcregrep.
658    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
659        can match an empty string. The code to do it in pcretest and pcredemo
660        needed fixing:
662        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
663            only one byte after an empty string match just before CRLF (this case
664            just got forgotten; "any" and "anycrlf" were OK).
666        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
667            mode when an empty string match preceded an ASCII character followed by
668            a non-ASCII character. (The code for advancing by one character rather
669            than one byte was nonsense.)
671        (c) The pcredemo.c sample program did not have any code at all to handle
672            the cases when CRLF is a valid newline sequence.
674    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
675        as a starting offset was within the subject string. There is now a new
676        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
677        negative or greater than the length of the string. In order to test this,
678        pcretest is extended to allow the setting of negative starting offsets.
680    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
681        starting offset points to the beginning of a UTF-8 character was
682        unnecessarily clumsy. I tidied it up.
684    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
685        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
687    17. Nobody had reported that the --include_dir option, which was added in
688        release 7.7 should have been called --include-dir (hyphen, not underscore)
689        for compatibility with GNU grep. I have changed it to --include-dir, but
690        left --include_dir as an undocumented synonym, and the same for
691        --exclude-dir, though that is not available in GNU grep, at least as of
692        release 2.5.4.
694    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
695        characters from a string of bytes) have been redefined so as not to use
696        loops, in order to improve performance in some environments. At the same
697        time, I abstracted some of the common code into auxiliary macros to save
698        repetition (this should not affect the compiled code).
700    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
701        compile-time error is now given if \c is not followed by an ASCII
702        character, that is, a byte less than 128. (In EBCDIC mode, the code is
703        different, and any byte value is allowed.)
705    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
706        START_OPTIMIZE option, which is now allowed at compile time - but just
707        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
708        to pcregrep and other applications that have no direct access to PCRE
709        options. The new /Y option in pcretest sets this option when calling
710        pcre_compile().
712    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
713        back references. Groups containing recursive back references were forced to
714        be atomic by that change, but in the case of named groups, the amount of
715        memory required was incorrectly computed, leading to "Failed: internal
716        error: code overflow". This has been fixed.
718    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
719        pcretest.c, to avoid build problems in some Borland environments.
722    Version 8.10 25-Jun-2010
723    ------------------------
725    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
726        THEN.
728    2.  (*ACCEPT) was not working when inside an atomic group.
730    3.  Inside a character class, \B is treated as a literal by default, but
731        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
732        causes the error). The code is unchanged, but I tidied the documentation.
734    4.  Inside a character class, PCRE always treated \R and \X as literals,
735        whereas Perl faults them if its -w option is set. I have changed PCRE so
736        that it faults them when PCRE_EXTRA is set.
738    5.  Added support for \N, which always matches any character other than
739        newline. (It is the same as "." when PCRE_DOTALL is not set.)
741    6.  When compiling pcregrep with newer versions of gcc which may have
742        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
743        declared with attribute warn_unused_result" were given. Just casting the
744        result to (void) does not stop the warnings; a more elaborate fudge is
745        needed. I've used a macro to implement this.
747    7.  Minor change to pcretest.c to avoid a compiler warning.
749    8.  Added four artifical Unicode properties to help with an option to make
750        \s etc use properties (see next item). The new properties are: Xan
751        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
753    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
754        use Unicode properties. (*UCP) at the start of a pattern can be used to set
755        this option. Modified pcretest to add /W to test this facility. Added
756        REG_UCP to make it available via the POSIX interface.
758    10. Added --line-buffered to pcregrep.
760    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
761        studied, and the match started with a letter with a code point greater than
762        127 whose first byte was different to the first byte of the other case of
763        the letter, the other case of this starting letter was not recognized
764        (#976).
766    12. If a pattern that was studied started with a repeated Unicode property
767        test, for example, \p{Nd}+, there was the theoretical possibility of
768        setting up an incorrect bitmap of starting bytes, but fortunately it could
769        not have actually happened in practice until change 8 above was made (it
770        added property types that matched character-matching opcodes).
772    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
773        possible starting bytes for non-anchored patterns.
775    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
776        \R, and also a number of cases that involve Unicode properties, both
777        explicit and implicit when PCRE_UCP is set.
779    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
780        input, it could crash or give wrong results if characters with values
781        greater than 0xc0 were present in the subject string. (Detail: it assumed
782        UTF-8 input when processing these items.)
784    16. Added a lot of (int) casts to avoid compiler warnings in systems where
785        size_t is 64-bit (#991).
787    17. Added a check for running out of memory when PCRE is compiled with
788        --disable-stack-for-recursion (#990).
790    18. If the last data line in a file for pcretest does not have a newline on
791        the end, a newline was missing in the output.
793    19. The default pcre_chartables.c file recognizes only ASCII characters (values
794        less than 128) in its various bitmaps. However, there is a facility for
795        generating tables according to the current locale when PCRE is compiled. It
796        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
797        characters, are recognized by isspace() and therefore were getting set in
798        these tables, and indeed these tables seem to approximate to ISO 8859. This
799        caused a problem in UTF-8 mode when pcre_study() was used to create a list
800        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
801        which of course cannot start UTF-8 characters. I have changed the code so
802        that only real ASCII characters (less than 128) and the correct starting
803        bytes for UTF-8 encodings are set for characters greater than 127 when in
804        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
805        altogether.)
807    20. Added the /T option to pcretest so as to be able to run tests with non-
808        standard character tables, thus making it possible to include the tests
809        used for 19 above in the standard set of tests.
811    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
812        reference to a subpattern the other side of a comment that contains an
813        opening parenthesis caused either an internal compiling error, or a
814        reference to the wrong subpattern.
817    Version 8.02 19-Mar-2010
818    ------------------------
820    1.  The Unicode data tables have been updated to Unicode 5.2.0.
822    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
823        configured.
825    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
826        original author of that file, following a query about its status.
828    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
829        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
831    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
832        quantifier applied to a forward-referencing subroutine call, could compile
833        incorrect code or give the error "internal error: previously-checked
834        referenced subpattern not found".
836    6.  Both MS Visual Studio and Symbian OS have problems with initializing
837        variables to point to external functions. For these systems, therefore,
838        pcre_malloc etc. are now initialized to local functions that call the
839        relevant global functions.
841    7.  There were two entries missing in the vectors called coptable and poptable
842        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
843        I've fixed the data, and added a kludgy way of testing at compile time that
844        the lengths are correct (equal to the number of opcodes).
846    8.  Following on from 7, I added a similar kludge to check the length of the
847        eint vector in pcreposix.c.
849    9.  Error texts for pcre_compile() are held as one long string to avoid too
850        much relocation at load time. To find a text, the string is searched,
851        counting zeros. There was no check for running off the end of the string,
852        which could happen if a new error number was added without updating the
853        string.
855    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
857    11. \K was not working if it appeared in an atomic group or in a group that
858        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
859        \K is "not well defined" if used in an assertion. PCRE now accepts it if
860        the assertion is positive, but not if it is negative.
862    12. Change 11 fortuitously reduced the size of the stack frame used in the
863        "match()" function of pcre_exec.c by one pointer. Forthcoming
864        implementation of support for (*MARK) will need an extra pointer on the
865        stack; I have reserved it now, so that the stack frame size does not
866        decrease.
868    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
869        item in branch that calls a recursion is a subroutine call - as in the
870        second branch in the above example - was incorrectly given the compile-
871        time error "recursive call could loop indefinitely" because pcre_compile()
872        was not correctly checking the subroutine for matching a non-empty string.
874    14. The checks for overrunning compiling workspace could trigger after an
875        overrun had occurred. This is a "should never occur" error, but it can be
876        triggered by pathological patterns such as hundreds of nested parentheses.
877        The checks now trigger 100 bytes before the end of the workspace.
879    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
882    Version 8.01 19-Jan-2010
883    ------------------------
885    1.  If a pattern contained a conditional subpattern with only one branch (in
886        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
887        computed the wrong minimum data length (which is of course zero for such
888        subpatterns). This could cause incorrect "no match" results.
890    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
891        the pattern is reset in the first branch, pcre_compile() failed with
892        "internal error: code overflow at offset...". This happened only when
893        the reset was to the original external option setting. (An optimization
894        abstracts leading options settings into an external setting, which was the
895        cause of this.)
897    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
898        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
899        assertion pattern did not match (meaning that the assertion was true), it
900        was incorrectly treated as false if the SKIP had been reached during the
901        matching. This also applied to assertions used as conditions.
903    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
904        assertion subpattern, including such a pattern used as a condition,
905        unpredictable results occurred, instead of the error return
908    5.  The C++ GlobalReplace function was not working like Perl for the special
909        situation when an empty string is matched. It now does the fancy magic
910        stuff that is necessary.
912    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
913        removed. (These were left over from very, very early versions of PCRE.)
915    7.  Some cosmetic changes to the code to make life easier when compiling it
916        as part of something else:
918        (a) Change DEBUG to PCRE_DEBUG.
920        (b) In pcre_compile(), rename the member of the "branch_chain" structure
921            called "current" as "current_branch", to prevent a collision with the
922            Linux macro when compiled as a kernel module.
924        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
925            prevent a collision with the Linux macro when compiled as a kernel
926            module.
928    8.  In pcre_compile() there are some checks for integer overflows that used to
929        cast potentially large values to (double). This has been changed to that
930        when building, a check for int64_t is made, and if it is found, it is used
931        instead, thus avoiding the use of floating point arithmetic. (There is no
932        other use of FP in PCRE.) If int64_t is not found, the fallback is to
933        double.
935    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
936        2005 (difference between two addresses compared to an unsigned value).
938    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
939        custom one, because of the following reported problem in Windows:
941          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
942              under Win32.
943          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
944              therefore missing the function definition.
945          - The compiler thus generates a "C" signature for the test function.
946          - The linker fails to find the "C" function.
947          - PCRE fails to configure if asked to do so against libbz2.
949    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
950        messages were output:
952          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
953          rerunning libtoolize, to keep the correct libtool macros in-tree.
954          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
956        I have done both of these things.
958    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
959        most of the time, it *can* run out if it is given a pattern that contains a
960        runaway infinite recursion. I updated the discussion in the pcrestack man
961        page.
963    13. Now that we have gone to the x.xx style of version numbers, the minor
964        version may start with zero. Using 08 or 09 is a bad idea because users
965        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
966        interpreted as invalid octal numbers. I've updated the previous comment in
967        configure.ac, and also added a check that gives an error if 08 or 09 are
968        used.
970    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
971        causing partial matching to fail when the end of the subject matched \W
972        in a UTF-8 pattern where \W was quantified with a minimum of 3.
974    15. There were some discrepancies between the declarations in pcre_internal.h
975        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
976        their definitions. The declarations used "const uschar *" and the
977        definitions used USPTR. Even though USPTR is normally defined as "const
978        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
979        reported that: "This difference in casting confuses some C++ compilers, for
980        example, SunCC recognizes above declarations as different functions and
981        generates broken code for hbpcre." I have changed the declarations to use
982        USPTR.
984    16. GNU libtool is named differently on some systems. The autogen.sh script now
985        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
986        (FreeBSD).
988    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
989        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
990        comment: "Figure out how to create a longlong from a string: strtoll and
991        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
992        instance, but it only takes 2 args instead of 3!"
994    18. A subtle bug concerned with back references has been fixed by a change of
995        specification, with a corresponding code fix. A pattern such as
996        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
997        refers, was giving matches when it shouldn't. For example, xa=xaaa would
998        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
999        same bug. Such groups have to be quantified to be useful, or contained
1000        inside another quantified group. (If there's no repetition, the reference
1001        can never match.) The problem arises because, having left the group and
1002        moved on to the rest of the pattern, a later failure that backtracks into
1003        the group uses the captured value from the final iteration of the group
1004        rather than the correct earlier one. I have fixed this in PCRE by forcing
1005        any group that contains a reference to itself to be an atomic group; that
1006        is, there cannot be any backtracking into it once it has completed. This is
1007        similar to recursive and subroutine calls.
1010    Version 8.00 19-Oct-09
1011    ----------------------
1013    1.  The table for translating pcre_compile() error codes into POSIX error codes
1014        was out-of-date, and there was no check on the pcre_compile() error code
1015        being within the table. This could lead to an OK return being given in
1016        error.
1018    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1019        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1020        in a Windows environment.
1022    3.  The pcregrep --count option prints the count for each file even when it is
1023        zero, as does GNU grep. However, pcregrep was also printing all files when
1024        --files-with-matches was added. Now, when both options are given, it prints
1025        counts only for those files that have at least one match. (GNU grep just
1026        prints the file name in this circumstance, but including the count seems
1027        more useful - otherwise, why use --count?) Also ensured that the
1028        combination -clh just lists non-zero counts, with no names.
1030    4.  The long form of the pcregrep -F option was incorrectly implemented as
1031        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1032        but it seems right to fix it, and I didn't think it was worth preserving
1033        the old behaviour.
1035    5.  The command line items --regex=pattern and --regexp=pattern were not
1036        recognized by pcregrep, which required --regex pattern or --regexp pattern
1037        (with a space rather than an '='). The man page documented the '=' forms,
1038        which are compatible with GNU grep; these now work.
1040    6.  No libpcreposix.pc file was created for pkg-config; there was just
1041        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1043    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1044        when UCP support is not needed, by modifying the Python script that
1045        generates it from Unicode data files. This should not matter if the module
1046        is correctly used as a library, but I received one complaint about 50K of
1047        unwanted data. My guess is that the person linked everything into his
1048        program rather than using a library. Anyway, it does no harm.
1050    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1051        was a minimum greater than 1 for a wide character in a possessive
1052        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1053        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1054        character. Chaos in the form of incorrect output or a compiling loop could
1055        result.
1057    9.  The restrictions on what a pattern can contain when partial matching is
1058        requested for pcre_exec() have been removed. All patterns can now be
1059        partially matched by this function. In addition, if there are at least two
1060        slots in the offset vector, the offset of the earliest inspected character
1061        for the match and the offset of the end of the subject are set in them when
1062        PCRE_ERROR_PARTIAL is returned.
1064    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1065        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1066        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1067        and may be more useful for multi-segment matching.
1069    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1070        used to be given if ever the end of the subject was reached; now it is
1071        given only if matching could not proceed because another character was
1072        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1073        string "Z", which now yields "no match" instead of "partial match". In the
1074        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1075        final character ended with (*FAIL).
1077    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1078        if the pattern had a "must contain" character that was already found in the
1079        earlier partial match, unless partial matching was again requested. For
1080        example, with the pattern /dog.(body)?/, the "must contain" character is
1081        "g". If the first part-match was for the string "dog", restarting with
1082        "sbody" failed. This bug has been fixed.
1084    13. The string returned by pcre_dfa_exec() after a partial match has been
1085        changed so that it starts at the first inspected character rather than the
1086        first character of the match. This makes a difference only if the pattern
1087        starts with a lookbehind assertion or \b or \B (\K is not supported by
1088        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1089        matching functions compatible, and I think it's the right thing to do.
1091    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1092        so that the demonstration program is easily available in environments where
1093        PCRE has not been installed from source.
1095    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1096        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1097        library.
1099    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1100        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1101        is not the first non-POSIX option to be added. Clearly some people find
1102        these options useful.
1104    17. If a caller to the POSIX matching function regexec() passes a non-zero
1105        value for nmatch with a NULL value for pmatch, the value of
1106        nmatch is forced to zero.
1108    18. RunGrepTest did not have a test for the availability of the -u option of
1109        the diff command, as RunTest does. It now checks in the same way as
1110        RunTest, and also checks for the -b option.
1112    19. If an odd number of negated classes containing just a single character
1113        interposed, within parentheses, between a forward reference to a named
1114        subpattern and the definition of the subpattern, compilation crashed with
1115        an internal error, complaining that it could not find the referenced
1116        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1117        [The bug was that it was starting one character too far in when skipping
1118        over the character class, thus treating the ] as data rather than
1119        terminating the class. This meant it could skip too much.]
1121    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1122        /g option in pcretest when the pattern contains \K, which makes it possible
1123        to have an empty string match not at the start, even when the pattern is
1124        anchored. Updated pcretest and pcredemo to use this option.
1126    21. If the maximum number of capturing subpatterns in a recursion was greater
1127        than the maximum at the outer level, the higher number was returned, but
1128        with unset values at the outer level. The correct (outer level) value is
1129        now given.
1131    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1132        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1133        make it do so. The string so far is captured, making this feature
1134        compatible with Perl.
1136    23. The tests have been re-organized, adding tests 11 and 12, to make it
1137        possible to check the Perl 5.10 features against Perl 5.10.
1139    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1140        pattern matches a fixed length string. PCRE did not allow this; now it
1141        does. Neither allows recursion.
1143    25. I finally figured out how to implement a request to provide the minimum
1144        length of subject string that was needed in order to match a given pattern.
1145        (It was back references and recursion that I had previously got hung up
1146        on.) This code has now been added to pcre_study(); it finds a lower bound
1147        to the length of subject needed. It is not necessarily the greatest lower
1148        bound, but using it to avoid searching strings that are too short does give
1149        some useful speed-ups. The value is available to calling programs via
1150        pcre_fullinfo().
1152    26. While implementing 25, I discovered to my embarrassment that pcretest had
1153        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1154        study optimizations had never been tested with that matching function.
1155        Oops. What is worse, even when it was passed study data, there was a bug in
1156        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1157        were also very few tests of studied patterns with pcre_dfa_exec().
1159    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1160        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1161        on the other side of the coin, they are no longer allowed to have different
1162        names, because these cannot be distinguished in PCRE, and this has caused
1163        confusion. (This is a difference from Perl.)
1165    28. When duplicate subpattern names are present (necessarily with different
1166        numbers, as required by 27 above), and a test is made by name in a
1167        conditional pattern, either for a subpattern having been matched, or for
1168        recursion in such a pattern, all the associated numbered subpatterns are
1169        tested, and the overall condition is true if the condition is true for any
1170        one of them. This is the way Perl works, and is also more like the way
1171        testing by number works.
1174    Version 7.9 11-Apr-09
1175    ---------------------
1177    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1178        (pcretest), all targets were linked against these libraries. This included
1179        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1180        libraries. This caused unwanted dependencies to be created. This problem
1181        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1182        pcretest is linked with readline.
1184    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1185        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1186        moved outside it again, because FALSE and TRUE are already defined in AIX,
1187        but BOOL is not.
1189    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1190        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1192    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1193        hyphens) following file names and line numbers when outputting matching
1194        lines. This is not true; no spaces are inserted. I have also clarified the
1195        wording for the --colour (or --color) option.
1197    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1198        was not coloured; this is different to GNU grep, so I have changed it to be
1199        the same.
1201    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1202        each matching line was coloured. Now it goes on to look for further matches
1203        of any of the test patterns, which is the same behaviour as GNU grep.
1205    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1206        doesn't make sense to accept an empty string match in pcregrep, so I have
1207        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1208        seems to be how GNU grep behaves.
1210    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1211        start or after a newline", because the conditional assertion was not being
1212        correctly handled. The rule now is that both the assertion and what follows
1213        in the first alternative must satisfy the test.
1215    9.  If auto-callout was enabled in a pattern with a conditional group whose
1216        condition was an assertion, PCRE could crash during matching, both with
1217        pcre_exec() and pcre_dfa_exec().
1219    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1220        used for matching.
1222    11. Unicode property support in character classes was not working for
1223        characters (bytes) greater than 127 when not in UTF-8 mode.
1225    12. Added the -M command line option to pcretest.
1227    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1229    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1231    16. Added comments and documentation about mis-use of no_arg in the C++
1232        wrapper.
1234    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1235        from Martin Jerabek that uses macro names for all relevant character and
1236        string constants.
1238    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1239        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1240        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1241        these, but not everybody uses configure.
1243    19. A conditional group that had only one branch was not being correctly
1244        recognized as an item that could match an empty string. This meant that an
1245        enclosing group might also not be so recognized, causing infinite looping
1246        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1247        with the subject "ab", where knowledge that the repeated group can match
1248        nothing is needed in order to break the loop.
1250    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1251        exec(), but without supplying a callout function, matching went wrong.
1253    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1254        leak if the size of the offset vector was greater than 30. When the vector
1255        is smaller, the saved offsets during recursion go onto a local stack
1256        vector, but for larger vectors malloc() is used. It was failing to free
1257        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1258        error, in fact).
1260    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1261        heapframe that is used only when UTF-8 support is enabled. This caused no
1262        problem, but was untidy.
1264    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1265        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1266        included within another project.
1268    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1269        slightly modified by me:
1271          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1272              not building pcregrep.
1274          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1275              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1277    25. Forward references, both numeric and by name, in patterns that made use of
1278        duplicate group numbers, could behave incorrectly or give incorrect errors,
1279        because when scanning forward to find the reference group, PCRE was not
1280        taking into account the duplicate group numbers. A pattern such as
1281        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1283    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1284        the feature of a custom pointer more persuasive (as requested by a user).
1286    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1287        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1288        pre-defined.
1290    28. Added support for (*UTF8) at the start of a pattern.
1292    29. Arrange for flags added by the "release type" setting in CMake to be shown
1293        in the configuration summary.
1296    Version 7.8 05-Sep-08
1297    ---------------------
1299    1.  Replaced UCP searching code with optimized version as implemented for Ad
1300        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1301        stage table and inline lookup instead of a function, giving speed ups of 2
1302        to 5 times on some simple patterns that I tested. Permission was given to
1303        distribute the MultiStage2.py script that generates the tables (it's not in
1304        the tarball, but is in the Subversion repository).
1306    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1307        scripts.
1309    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1310        a group with a zero qualifier. The result of the study could be incorrect,
1311        or the function might crash, depending on the pattern.
1313    4.  Caseless matching was not working for non-ASCII characters in back
1314        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1315        It now works when Unicode Property Support is available.
1317    5.  In pcretest, an escape such as \x{de} in the data was always generating
1318        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1319        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1320        truncation.
1322    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1324    7.  Added two (int) casts to pcregrep when printing the difference of two
1325        pointers, in case they are 64-bit values.
1327    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1328        test 2 if it fails.
1330    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1331        and a #define of that name to empty if it is not externally set. This is to
1332        allow users of MSVC to set it if necessary.
1334    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1335        the convenience functions in the pcre_get.c source file.
1337    11. An option change at the start of a pattern that had top-level alternatives
1338        could cause overwriting and/or a crash. This command provoked a crash in
1339        some environments:
1341          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1343        This potential security problem was recorded as CVE-2008-2371.
1345    12. For a pattern where the match had to start at the beginning or immediately
1346        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1347        pcre_dfa_exec() could read past the end of the passed subject if there was
1348        no match. To help with detecting such bugs (e.g. with valgrind), I modified
1349        pcretest so that it places the subject at the end of its malloc-ed buffer.
1351    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1352        exec() might read past the end of the data buffer in UTF-8 mode.
1354    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1355        the data contained the byte 0x85 as part of a UTF-8 character within its
1356        first line. This applied both to normal and DFA matching.
1358    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1359        /^[^d]*?$/8 failed to match "abc".
1361    16. Added a missing copyright notice to pcrecpp_internal.h.
1363    17. Make it more clear in the documentation that values returned from
1364        pcre_exec() in ovector are byte offsets, not character counts.
1366    18. Tidied a few places to stop certain compilers from issuing warnings.
1368    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1369        supplied by Stefan Weber. I made a further small update for 7.8 because
1370        there is a change of source arrangements: the pcre_searchfuncs.c module is
1371        replaced by pcre_ucd.c.
1374    Version 7.7 07-May-08
1375    ---------------------
1377    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
1378        a string to a long long, pretend we don't even have a long long." This is
1379        done by checking for the strtoq, strtoll, and _strtoi64 functions.
1381    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1382        pre-7.6 versions, which defined a global no_arg variable instead of putting
1383        it in the RE class. (See also #8 below.)
1385    3.  Remove a line of dead code, identified by coverity and reported by Nuno
1386        Lopes.
1388    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
1390        (1) The include/exclude patterns were being applied to the whole pathnames
1391            of files, instead of just to the final components.
1393        (2) If there was more than one level of directory, the subdirectories were
1394            skipped unless they satisfied the include/exclude conditions. This is
1395            inconsistent with GNU grep (and could even be seen as contrary to the
1396            pcregrep specification - which I improved to make it absolutely clear).
1397            The action now is always to scan all levels of directory, and just
1398            apply the include/exclude patterns to regular files.
1400    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1401        --exclude_dir in the tests to avoid scanning .svn directories.
1403    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
1404        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1405        doesn't support NULs in patterns.
1407    7.  Added some missing "const"s to declarations of static tables in
1408        pcre_compile.c and pcre_dfa_exec.c.
1410    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1411        caused by fix #2  above. (Subsequently also a second patch to fix the
1412        first patch. And a third patch - this was a messy problem.)
1414    9.  Applied Craig's patch to remove the use of push_back().
1416    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1417        matching function regexec().
1419    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1420        which, however, unlike Perl's \g{...}, are subroutine calls, not back
1421        references. PCRE supports relative numbers with this syntax (I don't think
1422        Oniguruma does).
1424    12. Previously, a group with a zero repeat such as (...){0} was completely
1425        omitted from the compiled regex. However, this means that if the group
1426        was called as a subroutine from elsewhere in the pattern, things went wrong
1427        (an internal error was given). Such groups are now left in the compiled
1428        pattern, with a new opcode that causes them to be skipped at execution
1429        time.
1431    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1432        to the way PCRE behaves:
1434        (a) A lone ] character is dis-allowed (Perl treats it as data).
1436        (b) A back reference to an unmatched subpattern matches an empty string
1437            (Perl fails the current match path).
1439        (c) A data ] in a character class must be notated as \] because if the
1440            first data character in a class is ], it defines an empty class. (In
1441            Perl it is not possible to have an empty class.) The empty class []
1442            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1443            The negative empty class [^] matches any one character, independently
1444            of the DOTALL setting.
1446    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1447        non-existent subpattern following a character class starting with ']' and
1448        containing () gave an internal compiling error instead of "reference to
1449        non-existent subpattern". Fortunately, when the pattern did exist, the
1450        compiled code was correct. (When scanning forwards to check for the
1451        existencd of the subpattern, it was treating the data ']' as terminating
1452        the class, so got the count wrong. When actually compiling, the reference
1453        was subsequently set up correctly.)
1455    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1456        it was being rejected as not supported by pcre_dfa_exec(), even though
1457        other assertions are supported. I have made pcre_dfa_exec() support
1458        (*FAIL).
1460    16. The implementation of 13c above involved the invention of a new opcode,
1461        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1462        cannot be changed at match time, I realized I could make a small
1463        improvement to matching performance by compiling OP_ALLANY instead of
1464        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1465        on the OP_ANY path.
1467    17. Compiling pcretest on Windows with readline support failed without the
1468        following two fixes: (1) Make the unistd.h include conditional on
1469        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1471    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1472        ncurses library to be included for pcretest when ReadLine support is
1473        requested, but also to allow for it to be overridden. This patch came from
1474        Daniel Bergström.
1476    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1477        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1478        any errors with the current Unicode tables. Thanks to Peter Kankowski for
1479        spotting this.
1482    Version 7.6 28-Jan-08
1483    ---------------------
1485    1.  A character class containing a very large number of characters with
1486        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1487        overflow.
1489    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
1490        HAVE_LONG_LONG is not defined.
1492    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
1493        bring them up to date and include new features. This patch includes:
1495        - Fixed PH's badly added libz and libbz2 support.
1496        - Fixed a problem with static linking.
1497        - Added pcredemo. [But later removed - see 7 below.]
1498        - Fixed dftables problem and added an option.
1499        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1500            HAVE_LONG_LONG.
1501        - Added readline support for pcretest.
1502        - Added an listing of the option settings after cmake has run.
1504    4.  A user submitted a patch to Makefile that makes it easy to create
1505        "pcre.dll" under mingw when using Configure/Make. I added stuff to
1506        Makefile.am that cause it to include this special target, without
1507        affecting anything else. Note that the same mingw target plus all
1508        the other distribution libraries and programs are now supported
1509        when configuring with CMake (see 6 below) instead of with
1510        Configure/Make.
1512    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1513        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1514        exported in the Windows port". It has not yet been confirmed that the patch
1515        solves the problem, but it does no harm.
1517    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1518        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1519        with CMake, and also correct the comment about stack recursion.
1521    7.  Remove the automatic building of pcredemo from the ./configure system and
1522        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1523        of a program that users should build themselves after PCRE is installed, so
1524        building it automatically is not really right. What is more, it gave
1525        trouble in some build environments.
1527    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
1530    Version 7.5 10-Jan-08
1531    ---------------------
1533    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
1534        values in parens when parsing an RE using the C++ wrapper."
1536    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
1537        Characters greater than 255 were excluded from the class instead of being
1538        included.
1540    3.  The same bug as (2) above applied to negated POSIX classes such as
1541        [:^space:].
1543    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
1544        defined or documented. It seems to have been a typo for PCRE_STATIC, so
1545        I have changed it.
1547    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
1548        first named subpattern) and a construct such as (?&a) would reference the
1549        first named subpattern whose name started with "a" (in other words, the
1550        length check was missing). Both these problems are fixed. "Subpattern name
1551        expected" is now given for (?&) (a zero-length name), and this patch also
1552        makes it give the same error for \k'' (previously it complained that that
1553        was a reference to a non-existent subpattern).
1555    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
1556        this is right because (?- can be followed by option settings as well as by
1557        digits. I have, however, made the messages clearer.
1559    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
1560        than the number used in the conditional) now cause a compile-time error.
1561        This is actually not compatible with Perl, which accepts such patterns, but
1562        treats the conditional as always being FALSE (as PCRE used to), but it
1563        seems to me that giving a diagnostic is better.
1565    8.  Change "alphameric" to the more common word "alphanumeric" in comments
1566        and messages.
1568    9.  Fix two occurrences of "backslash" in comments that should have been
1569        "backspace".
1571    10. Remove two redundant lines of code that can never be obeyed (their function
1572        was moved elsewhere).
1574    11. The program that makes PCRE's Unicode character property table had a bug
1575        which caused it to generate incorrect table entries for sequences of
1576        characters that have the same character type, but are in different scripts.
1577        It amalgamated them into a single range, with the script of the first of
1578        them. In other words, some characters were in the wrong script. There were
1579        thirteen such cases, affecting characters in the following ranges:
1581          U+002b0 - U+002c1
1582          U+0060c - U+0060d
1583          U+0061e - U+00612
1584          U+0064b - U+0065e
1585          U+0074d - U+0076d
1586          U+01800 - U+01805
1587          U+01d00 - U+01d77
1588          U+01d9b - U+01dbf
1589          U+0200b - U+0200f
1590          U+030fc - U+030fe
1591          U+03260 - U+0327f
1592          U+0fb46 - U+0fbb1
1593          U+10450 - U+1049d
1595    12. The -o option (show only the matching part of a line) for pcregrep was not
1596        compatible with GNU grep in that, if there was more than one match in a
1597        line, it showed only the first of them. It now behaves in the same way as
1598        GNU grep.
1600    13. If the -o and -v options were combined for pcregrep, it printed a blank
1601        line for every non-matching line. GNU grep prints nothing, and pcregrep now
1602        does the same. The return code can be used to tell if there were any
1603        non-matching lines.
1605    14. Added --file-offsets and --line-offsets to pcregrep.
1607    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
1608        infinitely looping recursion. The bug was that positive lookaheads were not
1609        being skipped when checking for a possible empty match (negative lookaheads
1610        and both kinds of lookbehind were skipped).
1612    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
1613        inclusion of <windows.h> to before rather than after the definition of
1614        INVALID_FILE_ATTRIBUTES (patch from David Byron).
1616    17. Specifying a possessive quantifier with a specific limit for a Unicode
1617        character property caused pcre_compile() to compile bad code, which led at
1618        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
1619        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
1620        caused the error; without that there was no problem.
1622    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
1624    19. Added --enable-pcretest-libreadline.
1626    20. In pcrecpp.cc, the variable 'count' was incremented twice in
1627        RE::GlobalReplace(). As a result, the number of replacements returned was
1628        double what it should be. I removed one of the increments, but Craig sent a
1629        later patch that removed the other one (the right fix) and added unit tests
1630        that check the return values (which was not done before).
1632    21. Several CMake things:
1634        (1) Arranged that, when cmake is used on Unix, the libraries end up with
1635            the names libpcre and libpcreposix, not just pcre and pcreposix.
1637        (2) The above change means that pcretest and pcregrep are now correctly
1638            linked with the newly-built libraries, not previously installed ones.
1642    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
1643        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
1644        UTF-8 newline character). The key issue is that the pattern starts .*;
1645        this means that the match must be either at the beginning, or after a
1646        newline. The bug was in the code for advancing after a failed match and
1647        checking that the new position followed a newline. It was not taking
1648        account of UTF-8 characters correctly.
1650    23. PCRE was behaving differently from Perl in the way it recognized POSIX
1651        character classes. PCRE was not treating the sequence [:...:] as a
1652        character class unless the ... were all letters. Perl, however, seems to
1653        allow any characters between [: and :], though of course it rejects as
1654        unknown any "names" that contain non-letters, because all the known class
1655        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
1656        for example, whereas PCRE did not - it did not recognize a POSIX character
1657        class. This seemed a bit dangerous, so the code has been changed to be
1658        closer to Perl. The behaviour is not identical to Perl, because PCRE will
1659        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
1660        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
1661        Perl does, and where it didn't before.
1663    24. Rewrite so as to remove the single use of %n from pcregrep because in some
1664        Windows environments %n is disabled by default.
1667    Version 7.4 21-Sep-07
1668    ---------------------
1670    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
1671        means that a class such as [\s] counted as "explicit reference to CR or
1672        LF". That isn't really right - the whole point of the change was to try to
1673        help when there was an actual mention of one of the two characters. So now
1674        the change happens only if \r or \n (or a literal CR or LF) character is
1675        encountered.
1677    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
1678        of both had grown to the point where there were only 3 bits left.
1679        Fortunately, there was spare space in the data structure, and so I have
1680        moved the internal flags into a new 16-bit field to free up more option
1681        bits.
1683    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
1684        but did not set the internal JCHANGED flag - either of these is enough to
1685        control the way the "get" function works - but the PCRE_INFO_JCHANGED
1686        facility is supposed to tell if (?J) was ever used, so now (?J) at the
1687        start sets both bits.
1689    4.  Added options (at build time, compile time, exec time) to change \R from
1690        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
1692    5.  doc/pcresyntax.html was missing from the distribution.
1694    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
1695        compatibility, even though it is no longer used.
1697    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
1698        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
1699        windows.h file is present (where different names are used). [This was
1700        reversed later after testing - see 16 below.]
1702    8.  Changed all #include <config.h> to #include "config.h". There were also
1703        some further <pcre.h> cases that I changed to "pcre.h".
1705    9.  When pcregrep was used with the --colour option, it missed the line ending
1706        sequence off the lines that it output.
1708    10. It was pointed out to me that arrays of string pointers cause lots of
1709        relocations when a shared library is dynamically loaded. A technique of
1710        using a single long string with a table of offsets can drastically reduce
1711        these. I have refactored PCRE in four places to do this. The result is
1712        dramatic:
1714          Originally:                          290
1715          After changing UCP table:            187
1716          After changing error message table:   43
1717          After changing table of "verbs"       36
1718          After changing table of Posix names   22
1720        Thanks to the folks working on Gregex for glib for this insight.
1722    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
1723        unicode-properties was also set.
1725    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
1727    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
1728        checked only for CRLF.
1730    14. Added casts to pcretest.c to avoid compiler warnings.
1732    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
1734    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
1735        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
1736        entirely. This removes changes made in 7 above.
1738    17. The CMake files have been updated, and there is now more information about
1739        building with CMake in the NON-UNIX-USE document.
1742    Version 7.3 28-Aug-07
1743    ---------------------
1745     1. In the rejigging of the build system that eventually resulted in 7.1, the
1746        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
1747        brackets there is not right, since it causes compilers to look for an
1748        installed pcre.h, not the version that is in the source that is being
1749        compiled (which of course may be different). I have changed it back to:
1751          #include "pcre.h"
1753        I have a vague recollection that the change was concerned with compiling in
1754        different directories, but in the new build system, that is taken care of
1755        by the VPATH setting the Makefile.
1757     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
1758        when the subject happened to end in the byte 0x85 (e.g. if the last
1759        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
1760        characters but of course it shouldn't be taken as a newline when it is part
1761        of another character. The bug was that, for an unlimited repeat of . in
1762        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
1763        characters when looking for a newline.
1765     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
1767     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
1768        in debug output.
1770     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
1771        long printing in the pcrecpp unittest when running under MinGW.
1773     6. ESC_K was left out of the EBCDIC table.
1775     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
1776        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
1777        limit also applies to "virtual nesting" when a pattern is recursive, and in
1778        this case 1000 isn't so big. I have been able to remove this limit at the
1779        expense of backing off one optimization in certain circumstances. Normally,
1780        when pcre_exec() would call its internal match() function recursively and
1781        immediately return the result unconditionally, it uses a "tail recursion"
1782        feature to save stack. However, when a subpattern that can match an empty
1783        string has an unlimited repetition quantifier, it no longer makes this
1784        optimization. That gives it a stack frame in which to save the data for
1785        checking that an empty string has been matched. Previously this was taken
1786        from the 1000-entry workspace that had been reserved. So now there is no
1787        explicit limit, but more stack is used.
1789     8. Applied Daniel's patches to solve problems with the import/export magic
1790        syntax that is required for Windows, and which was going wrong for the
1791        pcreposix and pcrecpp parts of the library. These were overlooked when this
1792        problem was solved for the main library.
1794     9. There were some crude static tests to avoid integer overflow when computing
1795        the size of patterns that contain repeated groups with explicit upper
1796        limits. As the maximum quantifier is 65535, the maximum group length was
1797        set at 30,000 so that the product of these two numbers did not overflow a
1798        32-bit integer. However, it turns out that people want to use groups that
1799        are longer than 30,000 bytes (though not repeat them that many times).
1800        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1801        made it possible to implement the integer overflow checks in a much more
1802        dynamic way, which I have now done. The artificial limitation on group
1803        length has been removed - we now have only the limit on the total length of
1804        the compiled pattern, which depends on the LINK_SIZE setting.
1806    10. Fixed a bug in the documentation for get/copy named substring when
1807        duplicate names are permitted. If none of the named substrings are set, the
1808        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1809        empty string.
1811    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1812        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1813        because the ] is interpreted as the first data character and the
1814        terminating ] is not found. PCRE has been made compatible with Perl in this
1815        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1816        cause memory overwriting.
1818    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1819        string has been matched (to stop an infinite loop). It was not recognizing
1820        a conditional subpattern that could match an empty string if that
1821        subpattern was within another subpattern. For example, it looped when
1822        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1823        condition was not nested. This bug has been fixed.
1825    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1826        past the start of the subject in the presence of bytes with the top bit
1827        set, for example "\x8aBCD".
1829    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1830        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1832    14. Optimized (?!) to (*FAIL).
1834    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1835        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1836        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1837        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1838        does: it's just the validity check that is more restrictive.
1840    16. Inserted checks for integer overflows during escape sequence (backslash)
1841        processing, and also fixed erroneous offset values for syntax errors during
1842        backslash processing.
1844    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1845        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1847    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1848        caused an overrun.
1850    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1851        something other than just ASCII characters) inside a group that had an
1852        unlimited repeat caused a loop at compile time (while checking to see
1853        whether the group could match an empty string).
1855    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1856        [\P{Any}] did so. (Error in the code for printing property names.)
1858    21. An orphan \E inside a character class could cause a crash.
1860    22. A repeated capturing bracket such as (A)? could cause a wild memory
1861        reference during compilation.
1863    23. There are several functions in pcre_compile() that scan along a compiled
1864        expression for various reasons (e.g. to see if it's fixed length for look
1865        behind). There were bugs in these functions when a repeated \p or \P was
1866        present in the pattern. These operators have additional parameters compared
1867        with \d, etc, and these were not being taken into account when moving along
1868        the compiled data. Specifically:
1870        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1871            length.
1873        (b) An item such as \pL+ within a repeated group could cause crashes or
1874            loops.
1876        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1877            "reference to non-existent subpattern" error.
1879        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1881    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1882        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1884    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1886    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1887        character were causing crashes (broken optimization).
1889    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1890        \p or \P) caused a compile-time loop.
1892    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1893        break. For example, the unstudied pattern [\r\n]A does not match the string
1894        "\r\nA" because change 7.0/46 below moves the current point on by two
1895        characters after failing to match at the start. However, the pattern \nA
1896        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1897        the same is true. There doesn't seem any very clean way out of this, but
1898        what I have chosen to do makes the common cases work: PCRE now takes note
1899        of whether there can be an explicit match for \r or \n anywhere in the
1900        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1901        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1902        pattern has explicit CR or LF references.
1904    29. Added (*CR) etc for changing newline setting at start of pattern.
1907    Version 7.2 19-Jun-07
1908    ---------------------
1910     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1911        which is apparently normally available under Windows.
1913     2. Re-jig the pcregrep tests with different newline settings in an attempt
1914        to make them independent of the local environment's newline setting.
1916     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1918     4. Some of the "internals" tests were previously cut out when the link size
1919        was not 2, because the output contained actual offsets. The recent new
1920        "Z" feature of pcretest means that these can be cut out, making the tests
1921        usable with all link sizes.
1923     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1924        stack recursion. This gives a massive performance boost under BSD, but just
1925        a small improvement under Linux. However, it saves one field in the frame
1926        in all cases.
1928     6. Added more features from the forthcoming Perl 5.10:
1930        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1931            recursion call. It refers to the nth most recently opened parentheses.
1933        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1934            to be opened parentheses.
1936        (c) Conditions that refer to capturing parentheses can be specified
1937            relatively, for example, (?(-2)... or (?(+3)...
1939        (d) \K resets the start of the current match so that everything before
1940            is not part of it.
1942        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1944        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1945            reference syntax.
1947        (g) (?| introduces a group in which the numbering of parentheses in each
1948            alternative starts with the same number.
1950        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1952     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1955     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1956        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1957        for detecting groups that can match an empty string.
1959     9. A pattern with a very large number of alternatives (more than several
1960        hundred) was running out of internal workspace during the pre-compile
1961        phase, where pcre_compile() figures out how much memory will be needed. A
1962        bit of new cunning has reduced the workspace needed for groups with
1963        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1964        workspace instead of running out of the 4096 that are available.
1966    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1968    11. Applied patch from Google to remove an optimization that didn't quite work.
1969        The report of the bug said:
1971          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1972          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1973          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1975    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1976        it matched the wrong number of bytes.
1979    Version 7.1 24-Apr-07
1980    ---------------------
1982     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1983        that is more "standard", making use of automake and other Autotools. There
1984        is some re-arrangement of the files and adjustment of comments consequent
1985        on this.
1987     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1988        for recursive directory scanning broke on some systems because the files
1989        are not scanned in any specific order and on different systems the order
1990        was different. A call to "sort" has been inserted into RunGrepTest for the
1991        approprate test as a short-term fix. In the longer term there may be an
1992        alternative.
1994     3. I had an email from Eric Raymond about problems translating some of PCRE's
1995        man pages to HTML (despite the fact that I distribute HTML pages, some
1996        people do their own conversions for various reasons). The problems
1997        concerned the use of low-level troff macros .br and .in. I have therefore
1998        removed all such uses from the man pages (some were redundant, some could
1999        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2000        HTML has been updated to handle .nf/.fi and to complain if it encounters
2001        .br or .in.
2003     4. Updated comments in configure.ac that get placed in config.h.in and also
2004        arranged for config.h to be included in the distribution, with the name
2005        config.h.generic, for the benefit of those who have to compile without
2006        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2008     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2009        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2010        makevp.bat for the current PCRE, using the additional files
2011        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2013     6. A Windows user reported a minor discrepancy with test 2, which turned out
2014        to be caused by a trailing space on an input line that had got lost in his
2015        copy. The trailing space was an accident, so I've just removed it.
2017     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2018        that is needed.
2020     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2021        as "const" (a) because they are and (b) because it helps the PHP
2022        maintainers who have recently made a script to detect big data structures
2023        in the php code that should be moved to the .rodata section. I remembered
2024        to update Builducptable as well, so it won't revert if ucptable.h is ever
2025        re-created.
2027     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2028        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2029        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2030        support is not required. This saves 1.5-2K of code, which is important in
2031        some applications.
2033        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2034        so as not to refer to the tables, even though these functions will never be
2035        called when UTF-8 support is disabled. Otherwise there are problems with a
2036        shared library.
2038    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2040        (a) It was defining its arguments as char * instead of void *.
2042        (b) It was assuming that all moves were upwards in memory; this was true
2043            a long time ago when I wrote it, but is no longer the case.
2045        The emulated memove() is provided for those environments that have neither
2046        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2047        is clearly not the case, as these two bugs were recently reported.
2049    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2050        and Detrail to create the HTML documentation, the .txt form of the man
2051        pages, and it removes trailing spaces from listed files. It also creates
2052        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2053        case, it wraps all the #defines with #ifndefs. This script should be run
2054        before "make dist".
2056    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2057        with Unicode property support.
2059        (a) For a maximizing quantifier, if the two different cases of the
2060            character were of different lengths in their UTF-8 codings (there are
2061            some cases like this - I found 11), and the matching function had to
2062            back up over a mixture of the two cases, it incorrectly assumed they
2063            were both the same length.
2065        (b) When PCRE was configured to use the heap rather than the stack for
2066            recursion during matching, it was not correctly preserving the data for
2067            the other case of a UTF-8 character when checking ahead for a match
2068            while processing a minimizing repeat. If the check also involved
2069            matching a wide character, but failed, corruption could cause an
2070            erroneous result when trying to check for a repeat of the original
2071            character.
2073    13. Some tidying changes to the testing mechanism:
2075        (a) The RunTest script now detects the internal link size and whether there
2076            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2077            values substituted by "configure". (The RunGrepTest script already did
2078            this for UTF-8.) The configure.ac script no longer substitutes the
2079            relevant variables.
2081        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2082            with length and offset values. This means that the output is different
2083            for different internal link sizes. Test 2 is skipped for link sizes
2084            other than 2 because of this, bypassing the problem. Unfortunately,
2085            there was also a test in test 3 (the locale tests) that used /B and
2086            failed for link sizes other than 2. Rather than cut the whole test out,
2087            I have added a new /Z option to pcretest that replaces the length and
2088            offset values with spaces. This is now used to make test 3 independent
2089            of link size. (Test 2 will be tidied up later.)
2091    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2092        segmentation fault instead of returning the appropriate error message.
2094    15. In multiline mode when the newline sequence was set to "any", the pattern
2095        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2096        This doesn't seem right; it now treats the CRLF combination as the line
2097        ending, and so does not match in that case. It's only a pattern such as ^$
2098        that would hit this one: something like ^ABC$ would have failed after \r
2099        and then tried again after \r\n.
2101    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2102        in an attempt to make files that differ only in their line terminators
2103        compare equal. This works on Linux.
2105    17. Under certain error circumstances pcregrep might try to free random memory
2106        as it exited. This is now fixed, thanks to valgrind.
2108    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2109        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2110        was because its rules for how to advance for /g after matching an empty
2111        string at the end of a line did not allow for this case. They now check for
2112        it specially.
2114    20. pcretest is supposed to handle patterns and data of any length, by
2115        extending its buffers when necessary. It was getting this wrong when the
2116        buffer for a data line had to be extended.
2118    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2119        CRLF as a newline sequence.
2121    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2122        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2123        I have nevertheless tidied it up.
2125    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2127    24. Added a man page for pcre-config.
2130    Version 7.0 19-Dec-06
2131    ---------------------
2133     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2134        moving to gcc 4.1.1.
2136     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2137        sys/time.h, which is documented as needed for this function. It doesn't
2138        seem to matter on Linux, but it showed up on some releases of OS X.
2140     3. It seems that there are systems where bytes whose values are greater than
2141        127 match isprint() in the "C" locale. The "C" locale should be the
2142        default when a C program starts up. In most systems, only ASCII printing
2143        characters match isprint(). This difference caused the output from pcretest
2144        to vary, making some of the tests fail. I have changed pcretest so that:
2146        (a) When it is outputting text in the compiled version of a pattern, bytes
2147            other than 32-126 are always shown as hex escapes.
2149        (b) When it is outputting text that is a matched part of a subject string,
2150            it does the same, unless a different locale has been set for the match
2151            (using the /L modifier). In this case, it uses isprint() to decide.
2153     4. Fixed a major bug that caused incorrect computation of the amount of memory
2154        required for a compiled pattern when options that changed within the
2155        pattern affected the logic of the preliminary scan that determines the
2156        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2157        that the computed length was too small. The symptoms of this bug were
2158        either the PCRE error "internal error: code overflow" from pcre_compile(),
2159        or a glibc crash with a message such as "pcretest: free(): invalid next
2160        size (fast)". Examples of patterns that provoked this bug (shown in
2161        pcretest format) are:
2163          /(?-x: )/x
2164          /(?x)(?-x: \s*#\s*)/
2165          /((?i)[\x{c0}])/8
2166          /(?i:[\x{c0}])/8
2168        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2169        is now done differently.
2171     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2172        wrapper classes; (b) implement a new function in the C++ scanner that is
2173        more efficient than the old way of doing things because it avoids levels of
2174        recursion in the regex matching; (c) add a paragraph to the documentation
2175        for the FullMatch() function.
2177     6. The escape sequence \n was being treated as whatever was defined as
2178        "newline". Not only was this contrary to the documentation, which states
2179        that \n is character 10 (hex 0A), but it also went horribly wrong when
2180        "newline" was defined as CRLF. This has been fixed.
2182     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2183        was being set to -1 for the "end of line" case (supposedly a value that no
2184        character can have). Though this value is never used (the check for end of
2185        line is "zero bytes in current character"), it caused compiler complaints.
2186        I've changed it to 0xffffffff.
2188     8. In pcre_version.c, the version string was being built by a sequence of
2189        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2190        string (as it is for production releases) called a macro with an empty
2191        argument. The C standard says the result of this is undefined. The gcc
2192        compiler treats it as an empty string (which was what was wanted) but it is
2193        reported that Visual C gives an error. The source has been hacked around to
2194        avoid this problem.
2196     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2197        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2198        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2199        of them did).
2201    10. Originally, pcretest opened its input and output without "b"; then I was
2202        told that "b" was needed in some environments, so it was added for release
2203        5.0 to both the input and output. (It makes no difference on Unix-like
2204        systems.) Later I was told that it is wrong for the input on Windows. I've
2205        now abstracted the modes into two macros, to make it easier to fiddle with
2206        them, and removed "b" from the input mode under Windows.
2208    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2210    12. Added -help and --help to pcretest as an official way of being reminded
2211        of the options.
2213    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2214        and pcrecpp.cc because they annoy compilers at high warning levels.
2216    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2218    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2219        scripts are not C programs :-) and which was not noticed because it works
2220        on Linux.
2222    16. pcretest is supposed to handle any length of pattern and data line (as one
2223        line or as a continued sequence of lines) by extending its input buffer if
2224        necessary. This feature was broken for very long pattern lines, leading to
2225        a string of junk being passed to pcre_compile() if the pattern was longer
2226        than about 50K.
2228    17. I have done a major re-factoring of the way pcre_compile() computes the
2229        amount of memory needed for a compiled pattern. Previously, there was code
2230        that made a preliminary scan of the pattern in order to do this. That was
2231        OK when PCRE was new, but as the facilities have expanded, it has become
2232        harder and harder to keep it in step with the real compile phase, and there
2233        have been a number of bugs (see for example, 4 above). I have now found a
2234        cunning way of running the real compile function in a "fake" mode that
2235        enables it to compute how much memory it would need, while actually only
2236        ever using a few hundred bytes of working memory and without too many
2237        tests of the mode. This should make future maintenance and development
2238        easier. A side effect of this work is that the limit of 200 on the nesting
2239        depth of parentheses has been removed (though this was never a serious
2240        limitation, I suspect). However, there is a downside: pcre_compile() now
2241        runs more slowly than before (30% or more, depending on the pattern). I
2242        hope this isn't a big issue. There is no effect on runtime performance.
2244    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2245        newline (only possible for the last line of a file) and it was a
2246        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2248    19. Added additional timing features to pcretest. (1) The -tm option now times
2249        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2250        separate command line item, by a number that specifies the number of
2251        repeats to use when timing. The default is 50000; this gives better
2252        precision, but takes uncomfortably long for very large patterns.
2254    20. Extended pcre_study() to be more clever in cases where a branch of a
2255        subpattern has no definite first character. For example, (a*|b*)[cd] would
2256        previously give no result from pcre_study(). Now it recognizes that the
2257        first character must be a, b, c, or d.
2259    21. There was an incorrect error "recursive call could loop indefinitely" if
2260        a subpattern (or the entire pattern) that was being tested for matching an
2261        empty string contained only one non-empty item after a nested subpattern.
2262        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2263        incorrectly, because the \d was being skipped in the check.
2265    22. The pcretest program now has a new pattern option /B and a command line
2266        option -b, which is equivalent to adding /B to every pattern. This causes
2267        it to show the compiled bytecode, without the additional information that
2268        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2269        is the same as /B/I).
2271    23. A new optimization is now able automatically to treat some sequences such
2272        as a*b as a*+b. More specifically, if something simple (such as a character
2273        or a simple class like \d) has an unlimited quantifier, and is followed by
2274        something that cannot possibly match the quantified thing, the quantifier
2275        is automatically "possessified".
2277    24. A recursive reference to a subpattern whose number was greater than 39
2278        went wrong under certain circumstances in UTF-8 mode. This bug could also
2279        have affected the operation of pcre_study().
2281    25. Realized that a little bit of performance could be had by replacing
2282        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2284    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2286    27. Possessive quantifiers such as a++ were previously implemented by turning
2287        them into atomic groups such as ($>a+). Now they have their own opcodes,
2288        which improves performance. This includes the automatically created ones
2289        from 23 above.
2291    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2292        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2293        the first matched character to be a colon. This applied both to named and
2294        numbered groups.
2296    29. The ucpinternal.h header file was missing its idempotency #ifdef.
2298    30. I was sent a "project" file called libpcre.a.dev which I understand makes
2299        building PCRE on Windows easier, so I have included it in the distribution.
2301    31. There is now a check in pcretest against a ridiculously large number being
2302        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2303        loop, the loop is abandoned.
2305    32. Forward references to subpatterns in conditions such as (?(2)...) where
2306        subpattern 2 is defined later cause pcre_compile() to search forwards in
2307        the pattern for the relevant set of parentheses. This search went wrong
2308        when there were unescaped parentheses in a character class, parentheses
2309        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2311    33. "Subroutine" calls and backreferences were previously restricted to
2312        referencing subpatterns earlier in the regex. This restriction has now
2313        been removed.
2315    34. Added a number of extra features that are going to be in Perl 5.10. On the
2316        whole, these are just syntactic alternatives for features that PCRE had
2317        previously implemented using the Python syntax or my own invention. The
2318        other formats are all retained for compatibility.
2320        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2321            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2322            also .NET compatible.
2324        (b) A recursion or subroutine call to a named group can now be defined as
2325            (?&name) as well as (?P>name).
2327        (c) A backreference to a named group can now be defined as \k<name> or
2328            \k'name' as well as (?P=name). The new forms, as well as being in Perl
2329            5.10, are also .NET compatible.
2331        (d) A conditional reference to a named group can now use the syntax
2332            (?(<name>) or (?('name') as well as (?(name).
2334        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2335            groups (named and numbered) that are never evaluated inline, but can be
2336            called as "subroutines" from elsewhere. In effect, the DEFINE condition
2337            is always false. There may be only one alternative in such a group.
2339        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2340            as the simple (?(R). The condition is true only if the most recent
2341            recursion is that of the given number or name. It does not search out
2342            through the entire recursion stack.
2344        (g) The escape \gN or \g{N} has been added, where N is a positive or
2345            negative number, specifying an absolute or relative reference.
2347    35. Tidied to get rid of some further signed/unsigned compiler warnings and
2348        some "unreachable code" warnings.
2350    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2351        things, this adds five new scripts.
2353    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2354        There were also incompatibilities regarding the handling of \Q..\E inside
2355        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2356        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2358    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2359        matches an empty string, and forcibly breaks the loop. There were bugs in
2360        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
2361        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
2362        separate and independent bugs (that affected different cases) have been
2363        fixed.
2365    39. Refactored the code to abolish the use of different opcodes for small
2366        capturing bracket numbers. This is a tidy that I avoided doing when I
2367        removed the limit on the number of capturing brackets for 3.5 back in 2001.
2368        The new approach is not only tidier, it makes it possible to reduce the
2369        memory needed to fix the previous bug (38).
2371    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2372        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2373        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2374        mode.
2376    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2377        report.
2379    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2380        copy construction and assignment in the C++ wrapper.
2382    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2383        couple of bugs that could have given wrong results in the "--newline=crlf"
2384        case.
2386    44. Added a number of casts and did some reorganization of signed/unsigned int
2387        variables following suggestions from Dair Grant. Also renamed the variable
2388        "this" as "item" because it is a C++ keyword.
2390    45. Arranged for dftables to add
2392          #include "pcre_internal.h"
2394        to pcre_chartables.c because without it, gcc 4.x may remove the array
2395        definition from the final binary if PCRE is built into a static library and
2396        dead code stripping is activated.
2398    46. For an unanchored pattern, if a match attempt fails at the start of a
2399        newline sequence, and the newline setting is CRLF or ANY, and the next two
2400        characters are CRLF, advance by two characters instead of one.
2403    Version 6.7 04-Jul-06
2404    ---------------------
2406     1. In order to handle tests when input lines are enormously long, pcretest has
2407        been re-factored so that it automatically extends its buffers when
2408        necessary. The code is crude, but this _is_ just a test program. The
2409        default size has been increased from 32K to 50K.
2411     2. The code in pcre_study() was using the value of the re argument before
2412        testing it for NULL. (Of course, in any sensible call of the function, it
2413        won't be NULL.)
2415     3. The memmove() emulation function in pcre_internal.h, which is used on
2416        systems that lack both memmove() and bcopy() - that is, hardly ever -
2417        was missing a "static" storage class specifier.
2419     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2420        containing an extended class (one that cannot be represented by a bitmap
2421        because it contains high-valued characters or Unicode property items, e.g.
2422        [\pZ]). Almost always one would set UTF-8 mode when processing such a
2423        pattern, but PCRE should not loop if you do not (it no longer does).
2424        [Detail: two cases were found: (a) a repeated subpattern containing an
2425        extended class; (b) a recursive reference to a subpattern that followed a
2426        previous extended class. It wasn't skipping over the extended class
2427        correctly when UTF-8 mode was not set.]
2429     5. A negated single-character class was not being recognized as fixed-length
2430        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2431        compile error "lookbehind assertion is not fixed length".
2433     6. The RunPerlTest auxiliary script was showing an unexpected difference
2434        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2435        write a Perl script that can interpret lines of an input file either as
2436        byte characters or as UTF-8, which is what "perltest" was being required to
2437        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2438        can't do is switch easily at run time between having the "use utf8;" pragma
2439        or not. In the end, I fudged it by using the RunPerlTest script to insert
2440        "use utf8;" explicitly for the UTF-8 tests.
2442     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2443        the end of the subject string, contrary to the documentation and to what
2444        Perl does. This was true of both matching functions. Now it matches only at
2445        the start of the subject and immediately after *internal* newlines.
2447     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2448        a pointer to an int instead of a pointer to an unsigned long int. This
2449        caused problems on 64-bit systems.
2451     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2452        instance of the 'standard' template library not being so standard".
2454    10. There was no check on the number of named subpatterns nor the maximum
2455        length of a subpattern name. The product of these values is used to compute
2456        the size of the memory block for a compiled pattern. By supplying a very
2457        long subpattern name and a large number of named subpatterns, the size
2458        computation could be caused to overflow. This is now prevented by limiting
2459        the length of names to 32 characters, and the number of named subpatterns
2460        to 10,000.
2462    11. Subpatterns that are repeated with specific counts have to be replicated in
2463        the compiled pattern. The size of memory for this was computed from the
2464        length of the subpattern and the repeat count. The latter is limited to
2465        65535, but there was no limit on the former, meaning that integer overflow
2466        could in principle occur. The compiled length of a repeated subpattern is
2467        now limited to 30,000 bytes in order to prevent this.
2469    12. Added the optional facility to have named substrings with the same name.
2471    13. Added the ability to use a named substring as a condition, using the
2472        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2473        are numbers (not recommended). Forward references are permitted.
2475    14. Added forward references in named backreferences (if you see what I mean).
2477    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2478        pattern could run off the end of the subject. For example, the pattern
2479        "(?s)(.{1,5})"8 did this with the subject "ab".
2481    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2482        PCRE_CASELESS was set when matching characters that were quantified with ?
2483        or *.
2485    17. A character class other than a single negated character that had a minimum
2486        but no maximum quantifier - for example [ab]{6,} - was not handled
2487        correctly by pce_dfa_exec(). It would match only one character.
2489    18. A valid (though odd) pattern that looked like a POSIX character
2490        class but used an invalid character after [ (for example [[,abc,]]) caused
2491        pcre_compile() to give the error "Failed: internal error: code overflow" or
2492        in some cases to crash with a glibc free() error. This could even happen if
2493        the pattern terminated after [[ but there just happened to be a sequence of
2494        letters, a binary zero, and a closing ] in the memory that followed.
2496    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2497        over the years. Originally (before any Unicode support), just the bottom 8
2498        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2499        output from "man perlunicode" includes this:
2501          The regular expression compiler produces polymorphic opcodes.  That
2502          is, the pattern adapts to the data and automatically switches to
2503          the Unicode character scheme when presented with Unicode data--or
2504          instead uses a traditional byte scheme when presented with byte
2505          data.
2507        Sadly, a wide octal escape does not cause a switch, and in a string with
2508        no other multibyte characters, these octal escapes are treated as before.
2509        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
2510        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2511        Unicode string.
2513        I have not perpetrated such confusion in PCRE. Up till now, it took just
2514        the bottom 8 bits, as in old Perl. I have now made octal escapes with
2515        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2516        translate to the appropriate multibyte character.
2518    29. Applied some refactoring to reduce the number of warnings from Microsoft
2519        and Borland compilers. This has included removing the fudge introduced
2520        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2521        a warning about an unused variable.
2523    21. PCRE has not included VT (character 0x0b) in the set of whitespace
2524        characters since release 4.0, because Perl (from release 5.004) does not.
2525        [Or at least, is documented not to: some releases seem to be in conflict
2526        with the documentation.] However, when a pattern was studied with
2527        pcre_study() and all its branches started with \s, PCRE still included VT
2528        as a possible starting character. Of course, this did no harm; it just
2529        caused an unnecessary match attempt.
2531    22. Removed a now-redundant internal flag bit that recorded the fact that case
2532        dependency changed within the pattern. This was once needed for "required
2533        byte" processing, but is no longer used. This recovers a now-scarce options
2534        bit. Also moved the least significant internal flag bit to the most-
2535        significant bit of the word, which was not previously used (hangover from
2536        the days when it was an int rather than a uint) to free up another bit for
2537        the future.
2539    23. Added support for CRLF line endings as well as CR and LF. As well as the
2540        default being selectable at build time, it can now be changed at runtime
2541        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
2542        specify that it is scanning data with non-default line endings.
2544    24. Changed the definition of CXXLINK to make it agree with the definition of
2545        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
2547    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
2548        recursions. This makes a big different to stack usage for some patterns.
2550    26. If a subpattern containing a named recursion or subroutine reference such
2551        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
2552        the space required for the compiled pattern went wrong and gave too small a
2553        value. Depending on the environment, this could lead to "Failed: internal
2554        error: code overflow at offset 49" or "glibc detected double free or
2555        corruption" errors.
2557    27. Applied patches from Google (a) to support the new newline modes and (b) to
2558        advance over multibyte UTF-8 characters in GlobalReplace.
2560    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
2561        difference for some implementation of PCRE in some Windows version.
2563    29. Added some extra testing facilities to pcretest:
2565        \q<number>   in a data line sets the "match limit" value
2566        \Q<number>   in a data line sets the "match recursion limt" value
2567        -S <number>  sets the stack size, where <number> is in megabytes
2569        The -S option isn't available for Windows.
2572    Version 6.6 06-Feb-06
2573    ---------------------
2575     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
2576        in pcreposix.h. I have copied the definition from pcre.h.
2578     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
2579        because pcre.h is no longer a built file.
2581     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
2582        not normally included in the compiled code.
2585    Version 6.5 01-Feb-06
2586    ---------------------
2588     1. When using the partial match feature with pcre_dfa_exec(), it was not
2589        anchoring the second and subsequent partial matches at the new starting
2590        point. This could lead to incorrect results. For example, with the pattern
2591        /1234/, partially matching against "123" and then "a4" gave a match.
2593     2. Changes to pcregrep:
2595        (a) All non-match returns from pcre_exec() were being treated as failures
2596            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
2597            error message is output. Some extra information is given for the
2598            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
2599            probably the only errors that are likely to be caused by users (by
2600            specifying a regex that has nested indefinite repeats, for instance).
2601            If there are more than 20 of these errors, pcregrep is abandoned.
2603        (b) A binary zero was treated as data while matching, but terminated the
2604            output line if it was written out. This has been fixed: binary zeroes
2605            are now no different to any other data bytes.
2607        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
2608            used to set a locale for matching. The --locale=xxxx long option has
2609            been added (no short equivalent) to specify a locale explicitly on the
2610            pcregrep command, overriding the environment variables.
2612        (d) When -B was used with -n, some line numbers in the output were one less
2613            than they should have been.
2615        (e) Added the -o (--only-matching) option.
2617        (f) If -A or -C was used with -c (count only), some lines of context were
2618            accidentally printed for the final match.
2620        (g) Added the -H (--with-filename) option.
2622        (h) The combination of options -rh failed to suppress file names for files
2623            that were found from directory arguments.
2625        (i) Added the -D (--devices) and -d (--directories) options.
2627        (j) Added the -F (--fixed-strings) option.
2629        (k) Allow "-" to be used as a file name for -f as well as for a data file.
2631        (l) Added the --colo(u)r option.
2633        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
2634            is not present by default.
2636     3. A nasty bug was discovered in the handling of recursive patterns, that is,
2637        items such as (?R) or (?1), when the recursion could match a number of
2638        alternatives. If it matched one of the alternatives, but subsequently,
2639        outside the recursion, there was a failure, the code tried to back up into
2640        the recursion. However, because of the way PCRE is implemented, this is not
2641        possible, and the result was an incorrect result from the match.
2643        In order to prevent this happening, the specification of recursion has
2644        been changed so that all such subpatterns are automatically treated as
2645        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
2647     4. I had overlooked the fact that, in some locales, there are characters for
2648        which isalpha() is true but neither isupper() nor islower() are true. In
2649        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
2650        and ordfeminine) are like this. This affected the treatment of \w and \W
2651        when they appeared in character classes, but not when they appeared outside
2652        a character class. The bit map for "word" characters is now created
2653        separately from the results of isalnum() instead of just taking it from the
2654        upper, lower, and digit maps. (Plus the underscore character, of course.)
2656     5. The above bug also affected the handling of POSIX character classes such as
2657        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
2658        permanent tables. Instead, the bit maps for such a class were previously
2659        created as the appropriate unions of the upper, lower, and digit bitmaps.
2660        Now they are created by subtraction from the [[:word:]] class, which has
2661        its own bitmap.
2663     6. The [[:blank:]] character class matches horizontal, but not vertical space.
2664        It is created by subtracting the vertical space characters (\x09, \x0a,
2665        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
2666        subtraction was done in the overall bitmap for a character class, meaning
2667        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
2668        be recognized. This bug has been fixed.
2670     7. Patches from the folks at Google:
2672          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
2673          real life, but is still worth protecting against".
2675          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
2676          regular expressions".
2678          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
2679          have it.
2681          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
2682          "configure" and the latter not, in order to fix a problem somebody had
2683          with compiling the Arg class on HP-UX.
2685          (e) Improve the error-handling of the C++ wrapper a little bit.
2687          (f) New tests for checking recursion limiting.
2689     8. The pcre_memmove() function, which is used only if the environment does not
2690        have a standard memmove() function (and is therefore rarely compiled),
2691        contained two bugs: (a) use of int instead of size_t, and (b) it was not
2692        returning a result (though PCRE never actually uses the result).
2694     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
2695        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
2696        returned instead of calling malloc() with an overflowing number that would
2697        most likely cause subsequent chaos.
2699    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
2701    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
2702        with this option is matched, the nmatch and pmatch options of regexec() are
2703        ignored.
2705    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
2706        provided in case anyone wants to the the POSIX interface with UTF-8
2707        strings.
2709    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
2710        C++ linking (needed for some HP-UX environments).
2712    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
2713        (unused parameter) and in the pcre_printint() function (omitted "default"
2714        switch label when the default is to do nothing).
2716    15. Added some code to make it possible, when PCRE is compiled as a C++
2717        library, to replace subject pointers for pcre_exec() with a smart pointer
2718        class, thus making it possible to process discontinuous strings.
2720    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
2721        much the same function. They were added by different people who were trying
2722        to make PCRE easy to compile on non-Unix systems. It has been suggested
2723        that PCRE_EXPORT be abolished now that there is more automatic apparatus
2724        for compiling on Windows systems. I have therefore replaced it with
2725        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
2726        defaults to "extern" for C or "extern C" for C++, which works fine on
2727        Unix-like systems. It is now possible to override the value of PCRE_DATA_
2728        SCOPE with something explicit in config.h. In addition:
2730        (a) pcreposix.h still had just "extern" instead of either of these macros;
2731            I have replaced it with PCRE_DATA_SCOPE.
2733        (b) Functions such as _pcre_xclass(), which are internal to the library,
2734            but external in the C sense, all had PCRE_EXPORT in their definitions.
2735            This is apparently wrong for the Windows case, so I have removed it.
2736            (It makes no difference on Unix-like systems.)
2738    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
2739        of recursive calls to match(). This is different to MATCH_LIMIT because
2740        that limits the total number of calls to match(), not all of which increase
2741        the depth of recursion. Limiting the recursion depth limits the amount of
2742        stack (or heap if NO_RECURSE is set) that is used. The default can be set
2743        when PCRE is compiled, and changed at run time. A patch from Google adds
2744        this functionality to the C++ interface.
2746    18. Changes to the handling of Unicode character properties:
2748        (a) Updated the table to Unicode 4.1.0.
2750        (b) Recognize characters that are not in the table as "Cn" (undefined).
2752        (c) I revised the way the table is implemented to a much improved format
2753            which includes recognition of ranges. It now supports the ranges that
2754            are defined in UnicodeData.txt, and it also amalgamates other
2755            characters into ranges. This has reduced the number of entries in the
2756            table from around 16,000 to around 3,000, thus reducing its size
2757            considerably. I realized I did not need to use a tree structure after
2758            all - a binary chop search is just as efficient. Having reduced the
2759            number of entries, I extended their size from 6 bytes to 8 bytes to
2760            allow for more data.
2762        (d) Added support for Unicode script names via properties such as \p{Han}.
2764    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
2765        matching that character.
2767    20. When matching a repeated Unicode property with a minimum greater than zero,
2768        (for example \pL{2,}), PCRE could look past the end of the subject if it
2769        reached it while seeking the minimum number of characters. This could
2770        happen only if some of the characters were more than one byte long, because
2771        there is a check for at least the minimum number of bytes.
2773    21. Refactored the implementation of \p and \P so as to be more general, to
2774        allow for more different types of property in future. This has changed the
2775        compiled form incompatibly. Anybody with saved compiled patterns that use
2776        \p or \P will have to recompile them.
2778    22. Added "Any" and "L&" to the supported property types.
2780    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
2781        but give a compile time error if the value is greater than 0xff.
2783    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
2784        accidentally not being installed or uninstalled.
2786    25. The pcre.h file was built from pcre.h.in, but the only changes that were
2787        made were to insert the current release number. This seemed silly, because
2788        it made things harder for people building PCRE on systems that don't run
2789        "configure". I have turned pcre.h into a distributed file, no longer built
2790        by "configure", with the version identification directly included. There is
2791        no longer a pcre.h.in file.
2793        However, this change necessitated a change to the pcre-config script as
2794        well. It is built from pcre-config.in, and one of the substitutions was the
2795        release number. I have updated configure.ac so that ./configure now finds
2796        the release number by grepping pcre.h.
2798    26. Added the ability to run the tests under valgrind.
2801    Version 6.4 05-Sep-05
2802    ---------------------
2804     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2805        "--" to be printed when multiple files were scanned, even when none of the
2806        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2807        consider it to be a bug, and have restored the previous behaviour.
2809     2. A couple of code tidies to get rid of compiler warnings.
2811     3. The pcretest program used to cheat by referring to symbols in the library
2812        whose names begin with _pcre_. These are internal symbols that are not
2813        really supposed to be visible externally, and in some environments it is
2814        possible to suppress them. The cheating is now confined to including
2815        certain files from the library's source, which is a bit cleaner.
2817     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2818        file's purpose clearer.
2820     5. Reorganized pcre_ucp_findchar().
2823    Version 6.3 15-Aug-05
2824    ---------------------
2826     1. The file libpcre.pc.in did not have general read permission in the tarball.
2828     2. There were some problems when building without C++ support:
2830        (a) If C++ support was not built, "make install" and "make test" still
2831            tried to test it.
2833        (b) There were problems when the value of CXX was explicitly set. Some
2834            changes have been made to try to fix these, and ...
2836        (c) --disable-cpp can now be used to explicitly disable C++ support.
2838        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2839            backslash in a target when C++ was disabled. This confuses some
2840            versions of "make", apparently. Using an intermediate variable solves
2841            this. (Same for CPP_LOBJ.)
2843     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2844        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2845        necessary on certain architectures.
2847     4. Added a setting of -export-symbols-regex to the link command to remove
2848        those symbols that are exported in the C sense, but actually are local
2849        within the library, and not documented. Their names all begin with
2850        "_pcre_". This is not a perfect job, because (a) we have to except some
2851        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2852        available (and never for static libraries). I have made a note to try to
2853        find a way round (a) in the future.
2856    Version 6.2 01-Aug-05
2857    ---------------------
2859     1. There was no test for integer overflow of quantifier values. A construction
2860        such as {1111111111111111} would give undefined results. What is worse, if
2861        a minimum quantifier for a parenthesized subpattern overflowed and became
2862        negative, the calculation of the memory size went wrong. This could have
2863        led to memory overwriting.
2865     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2867     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2868        operating environments where this matters.
2870     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2871        PCRE options from within the C++ wrapper.
2873     5. Named capturing subpatterns were not being correctly counted when a pattern
2874        was compiled. This caused two problems: (a) If there were more than 100
2875        such subpatterns, the calculation of the memory needed for the whole
2876        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2877        back references of the form \12, where the number was greater than 9, were
2878        not recognized as back references, even though there were sufficient
2879        previous subpatterns.
2881     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2882        versions of gcc, e.g. 2.95.4.
2885    Version 6.1 21-Jun-05
2886    ---------------------
2888     1. There was one reference to the variable "posix" in pcretest.c that was not
2889        surrounded by "#if !defined NOPOSIX".
2891     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2892        the cross-check on the old pcre_info() function, for the benefit of the
2893        cut-down version of PCRE that is currently imported into Exim.
2895     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2896        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2897        patterns that start (?i) but I don't think that matters. The use of (?i) is
2898        just an example; this all applies to the other options as well.
2900     4. Since libtool seems to echo the compile commands it is issuing, the output
2901        from "make" can be reduced a bit by putting "@" in front of each libtool
2902        compile command.
2904     5. Patch from the folks at Google for configure.in to be a bit more thorough
2905        in checking for a suitable C++ installation before trying to compile the
2906        C++ stuff. This should fix a reported problem when a compiler was present,
2907        but no suitable headers.
2909     6. The man pages all had just "PCRE" as their title. I have changed them to
2910        be the relevant file name. I have also arranged that these names are
2911        retained in the file doc/pcre.txt, which is a concatenation in text format
2912        of all the man pages except the little individual ones for each function.
2914     7. The NON-UNIX-USE file had not been updated for the different set of source
2915        files that come with release 6. I also added a few comments about the C++
2916        wrapper.
2919    Version 6.0 07-Jun-05
2920    ---------------------
2922     1. Some minor internal re-organization to help with my DFA experiments.
2924     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2925        didn't matter for the library itself when fully configured, but did matter
2926        when compiling without UCP support, or within Exim, where the ucp files are
2927        not imported.
2929     3. Refactoring of the library code to split up the various functions into
2930        different source modules. The addition of the new DFA matching code (see
2931        below) to a single monolithic source would have made it really too
2932        unwieldy, quite apart from causing all the code to be include in a
2933        statically linked application, when only some functions are used. This is
2934        relevant even without the DFA addition now that patterns can be compiled in
2935        one application and matched in another.
2937        The downside of splitting up is that there have to be some external
2938        functions and data tables that are used internally in different modules of
2939        the library but which are not part of the API. These have all had their
2940        names changed to start with "_pcre_" so that they are unlikely to clash
2941        with other external names.
2943     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2944        a different (DFA) algorithm. Although it is slower than the original
2945        function, it does have some advantages for certain types of matching
2946        problem.
2948     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2949        including restarting after a partial match.
2951     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2952        defined when compiling for Windows was sent to me. I have put it into the
2953        code, though I have no means of testing or verifying it.
2955     7. Added the pcre_refcount() auxiliary function.
2957     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2958        match before or at the first newline in the subject string. In pcretest,
2959        the /f option on a pattern can be used to set this.
2961     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2962        would behave wrongly. This has been present in PCRE since release 4.0.
2964    10. A number of changes to the pcregrep command:
2966        (a) Refactored how -x works; insert ^(...)$ instead of setting
2967            PCRE_ANCHORED and checking the length, in preparation for adding
2968            something similar for -w.
2970        (b) Added the -w (match as a word) option.
2972        (c) Refactored the way lines are read and buffered so as to have more
2973            than one at a time available.
2975        (d) Implemented a pcregrep test script.
2977        (e) Added the -M (multiline match) option. This allows patterns to match
2978            over several lines of the subject. The buffering ensures that at least
2979            8K, or the rest of the document (whichever is the shorter) is available
2980            for matching (and similarly the previous 8K for lookbehind assertions).
2982        (f) Changed the --help output so that it now says
2984              -w, --word-regex(p)
2986            instead of two lines, one with "regex" and the other with "regexp"
2987            because that confused at least one person since the short forms are the
2988            same. (This required a bit of code, as the output is generated
2989            automatically from a table. It wasn't just a text change.)
2991        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2992            option but starts with a hyphen. Could be a pattern or a path name
2993            starting with a hyphen, for instance.
2995        (h) "-" can be given as a file name to represent stdin.
2997        (i) When file names are being printed, "(standard input)" is used for
2998            the standard input, for compatibility with GNU grep. Previously
2999            "<stdin>" was used.
3001        (j) The option --label=xxx can be used to supply a name to be used for
3002            stdin when file names are being printed. There is no short form.
3004        (k) Re-factored the options decoding logic because we are going to add
3005            two more options that take data. Such options can now be given in four
3006            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3008        (l) Added the -A, -B, and -C options for requesting that lines of context
3009            around matches be printed.
3011        (m) Added the -L option to print the names of files that do not contain
3012            any matching lines, that is, the complement of -l.
3014        (n) The return code is 2 if any file cannot be opened, but pcregrep does
3015            continue to scan other files.
3017        (o) The -s option was incorrectly implemented. For compatibility with other
3018            greps, it now suppresses the error message for a non-existent or non-
3019            accessible file (but not the return code). There is a new option called
3020            -q that suppresses the output of matching lines, which was what -s was
3021            previously doing.
3023        (p) Added --include and --exclude options to specify files for inclusion
3024            and exclusion when recursing.
3026    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3027        Hopefully, it now does.
3029    12. Missing cast in pcre_study().
3031    13. Added an "uninstall" target to the makefile.
3033    14. Replaced "extern" in the function prototypes in Makefile.in with
3034        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3035        world, but is set differently for Windows.
3037    15. Added a second compiling function called pcre_compile2(). The only
3038        difference is that it has an extra argument, which is a pointer to an
3039        integer error code. When there is a compile-time failure, this is set
3040        non-zero, in addition to the error test pointer being set to point to an
3041        error message. The new argument may be NULL if no error number is required
3042        (but then you may as well call pcre_compile(), which is now just a
3043        wrapper). This facility is provided because some applications need a
3044        numeric error indication, but it has also enabled me to tidy up the way
3045        compile-time errors are handled in the POSIX wrapper.
3047    16. Added VPATH=.libs to the makefile; this should help when building with one
3048        prefix path and installing with another. (Or so I'm told by someone who
3049        knows more about this stuff than I do.)
3051    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3052        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3053        match everything, including newlines. This is not POSIX-compatible, but
3054        somebody wanted the feature. From pcretest it can be activated by using
3055        both the P and the s flags.
3057    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3059    19. libpcre.pc was being incorrectly installed as executable.
3061    20. A couple of places in pcretest check for end-of-line by looking for '\n';
3062        it now also looks for '\r' so that it will work unmodified on Windows.
3064    21. Added Google's contributed C++ wrapper to the distribution.
3066    22. Added some untidy missing memory free() calls in pcretest, to keep
3067        Electric Fence happy when testing.
3071    Version 5.0 13-Sep-04
3072    ---------------------
3074     1. Internal change: literal characters are no longer packed up into items
3075        containing multiple characters in a single byte-string. Each character
3076        is now matched using a separate opcode. However, there may be more than one
3077        byte in the character in UTF-8 mode.
3079     2. The pcre_callout_block structure has two new fields: pattern_position and
3080        next_item_length. These contain the offset in the pattern to the next match
3081        item, and its length, respectively.
3083     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3084        insertion of callouts before each pattern item. Added the /C option to
3085        pcretest to make use of this.
3087     4. On the advice of a Windows user, the lines
3089          #if defined(_WIN32) || defined(WIN32)
3090          _setmode( _fileno( stdout ), 0x8000 );
3091          #endif  /* defined(_WIN32) || defined(WIN32) */
3093        have been added to the source of pcretest. This apparently does useful
3094        magic in relation to line terminators.
3096     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3097        for the benefit of those environments where the "b" makes a difference.
3099     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3100        to know about it. I have put a hack into configure.in that adds in code
3101        to set GCC=yes if CC=icc. This seems to end up at a point in the
3102        generated configure script that is early enough to affect the setting of
3103        compiler options, which is what is needed, but I have no means of testing
3104        whether it really works. (The user who reported this had patched the
3105        generated configure script, which of course I cannot do.)
3107        LATER: After change 22 below (new libtool files), the configure script
3108        seems to know about icc (and also ecc). Therefore, I have commented out
3109        this hack in configure.in.
3111     7. Added support for pkg-config (2 patches were sent in).
3113     8. Negated POSIX character classes that used a combination of internal tables
3114        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3115        [[:^ascii]]. Typically, they would match almost any characters. The other
3116        POSIX classes were not broken in this way.
3118     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3119        to find the match, as PCRE was deluded into thinking that the match had to
3120        start at the start point or following a newline. The same bug applied to
3121        patterns with negative forward assertions or any backward assertions
3122        preceding ".*" at the start, unless the pattern required a fixed first
3123        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3125    10. In UTF-8 mode, when moving forwards in the subject after a failed match
3126        starting at the last subject character, bytes beyond the end of the subject
3127        string were read.
3129    11. Renamed the variable "class" as "classbits" to make life easier for C++
3130        users. (Previously there was a macro definition, but it apparently wasn't
3131        enough.)
3133    12. Added the new field "tables" to the extra data so that tables can be passed
3134        in at exec time, or the internal tables can be re-selected. This allows
3135        a compiled regex to be saved and re-used at a later time by a different
3136        program that might have everything at different addresses.
3138    13. Modified the pcre-config script so that, when run on Solaris, it shows a
3139        -R library as well as a -L library.
3141    14. The debugging options of pcretest (-d on the command line or D on a
3142        pattern) showed incorrect output for anything following an extended class
3143        that contained multibyte characters and which was followed by a quantifier.
3145    15. Added optional support for general category Unicode character properties
3146        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3147        support. It adds about 90K to the size of the library. The meanings of the
3148        inbuilt class escapes such as \d and \s have NOT been changed.
3150    16. Updated pcredemo.c to include calls to free() to release the memory for the
3151        compiled pattern.
3153    17. The generated file chartables.c was being created in the source directory
3154        instead of in the building directory. This caused the build to fail if the
3155        source directory was different from the building directory, and was
3156        read-only.
3158    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3159        file. No doubt somebody will tell me if they don't make sense... Also added
3160        Dan Mooney's comments about building on OpenVMS.
3162    19. Added support for partial matching via the PCRE_PARTIAL option for
3163        pcre_exec() and the \P data escape in pcretest.
3165    20. Extended pcretest with 3 new pattern features:
3167        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
3168              write the compiled pattern to the file whose name is "rest-of-line".
3169              This is a straight binary dump of the data, with the saved pointer to
3170              the character tables forced to be NULL. The study data, if any, is
3171              written too. After writing, pcretest reads a new pattern.
3173        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3174              compiled pattern from the given file. There must not be any
3175              occurrences of "<" in the file name (pretty unlikely); if there are,
3176              pcretest will instead treat the initial "<" as a pattern delimiter.
3177              After reading in the pattern, pcretest goes on to read data lines as
3178              usual.
3180        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3181              and 16-bit fields in a compiled pattern, to simulate a pattern that
3182              was compiled on a host of opposite endianness.
3184    21. The pcre-exec() function can now cope with patterns that were compiled on
3185        hosts of opposite endianness, with this restriction:
3187          As for any compiled expression that is saved and used later, the tables
3188          pointer field cannot be preserved; the extra_data field in the arguments
3189          to pcre_exec() should be used to pass in a tables address if a value
3190          other than the default internal tables were used at compile time.
3192    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3193        now diagnosed as an error. Previously, most of the time, a negative number
3194        would have been treated as zero, but if in addition "ovector" was passed as
3195        NULL, a crash could occur.
3197    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3198        new versions from the libtool 1.5 distribution (the last one is a copy of
3199        a file called libtool.m4). This seems to have fixed the need to patch
3200        "configure" to support Darwin 1.3 (which I used to do). However, I still
3201        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3202        workstation).
3204    24. Changed the PCRE licence to be the more standard "BSD" licence.
3207    Version 4.5 01-Dec-03
3208    ---------------------
3210     1. There has been some re-arrangement of the code for the match() function so
3211        that it can be compiled in a version that does not call itself recursively.
3212        Instead, it keeps those local variables that need separate instances for
3213        each "recursion" in a frame on the heap, and gets/frees frames whenever it
3214        needs to "recurse". Keeping track of where control must go is done by means
3215        of setjmp/longjmp. The whole thing is implemented by a set of macros that
3216        hide most of the details from the main code, and operates only if
3217        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3218        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3219        operating.
3221        To make it easier for callers to provide specially tailored get/free
3222        functions for this usage, two new functions, pcre_stack_malloc, and
3223        pcre_stack_free, are used. They are always called in strict stacking order,
3224        and the size of block requested is always the same.
3226        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3227        PCRE has been compiled to use the stack or the heap for recursion. The
3228        -C option of pcretest uses this to show which version is compiled.
3230        A new data escape \S, is added to pcretest; it causes the amounts of store
3231        obtained and freed by both kinds of malloc/free at match time to be added
3232        to the output.
3234     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3235        what's available on my current Linux desktop machine.
3237     3. When matching a UTF-8 string, the test for a valid string at the start has
3238        been extended. If start_offset is not zero, PCRE now checks that it points
3239        to a byte that is the start of a UTF-8 character. If not, it returns
3240        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3241        this is necessary because there may be backward assertions in the pattern.
3242        When matching the same subject several times, it may save resources to use
3243        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3245     4. The code for checking the validity of UTF-8 strings has been tightened so
3246        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3247        containing "overlong sequences".
3249     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3250        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3251        should have been "&", but it just so happened that all the cases this let
3252        through by mistake were picked up later in the function.
3254     6. I had used a variable called "isblank" - this is a C99 function, causing
3255        some compilers to warn. To avoid this, I renamed it (as "blankclass").
3257     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3258        prompting; (b) run "./pcretest /dev/null" at the start of the test script
3259        so the version is shown; (c) stop "make test" echoing "./RunTest".
3261     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3263     9. The prototype for memmove() for systems that don't have it was using
3264        size_t, but the inclusion of the header that defines size_t was later. I've
3265        moved the #includes for the C headers earlier to avoid this.
3267    10. Added some adjustments to the code to make it easier to compiler on certain
3268        special systems:
3270          (a) Some "const" qualifiers were missing.
3271          (b) Added the macro EXPORT before all exported functions; by default this
3272              is defined to be empty.
3273          (c) Changed the dftables auxiliary program (that builds chartables.c) so
3274              that it reads its output file name as an argument instead of writing
3275              to the standard output and assuming this can be redirected.
3277    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3278        class containing characters with values greater than 255, PCRE compilation
3279        went into a loop.
3281    12. A recursive reference to a subpattern that was within another subpattern
3282        that had a minimum quantifier of zero caused PCRE to crash. For example,
3283        (x(y(?2))z)? provoked this bug with a subject that got as far as the
3284        recursion. If the recursively-called subpattern itself had a zero repeat,
3285        that was OK.
3287    13. In pcretest, the buffer for reading a data line was set at 30K, but the
3288        buffer into which it was copied (for escape processing) was still set at
3289        1024, so long lines caused crashes.
3291    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3292        "internal error: code overflow...". This applied to any character class
3293        that was followed by a possessive quantifier.
3295    15. Modified the Makefile to add libpcre.la as a prerequisite for
3296        libpcreposix.la because I was told this is needed for a parallel build to
3297        work.
3299    16. If a pattern that contained .* following optional items at the start was
3300        studied, the wrong optimizing data was generated, leading to matching
3301        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3302        matching string must start with a or b or c. The correct conclusion for
3303        this pattern is that a match can start with any character.
3306    Version 4.4 13-Aug-03
3307    ---------------------
3309     1. In UTF-8 mode, a character class containing characters with values between
3310        127 and 255 was not handled correctly if the compiled pattern was studied.
3311        In fixing this, I have also improved the studying algorithm for such
3312        classes (slightly).
3314     2. Three internal functions had redundant arguments passed to them. Removal
3315        might give a very teeny performance improvement.
3317     3. Documentation bug: the value of the capture_top field in a callout is *one
3318        more than* the number of the hightest numbered captured substring.
3320     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3321        in incorrectly linking with a previously installed version. They now link
3322        explicitly with libpcre.la.
3324     5. configure.in no longer needs to recognize Cygwin specially.
3326     6. A problem in pcre.in for Windows platforms is fixed.
3328     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3329        pcretest, it used to include the size of the study block as part of its
3330        output. Unfortunately, the structure contains a field that has a different
3331        size on different hardware architectures. This meant that the tests that
3332        showed this size failed. As the block is currently always of a fixed size,
3333        this information isn't actually particularly useful in pcretest output, so
3334        I have just removed it.
3336     8. Three pre-processor statements accidentally did not start in column 1.
3337        Sadly, there are *still* compilers around that complain, even though
3338        standard C has not required this for well over a decade. Sigh.
3340     9. In pcretest, the code for checking callouts passed small integers in the
3341        callout_data field, which is a void * field. However, some picky compilers
3342        complained about the casts involved for this on 64-bit systems. Now
3343        pcretest passes the address of the small integer instead, which should get
3344        rid of the warnings.
3346    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3347        both compile and run time, and gives an error if an invalid UTF-8 sequence
3348        is found. There is a option for disabling this check in cases where the
3349        string is known to be correct and/or the maximum performance is wanted.
3351    11. In response to a bug report, I changed one line in Makefile.in from
3353            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3354        to
3355            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3357        to look similar to other lines, but I have no way of telling whether this
3358        is the right thing to do, as I do not use Windows. No doubt I'll get told
3359        if it's wrong...
3362    Version 4.3 21-May-03
3363    ---------------------
3365    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3366       Makefile.
3368    2. Some refactoring to improve the quality of the code:
3370       (i)   The utf8_table... variables are now declared "const".
3372       (ii)  The code for \cx, which used the "case flipping" table to upper case
3373             lower case letters, now just substracts 32. This is ASCII-specific,
3374             but the whole concept of \cx is ASCII-specific, so it seems
3375             reasonable.
3377       (iii) PCRE was using its character types table to recognize decimal and
3378             hexadecimal digits in the pattern. This is silly, because it handles
3379             only 0-9, a-f, and A-F, but the character types table is locale-
3380             specific, which means strange things might happen. A private
3381             table is now used for this - though it costs 256 bytes, a table is
3382             much faster than multiple explicit tests. Of course, the standard
3383             character types table is still used for matching digits in subject
3384             strings against \d.
3386       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3387             ending in _t are). So I've renamed it as ESC_tee.
3389    3. The first argument for regexec() in the POSIX wrapper should have been
3390       defined as "const".
3392    4. Changed pcretest to use malloc() for its buffers so that they can be
3393       Electric Fenced for debugging.
3395    5. There were several places in the code where, in UTF-8 mode, PCRE would try
3396       to read one or more bytes before the start of the subject string. Often this
3397       had no effect on PCRE's behaviour, but in some circumstances it could
3398       provoke a segmentation fault.
3400    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3401       to try to read one or more bytes before the start of the subject string.
3403    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3404       UTF-8 support could misbehave in various ways if the subject string
3405       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3406       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3407       back over UTF-8 characters.)
3410    Version 4.2 14-Apr-03
3411    ---------------------
3413    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3415    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3416         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3417         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3418         [WIN_PREFIX]: new variable, "cyg" for Cygwin
3419         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3420           and BUILD_EXEEXT
3421         Note: automatic setting of the BUILD variables is not yet working
3422         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3423           compile-time but not at link-time
3424         [LINK]: use for linking executables only
3425         make different versions for Windows and non-Windows
3426         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3427           libraries
3428         [LINK_FOR_BUILD]: new variable
3429         [OBJEXT]: use throughout
3430         [EXEEXT]: use throughout
3431         <winshared>: new target
3432         <wininstall>: new target
3433         <dftables.o>: use native compiler
3434         <dftables>: use native linker
3435         <install>: handle Windows platform correctly
3436         <clean>: ditto
3437         <check>: ditto
3438         copy DLL to top builddir before testing
3440       As part of these changes, -no-undefined was removed again. This was reported
3441       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3442       in any case.
3444    3. Some tidies to get rid of compiler warnings:
3446       . In the match_data structure, match_limit was an unsigned long int, whereas
3447         match_call_count was an int. I've made them both unsigned long ints.
3449       . In pcretest the fact that a const uschar * doesn't automatically cast to
3450         a void * provoked a warning.
3452       . Turning on some more compiler warnings threw up some "shadow" variables
3453         and a few more missing casts.
3455    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3456       option, a class that contained a single character with a value between 128
3457       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3459    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3460       option, a class that contained several characters, but with at least one
3461       whose value was between 128 and 255 caused PCRE to crash.
3464    Version 4.1 12-Mar-03
3465    ---------------------
3467    1. Compiling with gcc -pedantic found a couple of places where casts were
3468    needed, and a string in dftables.c that was longer than standard compilers are
3469    required to support.
3471    2. Compiling with Sun's compiler found a few more places where the code could
3472    be tidied up in order to avoid warnings.
3474    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3475    first of these names is deprecated in the latest Autoconf in favour of the name
3476    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3477    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3478    analogy I have changed it to CFLAGS_FOR_BUILD.
3480    4. Added -no-undefined to the linking command in the Makefile, because this is
3481    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3482    linking step for the pcreposix library.
3484    5. PCRE was failing to diagnose the case of two named groups with the same
3485    name.
3487    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3488    literal character that is needed in the subject for a match, and scans along to
3489    ensure that it is present before embarking on the full matching process. This
3490    saves time in cases of nested unlimited repeats that are never going to match.
3491    Problem: the scan can take a lot of time if the subject is very long (e.g.
3492    megabytes), thus penalizing straightforward matches. It is now done only if the
3493    amount of subject to be scanned is less than 1000 bytes.
3495    7. A lesser problem with the same optimization is that it was recording the
3496    first character of an anchored pattern as "needed", thus provoking a search
3497    right along the subject, even when the first match of the pattern was going to
3498    fail. The "needed" character is now not set for anchored patterns, unless it
3499    follows something in the pattern that is of non-fixed length. Thus, it still
3500    fulfils its original purpose of finding quick non-matches in cases of nested
3501    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3504    Version 4.0 17-Feb-03
3505    ---------------------
3507    1. If a comment in an extended regex that started immediately after a meta-item
3508    extended to the end of string, PCRE compiled incorrect data. This could lead to
3509    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3511    2. Moved to autoconf 2.53 and libtool 1.4.2.
3513    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3514    the special perltest8 script is no longer needed - all the tests can be run
3515    from a single perltest script.
3517    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3518    by \s. It has now been removed in PCRE. This means it isn't recognized as
3519    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3520    class [:space:] *does* include VT, thereby creating a mess.
3522    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3523    space and tab.
3525    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3526    its new features into the main test script, reducing the number of scripts.
3528    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3529    were backward compatible, and made the (?i) apply to the whole pattern, as if
3530    /i were given. Now it behaves more logically, and applies the option setting
3531    only to what follows. PCRE has been changed to follow suit. However, if it
3532    finds options settings right at the start of the pattern, it extracts them into
3533    the global options, as before. Thus, they show up in the info data.
3535    8. Added support for the \Q...\E escape sequence. Characters in between are
3536    treated as literals. This is slightly different from Perl in that $ and @ are
3537    also handled as literals inside the quotes. In Perl, they will cause variable
3538    interpolation. Note the following examples:
3540        Pattern            PCRE matches      Perl matches
3542        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
3543        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
3544        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
3546    For compatibility with Perl, \Q...\E sequences are recognized inside character
3547    classes as well as outside them.
3549    9. Re-organized 3 code statements in pcretest to avoid "overflow in
3550    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
3551    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
3552    signed/unsigned warnings.
3554    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
3555    option for pcretest, so I've replaced it by a simple function that does just
3556    that job.
3558    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
3559    "pcregrep -".
3561    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
3562    Java package. This provides some syntactic sugar for simple cases of what my
3563    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
3564    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
3565    item, you can use this simplified notation. Note that only makes sense with
3566    greedy quantifiers. Consequently, the use of the possessive quantifier forces
3567    greediness, whatever the setting of the PCRE_UNGREEDY option.
3569    13. A change of greediness default within a pattern was not taking effect at
3570    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
3571    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
3572    was abstracted outside.
3574    14. PCRE now supports the \G assertion. It is true when the current matching
3575    position is at the start point of the match. This differs from \A when the
3576    starting offset is non-zero. Used with the /g option of pcretest (or similar
3577    code), it works in the same way as it does for Perl's /g option. If all
3578    alternatives of a regex begin with \G, the expression is anchored to the start
3579    match position, and the "anchored" flag is set in the compiled expression.
3581    15. Some bugs concerning the handling of certain option changes within patterns
3582    have been fixed. These applied to options other than (?ims). For example,
3583    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
3584    been the other way round. Some of this was related to change 7 above.
3586    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
3587    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
3588    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
3589    POSIX classes only within a class (e.g. /[[:alpha:]]/).
3591    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
3592    mode. Unlike ".", it always matches newline, whatever the setting of
3593    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
3594    assertions. Perl allows it, but it doesn't (in general) work because it can't
3595    calculate the length of the lookbehind. At least, that's the case for Perl
3596    5.8.0 - I've been told they are going to document that it doesn't work in
3597    future.
3599    18. Added an error diagnosis for escapes that PCRE does not support: these are<