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

Diff of /code/trunk/ChangeLog

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

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