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

Diff of /code/trunk/ChangeLog

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

revision 29 by nigel, Sat Feb 24 21:38:53 2007 UTC revision 514 by ph10, Mon May 3 12:54:22 2010 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 8.10 03 May-2010
5    ------------------------
6    
7    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
8        THEN.
9    
10    2.  (*ACCEPT) was not working when inside an atomic group.
11    
12    3.  Inside a character class, \B is treated as a literal by default, but
13        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
14        causes the error). The code is unchanged, but I tidied the documentation.
15    
16    4.  Inside a character class, PCRE always treated \R and \X as literals,
17        whereas Perl faults them if its -w option is set. I have changed PCRE so
18        that it faults them when PCRE_EXTRA is set.
19    
20    5.  Added support for \N, which always matches any character other than
21        newline. (It is the same as "." when PCRE_DOTALL is not set.)
22    
23    
24    Version 8.02 19-Mar-2010
25    ------------------------
26    
27    1.  The Unicode data tables have been updated to Unicode 5.2.0.
28    
29    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
30        configured.
31    
32    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
33        original author of that file, following a query about its status.
34    
35    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
36        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
37    
38    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
39        quantifier applied to a forward-referencing subroutine call, could compile
40        incorrect code or give the error "internal error: previously-checked
41        referenced subpattern not found".
42    
43    6.  Both MS Visual Studio and Symbian OS have problems with initializing
44        variables to point to external functions. For these systems, therefore,
45        pcre_malloc etc. are now initialized to local functions that call the
46        relevant global functions.
47    
48    7.  There were two entries missing in the vectors called coptable and poptable
49        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
50        I've fixed the data, and added a kludgy way of testing at compile time that
51        the lengths are correct (equal to the number of opcodes).
52    
53    8.  Following on from 7, I added a similar kludge to check the length of the
54        eint vector in pcreposix.c.
55    
56    9.  Error texts for pcre_compile() are held as one long string to avoid too
57        much relocation at load time. To find a text, the string is searched,
58        counting zeros. There was no check for running off the end of the string,
59        which could happen if a new error number was added without updating the
60        string.
61    
62    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
63    
64    11. \K was not working if it appeared in an atomic group or in a group that
65        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
66        \K is "not well defined" if used in an assertion. PCRE now accepts it if
67        the assertion is positive, but not if it is negative.
68    
69    12. Change 11 fortuitously reduced the size of the stack frame used in the
70        "match()" function of pcre_exec.c by one pointer. Forthcoming
71        implementation of support for (*MARK) will need an extra pointer on the
72        stack; I have reserved it now, so that the stack frame size does not
73        decrease.
74    
75    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
76        item in branch that calls a recursion is a subroutine call - as in the
77        second branch in the above example - was incorrectly given the compile-
78        time error "recursive call could loop indefinitely" because pcre_compile()
79        was not correctly checking the subroutine for matching a non-empty string.
80    
81    14. The checks for overrunning compiling workspace could trigger after an
82        overrun had occurred. This is a "should never occur" error, but it can be
83        triggered by pathological patterns such as hundreds of nested parentheses.
84        The checks now trigger 100 bytes before the end of the workspace.
85    
86    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
87    
88    
89    Version 8.01 19-Jan-2010
90    ------------------------
91    
92    1.  If a pattern contained a conditional subpattern with only one branch (in
93        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
94        computed the wrong minimum data length (which is of course zero for such
95        subpatterns). This could cause incorrect "no match" results.
96    
97    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
98        the pattern is reset in the first branch, pcre_compile() failed with
99        "internal error: code overflow at offset...". This happened only when
100        the reset was to the original external option setting. (An optimization
101        abstracts leading options settings into an external setting, which was the
102        cause of this.)
103    
104    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
105        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
106        assertion pattern did not match (meaning that the assertion was true), it
107        was incorrectly treated as false if the SKIP had been reached during the
108        matching. This also applied to assertions used as conditions.
109    
110    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
111        assertion subpattern, including such a pattern used as a condition,
112        unpredictable results occurred, instead of the error return
113        PCRE_ERROR_DFA_UITEM.
114    
115    5.  The C++ GlobalReplace function was not working like Perl for the special
116        situation when an empty string is matched. It now does the fancy magic
117        stuff that is necessary.
118    
119    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
120        removed. (These were left over from very, very early versions of PCRE.)
121    
122    7.  Some cosmetic changes to the code to make life easier when compiling it
123        as part of something else:
124    
125        (a) Change DEBUG to PCRE_DEBUG.
126    
127        (b) In pcre_compile(), rename the member of the "branch_chain" structure
128            called "current" as "current_branch", to prevent a collision with the
129            Linux macro when compiled as a kernel module.
130    
131        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
132            prevent a collision with the Linux macro when compiled as a kernel
133            module.
134    
135    8.  In pcre_compile() there are some checks for integer overflows that used to
136        cast potentially large values to (double). This has been changed to that
137        when building, a check for int64_t is made, and if it is found, it is used
138        instead, thus avoiding the use of floating point arithmetic. (There is no
139        other use of FP in PCRE.) If int64_t is not found, the fallback is to
140        double.
141    
142    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
143        2005 (difference between two addresses compared to an unsigned value).
144    
145    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
146        custom one, because of the following reported problem in Windows:
147    
148          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
149              under Win32.
150          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
151              therefore missing the function definition.
152          - The compiler thus generates a "C" signature for the test function.
153          - The linker fails to find the "C" function.
154          - PCRE fails to configure if asked to do so against libbz2.
155    
156    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
157        messages were output:
158    
159          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
160          rerunning libtoolize, to keep the correct libtool macros in-tree.
161          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
162    
163        I have done both of these things.
164    
165    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
166        most of the time, it *can* run out if it is given a pattern that contains a
167        runaway infinite recursion. I updated the discussion in the pcrestack man
168        page.
169    
170    13. Now that we have gone to the x.xx style of version numbers, the minor
171        version may start with zero. Using 08 or 09 is a bad idea because users
172        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
173        interpreted as invalid octal numbers. I've updated the previous comment in
174        configure.ac, and also added a check that gives an error if 08 or 09 are
175        used.
176    
177    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
178        causing partial matching to fail when the end of the subject matched \W
179        in a UTF-8 pattern where \W was quantified with a minimum of 3.
180    
181    15. There were some discrepancies between the declarations in pcre_internal.h
182        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
183        their definitions. The declarations used "const uschar *" and the
184        definitions used USPTR. Even though USPTR is normally defined as "const
185        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
186        reported that: "This difference in casting confuses some C++ compilers, for
187        example, SunCC recognizes above declarations as different functions and
188        generates broken code for hbpcre." I have changed the declarations to use
189        USPTR.
190    
191    16. GNU libtool is named differently on some systems. The autogen.sh script now
192        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
193        (FreeBSD).
194    
195    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
196        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
197        comment: "Figure out how to create a longlong from a string: strtoll and
198        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
199        instance, but it only takes 2 args instead of 3!"
200    
201    18. A subtle bug concerned with back references has been fixed by a change of
202        specification, with a corresponding code fix. A pattern such as
203        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
204        refers, was giving matches when it shouldn't. For example, xa=xaaa would
205        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
206        same bug. Such groups have to be quantified to be useful, or contained
207        inside another quantified group. (If there's no repetition, the reference
208        can never match.) The problem arises because, having left the group and
209        moved on to the rest of the pattern, a later failure that backtracks into
210        the group uses the captured value from the final iteration of the group
211        rather than the correct earlier one. I have fixed this in PCRE by forcing
212        any group that contains a reference to itself to be an atomic group; that
213        is, there cannot be any backtracking into it once it has completed. This is
214        similar to recursive and subroutine calls.
215    
216    
217    Version 8.00 19-Oct-09
218    ----------------------
219    
220    1.  The table for translating pcre_compile() error codes into POSIX error codes
221        was out-of-date, and there was no check on the pcre_compile() error code
222        being within the table. This could lead to an OK return being given in
223        error.
224    
225    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
226        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
227        in a Windows environment.
228    
229    3.  The pcregrep --count option prints the count for each file even when it is
230        zero, as does GNU grep. However, pcregrep was also printing all files when
231        --files-with-matches was added. Now, when both options are given, it prints
232        counts only for those files that have at least one match. (GNU grep just
233        prints the file name in this circumstance, but including the count seems
234        more useful - otherwise, why use --count?) Also ensured that the
235        combination -clh just lists non-zero counts, with no names.
236    
237    4.  The long form of the pcregrep -F option was incorrectly implemented as
238        --fixed_strings instead of --fixed-strings. This is an incompatible change,
239        but it seems right to fix it, and I didn't think it was worth preserving
240        the old behaviour.
241    
242    5.  The command line items --regex=pattern and --regexp=pattern were not
243        recognized by pcregrep, which required --regex pattern or --regexp pattern
244        (with a space rather than an '='). The man page documented the '=' forms,
245        which are compatible with GNU grep; these now work.
246    
247    6.  No libpcreposix.pc file was created for pkg-config; there was just
248        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
249    
250    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
251        when UCP support is not needed, by modifying the Python script that
252        generates it from Unicode data files. This should not matter if the module
253        is correctly used as a library, but I received one complaint about 50K of
254        unwanted data. My guess is that the person linked everything into his
255        program rather than using a library. Anyway, it does no harm.
256    
257    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
258        was a minimum greater than 1 for a wide character in a possessive
259        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
260        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
261        character. Chaos in the form of incorrect output or a compiling loop could
262        result.
263    
264    9.  The restrictions on what a pattern can contain when partial matching is
265        requested for pcre_exec() have been removed. All patterns can now be
266        partially matched by this function. In addition, if there are at least two
267        slots in the offset vector, the offset of the earliest inspected character
268        for the match and the offset of the end of the subject are set in them when
269        PCRE_ERROR_PARTIAL is returned.
270    
271    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
272        synonymous with PCRE_PARTIAL, for backwards compatibility, and
273        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
274        and may be more useful for multi-segment matching.
275    
276    11. Partial matching with pcre_exec() is now more intuitive. A partial match
277        used to be given if ever the end of the subject was reached; now it is
278        given only if matching could not proceed because another character was
279        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
280        string "Z", which now yields "no match" instead of "partial match". In the
281        case of pcre_dfa_exec(), "no match" is given if every matching path for the
282        final character ended with (*FAIL).
283    
284    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
285        if the pattern had a "must contain" character that was already found in the
286        earlier partial match, unless partial matching was again requested. For
287        example, with the pattern /dog.(body)?/, the "must contain" character is
288        "g". If the first part-match was for the string "dog", restarting with
289        "sbody" failed. This bug has been fixed.
290    
291    13. The string returned by pcre_dfa_exec() after a partial match has been
292        changed so that it starts at the first inspected character rather than the
293        first character of the match. This makes a difference only if the pattern
294        starts with a lookbehind assertion or \b or \B (\K is not supported by
295        pcre_dfa_exec()). It's an incompatible change, but it makes the two
296        matching functions compatible, and I think it's the right thing to do.
297    
298    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
299        so that the demonstration program is easily available in environments where
300        PCRE has not been installed from source.
301    
302    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
303        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
304        library.
305    
306    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
307        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
308        is not the first non-POSIX option to be added. Clearly some people find
309        these options useful.
310    
311    17. If a caller to the POSIX matching function regexec() passes a non-zero
312        value for nmatch with a NULL value for pmatch, the value of
313        nmatch is forced to zero.
314    
315    18. RunGrepTest did not have a test for the availability of the -u option of
316        the diff command, as RunTest does. It now checks in the same way as
317        RunTest, and also checks for the -b option.
318    
319    19. If an odd number of negated classes containing just a single character
320        interposed, within parentheses, between a forward reference to a named
321        subpattern and the definition of the subpattern, compilation crashed with
322        an internal error, complaining that it could not find the referenced
323        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
324        [The bug was that it was starting one character too far in when skipping
325        over the character class, thus treating the ] as data rather than
326        terminating the class. This meant it could skip too much.]
327    
328    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
329        /g option in pcretest when the pattern contains \K, which makes it possible
330        to have an empty string match not at the start, even when the pattern is
331        anchored. Updated pcretest and pcredemo to use this option.
332    
333    21. If the maximum number of capturing subpatterns in a recursion was greater
334        than the maximum at the outer level, the higher number was returned, but
335        with unset values at the outer level. The correct (outer level) value is
336        now given.
337    
338    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
339        PCRE did not set those parentheses (unlike Perl). I have now found a way to
340        make it do so. The string so far is captured, making this feature
341        compatible with Perl.
342    
343    23. The tests have been re-organized, adding tests 11 and 12, to make it
344        possible to check the Perl 5.10 features against Perl 5.10.
345    
346    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
347        pattern matches a fixed length string. PCRE did not allow this; now it
348        does. Neither allows recursion.
349    
350    25. I finally figured out how to implement a request to provide the minimum
351        length of subject string that was needed in order to match a given pattern.
352        (It was back references and recursion that I had previously got hung up
353        on.) This code has now been added to pcre_study(); it finds a lower bound
354        to the length of subject needed. It is not necessarily the greatest lower
355        bound, but using it to avoid searching strings that are too short does give
356        some useful speed-ups. The value is available to calling programs via
357        pcre_fullinfo().
358    
359    26. While implementing 25, I discovered to my embarrassment that pcretest had
360        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
361        study optimizations had never been tested with that matching function.
362        Oops. What is worse, even when it was passed study data, there was a bug in
363        pcre_dfa_exec() that meant it never actually used it. Double oops. There
364        were also very few tests of studied patterns with pcre_dfa_exec().
365    
366    27. If (?| is used to create subpatterns with duplicate numbers, they are now
367        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
368        on the other side of the coin, they are no longer allowed to have different
369        names, because these cannot be distinguished in PCRE, and this has caused
370        confusion. (This is a difference from Perl.)
371    
372    28. When duplicate subpattern names are present (necessarily with different
373        numbers, as required by 27 above), and a test is made by name in a
374        conditional pattern, either for a subpattern having been matched, or for
375        recursion in such a pattern, all the associated numbered subpatterns are
376        tested, and the overall condition is true if the condition is true for any
377        one of them. This is the way Perl works, and is also more like the way
378        testing by number works.
379    
380    
381    Version 7.9 11-Apr-09
382    ---------------------
383    
384    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
385        (pcretest), all targets were linked against these libraries. This included
386        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
387        libraries. This caused unwanted dependencies to be created. This problem
388        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
389        pcretest is linked with readline.
390    
391    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
392        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
393        moved outside it again, because FALSE and TRUE are already defined in AIX,
394        but BOOL is not.
395    
396    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
397        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
398    
399    4.  The pcregrep documentation said spaces were inserted as well as colons (or
400        hyphens) following file names and line numbers when outputting matching
401        lines. This is not true; no spaces are inserted. I have also clarified the
402        wording for the --colour (or --color) option.
403    
404    5.  In pcregrep, when --colour was used with -o, the list of matching strings
405        was not coloured; this is different to GNU grep, so I have changed it to be
406        the same.
407    
408    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
409        each matching line was coloured. Now it goes on to look for further matches
410        of any of the test patterns, which is the same behaviour as GNU grep.
411    
412    7.  A pattern that could match an empty string could cause pcregrep to loop; it
413        doesn't make sense to accept an empty string match in pcregrep, so I have
414        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
415        seems to be how GNU grep behaves.
416    
417    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
418        start or after a newline", because the conditional assertion was not being
419        correctly handled. The rule now is that both the assertion and what follows
420        in the first alternative must satisfy the test.
421    
422    9.  If auto-callout was enabled in a pattern with a conditional group whose
423        condition was an assertion, PCRE could crash during matching, both with
424        pcre_exec() and pcre_dfa_exec().
425    
426    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
427        used for matching.
428    
429    11. Unicode property support in character classes was not working for
430        characters (bytes) greater than 127 when not in UTF-8 mode.
431    
432    12. Added the -M command line option to pcretest.
433    
434    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
435    
436    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
437    
438    16. Added comments and documentation about mis-use of no_arg in the C++
439        wrapper.
440    
441    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
442        from Martin Jerabek that uses macro names for all relevant character and
443        string constants.
444    
445    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
446        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
447        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
448        these, but not everybody uses configure.
449    
450    19. A conditional group that had only one branch was not being correctly
451        recognized as an item that could match an empty string. This meant that an
452        enclosing group might also not be so recognized, causing infinite looping
453        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
454        with the subject "ab", where knowledge that the repeated group can match
455        nothing is needed in order to break the loop.
456    
457    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
458        exec(), but without supplying a callout function, matching went wrong.
459    
460    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
461        leak if the size of the offset vector was greater than 30. When the vector
462        is smaller, the saved offsets during recursion go onto a local stack
463        vector, but for larger vectors malloc() is used. It was failing to free
464        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
465        error, in fact).
466    
467    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
468        heapframe that is used only when UTF-8 support is enabled. This caused no
469        problem, but was untidy.
470    
471    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
472        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
473        included within another project.
474    
475    24. Steven Van Ingelgem's patches to add more options to the CMake support,
476        slightly modified by me:
477    
478          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
479              not building pcregrep.
480    
481          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
482              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
483    
484    25. Forward references, both numeric and by name, in patterns that made use of
485        duplicate group numbers, could behave incorrectly or give incorrect errors,
486        because when scanning forward to find the reference group, PCRE was not
487        taking into account the duplicate group numbers. A pattern such as
488        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
489    
490    26. Changed a few more instances of "const unsigned char *" to USPTR, making
491        the feature of a custom pointer more persuasive (as requested by a user).
492    
493    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
494        pcretest.c, inside #ifndefs, because it seems they are sometimes already
495        pre-defined.
496    
497    28. Added support for (*UTF8) at the start of a pattern.
498    
499    29. Arrange for flags added by the "release type" setting in CMake to be shown
500        in the configuration summary.
501    
502    
503    Version 7.8 05-Sep-08
504    ---------------------
505    
506    1.  Replaced UCP searching code with optimized version as implemented for Ad
507        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
508        stage table and inline lookup instead of a function, giving speed ups of 2
509        to 5 times on some simple patterns that I tested. Permission was given to
510        distribute the MultiStage2.py script that generates the tables (it's not in
511        the tarball, but is in the Subversion repository).
512    
513    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
514        scripts.
515    
516    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
517        a group with a zero qualifier. The result of the study could be incorrect,
518        or the function might crash, depending on the pattern.
519    
520    4.  Caseless matching was not working for non-ASCII characters in back
521        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
522        It now works when Unicode Property Support is available.
523    
524    5.  In pcretest, an escape such as \x{de} in the data was always generating
525        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
526        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
527        truncation.
528    
529    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
530    
531    7.  Added two (int) casts to pcregrep when printing the difference of two
532        pointers, in case they are 64-bit values.
533    
534    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
535        test 2 if it fails.
536    
537    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
538        and a #define of that name to empty if it is not externally set. This is to
539        allow users of MSVC to set it if necessary.
540    
541    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
542        the convenience functions in the pcre_get.c source file.
543    
544    11. An option change at the start of a pattern that had top-level alternatives
545        could cause overwriting and/or a crash. This command provoked a crash in
546        some environments:
547    
548          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
549    
550        This potential security problem was recorded as CVE-2008-2371.
551    
552    12. For a pattern where the match had to start at the beginning or immediately
553        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
554        pcre_dfa_exec() could read past the end of the passed subject if there was
555        no match. To help with detecting such bugs (e.g. with valgrind), I modified
556        pcretest so that it places the subject at the end of its malloc-ed buffer.
557    
558    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
559        exec() might read past the end of the data buffer in UTF-8 mode.
560    
561    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
562        the data contained the byte 0x85 as part of a UTF-8 character within its
563        first line. This applied both to normal and DFA matching.
564    
565    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
566        /^[^d]*?$/8 failed to match "abc".
567    
568    16. Added a missing copyright notice to pcrecpp_internal.h.
569    
570    17. Make it more clear in the documentation that values returned from
571        pcre_exec() in ovector are byte offsets, not character counts.
572    
573    18. Tidied a few places to stop certain compilers from issuing warnings.
574    
575    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
576        supplied by Stefan Weber. I made a further small update for 7.8 because
577        there is a change of source arrangements: the pcre_searchfuncs.c module is
578        replaced by pcre_ucd.c.
579    
580    
581    Version 7.7 07-May-08
582    ---------------------
583    
584    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
585        a string to a long long, pretend we don't even have a long long." This is
586        done by checking for the strtoq, strtoll, and _strtoi64 functions.
587    
588    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
589        pre-7.6 versions, which defined a global no_arg variable instead of putting
590        it in the RE class. (See also #8 below.)
591    
592    3.  Remove a line of dead code, identified by coverity and reported by Nuno
593        Lopes.
594    
595    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
596    
597        (1) The include/exclude patterns were being applied to the whole pathnames
598            of files, instead of just to the final components.
599    
600        (2) If there was more than one level of directory, the subdirectories were
601            skipped unless they satisfied the include/exclude conditions. This is
602            inconsistent with GNU grep (and could even be seen as contrary to the
603            pcregrep specification - which I improved to make it absolutely clear).
604            The action now is always to scan all levels of directory, and just
605            apply the include/exclude patterns to regular files.
606    
607    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
608        --exclude_dir in the tests to avoid scanning .svn directories.
609    
610    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
611        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
612        doesn't support NULs in patterns.
613    
614    7.  Added some missing "const"s to declarations of static tables in
615        pcre_compile.c and pcre_dfa_exec.c.
616    
617    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
618        caused by fix #2  above. (Subsequently also a second patch to fix the
619        first patch. And a third patch - this was a messy problem.)
620    
621    9.  Applied Craig's patch to remove the use of push_back().
622    
623    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
624        matching function regexec().
625    
626    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
627        which, however, unlike Perl's \g{...}, are subroutine calls, not back
628        references. PCRE supports relative numbers with this syntax (I don't think
629        Oniguruma does).
630    
631    12. Previously, a group with a zero repeat such as (...){0} was completely
632        omitted from the compiled regex. However, this means that if the group
633        was called as a subroutine from elsewhere in the pattern, things went wrong
634        (an internal error was given). Such groups are now left in the compiled
635        pattern, with a new opcode that causes them to be skipped at execution
636        time.
637    
638    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
639        to the way PCRE behaves:
640    
641        (a) A lone ] character is dis-allowed (Perl treats it as data).
642    
643        (b) A back reference to an unmatched subpattern matches an empty string
644            (Perl fails the current match path).
645    
646        (c) A data ] in a character class must be notated as \] because if the
647            first data character in a class is ], it defines an empty class. (In
648            Perl it is not possible to have an empty class.) The empty class []
649            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
650            The negative empty class [^] matches any one character, independently
651            of the DOTALL setting.
652    
653    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
654        non-existent subpattern following a character class starting with ']' and
655        containing () gave an internal compiling error instead of "reference to
656        non-existent subpattern". Fortunately, when the pattern did exist, the
657        compiled code was correct. (When scanning forwards to check for the
658        existencd of the subpattern, it was treating the data ']' as terminating
659        the class, so got the count wrong. When actually compiling, the reference
660        was subsequently set up correctly.)
661    
662    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
663        it was being rejected as not supported by pcre_dfa_exec(), even though
664        other assertions are supported. I have made pcre_dfa_exec() support
665        (*FAIL).
666    
667    16. The implementation of 13c above involved the invention of a new opcode,
668        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
669        cannot be changed at match time, I realized I could make a small
670        improvement to matching performance by compiling OP_ALLANY instead of
671        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
672        on the OP_ANY path.
673    
674    17. Compiling pcretest on Windows with readline support failed without the
675        following two fixes: (1) Make the unistd.h include conditional on
676        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
677    
678    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
679        ncurses library to be included for pcretest when ReadLine support is
680        requested, but also to allow for it to be overridden. This patch came from
681        Daniel Bergström.
682    
683    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
684        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
685        any errors with the current Unicode tables. Thanks to Peter Kankowski for
686        spotting this.
687    
688    
689    Version 7.6 28-Jan-08
690    ---------------------
691    
692    1.  A character class containing a very large number of characters with
693        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
694        overflow.
695    
696    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
697        HAVE_LONG_LONG is not defined.
698    
699    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
700        bring them up to date and include new features. This patch includes:
701    
702        - Fixed PH's badly added libz and libbz2 support.
703        - Fixed a problem with static linking.
704        - Added pcredemo. [But later removed - see 7 below.]
705        - Fixed dftables problem and added an option.
706        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
707            HAVE_LONG_LONG.
708        - Added readline support for pcretest.
709        - Added an listing of the option settings after cmake has run.
710    
711    4.  A user submitted a patch to Makefile that makes it easy to create
712        "pcre.dll" under mingw when using Configure/Make. I added stuff to
713        Makefile.am that cause it to include this special target, without
714        affecting anything else. Note that the same mingw target plus all
715        the other distribution libraries and programs are now supported
716        when configuring with CMake (see 6 below) instead of with
717        Configure/Make.
718    
719    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
720        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
721        exported in the Windows port". It has not yet been confirmed that the patch
722        solves the problem, but it does no harm.
723    
724    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
725        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
726        with CMake, and also correct the comment about stack recursion.
727    
728    7.  Remove the automatic building of pcredemo from the ./configure system and
729        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
730        of a program that users should build themselves after PCRE is installed, so
731        building it automatically is not really right. What is more, it gave
732        trouble in some build environments.
733    
734    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
735    
736    
737    Version 7.5 10-Jan-08
738    ---------------------
739    
740    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
741        values in parens when parsing an RE using the C++ wrapper."
742    
743    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
744        Characters greater than 255 were excluded from the class instead of being
745        included.
746    
747    3.  The same bug as (2) above applied to negated POSIX classes such as
748        [:^space:].
749    
750    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
751        defined or documented. It seems to have been a typo for PCRE_STATIC, so
752        I have changed it.
753    
754    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
755        first named subpattern) and a construct such as (?&a) would reference the
756        first named subpattern whose name started with "a" (in other words, the
757        length check was missing). Both these problems are fixed. "Subpattern name
758        expected" is now given for (?&) (a zero-length name), and this patch also
759        makes it give the same error for \k'' (previously it complained that that
760        was a reference to a non-existent subpattern).
761    
762    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
763        this is right because (?- can be followed by option settings as well as by
764        digits. I have, however, made the messages clearer.
765    
766    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
767        than the number used in the conditional) now cause a compile-time error.
768        This is actually not compatible with Perl, which accepts such patterns, but
769        treats the conditional as always being FALSE (as PCRE used to), but it
770        seems to me that giving a diagnostic is better.
771    
772    8.  Change "alphameric" to the more common word "alphanumeric" in comments
773        and messages.
774    
775    9.  Fix two occurrences of "backslash" in comments that should have been
776        "backspace".
777    
778    10. Remove two redundant lines of code that can never be obeyed (their function
779        was moved elsewhere).
780    
781    11. The program that makes PCRE's Unicode character property table had a bug
782        which caused it to generate incorrect table entries for sequences of
783        characters that have the same character type, but are in different scripts.
784        It amalgamated them into a single range, with the script of the first of
785        them. In other words, some characters were in the wrong script. There were
786        thirteen such cases, affecting characters in the following ranges:
787    
788          U+002b0 - U+002c1
789          U+0060c - U+0060d
790          U+0061e - U+00612
791          U+0064b - U+0065e
792          U+0074d - U+0076d
793          U+01800 - U+01805
794          U+01d00 - U+01d77
795          U+01d9b - U+01dbf
796          U+0200b - U+0200f
797          U+030fc - U+030fe
798          U+03260 - U+0327f
799          U+0fb46 - U+0fbb1
800          U+10450 - U+1049d
801    
802    12. The -o option (show only the matching part of a line) for pcregrep was not
803        compatible with GNU grep in that, if there was more than one match in a
804        line, it showed only the first of them. It now behaves in the same way as
805        GNU grep.
806    
807    13. If the -o and -v options were combined for pcregrep, it printed a blank
808        line for every non-matching line. GNU grep prints nothing, and pcregrep now
809        does the same. The return code can be used to tell if there were any
810        non-matching lines.
811    
812    14. Added --file-offsets and --line-offsets to pcregrep.
813    
814    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
815        infinitely looping recursion. The bug was that positive lookaheads were not
816        being skipped when checking for a possible empty match (negative lookaheads
817        and both kinds of lookbehind were skipped).
818    
819    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
820        inclusion of <windows.h> to before rather than after the definition of
821        INVALID_FILE_ATTRIBUTES (patch from David Byron).
822    
823    17. Specifying a possessive quantifier with a specific limit for a Unicode
824        character property caused pcre_compile() to compile bad code, which led at
825        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
826        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
827        caused the error; without that there was no problem.
828    
829    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
830    
831    19. Added --enable-pcretest-libreadline.
832    
833    20. In pcrecpp.cc, the variable 'count' was incremented twice in
834        RE::GlobalReplace(). As a result, the number of replacements returned was
835        double what it should be. I removed one of the increments, but Craig sent a
836        later patch that removed the other one (the right fix) and added unit tests
837        that check the return values (which was not done before).
838    
839    21. Several CMake things:
840    
841        (1) Arranged that, when cmake is used on Unix, the libraries end up with
842            the names libpcre and libpcreposix, not just pcre and pcreposix.
843    
844        (2) The above change means that pcretest and pcregrep are now correctly
845            linked with the newly-built libraries, not previously installed ones.
846    
847        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
848    
849    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
850        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
851        UTF-8 newline character). The key issue is that the pattern starts .*;
852        this means that the match must be either at the beginning, or after a
853        newline. The bug was in the code for advancing after a failed match and
854        checking that the new position followed a newline. It was not taking
855        account of UTF-8 characters correctly.
856    
857    23. PCRE was behaving differently from Perl in the way it recognized POSIX
858        character classes. PCRE was not treating the sequence [:...:] as a
859        character class unless the ... were all letters. Perl, however, seems to
860        allow any characters between [: and :], though of course it rejects as
861        unknown any "names" that contain non-letters, because all the known class
862        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
863        for example, whereas PCRE did not - it did not recognize a POSIX character
864        class. This seemed a bit dangerous, so the code has been changed to be
865        closer to Perl. The behaviour is not identical to Perl, because PCRE will
866        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
867        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
868        Perl does, and where it didn't before.
869    
870    24. Rewrite so as to remove the single use of %n from pcregrep because in some
871        Windows environments %n is disabled by default.
872    
873    
874    Version 7.4 21-Sep-07
875    ---------------------
876    
877    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
878        means that a class such as [\s] counted as "explicit reference to CR or
879        LF". That isn't really right - the whole point of the change was to try to
880        help when there was an actual mention of one of the two characters. So now
881        the change happens only if \r or \n (or a literal CR or LF) character is
882        encountered.
883    
884    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
885        of both had grown to the point where there were only 3 bits left.
886        Fortunately, there was spare space in the data structure, and so I have
887        moved the internal flags into a new 16-bit field to free up more option
888        bits.
889    
890    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
891        but did not set the internal JCHANGED flag - either of these is enough to
892        control the way the "get" function works - but the PCRE_INFO_JCHANGED
893        facility is supposed to tell if (?J) was ever used, so now (?J) at the
894        start sets both bits.
895    
896    4.  Added options (at build time, compile time, exec time) to change \R from
897        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
898    
899    5.  doc/pcresyntax.html was missing from the distribution.
900    
901    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
902        compatibility, even though it is no longer used.
903    
904    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
905        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
906        windows.h file is present (where different names are used). [This was
907        reversed later after testing - see 16 below.]
908    
909    8.  Changed all #include <config.h> to #include "config.h". There were also
910        some further <pcre.h> cases that I changed to "pcre.h".
911    
912    9.  When pcregrep was used with the --colour option, it missed the line ending
913        sequence off the lines that it output.
914    
915    10. It was pointed out to me that arrays of string pointers cause lots of
916        relocations when a shared library is dynamically loaded. A technique of
917        using a single long string with a table of offsets can drastically reduce
918        these. I have refactored PCRE in four places to do this. The result is
919        dramatic:
920    
921          Originally:                          290
922          After changing UCP table:            187
923          After changing error message table:   43
924          After changing table of "verbs"       36
925          After changing table of Posix names   22
926    
927        Thanks to the folks working on Gregex for glib for this insight.
928    
929    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
930        unicode-properties was also set.
931    
932    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
933    
934    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
935        checked only for CRLF.
936    
937    14. Added casts to pcretest.c to avoid compiler warnings.
938    
939    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
940    
941    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
942        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
943        entirely. This removes changes made in 7 above.
944    
945    17. The CMake files have been updated, and there is now more information about
946        building with CMake in the NON-UNIX-USE document.
947    
948    
949    Version 7.3 28-Aug-07
950    ---------------------
951    
952     1. In the rejigging of the build system that eventually resulted in 7.1, the
953        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
954        brackets there is not right, since it causes compilers to look for an
955        installed pcre.h, not the version that is in the source that is being
956        compiled (which of course may be different). I have changed it back to:
957    
958          #include "pcre.h"
959    
960        I have a vague recollection that the change was concerned with compiling in
961        different directories, but in the new build system, that is taken care of
962        by the VPATH setting the Makefile.
963    
964     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
965        when the subject happened to end in the byte 0x85 (e.g. if the last
966        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
967        characters but of course it shouldn't be taken as a newline when it is part
968        of another character. The bug was that, for an unlimited repeat of . in
969        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
970        characters when looking for a newline.
971    
972     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
973    
974     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
975        in debug output.
976    
977     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
978        long printing in the pcrecpp unittest when running under MinGW.
979    
980     6. ESC_K was left out of the EBCDIC table.
981    
982     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
983        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
984        limit also applies to "virtual nesting" when a pattern is recursive, and in
985        this case 1000 isn't so big. I have been able to remove this limit at the
986        expense of backing off one optimization in certain circumstances. Normally,
987        when pcre_exec() would call its internal match() function recursively and
988        immediately return the result unconditionally, it uses a "tail recursion"
989        feature to save stack. However, when a subpattern that can match an empty
990        string has an unlimited repetition quantifier, it no longer makes this
991        optimization. That gives it a stack frame in which to save the data for
992        checking that an empty string has been matched. Previously this was taken
993        from the 1000-entry workspace that had been reserved. So now there is no
994        explicit limit, but more stack is used.
995    
996     8. Applied Daniel's patches to solve problems with the import/export magic
997        syntax that is required for Windows, and which was going wrong for the
998        pcreposix and pcrecpp parts of the library. These were overlooked when this
999        problem was solved for the main library.
1000    
1001     9. There were some crude static tests to avoid integer overflow when computing
1002        the size of patterns that contain repeated groups with explicit upper
1003        limits. As the maximum quantifier is 65535, the maximum group length was
1004        set at 30,000 so that the product of these two numbers did not overflow a
1005        32-bit integer. However, it turns out that people want to use groups that
1006        are longer than 30,000 bytes (though not repeat them that many times).
1007        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
1008        made it possible to implement the integer overflow checks in a much more
1009        dynamic way, which I have now done. The artificial limitation on group
1010        length has been removed - we now have only the limit on the total length of
1011        the compiled pattern, which depends on the LINK_SIZE setting.
1012    
1013    10. Fixed a bug in the documentation for get/copy named substring when
1014        duplicate names are permitted. If none of the named substrings are set, the
1015        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
1016        empty string.
1017    
1018    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
1019        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
1020        because the ] is interpreted as the first data character and the
1021        terminating ] is not found. PCRE has been made compatible with Perl in this
1022        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
1023        cause memory overwriting.
1024    
1025    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
1026        string has been matched (to stop an infinite loop). It was not recognizing
1027        a conditional subpattern that could match an empty string if that
1028        subpattern was within another subpattern. For example, it looped when
1029        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
1030        condition was not nested. This bug has been fixed.
1031    
1032    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
1033        past the start of the subject in the presence of bytes with the top bit
1034        set, for example "\x8aBCD".
1035    
1036    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
1037        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
1038    
1039    14. Optimized (?!) to (*FAIL).
1040    
1041    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
1042        This restricts code points to be within the range 0 to 0x10FFFF, excluding
1043        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
1044        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
1045        does: it's just the validity check that is more restrictive.
1046    
1047    16. Inserted checks for integer overflows during escape sequence (backslash)
1048        processing, and also fixed erroneous offset values for syntax errors during
1049        backslash processing.
1050    
1051    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
1052        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
1053    
1054    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
1055        caused an overrun.
1056    
1057    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
1058        something other than just ASCII characters) inside a group that had an
1059        unlimited repeat caused a loop at compile time (while checking to see
1060        whether the group could match an empty string).
1061    
1062    20. Debugging a pattern containing \p or \P could cause a crash. For example,
1063        [\P{Any}] did so. (Error in the code for printing property names.)
1064    
1065    21. An orphan \E inside a character class could cause a crash.
1066    
1067    22. A repeated capturing bracket such as (A)? could cause a wild memory
1068        reference during compilation.
1069    
1070    23. There are several functions in pcre_compile() that scan along a compiled
1071        expression for various reasons (e.g. to see if it's fixed length for look
1072        behind). There were bugs in these functions when a repeated \p or \P was
1073        present in the pattern. These operators have additional parameters compared
1074        with \d, etc, and these were not being taken into account when moving along
1075        the compiled data. Specifically:
1076    
1077        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
1078            length.
1079    
1080        (b) An item such as \pL+ within a repeated group could cause crashes or
1081            loops.
1082    
1083        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
1084            "reference to non-existent subpattern" error.
1085    
1086        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
1087    
1088    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
1089        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
1090    
1091    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
1092    
1093    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
1094        character were causing crashes (broken optimization).
1095    
1096    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
1097        \p or \P) caused a compile-time loop.
1098    
1099    28. More problems have arisen in unanchored patterns when CRLF is a valid line
1100        break. For example, the unstudied pattern [\r\n]A does not match the string
1101        "\r\nA" because change 7.0/46 below moves the current point on by two
1102        characters after failing to match at the start. However, the pattern \nA
1103        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
1104        the same is true. There doesn't seem any very clean way out of this, but
1105        what I have chosen to do makes the common cases work: PCRE now takes note
1106        of whether there can be an explicit match for \r or \n anywhere in the
1107        pattern, and if so, 7.0/46 no longer applies. As part of this change,
1108        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
1109        pattern has explicit CR or LF references.
1110    
1111    29. Added (*CR) etc for changing newline setting at start of pattern.
1112    
1113    
1114    Version 7.2 19-Jun-07
1115    ---------------------
1116    
1117     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
1118        which is apparently normally available under Windows.
1119    
1120     2. Re-jig the pcregrep tests with different newline settings in an attempt
1121        to make them independent of the local environment's newline setting.
1122    
1123     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
1124    
1125     4. Some of the "internals" tests were previously cut out when the link size
1126        was not 2, because the output contained actual offsets. The recent new
1127        "Z" feature of pcretest means that these can be cut out, making the tests
1128        usable with all link sizes.
1129    
1130     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
1131        stack recursion. This gives a massive performance boost under BSD, but just
1132        a small improvement under Linux. However, it saves one field in the frame
1133        in all cases.
1134    
1135     6. Added more features from the forthcoming Perl 5.10:
1136    
1137        (a) (?-n) (where n is a string of digits) is a relative subroutine or
1138            recursion call. It refers to the nth most recently opened parentheses.
1139    
1140        (b) (?+n) is also a relative subroutine call; it refers to the nth next
1141            to be opened parentheses.
1142    
1143        (c) Conditions that refer to capturing parentheses can be specified
1144            relatively, for example, (?(-2)... or (?(+3)...
1145    
1146        (d) \K resets the start of the current match so that everything before
1147            is not part of it.
1148    
1149        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
1150    
1151        (f) \g{name} is another synonym - part of Perl 5.10's unification of
1152            reference syntax.
1153    
1154        (g) (?| introduces a group in which the numbering of parentheses in each
1155            alternative starts with the same number.
1156    
1157        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
1158    
1159     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
1160        PCRE_INFO_JCHANGED.
1161    
1162     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
1163        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
1164        for detecting groups that can match an empty string.
1165    
1166     9. A pattern with a very large number of alternatives (more than several
1167        hundred) was running out of internal workspace during the pre-compile
1168        phase, where pcre_compile() figures out how much memory will be needed. A
1169        bit of new cunning has reduced the workspace needed for groups with
1170        alternatives. The 1000-alternative test pattern now uses 12 bytes of
1171        workspace instead of running out of the 4096 that are available.
1172    
1173    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
1174    
1175    11. Applied patch from Google to remove an optimization that didn't quite work.
1176        The report of the bug said:
1177    
1178          pcrecpp::RE("a*").FullMatch("aaa") matches, while
1179          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
1180          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
1181    
1182    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
1183        it matched the wrong number of bytes.
1184    
1185    
1186    Version 7.1 24-Apr-07
1187    ---------------------
1188    
1189     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
1190        that is more "standard", making use of automake and other Autotools. There
1191        is some re-arrangement of the files and adjustment of comments consequent
1192        on this.
1193    
1194     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
1195        for recursive directory scanning broke on some systems because the files
1196        are not scanned in any specific order and on different systems the order
1197        was different. A call to "sort" has been inserted into RunGrepTest for the
1198        approprate test as a short-term fix. In the longer term there may be an
1199        alternative.
1200    
1201     3. I had an email from Eric Raymond about problems translating some of PCRE's
1202        man pages to HTML (despite the fact that I distribute HTML pages, some
1203        people do their own conversions for various reasons). The problems
1204        concerned the use of low-level troff macros .br and .in. I have therefore
1205        removed all such uses from the man pages (some were redundant, some could
1206        be replaced by .nf/.fi pairs). The 132html script that I use to generate
1207        HTML has been updated to handle .nf/.fi and to complain if it encounters
1208        .br or .in.
1209    
1210     4. Updated comments in configure.ac that get placed in config.h.in and also
1211        arranged for config.h to be included in the distribution, with the name
1212        config.h.generic, for the benefit of those who have to compile without
1213        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
1214    
1215     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
1216        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
1217        makevp.bat for the current PCRE, using the additional files
1218        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
1219    
1220     6. A Windows user reported a minor discrepancy with test 2, which turned out
1221        to be caused by a trailing space on an input line that had got lost in his
1222        copy. The trailing space was an accident, so I've just removed it.
1223    
1224     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
1225        that is needed.
1226    
1227     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
1228        as "const" (a) because they are and (b) because it helps the PHP
1229        maintainers who have recently made a script to detect big data structures
1230        in the php code that should be moved to the .rodata section. I remembered
1231        to update Builducptable as well, so it won't revert if ucptable.h is ever
1232        re-created.
1233    
1234     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
1235        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
1236        order to be able to cut out the UTF-8 tables in the latter when UTF-8
1237        support is not required. This saves 1.5-2K of code, which is important in
1238        some applications.
1239    
1240        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
1241        so as not to refer to the tables, even though these functions will never be
1242        called when UTF-8 support is disabled. Otherwise there are problems with a
1243        shared library.
1244    
1245    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
1246    
1247        (a) It was defining its arguments as char * instead of void *.
1248    
1249        (b) It was assuming that all moves were upwards in memory; this was true
1250            a long time ago when I wrote it, but is no longer the case.
1251    
1252        The emulated memove() is provided for those environments that have neither
1253        memmove() nor bcopy(). I didn't think anyone used it these days, but that
1254        is clearly not the case, as these two bugs were recently reported.
1255    
1256    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
1257        and Detrail to create the HTML documentation, the .txt form of the man
1258        pages, and it removes trailing spaces from listed files. It also creates
1259        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
1260        case, it wraps all the #defines with #ifndefs. This script should be run
1261        before "make dist".
1262    
1263    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
1264        with Unicode property support.
1265    
1266        (a) For a maximizing quantifier, if the two different cases of the
1267            character were of different lengths in their UTF-8 codings (there are
1268            some cases like this - I found 11), and the matching function had to
1269            back up over a mixture of the two cases, it incorrectly assumed they
1270            were both the same length.
1271    
1272        (b) When PCRE was configured to use the heap rather than the stack for
1273            recursion during matching, it was not correctly preserving the data for
1274            the other case of a UTF-8 character when checking ahead for a match
1275            while processing a minimizing repeat. If the check also involved
1276            matching a wide character, but failed, corruption could cause an
1277            erroneous result when trying to check for a repeat of the original
1278            character.
1279    
1280    13. Some tidying changes to the testing mechanism:
1281    
1282        (a) The RunTest script now detects the internal link size and whether there
1283            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
1284            values substituted by "configure". (The RunGrepTest script already did
1285            this for UTF-8.) The configure.ac script no longer substitutes the
1286            relevant variables.
1287    
1288        (b) The debugging options /B and /D in pcretest show the compiled bytecode
1289            with length and offset values. This means that the output is different
1290            for different internal link sizes. Test 2 is skipped for link sizes
1291            other than 2 because of this, bypassing the problem. Unfortunately,
1292            there was also a test in test 3 (the locale tests) that used /B and
1293            failed for link sizes other than 2. Rather than cut the whole test out,
1294            I have added a new /Z option to pcretest that replaces the length and
1295            offset values with spaces. This is now used to make test 3 independent
1296            of link size. (Test 2 will be tidied up later.)
1297    
1298    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
1299        segmentation fault instead of returning the appropriate error message.
1300    
1301    15. In multiline mode when the newline sequence was set to "any", the pattern
1302        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
1303        This doesn't seem right; it now treats the CRLF combination as the line
1304        ending, and so does not match in that case. It's only a pattern such as ^$
1305        that would hit this one: something like ^ABC$ would have failed after \r
1306        and then tried again after \r\n.
1307    
1308    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
1309        in an attempt to make files that differ only in their line terminators
1310        compare equal. This works on Linux.
1311    
1312    17. Under certain error circumstances pcregrep might try to free random memory
1313        as it exited. This is now fixed, thanks to valgrind.
1314    
1315    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
1316        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
1317        was because its rules for how to advance for /g after matching an empty
1318        string at the end of a line did not allow for this case. They now check for
1319        it specially.
1320    
1321    20. pcretest is supposed to handle patterns and data of any length, by
1322        extending its buffers when necessary. It was getting this wrong when the
1323        buffer for a data line had to be extended.
1324    
1325    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
1326        CRLF as a newline sequence.
1327    
1328    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
1329        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
1330        I have nevertheless tidied it up.
1331    
1332    23. Added some casts to kill warnings from HP-UX ia64 compiler.
1333    
1334    24. Added a man page for pcre-config.
1335    
1336    
1337    Version 7.0 19-Dec-06
1338    ---------------------
1339    
1340     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
1341        moving to gcc 4.1.1.
1342    
1343     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
1344        sys/time.h, which is documented as needed for this function. It doesn't
1345        seem to matter on Linux, but it showed up on some releases of OS X.
1346    
1347     3. It seems that there are systems where bytes whose values are greater than
1348        127 match isprint() in the "C" locale. The "C" locale should be the
1349        default when a C program starts up. In most systems, only ASCII printing
1350        characters match isprint(). This difference caused the output from pcretest
1351        to vary, making some of the tests fail. I have changed pcretest so that:
1352    
1353        (a) When it is outputting text in the compiled version of a pattern, bytes
1354            other than 32-126 are always shown as hex escapes.
1355    
1356        (b) When it is outputting text that is a matched part of a subject string,
1357            it does the same, unless a different locale has been set for the match
1358            (using the /L modifier). In this case, it uses isprint() to decide.
1359    
1360     4. Fixed a major bug that caused incorrect computation of the amount of memory
1361        required for a compiled pattern when options that changed within the
1362        pattern affected the logic of the preliminary scan that determines the
1363        length. The relevant options are -x, and -i in UTF-8 mode. The result was
1364        that the computed length was too small. The symptoms of this bug were
1365        either the PCRE error "internal error: code overflow" from pcre_compile(),
1366        or a glibc crash with a message such as "pcretest: free(): invalid next
1367        size (fast)". Examples of patterns that provoked this bug (shown in
1368        pcretest format) are:
1369    
1370          /(?-x: )/x
1371          /(?x)(?-x: \s*#\s*)/
1372          /((?i)[\x{c0}])/8
1373          /(?i:[\x{c0}])/8
1374    
1375        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
1376        is now done differently.
1377    
1378     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
1379        wrapper classes; (b) implement a new function in the C++ scanner that is
1380        more efficient than the old way of doing things because it avoids levels of
1381        recursion in the regex matching; (c) add a paragraph to the documentation
1382        for the FullMatch() function.
1383    
1384     6. The escape sequence \n was being treated as whatever was defined as
1385        "newline". Not only was this contrary to the documentation, which states
1386        that \n is character 10 (hex 0A), but it also went horribly wrong when
1387        "newline" was defined as CRLF. This has been fixed.
1388    
1389     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
1390        was being set to -1 for the "end of line" case (supposedly a value that no
1391        character can have). Though this value is never used (the check for end of
1392        line is "zero bytes in current character"), it caused compiler complaints.
1393        I've changed it to 0xffffffff.
1394    
1395     8. In pcre_version.c, the version string was being built by a sequence of
1396        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
1397        string (as it is for production releases) called a macro with an empty
1398        argument. The C standard says the result of this is undefined. The gcc
1399        compiler treats it as an empty string (which was what was wanted) but it is
1400        reported that Visual C gives an error. The source has been hacked around to
1401        avoid this problem.
1402    
1403     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
1404        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
1405        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
1406        of them did).
1407    
1408    10. Originally, pcretest opened its input and output without "b"; then I was
1409        told that "b" was needed in some environments, so it was added for release
1410        5.0 to both the input and output. (It makes no difference on Unix-like
1411        systems.) Later I was told that it is wrong for the input on Windows. I've
1412        now abstracted the modes into two macros, to make it easier to fiddle with
1413        them, and removed "b" from the input mode under Windows.
1414    
1415    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
1416    
1417    12. Added -help and --help to pcretest as an official way of being reminded
1418        of the options.
1419    
1420    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
1421        and pcrecpp.cc because they annoy compilers at high warning levels.
1422    
1423    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
1424    
1425    15. Fixed an occurrence of == in configure.ac that should have been = (shell
1426        scripts are not C programs :-) and which was not noticed because it works
1427        on Linux.
1428    
1429    16. pcretest is supposed to handle any length of pattern and data line (as one
1430        line or as a continued sequence of lines) by extending its input buffer if
1431        necessary. This feature was broken for very long pattern lines, leading to
1432        a string of junk being passed to pcre_compile() if the pattern was longer
1433        than about 50K.
1434    
1435    17. I have done a major re-factoring of the way pcre_compile() computes the
1436        amount of memory needed for a compiled pattern. Previously, there was code
1437        that made a preliminary scan of the pattern in order to do this. That was
1438        OK when PCRE was new, but as the facilities have expanded, it has become
1439        harder and harder to keep it in step with the real compile phase, and there
1440        have been a number of bugs (see for example, 4 above). I have now found a
1441        cunning way of running the real compile function in a "fake" mode that
1442        enables it to compute how much memory it would need, while actually only
1443        ever using a few hundred bytes of working memory and without too many
1444        tests of the mode. This should make future maintenance and development
1445        easier. A side effect of this work is that the limit of 200 on the nesting
1446        depth of parentheses has been removed (though this was never a serious
1447        limitation, I suspect). However, there is a downside: pcre_compile() now
1448        runs more slowly than before (30% or more, depending on the pattern). I
1449        hope this isn't a big issue. There is no effect on runtime performance.
1450    
1451    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
1452        newline (only possible for the last line of a file) and it was a
1453        pattern that set a locale (followed by /Lsomething), pcretest crashed.
1454    
1455    19. Added additional timing features to pcretest. (1) The -tm option now times
1456        matching only, not compiling. (2) Both -t and -tm can be followed, as a
1457        separate command line item, by a number that specifies the number of
1458        repeats to use when timing. The default is 50000; this gives better
1459        precision, but takes uncomfortably long for very large patterns.
1460    
1461    20. Extended pcre_study() to be more clever in cases where a branch of a
1462        subpattern has no definite first character. For example, (a*|b*)[cd] would
1463        previously give no result from pcre_study(). Now it recognizes that the
1464        first character must be a, b, c, or d.
1465    
1466    21. There was an incorrect error "recursive call could loop indefinitely" if
1467        a subpattern (or the entire pattern) that was being tested for matching an
1468        empty string contained only one non-empty item after a nested subpattern.
1469        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
1470        incorrectly, because the \d was being skipped in the check.
1471    
1472    22. The pcretest program now has a new pattern option /B and a command line
1473        option -b, which is equivalent to adding /B to every pattern. This causes
1474        it to show the compiled bytecode, without the additional information that
1475        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
1476        is the same as /B/I).
1477    
1478    23. A new optimization is now able automatically to treat some sequences such
1479        as a*b as a*+b. More specifically, if something simple (such as a character
1480        or a simple class like \d) has an unlimited quantifier, and is followed by
1481        something that cannot possibly match the quantified thing, the quantifier
1482        is automatically "possessified".
1483    
1484    24. A recursive reference to a subpattern whose number was greater than 39
1485        went wrong under certain circumstances in UTF-8 mode. This bug could also
1486        have affected the operation of pcre_study().
1487    
1488    25. Realized that a little bit of performance could be had by replacing
1489        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
1490    
1491    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
1492    
1493    27. Possessive quantifiers such as a++ were previously implemented by turning
1494        them into atomic groups such as ($>a+). Now they have their own opcodes,
1495        which improves performance. This includes the automatically created ones
1496        from 23 above.
1497    
1498    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
1499        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
1500        the first matched character to be a colon. This applied both to named and
1501        numbered groups.
1502    
1503    29. The ucpinternal.h header file was missing its idempotency #ifdef.
1504    
1505    30. I was sent a "project" file called libpcre.a.dev which I understand makes
1506        building PCRE on Windows easier, so I have included it in the distribution.
1507    
1508    31. There is now a check in pcretest against a ridiculously large number being
1509        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
1510        loop, the loop is abandoned.
1511    
1512    32. Forward references to subpatterns in conditions such as (?(2)...) where
1513        subpattern 2 is defined later cause pcre_compile() to search forwards in
1514        the pattern for the relevant set of parentheses. This search went wrong
1515        when there were unescaped parentheses in a character class, parentheses
1516        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
1517    
1518    33. "Subroutine" calls and backreferences were previously restricted to
1519        referencing subpatterns earlier in the regex. This restriction has now
1520        been removed.
1521    
1522    34. Added a number of extra features that are going to be in Perl 5.10. On the
1523        whole, these are just syntactic alternatives for features that PCRE had
1524        previously implemented using the Python syntax or my own invention. The
1525        other formats are all retained for compatibility.
1526    
1527        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
1528            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
1529            also .NET compatible.
1530    
1531        (b) A recursion or subroutine call to a named group can now be defined as
1532            (?&name) as well as (?P>name).
1533    
1534        (c) A backreference to a named group can now be defined as \k<name> or
1535            \k'name' as well as (?P=name). The new forms, as well as being in Perl
1536            5.10, are also .NET compatible.
1537    
1538        (d) A conditional reference to a named group can now use the syntax
1539            (?(<name>) or (?('name') as well as (?(name).
1540    
1541        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
1542            groups (named and numbered) that are never evaluated inline, but can be
1543            called as "subroutines" from elsewhere. In effect, the DEFINE condition
1544            is always false. There may be only one alternative in such a group.
1545    
1546        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
1547            as the simple (?(R). The condition is true only if the most recent
1548            recursion is that of the given number or name. It does not search out
1549            through the entire recursion stack.
1550    
1551        (g) The escape \gN or \g{N} has been added, where N is a positive or
1552            negative number, specifying an absolute or relative reference.
1553    
1554    35. Tidied to get rid of some further signed/unsigned compiler warnings and
1555        some "unreachable code" warnings.
1556    
1557    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
1558        things, this adds five new scripts.
1559    
1560    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
1561        There were also incompatibilities regarding the handling of \Q..\E inside
1562        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
1563        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
1564    
1565    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
1566        matches an empty string, and forcibly breaks the loop. There were bugs in
1567        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
1568        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
1569        separate and independent bugs (that affected different cases) have been
1570        fixed.
1571    
1572    39. Refactored the code to abolish the use of different opcodes for small
1573        capturing bracket numbers. This is a tidy that I avoided doing when I
1574        removed the limit on the number of capturing brackets for 3.5 back in 2001.
1575        The new approach is not only tidier, it makes it possible to reduce the
1576        memory needed to fix the previous bug (38).
1577    
1578    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
1579        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
1580        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
1581        mode.
1582    
1583    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
1584        report.
1585    
1586    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
1587        copy construction and assignment in the C++ wrapper.
1588    
1589    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
1590        couple of bugs that could have given wrong results in the "--newline=crlf"
1591        case.
1592    
1593    44. Added a number of casts and did some reorganization of signed/unsigned int
1594        variables following suggestions from Dair Grant. Also renamed the variable
1595        "this" as "item" because it is a C++ keyword.
1596    
1597    45. Arranged for dftables to add
1598    
1599          #include "pcre_internal.h"
1600    
1601        to pcre_chartables.c because without it, gcc 4.x may remove the array
1602        definition from the final binary if PCRE is built into a static library and
1603        dead code stripping is activated.
1604    
1605    46. For an unanchored pattern, if a match attempt fails at the start of a
1606        newline sequence, and the newline setting is CRLF or ANY, and the next two
1607        characters are CRLF, advance by two characters instead of one.
1608    
1609    
1610    Version 6.7 04-Jul-06
1611    ---------------------
1612    
1613     1. In order to handle tests when input lines are enormously long, pcretest has
1614        been re-factored so that it automatically extends its buffers when
1615        necessary. The code is crude, but this _is_ just a test program. The
1616        default size has been increased from 32K to 50K.
1617    
1618     2. The code in pcre_study() was using the value of the re argument before
1619        testing it for NULL. (Of course, in any sensible call of the function, it
1620        won't be NULL.)
1621    
1622     3. The memmove() emulation function in pcre_internal.h, which is used on
1623        systems that lack both memmove() and bcopy() - that is, hardly ever -
1624        was missing a "static" storage class specifier.
1625    
1626     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
1627        containing an extended class (one that cannot be represented by a bitmap
1628        because it contains high-valued characters or Unicode property items, e.g.
1629        [\pZ]). Almost always one would set UTF-8 mode when processing such a
1630        pattern, but PCRE should not loop if you do not (it no longer does).
1631        [Detail: two cases were found: (a) a repeated subpattern containing an
1632        extended class; (b) a recursive reference to a subpattern that followed a
1633        previous extended class. It wasn't skipping over the extended class
1634        correctly when UTF-8 mode was not set.]
1635    
1636     5. A negated single-character class was not being recognized as fixed-length
1637        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
1638        compile error "lookbehind assertion is not fixed length".
1639    
1640     6. The RunPerlTest auxiliary script was showing an unexpected difference
1641        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
1642        write a Perl script that can interpret lines of an input file either as
1643        byte characters or as UTF-8, which is what "perltest" was being required to
1644        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
1645        can't do is switch easily at run time between having the "use utf8;" pragma
1646        or not. In the end, I fudged it by using the RunPerlTest script to insert
1647        "use utf8;" explicitly for the UTF-8 tests.
1648    
1649     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
1650        the end of the subject string, contrary to the documentation and to what
1651        Perl does. This was true of both matching functions. Now it matches only at
1652        the start of the subject and immediately after *internal* newlines.
1653    
1654     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
1655        a pointer to an int instead of a pointer to an unsigned long int. This
1656        caused problems on 64-bit systems.
1657    
1658     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
1659        instance of the 'standard' template library not being so standard".
1660    
1661    10. There was no check on the number of named subpatterns nor the maximum
1662        length of a subpattern name. The product of these values is used to compute
1663        the size of the memory block for a compiled pattern. By supplying a very
1664        long subpattern name and a large number of named subpatterns, the size
1665        computation could be caused to overflow. This is now prevented by limiting
1666        the length of names to 32 characters, and the number of named subpatterns
1667        to 10,000.
1668    
1669    11. Subpatterns that are repeated with specific counts have to be replicated in
1670        the compiled pattern. The size of memory for this was computed from the
1671        length of the subpattern and the repeat count. The latter is limited to
1672        65535, but there was no limit on the former, meaning that integer overflow
1673        could in principle occur. The compiled length of a repeated subpattern is
1674        now limited to 30,000 bytes in order to prevent this.
1675    
1676    12. Added the optional facility to have named substrings with the same name.
1677    
1678    13. Added the ability to use a named substring as a condition, using the
1679        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
1680        are numbers (not recommended). Forward references are permitted.
1681    
1682    14. Added forward references in named backreferences (if you see what I mean).
1683    
1684    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
1685        pattern could run off the end of the subject. For example, the pattern
1686        "(?s)(.{1,5})"8 did this with the subject "ab".
1687    
1688    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
1689        PCRE_CASELESS was set when matching characters that were quantified with ?
1690        or *.
1691    
1692    17. A character class other than a single negated character that had a minimum
1693        but no maximum quantifier - for example [ab]{6,} - was not handled
1694        correctly by pce_dfa_exec(). It would match only one character.
1695    
1696    18. A valid (though odd) pattern that looked like a POSIX character
1697        class but used an invalid character after [ (for example [[,abc,]]) caused
1698        pcre_compile() to give the error "Failed: internal error: code overflow" or
1699        in some cases to crash with a glibc free() error. This could even happen if
1700        the pattern terminated after [[ but there just happened to be a sequence of
1701        letters, a binary zero, and a closing ] in the memory that followed.
1702    
1703    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
1704        over the years. Originally (before any Unicode support), just the bottom 8
1705        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
1706        output from "man perlunicode" includes this:
1707    
1708          The regular expression compiler produces polymorphic opcodes.  That
1709          is, the pattern adapts to the data and automatically switches to
1710          the Unicode character scheme when presented with Unicode data--or
1711          instead uses a traditional byte scheme when presented with byte
1712          data.
1713    
1714        Sadly, a wide octal escape does not cause a switch, and in a string with
1715        no other multibyte characters, these octal escapes are treated as before.
1716        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
1717        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
1718        Unicode string.
1719    
1720        I have not perpetrated such confusion in PCRE. Up till now, it took just
1721        the bottom 8 bits, as in old Perl. I have now made octal escapes with
1722        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
1723        translate to the appropriate multibyte character.
1724    
1725    29. Applied some refactoring to reduce the number of warnings from Microsoft
1726        and Borland compilers. This has included removing the fudge introduced
1727        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
1728        a warning about an unused variable.
1729    
1730    21. PCRE has not included VT (character 0x0b) in the set of whitespace
1731        characters since release 4.0, because Perl (from release 5.004) does not.
1732        [Or at least, is documented not to: some releases seem to be in conflict
1733        with the documentation.] However, when a pattern was studied with
1734        pcre_study() and all its branches started with \s, PCRE still included VT
1735        as a possible starting character. Of course, this did no harm; it just
1736        caused an unnecessary match attempt.
1737    
1738    22. Removed a now-redundant internal flag bit that recorded the fact that case
1739        dependency changed within the pattern. This was once needed for "required
1740        byte" processing, but is no longer used. This recovers a now-scarce options
1741        bit. Also moved the least significant internal flag bit to the most-
1742        significant bit of the word, which was not previously used (hangover from
1743        the days when it was an int rather than a uint) to free up another bit for
1744        the future.
1745    
1746    23. Added support for CRLF line endings as well as CR and LF. As well as the
1747        default being selectable at build time, it can now be changed at runtime
1748        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
1749        specify that it is scanning data with non-default line endings.
1750    
1751    24. Changed the definition of CXXLINK to make it agree with the definition of
1752        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
1753    
1754    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
1755        recursions. This makes a big different to stack usage for some patterns.
1756    
1757    26. If a subpattern containing a named recursion or subroutine reference such
1758        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
1759        the space required for the compiled pattern went wrong and gave too small a
1760        value. Depending on the environment, this could lead to "Failed: internal
1761        error: code overflow at offset 49" or "glibc detected double free or
1762        corruption" errors.
1763    
1764    27. Applied patches from Google (a) to support the new newline modes and (b) to
1765        advance over multibyte UTF-8 characters in GlobalReplace.
1766    
1767    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
1768        difference for some implementation of PCRE in some Windows version.
1769    
1770    29. Added some extra testing facilities to pcretest:
1771    
1772        \q<number>   in a data line sets the "match limit" value
1773        \Q<number>   in a data line sets the "match recursion limt" value
1774        -S <number>  sets the stack size, where <number> is in megabytes
1775    
1776        The -S option isn't available for Windows.
1777    
1778    
1779    Version 6.6 06-Feb-06
1780    ---------------------
1781    
1782     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
1783        in pcreposix.h. I have copied the definition from pcre.h.
1784    
1785     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
1786        because pcre.h is no longer a built file.
1787    
1788     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
1789        not normally included in the compiled code.
1790    
1791    
1792    Version 6.5 01-Feb-06
1793    ---------------------
1794    
1795     1. When using the partial match feature with pcre_dfa_exec(), it was not
1796        anchoring the second and subsequent partial matches at the new starting
1797        point. This could lead to incorrect results. For example, with the pattern
1798        /1234/, partially matching against "123" and then "a4" gave a match.
1799    
1800     2. Changes to pcregrep:
1801    
1802        (a) All non-match returns from pcre_exec() were being treated as failures
1803            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
1804            error message is output. Some extra information is given for the
1805            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
1806            probably the only errors that are likely to be caused by users (by
1807            specifying a regex that has nested indefinite repeats, for instance).
1808            If there are more than 20 of these errors, pcregrep is abandoned.
1809    
1810        (b) A binary zero was treated as data while matching, but terminated the
1811            output line if it was written out. This has been fixed: binary zeroes
1812            are now no different to any other data bytes.
1813    
1814        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
1815            used to set a locale for matching. The --locale=xxxx long option has
1816            been added (no short equivalent) to specify a locale explicitly on the
1817            pcregrep command, overriding the environment variables.
1818    
1819        (d) When -B was used with -n, some line numbers in the output were one less
1820            than they should have been.
1821    
1822        (e) Added the -o (--only-matching) option.
1823    
1824        (f) If -A or -C was used with -c (count only), some lines of context were
1825            accidentally printed for the final match.
1826    
1827        (g) Added the -H (--with-filename) option.
1828    
1829        (h) The combination of options -rh failed to suppress file names for files
1830            that were found from directory arguments.
1831    
1832        (i) Added the -D (--devices) and -d (--directories) options.
1833    
1834        (j) Added the -F (--fixed-strings) option.
1835    
1836        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1837    
1838        (l) Added the --colo(u)r option.
1839    
1840        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1841            is not present by default.
1842    
1843     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1844        items such as (?R) or (?1), when the recursion could match a number of
1845        alternatives. If it matched one of the alternatives, but subsequently,
1846        outside the recursion, there was a failure, the code tried to back up into
1847        the recursion. However, because of the way PCRE is implemented, this is not
1848        possible, and the result was an incorrect result from the match.
1849    
1850        In order to prevent this happening, the specification of recursion has
1851        been changed so that all such subpatterns are automatically treated as
1852        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1853    
1854     4. I had overlooked the fact that, in some locales, there are characters for
1855        which isalpha() is true but neither isupper() nor islower() are true. In
1856        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1857        and ordfeminine) are like this. This affected the treatment of \w and \W
1858        when they appeared in character classes, but not when they appeared outside
1859        a character class. The bit map for "word" characters is now created
1860        separately from the results of isalnum() instead of just taking it from the
1861        upper, lower, and digit maps. (Plus the underscore character, of course.)
1862    
1863     5. The above bug also affected the handling of POSIX character classes such as
1864        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1865        permanent tables. Instead, the bit maps for such a class were previously
1866        created as the appropriate unions of the upper, lower, and digit bitmaps.
1867        Now they are created by subtraction from the [[:word:]] class, which has
1868        its own bitmap.
1869    
1870     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1871        It is created by subtracting the vertical space characters (\x09, \x0a,
1872        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1873        subtraction was done in the overall bitmap for a character class, meaning
1874        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1875        be recognized. This bug has been fixed.
1876    
1877     7. Patches from the folks at Google:
1878    
1879          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1880          real life, but is still worth protecting against".
1881    
1882          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1883          regular expressions".
1884    
1885          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1886          have it.
1887    
1888          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1889          "configure" and the latter not, in order to fix a problem somebody had
1890          with compiling the Arg class on HP-UX.
1891    
1892          (e) Improve the error-handling of the C++ wrapper a little bit.
1893    
1894          (f) New tests for checking recursion limiting.
1895    
1896     8. The pcre_memmove() function, which is used only if the environment does not
1897        have a standard memmove() function (and is therefore rarely compiled),
1898        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1899        returning a result (though PCRE never actually uses the result).
1900    
1901     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1902        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1903        returned instead of calling malloc() with an overflowing number that would
1904        most likely cause subsequent chaos.
1905    
1906    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1907    
1908    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1909        with this option is matched, the nmatch and pmatch options of regexec() are
1910        ignored.
1911    
1912    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1913        provided in case anyone wants to the the POSIX interface with UTF-8
1914        strings.
1915    
1916    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1917        C++ linking (needed for some HP-UX environments).
1918    
1919    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1920        (unused parameter) and in the pcre_printint() function (omitted "default"
1921        switch label when the default is to do nothing).
1922    
1923    15. Added some code to make it possible, when PCRE is compiled as a C++
1924        library, to replace subject pointers for pcre_exec() with a smart pointer
1925        class, thus making it possible to process discontinuous strings.
1926    
1927    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1928        much the same function. They were added by different people who were trying
1929        to make PCRE easy to compile on non-Unix systems. It has been suggested
1930        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1931        for compiling on Windows systems. I have therefore replaced it with
1932        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1933        defaults to "extern" for C or "extern C" for C++, which works fine on
1934        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1935        SCOPE with something explicit in config.h. In addition:
1936    
1937        (a) pcreposix.h still had just "extern" instead of either of these macros;
1938            I have replaced it with PCRE_DATA_SCOPE.
1939    
1940        (b) Functions such as _pcre_xclass(), which are internal to the library,
1941            but external in the C sense, all had PCRE_EXPORT in their definitions.
1942            This is apparently wrong for the Windows case, so I have removed it.
1943            (It makes no difference on Unix-like systems.)
1944    
1945    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1946        of recursive calls to match(). This is different to MATCH_LIMIT because
1947        that limits the total number of calls to match(), not all of which increase
1948        the depth of recursion. Limiting the recursion depth limits the amount of
1949        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1950        when PCRE is compiled, and changed at run time. A patch from Google adds
1951        this functionality to the C++ interface.
1952    
1953    18. Changes to the handling of Unicode character properties:
1954    
1955        (a) Updated the table to Unicode 4.1.0.
1956    
1957        (b) Recognize characters that are not in the table as "Cn" (undefined).
1958    
1959        (c) I revised the way the table is implemented to a much improved format
1960            which includes recognition of ranges. It now supports the ranges that
1961            are defined in UnicodeData.txt, and it also amalgamates other
1962            characters into ranges. This has reduced the number of entries in the
1963            table from around 16,000 to around 3,000, thus reducing its size
1964            considerably. I realized I did not need to use a tree structure after
1965            all - a binary chop search is just as efficient. Having reduced the
1966            number of entries, I extended their size from 6 bytes to 8 bytes to
1967            allow for more data.
1968    
1969        (d) Added support for Unicode script names via properties such as \p{Han}.
1970    
1971    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1972        matching that character.
1973    
1974    20. When matching a repeated Unicode property with a minimum greater than zero,
1975        (for example \pL{2,}), PCRE could look past the end of the subject if it
1976        reached it while seeking the minimum number of characters. This could
1977        happen only if some of the characters were more than one byte long, because
1978        there is a check for at least the minimum number of bytes.
1979    
1980    21. Refactored the implementation of \p and \P so as to be more general, to
1981        allow for more different types of property in future. This has changed the
1982        compiled form incompatibly. Anybody with saved compiled patterns that use
1983        \p or \P will have to recompile them.
1984    
1985    22. Added "Any" and "L&" to the supported property types.
1986    
1987    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1988        but give a compile time error if the value is greater than 0xff.
1989    
1990    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1991        accidentally not being installed or uninstalled.
1992    
1993    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1994        made were to insert the current release number. This seemed silly, because
1995        it made things harder for people building PCRE on systems that don't run
1996        "configure". I have turned pcre.h into a distributed file, no longer built
1997        by "configure", with the version identification directly included. There is
1998        no longer a pcre.h.in file.
1999    
2000        However, this change necessitated a change to the pcre-config script as
2001        well. It is built from pcre-config.in, and one of the substitutions was the
2002        release number. I have updated configure.ac so that ./configure now finds
2003        the release number by grepping pcre.h.
2004    
2005    26. Added the ability to run the tests under valgrind.
2006    
2007    
2008    Version 6.4 05-Sep-05
2009    ---------------------
2010    
2011     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
2012        "--" to be printed when multiple files were scanned, even when none of the
2013        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
2014        consider it to be a bug, and have restored the previous behaviour.
2015    
2016     2. A couple of code tidies to get rid of compiler warnings.
2017    
2018     3. The pcretest program used to cheat by referring to symbols in the library
2019        whose names begin with _pcre_. These are internal symbols that are not
2020        really supposed to be visible externally, and in some environments it is
2021        possible to suppress them. The cheating is now confined to including
2022        certain files from the library's source, which is a bit cleaner.
2023    
2024     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
2025        file's purpose clearer.
2026    
2027     5. Reorganized pcre_ucp_findchar().
2028    
2029    
2030    Version 6.3 15-Aug-05
2031    ---------------------
2032    
2033     1. The file libpcre.pc.in did not have general read permission in the tarball.
2034    
2035     2. There were some problems when building without C++ support:
2036    
2037        (a) If C++ support was not built, "make install" and "make test" still
2038            tried to test it.
2039    
2040        (b) There were problems when the value of CXX was explicitly set. Some
2041            changes have been made to try to fix these, and ...
2042    
2043        (c) --disable-cpp can now be used to explicitly disable C++ support.
2044    
2045        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
2046            backslash in a target when C++ was disabled. This confuses some
2047            versions of "make", apparently. Using an intermediate variable solves
2048            this. (Same for CPP_LOBJ.)
2049    
2050     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
2051        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
2052        necessary on certain architectures.
2053    
2054     4. Added a setting of -export-symbols-regex to the link command to remove
2055        those symbols that are exported in the C sense, but actually are local
2056        within the library, and not documented. Their names all begin with
2057        "_pcre_". This is not a perfect job, because (a) we have to except some
2058        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
2059        available (and never for static libraries). I have made a note to try to
2060        find a way round (a) in the future.
2061    
2062    
2063    Version 6.2 01-Aug-05
2064    ---------------------
2065    
2066     1. There was no test for integer overflow of quantifier values. A construction
2067        such as {1111111111111111} would give undefined results. What is worse, if
2068        a minimum quantifier for a parenthesized subpattern overflowed and became
2069        negative, the calculation of the memory size went wrong. This could have
2070        led to memory overwriting.
2071    
2072     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
2073    
2074     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
2075        operating environments where this matters.
2076    
2077     4. Applied Giuseppe Maxia's patch to add additional features for controlling
2078        PCRE options from within the C++ wrapper.
2079    
2080     5. Named capturing subpatterns were not being correctly counted when a pattern
2081        was compiled. This caused two problems: (a) If there were more than 100
2082        such subpatterns, the calculation of the memory needed for the whole
2083        compiled pattern went wrong, leading to an overflow error. (b) Numerical
2084        back references of the form \12, where the number was greater than 9, were
2085        not recognized as back references, even though there were sufficient
2086        previous subpatterns.
2087    
2088     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
2089        versions of gcc, e.g. 2.95.4.
2090    
2091    
2092    Version 6.1 21-Jun-05
2093    ---------------------
2094    
2095     1. There was one reference to the variable "posix" in pcretest.c that was not
2096        surrounded by "#if !defined NOPOSIX".
2097    
2098     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
2099        the cross-check on the old pcre_info() function, for the benefit of the
2100        cut-down version of PCRE that is currently imported into Exim.
2101    
2102     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
2103        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
2104        patterns that start (?i) but I don't think that matters. The use of (?i) is
2105        just an example; this all applies to the other options as well.
2106    
2107     4. Since libtool seems to echo the compile commands it is issuing, the output
2108        from "make" can be reduced a bit by putting "@" in front of each libtool
2109        compile command.
2110    
2111     5. Patch from the folks at Google for configure.in to be a bit more thorough
2112        in checking for a suitable C++ installation before trying to compile the
2113        C++ stuff. This should fix a reported problem when a compiler was present,
2114        but no suitable headers.
2115    
2116     6. The man pages all had just "PCRE" as their title. I have changed them to
2117        be the relevant file name. I have also arranged that these names are
2118        retained in the file doc/pcre.txt, which is a concatenation in text format
2119        of all the man pages except the little individual ones for each function.
2120    
2121     7. The NON-UNIX-USE file had not been updated for the different set of source
2122        files that come with release 6. I also added a few comments about the C++
2123        wrapper.
2124    
2125    
2126    Version 6.0 07-Jun-05
2127    ---------------------
2128    
2129     1. Some minor internal re-organization to help with my DFA experiments.
2130    
2131     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
2132        didn't matter for the library itself when fully configured, but did matter
2133        when compiling without UCP support, or within Exim, where the ucp files are
2134        not imported.
2135    
2136     3. Refactoring of the library code to split up the various functions into
2137        different source modules. The addition of the new DFA matching code (see
2138        below) to a single monolithic source would have made it really too
2139        unwieldy, quite apart from causing all the code to be include in a
2140        statically linked application, when only some functions are used. This is
2141        relevant even without the DFA addition now that patterns can be compiled in
2142        one application and matched in another.
2143    
2144        The downside of splitting up is that there have to be some external
2145        functions and data tables that are used internally in different modules of
2146        the library but which are not part of the API. These have all had their
2147        names changed to start with "_pcre_" so that they are unlikely to clash
2148        with other external names.
2149    
2150     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
2151        a different (DFA) algorithm. Although it is slower than the original
2152        function, it does have some advantages for certain types of matching
2153        problem.
2154    
2155     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
2156        including restarting after a partial match.
2157    
2158     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
2159        defined when compiling for Windows was sent to me. I have put it into the
2160        code, though I have no means of testing or verifying it.
2161    
2162     7. Added the pcre_refcount() auxiliary function.
2163    
2164     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
2165        match before or at the first newline in the subject string. In pcretest,
2166        the /f option on a pattern can be used to set this.
2167    
2168     9. A repeated \w when used in UTF-8 mode with characters greater than 256
2169        would behave wrongly. This has been present in PCRE since release 4.0.
2170    
2171    10. A number of changes to the pcregrep command:
2172    
2173        (a) Refactored how -x works; insert ^(...)$ instead of setting
2174            PCRE_ANCHORED and checking the length, in preparation for adding
2175            something similar for -w.
2176    
2177        (b) Added the -w (match as a word) option.
2178    
2179        (c) Refactored the way lines are read and buffered so as to have more
2180            than one at a time available.
2181    
2182        (d) Implemented a pcregrep test script.
2183    
2184        (e) Added the -M (multiline match) option. This allows patterns to match
2185            over several lines of the subject. The buffering ensures that at least
2186            8K, or the rest of the document (whichever is the shorter) is available
2187            for matching (and similarly the previous 8K for lookbehind assertions).
2188    
2189        (f) Changed the --help output so that it now says
2190    
2191              -w, --word-regex(p)
2192    
2193            instead of two lines, one with "regex" and the other with "regexp"
2194            because that confused at least one person since the short forms are the
2195            same. (This required a bit of code, as the output is generated
2196            automatically from a table. It wasn't just a text change.)
2197    
2198        (g) -- can be used to terminate pcregrep options if the next thing isn't an
2199            option but starts with a hyphen. Could be a pattern or a path name
2200            starting with a hyphen, for instance.
2201    
2202        (h) "-" can be given as a file name to represent stdin.
2203    
2204        (i) When file names are being printed, "(standard input)" is used for
2205            the standard input, for compatibility with GNU grep. Previously
2206            "<stdin>" was used.
2207    
2208        (j) The option --label=xxx can be used to supply a name to be used for
2209            stdin when file names are being printed. There is no short form.
2210    
2211        (k) Re-factored the options decoding logic because we are going to add
2212            two more options that take data. Such options can now be given in four
2213            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
2214    
2215        (l) Added the -A, -B, and -C options for requesting that lines of context
2216            around matches be printed.
2217    
2218        (m) Added the -L option to print the names of files that do not contain
2219            any matching lines, that is, the complement of -l.
2220    
2221        (n) The return code is 2 if any file cannot be opened, but pcregrep does
2222            continue to scan other files.
2223    
2224        (o) The -s option was incorrectly implemented. For compatibility with other
2225            greps, it now suppresses the error message for a non-existent or non-
2226            accessible file (but not the return code). There is a new option called
2227            -q that suppresses the output of matching lines, which was what -s was
2228            previously doing.
2229    
2230        (p) Added --include and --exclude options to specify files for inclusion
2231            and exclusion when recursing.
2232    
2233    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
2234        Hopefully, it now does.
2235    
2236    12. Missing cast in pcre_study().
2237    
2238    13. Added an "uninstall" target to the makefile.
2239    
2240    14. Replaced "extern" in the function prototypes in Makefile.in with
2241        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
2242        world, but is set differently for Windows.
2243    
2244    15. Added a second compiling function called pcre_compile2(). The only
2245        difference is that it has an extra argument, which is a pointer to an
2246        integer error code. When there is a compile-time failure, this is set
2247        non-zero, in addition to the error test pointer being set to point to an
2248        error message. The new argument may be NULL if no error number is required
2249        (but then you may as well call pcre_compile(), which is now just a
2250        wrapper). This facility is provided because some applications need a
2251        numeric error indication, but it has also enabled me to tidy up the way
2252        compile-time errors are handled in the POSIX wrapper.
2253    
2254    16. Added VPATH=.libs to the makefile; this should help when building with one
2255        prefix path and installing with another. (Or so I'm told by someone who
2256        knows more about this stuff than I do.)
2257    
2258    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
2259        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
2260        match everything, including newlines. This is not POSIX-compatible, but
2261        somebody wanted the feature. From pcretest it can be activated by using
2262        both the P and the s flags.
2263    
2264    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
2265    
2266    19. libpcre.pc was being incorrectly installed as executable.
2267    
2268    20. A couple of places in pcretest check for end-of-line by looking for '\n';
2269        it now also looks for '\r' so that it will work unmodified on Windows.
2270    
2271    21. Added Google's contributed C++ wrapper to the distribution.
2272    
2273    22. Added some untidy missing memory free() calls in pcretest, to keep
2274        Electric Fence happy when testing.
2275    
2276    
2277    
2278    Version 5.0 13-Sep-04
2279    ---------------------
2280    
2281     1. Internal change: literal characters are no longer packed up into items
2282        containing multiple characters in a single byte-string. Each character
2283        is now matched using a separate opcode. However, there may be more than one
2284        byte in the character in UTF-8 mode.
2285    
2286     2. The pcre_callout_block structure has two new fields: pattern_position and
2287        next_item_length. These contain the offset in the pattern to the next match
2288        item, and its length, respectively.
2289    
2290     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
2291        insertion of callouts before each pattern item. Added the /C option to
2292        pcretest to make use of this.
2293    
2294     4. On the advice of a Windows user, the lines
2295    
2296          #if defined(_WIN32) || defined(WIN32)
2297          _setmode( _fileno( stdout ), 0x8000 );
2298          #endif  /* defined(_WIN32) || defined(WIN32) */
2299    
2300        have been added to the source of pcretest. This apparently does useful
2301        magic in relation to line terminators.
2302    
2303     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
2304        for the benefit of those environments where the "b" makes a difference.
2305    
2306     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
2307        to know about it. I have put a hack into configure.in that adds in code
2308        to set GCC=yes if CC=icc. This seems to end up at a point in the
2309        generated configure script that is early enough to affect the setting of
2310        compiler options, which is what is needed, but I have no means of testing
2311        whether it really works. (The user who reported this had patched the
2312        generated configure script, which of course I cannot do.)
2313    
2314        LATER: After change 22 below (new libtool files), the configure script
2315        seems to know about icc (and also ecc). Therefore, I have commented out
2316        this hack in configure.in.
2317    
2318     7. Added support for pkg-config (2 patches were sent in).
2319    
2320     8. Negated POSIX character classes that used a combination of internal tables
2321        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
2322        [[:^ascii]]. Typically, they would match almost any characters. The other
2323        POSIX classes were not broken in this way.
2324    
2325     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
2326        to find the match, as PCRE was deluded into thinking that the match had to
2327        start at the start point or following a newline. The same bug applied to
2328        patterns with negative forward assertions or any backward assertions
2329        preceding ".*" at the start, unless the pattern required a fixed first
2330        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
2331    
2332    10. In UTF-8 mode, when moving forwards in the subject after a failed match
2333        starting at the last subject character, bytes beyond the end of the subject
2334        string were read.
2335    
2336    11. Renamed the variable "class" as "classbits" to make life easier for C++
2337        users. (Previously there was a macro definition, but it apparently wasn't
2338        enough.)
2339    
2340    12. Added the new field "tables" to the extra data so that tables can be passed
2341        in at exec time, or the internal tables can be re-selected. This allows
2342        a compiled regex to be saved and re-used at a later time by a different
2343        program that might have everything at different addresses.
2344    
2345    13. Modified the pcre-config script so that, when run on Solaris, it shows a
2346        -R library as well as a -L library.
2347    
2348    14. The debugging options of pcretest (-d on the command line or D on a
2349        pattern) showed incorrect output for anything following an extended class
2350        that contained multibyte characters and which was followed by a quantifier.
2351    
2352    15. Added optional support for general category Unicode character properties
2353        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
2354        support. It adds about 90K to the size of the library. The meanings of the
2355        inbuilt class escapes such as \d and \s have NOT been changed.
2356    
2357    16. Updated pcredemo.c to include calls to free() to release the memory for the
2358        compiled pattern.
2359    
2360    17. The generated file chartables.c was being created in the source directory
2361        instead of in the building directory. This caused the build to fail if the
2362        source directory was different from the building directory, and was
2363        read-only.
2364    
2365    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
2366        file. No doubt somebody will tell me if they don't make sense... Also added
2367        Dan Mooney's comments about building on OpenVMS.
2368    
2369    19. Added support for partial matching via the PCRE_PARTIAL option for
2370        pcre_exec() and the \P data escape in pcretest.
2371    
2372    20. Extended pcretest with 3 new pattern features:
2373    
2374        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
2375              write the compiled pattern to the file whose name is "rest-of-line".
2376              This is a straight binary dump of the data, with the saved pointer to
2377              the character tables forced to be NULL. The study data, if any, is
2378              written too. After writing, pcretest reads a new pattern.
2379    
2380        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
2381              compiled pattern from the given file. There must not be any
2382              occurrences of "<" in the file name (pretty unlikely); if there are,
2383              pcretest will instead treat the initial "<" as a pattern delimiter.
2384              After reading in the pattern, pcretest goes on to read data lines as
2385              usual.
2386    
2387        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
2388              and 16-bit fields in a compiled pattern, to simulate a pattern that
2389              was compiled on a host of opposite endianness.
2390    
2391    21. The pcre-exec() function can now cope with patterns that were compiled on
2392        hosts of opposite endianness, with this restriction:
2393    
2394          As for any compiled expression that is saved and used later, the tables
2395          pointer field cannot be preserved; the extra_data field in the arguments
2396          to pcre_exec() should be used to pass in a tables address if a value
2397          other than the default internal tables were used at compile time.
2398    
2399    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
2400        now diagnosed as an error. Previously, most of the time, a negative number
2401        would have been treated as zero, but if in addition "ovector" was passed as
2402        NULL, a crash could occur.
2403    
2404    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
2405        new versions from the libtool 1.5 distribution (the last one is a copy of
2406        a file called libtool.m4). This seems to have fixed the need to patch
2407        "configure" to support Darwin 1.3 (which I used to do). However, I still
2408        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
2409        workstation).
2410    
2411    24. Changed the PCRE licence to be the more standard "BSD" licence.
2412    
2413    
2414    Version 4.5 01-Dec-03
2415    ---------------------
2416    
2417     1. There has been some re-arrangement of the code for the match() function so
2418        that it can be compiled in a version that does not call itself recursively.
2419        Instead, it keeps those local variables that need separate instances for
2420        each "recursion" in a frame on the heap, and gets/frees frames whenever it
2421        needs to "recurse". Keeping track of where control must go is done by means
2422        of setjmp/longjmp. The whole thing is implemented by a set of macros that
2423        hide most of the details from the main code, and operates only if
2424        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
2425        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
2426        operating.
2427    
2428        To make it easier for callers to provide specially tailored get/free
2429        functions for this usage, two new functions, pcre_stack_malloc, and
2430        pcre_stack_free, are used. They are always called in strict stacking order,
2431        and the size of block requested is always the same.
2432    
2433        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
2434        PCRE has been compiled to use the stack or the heap for recursion. The
2435        -C option of pcretest uses this to show which version is compiled.
2436    
2437        A new data escape \S, is added to pcretest; it causes the amounts of store
2438        obtained and freed by both kinds of malloc/free at match time to be added
2439        to the output.
2440    
2441     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
2442        what's available on my current Linux desktop machine.
2443    
2444     3. When matching a UTF-8 string, the test for a valid string at the start has
2445        been extended. If start_offset is not zero, PCRE now checks that it points
2446        to a byte that is the start of a UTF-8 character. If not, it returns
2447        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
2448        this is necessary because there may be backward assertions in the pattern.
2449        When matching the same subject several times, it may save resources to use
2450        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
2451    
2452     4. The code for checking the validity of UTF-8 strings has been tightened so
2453        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
2454        containing "overlong sequences".
2455    
2456     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
2457        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
2458        should have been "&", but it just so happened that all the cases this let
2459        through by mistake were picked up later in the function.
2460    
2461     6. I had used a variable called "isblank" - this is a C99 function, causing
2462        some compilers to warn. To avoid this, I renamed it (as "blankclass").
2463    
2464     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
2465        prompting; (b) run "./pcretest /dev/null" at the start of the test script
2466        so the version is shown; (c) stop "make test" echoing "./RunTest".
2467    
2468     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
2469    
2470     9. The prototype for memmove() for systems that don't have it was using
2471        size_t, but the inclusion of the header that defines size_t was later. I've
2472        moved the #includes for the C headers earlier to avoid this.
2473    
2474    10. Added some adjustments to the code to make it easier to compiler on certain
2475        special systems:
2476    
2477          (a) Some "const" qualifiers were missing.
2478          (b) Added the macro EXPORT before all exported functions; by default this
2479              is defined to be empty.
2480          (c) Changed the dftables auxiliary program (that builds chartables.c) so
2481              that it reads its output file name as an argument instead of writing
2482              to the standard output and assuming this can be redirected.
2483    
2484    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
2485        class containing characters with values greater than 255, PCRE compilation
2486        went into a loop.
2487    
2488    12. A recursive reference to a subpattern that was within another subpattern
2489        that had a minimum quantifier of zero caused PCRE to crash. For example,
2490        (x(y(?2))z)? provoked this bug with a subject that got as far as the
2491        recursion. If the recursively-called subpattern itself had a zero repeat,
2492        that was OK.
2493    
2494    13. In pcretest, the buffer for reading a data line was set at 30K, but the
2495        buffer into which it was copied (for escape processing) was still set at
2496        1024, so long lines caused crashes.
2497    
2498    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
2499        "internal error: code overflow...". This applied to any character class
2500        that was followed by a possessive quantifier.
2501    
2502    15. Modified the Makefile to add libpcre.la as a prerequisite for
2503        libpcreposix.la because I was told this is needed for a parallel build to
2504        work.
2505    
2506    16. If a pattern that contained .* following optional items at the start was
2507        studied, the wrong optimizing data was generated, leading to matching
2508        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
2509        matching string must start with a or b or c. The correct conclusion for
2510        this pattern is that a match can start with any character.
2511    
2512    
2513    Version 4.4 13-Aug-03
2514    ---------------------
2515    
2516     1. In UTF-8 mode, a character class containing characters with values between
2517        127 and 255 was not handled correctly if the compiled pattern was studied.
2518        In fixing this, I have also improved the studying algorithm for such
2519        classes (slightly).
2520    
2521     2. Three internal functions had redundant arguments passed to them. Removal
2522        might give a very teeny performance improvement.
2523    
2524     3. Documentation bug: the value of the capture_top field in a callout is *one
2525        more than* the number of the hightest numbered captured substring.
2526    
2527     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
2528        in incorrectly linking with a previously installed version. They now link
2529        explicitly with libpcre.la.
2530    
2531     5. configure.in no longer needs to recognize Cygwin specially.
2532    
2533     6. A problem in pcre.in for Windows platforms is fixed.
2534    
2535     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
2536        pcretest, it used to include the size of the study block as part of its
2537        output. Unfortunately, the structure contains a field that has a different
2538        size on different hardware architectures. This meant that the tests that
2539        showed this size failed. As the block is currently always of a fixed size,
2540        this information isn't actually particularly useful in pcretest output, so
2541        I have just removed it.
2542    
2543     8. Three pre-processor statements accidentally did not start in column 1.
2544        Sadly, there are *still* compilers around that complain, even though
2545        standard C has not required this for well over a decade. Sigh.
2546    
2547     9. In pcretest, the code for checking callouts passed small integers in the
2548        callout_data field, which is a void * field. However, some picky compilers
2549        complained about the casts involved for this on 64-bit systems. Now
2550        pcretest passes the address of the small integer instead, which should get
2551        rid of the warnings.
2552    
2553    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
2554        both compile and run time, and gives an error if an invalid UTF-8 sequence
2555        is found. There is a option for disabling this check in cases where the
2556        string is known to be correct and/or the maximum performance is wanted.
2557    
2558    11. In response to a bug report, I changed one line in Makefile.in from
2559    
2560            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
2561        to
2562            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
2563    
2564        to look similar to other lines, but I have no way of telling whether this
2565        is the right thing to do, as I do not use Windows. No doubt I'll get told
2566        if it's wrong...
2567    
2568    
2569    Version 4.3 21-May-03
2570    ---------------------
2571    
2572    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
2573       Makefile.
2574    
2575    2. Some refactoring to improve the quality of the code:
2576    
2577       (i)   The utf8_table... variables are now declared "const".
2578    
2579       (ii)  The code for \cx, which used the "case flipping" table to upper case
2580             lower case letters, now just substracts 32. This is ASCII-specific,
2581             but the whole concept of \cx is ASCII-specific, so it seems
2582             reasonable.
2583    
2584       (iii) PCRE was using its character types table to recognize decimal and
2585             hexadecimal digits in the pattern. This is silly, because it handles
2586             only 0-9, a-f, and A-F, but the character types table is locale-
2587             specific, which means strange things might happen. A private
2588             table is now used for this - though it costs 256 bytes, a table is
2589             much faster than multiple explicit tests. Of course, the standard
2590             character types table is still used for matching digits in subject
2591             strings against \d.
2592    
2593       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
2594             ending in _t are). So I've renamed it as ESC_tee.
2595    
2596    3. The first argument for regexec() in the POSIX wrapper should have been
2597       defined as "const".
2598    
2599    4. Changed pcretest to use malloc() for its buffers so that they can be
2600       Electric Fenced for debugging.
2601    
2602    5. There were several places in the code where, in UTF-8 mode, PCRE would try
2603       to read one or more bytes before the start of the subject string. Often this
2604       had no effect on PCRE's behaviour, but in some circumstances it could
2605       provoke a segmentation fault.
2606    
2607    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
2608       to try to read one or more bytes before the start of the subject string.
2609    
2610    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
2611       UTF-8 support could misbehave in various ways if the subject string
2612       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
2613       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
2614       back over UTF-8 characters.)
2615    
2616    
2617    Version 4.2 14-Apr-03
2618    ---------------------
2619    
2620    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
2621    
2622    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
2623         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
2624         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
2625         [WIN_PREFIX]: new variable, "cyg" for Cygwin
2626         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
2627           and BUILD_EXEEXT
2628         Note: automatic setting of the BUILD variables is not yet working
2629         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
2630           compile-time but not at link-time
2631         [LINK]: use for linking executables only
2632         make different versions for Windows and non-Windows
2633         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
2634           libraries
2635         [LINK_FOR_BUILD]: new variable
2636         [OBJEXT]: use throughout
2637         [EXEEXT]: use throughout
2638         <winshared>: new target
2639         <wininstall>: new target
2640         <dftables.o>: use native compiler
2641         <dftables>: use native linker
2642         <install>: handle Windows platform correctly
2643         <clean>: ditto
2644         <check>: ditto
2645         copy DLL to top builddir before testing
2646    
2647       As part of these changes, -no-undefined was removed again. This was reported
2648       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
2649       in any case.
2650    
2651    3. Some tidies to get rid of compiler warnings:
2652    
2653       . In the match_data structure, match_limit was an unsigned long int, whereas
2654         match_call_count was an int. I've made them both unsigned long ints.
2655    
2656       . In pcretest the fact that a const uschar * doesn't automatically cast to
2657         a void * provoked a warning.
2658    
2659       . Turning on some more compiler warnings threw up some "shadow" variables
2660         and a few more missing casts.
2661    
2662    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
2663       option, a class that contained a single character with a value between 128
2664       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
2665    
2666    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
2667       option, a class that contained several characters, but with at least one
2668       whose value was between 128 and 255 caused PCRE to crash.
2669    
2670    
2671    Version 4.1 12-Mar-03
2672    ---------------------
2673    
2674    1. Compiling with gcc -pedantic found a couple of places where casts were
2675    needed, and a string in dftables.c that was longer than standard compilers are
2676    required to support.
2677    
2678    2. Compiling with Sun's compiler found a few more places where the code could
2679    be tidied up in order to avoid warnings.
2680    
2681    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
2682    first of these names is deprecated in the latest Autoconf in favour of the name
2683    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
2684    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
2685    analogy I have changed it to CFLAGS_FOR_BUILD.
2686    
2687    4. Added -no-undefined to the linking command in the Makefile, because this is
2688    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
2689    linking step for the pcreposix library.
2690    
2691    5. PCRE was failing to diagnose the case of two named groups with the same
2692    name.
2693    
2694    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
2695    literal character that is needed in the subject for a match, and scans along to
2696    ensure that it is present before embarking on the full matching process. This
2697    saves time in cases of nested unlimited repeats that are never going to match.
2698    Problem: the scan can take a lot of time if the subject is very long (e.g.
2699    megabytes), thus penalizing straightforward matches. It is now done only if the
2700    amount of subject to be scanned is less than 1000 bytes.
2701    
2702    7. A lesser problem with the same optimization is that it was recording the
2703    first character of an anchored pattern as "needed", thus provoking a search
2704    right along the subject, even when the first match of the pattern was going to
2705    fail. The "needed" character is now not set for anchored patterns, unless it
2706    follows something in the pattern that is of non-fixed length. Thus, it still
2707    fulfils its original purpose of finding quick non-matches in cases of nested
2708    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
2709    
2710    
2711    Version 4.0 17-Feb-03
2712    ---------------------
2713    
2714    1. If a comment in an extended regex that started immediately after a meta-item
2715    extended to the end of string, PCRE compiled incorrect data. This could lead to
2716    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
2717    
2718    2. Moved to autoconf 2.53 and libtool 1.4.2.
2719    
2720    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
2721    the special perltest8 script is no longer needed - all the tests can be run
2722    from a single perltest script.
2723    
2724    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
2725    by \s. It has now been removed in PCRE. This means it isn't recognized as
2726    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
2727    class [:space:] *does* include VT, thereby creating a mess.
2728    
2729    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
2730    space and tab.
2731    
2732    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
2733    its new features into the main test script, reducing the number of scripts.
2734    
2735    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
2736    were backward compatible, and made the (?i) apply to the whole pattern, as if
2737    /i were given. Now it behaves more logically, and applies the option setting
2738    only to what follows. PCRE has been changed to follow suit. However, if it
2739    finds options settings right at the start of the pattern, it extracts them into
2740    the global options, as before. Thus, they show up in the info data.
2741    
2742    8. Added support for the \Q...\E escape sequence. Characters in between are
2743    treated as literals. This is slightly different from Perl in that $ and @ are
2744    also handled as literals inside the quotes. In Perl, they will cause variable
2745    interpolation. Note the following examples:
2746    
2747        Pattern            PCRE matches      Perl matches
2748    
2749        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
2750        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
2751        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
2752    
2753    For compatibility with Perl, \Q...\E sequences are recognized inside character
2754    classes as well as outside them.
2755    
2756    9. Re-organized 3 code statements in pcretest to avoid "overflow in
2757    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
2758    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
2759    signed/unsigned warnings.
2760    
2761    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
2762    option for pcretest, so I've replaced it by a simple function that does just
2763    that job.
2764    
2765    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
2766    "pcregrep -".
2767    
2768    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
2769    Java package. This provides some syntactic sugar for simple cases of what my
2770    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
2771    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
2772    item, you can use this simplified notation. Note that only makes sense with
2773    greedy quantifiers. Consequently, the use of the possessive quantifier forces
2774    greediness, whatever the setting of the PCRE_UNGREEDY option.
2775    
2776    13. A change of greediness default within a pattern was not taking effect at
2777    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
2778    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
2779    was abstracted outside.
2780    
2781    14. PCRE now supports the \G assertion. It is true when the current matching
2782    position is at the start point of the match. This differs from \A when the
2783    starting offset is non-zero. Used with the /g option of pcretest (or similar
2784    code), it works in the same way as it does for Perl's /g option. If all
2785    alternatives of a regex begin with \G, the expression is anchored to the start
2786    match position, and the "anchored" flag is set in the compiled expression.
2787    
2788    15. Some bugs concerning the handling of certain option changes within patterns
2789    have been fixed. These applied to options other than (?ims). For example,
2790    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
2791    been the other way round. Some of this was related to change 7 above.
2792    
2793    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
2794    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
2795    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
2796    POSIX classes only within a class (e.g. /[[:alpha:]]/).
2797    
2798    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
2799    mode. Unlike ".", it always matches newline, whatever the setting of
2800    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
2801    assertions. Perl allows it, but it doesn't (in general) work because it can't
2802    calculate the length of the lookbehind. At least, that's the case for Perl
2803    5.8.0 - I've been told they are going to document that it doesn't work in
2804    future.
2805    
2806    18. Added an error diagnosis for escapes that PCRE does not support: these are
2807    \L, \l, \N, \P, \p, \U, \u, and \X.
2808    
2809    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
2810    reading past the end of the pattern in cases such as /[abcd/.
2811    
2812    20. PCRE was getting more memory than necessary for patterns with classes that
2813    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
2814    
2815    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
2816    compiling PCRE for use with Virtual Pascal.
2817    
2818    22. Small fix to the Makefile to make it work properly if the build is done
2819    outside the source tree.
2820    
2821    23. Added a new extension: a condition to go with recursion. If a conditional
2822    subpattern starts with (?(R) the "true" branch is used if recursion has
2823    happened, whereas the "false" branch is used only at the top level.
2824    
2825    24. When there was a very long string of literal characters (over 255 bytes
2826    without UTF support, over 250 bytes with UTF support), the computation of how
2827    much memory was required could be incorrect, leading to segfaults or other
2828    strange effects.
2829    
2830    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2831    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2832    there was a subsequent back reference to those brackets. This meant that, for
2833    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2834    possible to check for precisely this case. All we can do is abandon the
2835    optimization if .* occurs inside capturing brackets when there are any back
2836    references whatsoever. (See below for a better fix that came later.)
2837    
2838    26. The handling of the optimization for finding the first character of a
2839    non-anchored pattern, and for finding a character that is required later in the
2840    match were failing in some cases. This didn't break the matching; it just
2841    failed to optimize when it could. The way this is done has been re-implemented.
2842    
2843    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2844    
2845    28. Added a new feature that provides some of the functionality that Perl
2846    provides with (?{...}). The facility is termed a "callout". The way it is done
2847    in PCRE is for the caller to provide an optional function, by setting
2848    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2849    global variable. By default it is unset, which disables all calling out. To get
2850    the function called, the regex must include (?C) at appropriate points. This
2851    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2852    This provides a means of identifying different callout points. When PCRE
2853    reaches such a point in the regex, if pcre_callout has been set, the external
2854    function is called. It is provided with data in a structure called
2855    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2856    matching continues; if it returns a non-zero value, the match at the current
2857    point fails. However, backtracking will occur if possible. [This was changed
2858    later and other features added - see item 49 below.]
2859    
2860    29. pcretest is upgraded to test the callout functionality. It provides a
2861    callout function that displays information. By default, it shows the start of
2862    the match and the current position in the text. There are some new data escapes
2863    to vary what happens:
2864    
2865        \C+         in addition, show current contents of captured substrings
2866        \C-         do not supply a callout function
2867        \C!n        return 1 when callout number n is reached
2868        \C!n!m      return 1 when callout number n is reached for the mth time
2869    
2870    30. If pcregrep was called with the -l option and just a single file name, it
2871    output "<stdin>" if a match was found, instead of the file name.
2872    
2873    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2874    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2875    pcre_exec(). This saves a malloc/free per call. The default value of
2876    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2877    when configuring.
2878    
2879    32. The default maximum size of a compiled pattern is 64K. There have been a
2880    few cases of people hitting this limit. The code now uses macros to handle the
2881    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2882    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2883    configuring. Tests 2 and 5 work only with 2-byte links because they output
2884    debugging information about compiled patterns.
2885    
2886    33. Internal code re-arrangements:
2887    
2888    (a) Moved the debugging function for printing out a compiled regex into
2889        its own source file (printint.c) and used #include to pull it into
2890        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2891        separate copies.
2892    
2893    (b) Defined the list of op-code names for debugging as a macro in
2894        internal.h so that it is next to the definition of the opcodes.
2895    
2896    (c) Defined a table of op-code lengths for simpler skipping along compiled
2897        code. This is again a macro in internal.h so that it is next to the
2898        definition of the opcodes.
2899    
2900    34. Added support for recursive calls to individual subpatterns, along the
2901    lines of Robin Houston's patch (but implemented somewhat differently).
2902    
2903    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2904    allow it to read and process whole directories in Win32. This code was
2905    contributed by Lionel Fourquaux; it has not been tested by me.
2906    
2907    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2908    used to name a group. Names consist of alphanumerics and underscores, and must
2909    be unique. Back references use the syntax (?P=name) and recursive calls use
2910    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2911    numbers. The function pcre_fullinfo() can be used after compilation to extract
2912    a name/number map. There are three relevant calls:
2913    
2914      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2915      PCRE_INFO_NAMECOUNT            yields the number of entries
2916      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2917    
2918    The map is a vector of fixed-size entries. The size of each entry depends on
2919    the length of the longest name used. The first two bytes of each entry are the
2920    group number, most significant byte first. There follows the corresponding
2921    name, zero terminated. The names are in alphabetical order.
2922    
2923    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2924    case instead of 255. Making it the same both with and without UTF-8 support
2925    means that the same test output works with both.
2926    
2927    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2928    calling malloc() with a zero argument.
2929    
2930    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2931    optimization. I've improved things by keeping a bitmap of backreferences with
2932    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2933    fact referenced, the optimization can be applied. It is unlikely that a
2934    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2935    the match to follow \n) will appear inside brackets with a number greater than
2936    31, but if it does, any back reference > 31 suppresses the optimization.
2937    
2938    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2939    of disabling numbered capturing parentheses. Any opening parenthesis that is
2940    not followed by ? behaves as if it were followed by ?: but named parentheses
2941    can still be used for capturing (and they will acquire numbers in the usual
2942    way).
2943    
2944    41. Redesigned the return codes from the match() function into yes/no/error so
2945    that errors can be passed back from deep inside the nested calls. A malloc
2946    failure while inside a recursive subpattern call now causes the
2947    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2948    
2949    42. It is now possible to set a limit on the number of times the match()
2950    function is called in a call to pcre_exec(). This facility makes it possible to
2951    limit the amount of recursion and backtracking, though not in a directly
2952    obvious way, because the match() function is used in a number of different
2953    circumstances. The count starts from zero for each position in the subject
2954    string (for non-anchored patterns). The default limit is, for compatibility, a
2955    large number, namely 10 000 000. You can change this in two ways:
2956    
2957    (a) When configuring PCRE before making, you can use --with-match-limit=n
2958        to set a default value for the compiled library.
2959    
2960    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2961        a different value is set. See 45 below.
2962    
2963    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2964    
2965    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2966    of things that can be changed at compile time. The first argument specifies
2967    what is wanted and the second points to where the information is to be placed.
2968    The current list of available information is:
2969    
2970      PCRE_CONFIG_UTF8
2971    
2972    The output is an integer that is set to one if UTF-8 support is available;
2973    otherwise it is set to zero.
2974    
2975      PCRE_CONFIG_NEWLINE
2976    
2977    The output is an integer that it set to the value of the code that is used for
2978    newline. It is either LF (10) or CR (13).
2979    
2980      PCRE_CONFIG_LINK_SIZE
2981    
2982    The output is an integer that contains the number of bytes used for internal
2983    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2984    
2985      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2986    
2987    The output is an integer that contains the threshold above which the POSIX
2988    interface uses malloc() for output vectors. See item 31 above.
2989    
2990      PCRE_CONFIG_MATCH_LIMIT
2991    
2992    The output is an unsigned integer that contains the default limit of the number
2993    of match() calls in a pcre_exec() execution. See 42 above.
2994    
2995    44. pcretest has been upgraded by the addition of the -C option. This causes it
2996    to extract all the available output from the new pcre_config() function, and to
2997    output it. The program then exits immediately.
2998    
2999    45. A need has arisen to pass over additional data with calls to pcre_exec() in
3000    order to support additional features. One way would have been to define
3001    pcre_exec2() (for example) with extra arguments, but this would not have been
3002    extensible, and would also have required all calls to the original function to
3003    be mapped to the new one. Instead, I have chosen to extend the mechanism that
3004    is used for passing in "extra" data from pcre_study().
3005    
3006    The pcre_extra structure is now exposed and defined in pcre.h. It currently
3007    contains the following fields:
3008    
3009      flags         a bitmap indicating which of the following fields are set
3010      study_data    opaque data from pcre_study()
3011      match_limit   a way of specifying a limit on match() calls for a specific
3012                      call to pcre_exec()
3013      callout_data  data for callouts (see 49 below)
3014    
3015    The flag bits are also defined in pcre.h, and are
3016    
3017      PCRE_EXTRA_STUDY_DATA
3018      PCRE_EXTRA_MATCH_LIMIT
3019      PCRE_EXTRA_CALLOUT_DATA
3020    
3021    The pcre_study() function now returns one of these new pcre_extra blocks, with
3022    the actual study data pointed to by the study_data field, and the
3023    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
3024    before. That is, this change is entirely upwards-compatible and requires no
3025    change to existing code.
3026    
3027    If you want to pass in additional data to pcre_exec(), you can either place it
3028    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
3029    block.
3030    
3031    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
3032    data string contains the escape sequence \M, pcretest calls pcre_exec() several
3033    times with different match limits, until it finds the minimum value needed for
3034    pcre_exec() to complete. The value is then output. This can be instructive; for
3035    most simple matches the number is quite small, but for pathological cases it
3036    gets very large very quickly.
3037    
3038    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
3039    returns the size of the data block pointed to by the study_data field in a
3040    pcre_extra block, that is, the value that was passed as the argument to
3041    pcre_malloc() when PCRE was getting memory in which to place the information
3042    created by pcre_study(). The fourth argument should point to a size_t variable.
3043    pcretest has been extended so that this information is shown after a successful
3044    pcre_study() call when information about the compiled regex is being displayed.
3045    
3046    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
3047    because what follows is always an absolute path. (Later: it turns out that this
3048    is more than cosmetic for MinGW, because it doesn't like empty path
3049    components.)
3050    
3051    49. Some changes have been made to the callout feature (see 28 above):
3052    
3053    (i)  A callout function now has three choices for what it returns:
3054    
3055           0  =>  success, carry on matching
3056         > 0  =>  failure at this point, but backtrack if possible
3057         < 0  =>  serious error, return this value from pcre_exec()
3058    
3059         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
3060         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
3061         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
3062         use by callout functions. It will never be used by PCRE itself.
3063    
3064    (ii) The pcre_extra structure (see 45 above) has a void * field called
3065         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
3066         pcre_callout_block structure has a field of the same name. The contents of
3067         the field passed in the pcre_extra structure are passed to the callout
3068         function in the corresponding field in the callout block. This makes it
3069         easier to use the same callout-containing regex from multiple threads. For
3070         testing, the pcretest program has a new data escape
3071    
3072           \C*n        pass the number n (may be negative) as callout_data
3073    
3074         If the callout function in pcretest receives a non-zero value as
3075         callout_data, it returns that value.
3076    
3077    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
3078    there were some redundant $(CFLAGS) in commands that are now specified as
3079    $(LINK), which already includes $(CFLAGS).
3080    
3081    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
3082    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
3083    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
3084    one-byte characters throughout. Note that case-insensitive matching applies
3085    only to characters whose values are less than 256. PCRE doesn't support the
3086    notion of cases for higher-valued characters.
3087    
3088    (i)   A character class whose characters are all within 0-255 is handled as
3089          a bit map, and the map is inverted for negative classes. Previously, a
3090          character > 255 always failed to match such a class; however it should
3091          match if the class was a negative one (e.g. [^ab]). This has been fixed.
3092    
3093    (ii)  A negated character class with a single character < 255 is coded as
3094          "not this character" (OP_NOT). This wasn't working properly when the test
3095          character was multibyte, either singly or repeated.
3096    
3097    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
3098          mode, for example: \x{100}{2,3}.
3099    
3100    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
3101          singly or repeated) now correctly test multibyte characters. However,
3102          PCRE doesn't recognize any characters with values greater than 255 as
3103          digits, spaces, or word characters. Such characters always match \D, \S,
3104          and \W, and never match \d, \s, or \w.
3105    
3106    (v)   Classes may now contain characters and character ranges with values
3107          greater than 255. For example: [ab\x{100}-\x{400}].
3108    
3109    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
3110          PCRE in UTF-8 mode.
3111    
3112    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
3113    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
3114    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
3115    value.)
3116    
3117    53. The single man page has become too large. I have therefore split it up into
3118    a number of separate man pages. These also give rise to individual HTML pages;
3119    these are now put in a separate directory, and there is an index.html page that
3120    lists them all. Some hyperlinking between the pages has been installed.
3121    
3122    54. Added convenience functions for handling named capturing parentheses.
3123    
3124    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
3125    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
3126    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
3127    are faulted.
3128    
3129    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
3130    calling configure. These values are used when compiling the dftables.c program
3131    which is run to generate the source of the default character tables. They
3132    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
3133    you will need to set these values.
3134    
3135    57. Updated the building process for Windows DLL, as provided by Fred Cox.
3136    
3137    
3138    Version 3.9 02-Jan-02
3139    ---------------------
3140    
3141    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
3142    
3143    2. If --disable-static was given, the building process failed when trying to
3144    build pcretest and pcregrep. (For some reason it was using libtool to compile
3145    them, which is not right, as they aren't part of the library.)
3146    
3147    
3148    Version 3.8 18-Dec-01
3149    ---------------------
3150    
3151    1. The experimental UTF-8 code was completely screwed up. It was packing the
3152    bytes in the wrong order. How dumb can you get?
3153    
3154    
3155    Version 3.7 29-Oct-01
3156    ---------------------
3157    
3158    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
3159    This caused pcretest, when used on the test data, to segfault. Unfortunately,
3160    this didn't happen under Solaris 8, where I normally test things.
3161    
3162    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
3163    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
3164    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
3165    this fix an hour or so after the initial 3.7 release.)
3166    
3167    
3168    Version 3.6 23-Oct-01
3169    ---------------------
3170    
3171    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
3172    offsets passed as NULL with zero offset count.
3173    
3174    2. The config.guess and config.sub files had not been updated when I moved to
3175    the latest autoconf.
3176    
3177    
3178    Version 3.5 15-Aug-01
3179    ---------------------
3180    
3181    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
3182    had been forgotten.
3183    
3184    2. By using declared but undefined structures, we can avoid using "void"
3185    definitions in pcre.h while keeping the internal definitions of the structures
3186    private.
3187    
3188    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
3189    user point of view, this means that both static and shared libraries are built
3190    by default, but this can be individually controlled. More of the work of
3191    handling this static/shared cases is now inside libtool instead of PCRE's make
3192    file.
3193    
3194    4. The pcretest utility is now installed along with pcregrep because it is
3195    useful for users (to test regexs) and by doing this, it automatically gets
3196    relinked by libtool. The documentation has been turned into a man page, so
3197    there are now .1, .txt, and .html versions in /doc.
3198    
3199    5. Upgrades to pcregrep:
3200       (i)   Added long-form option names like gnu grep.
3201       (ii)  Added --help to list all options with an explanatory phrase.
3202       (iii) Added -r, --recursive to recurse into sub-directories.
3203       (iv)  Added -f, --file to read patterns from a file.
3204    
3205    6. pcre_exec() was referring to its "code" argument before testing that
3206    argument for NULL (and giving an error if it was NULL).
3207    
3208    7. Upgraded Makefile.in to allow for compiling in a different directory from
3209    the source directory.
3210    
3211    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
3212    options bits, the pointer it was passed was to an int instead of to an unsigned
3213    long int. This mattered only on 64-bit systems.
3214    
3215    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
3216    generated) instead of pcre.in, which it its source. Also made the same change
3217    in several of the .c files.
3218    
3219    10. A new release of gcc defines printf() as a macro, which broke pcretest
3220    because it had an ifdef in the middle of a string argument for printf(). Fixed
3221    by using separate calls to printf().
3222    
3223    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
3224    script, to force use of CR or LF instead of \n in the source. On non-Unix
3225    systems, the value can be set in config.h.
3226    
3227    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
3228    absolute limit. Changed the text of the error message to make this clear, and
3229    likewise updated the man page.
3230    
3231    13. The limit of 99 on the number of capturing subpatterns has been removed.
3232    The new limit is 65535, which I hope will not be a "real" limit.
3233    
3234    
3235    Version 3.4 22-Aug-00
3236    ---------------------
3237    
3238    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
3239    
3240    2. Diagnose condition (?(0) as an error instead of crashing on matching.
3241    
3242    
3243    Version 3.3 01-Aug-00
3244    ---------------------
3245    
3246    1. If an octal character was given, but the value was greater than \377, it
3247    was not getting masked to the least significant bits, as documented. This could
3248    lead to crashes in some systems.
3249    
3250    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
3251    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
3252    
3253    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
3254    These just pass their arguments on to (pcre_free)(), but they are provided
3255    because some uses of PCRE bind it to non-C systems that can call its functions,
3256    but cannot call free() or pcre_free() directly.
3257    
3258    4. Add "make test" as a synonym for "make check". Corrected some comments in
3259    the Makefile.
3260    
3261    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
3262    Makefile.
3263    
3264    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
3265    command called pgrep for grepping around the active processes.
3266    
3267    7. Added the beginnings of support for UTF-8 character strings.
3268    
3269    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
3270    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
3271    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
3272    out for the ar command.)
3273    
3274    
3275    Version 3.2 12-May-00
3276    ---------------------
3277    
3278    This is purely a bug fixing release.
3279    
3280    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
3281    of ZA. This was just one example of several cases that could provoke this bug,
3282    which was introduced by change 9 of version 2.00. The code for breaking
3283    infinite loops after an iteration that matches an empty string was't working
3284    correctly.
3285    
3286    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
3287    when matched against abbab (for example). After matching an empty string, it
3288    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
3289    caused it to match further down the string than it should.
3290    
3291    3. The code contained an inclusion of sys/types.h. It isn't clear why this
3292    was there because it doesn't seem to be needed, and it causes trouble on some
3293    systems, as it is not a Standard C header. It has been removed.
3294    
3295    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
3296    were reported on the Macintosh. The changes were from
3297    
3298      while ((c = *(++ptr)) != 0 && c != '\n');
3299    to
3300      while ((c = *(++ptr)) != 0 && c != '\n') ;
3301    
3302    Totally extraordinary, but if that's what it takes...
3303    
3304    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
3305    available. Added HAVE_BCOPY and an autoconf test for it; if neither
3306    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
3307    assumes the way PCRE uses memmove() (always moving upwards).
3308    
3309    6. PCRE is being used in one environment where strchr() is not available. There
3310    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
3311    faster code anyway.
3312    
3313    
3314    Version 3.1 09-Feb-00
3315    ---------------------
3316    
3317    The only change in this release is the fixing of some bugs in Makefile.in for
3318    the "install" target:
3319    
3320    (1) It was failing to install pcreposix.h.
3321    
3322    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
3323    
3324    
3325    Version 3.0 01-Feb-00
3326    ---------------------
3327    
3328    1. Add support for the /+ modifier to perltest (to output $` like it does in
3329    pcretest).
3330    
3331    2. Add support for the /g modifier to perltest.
3332    
3333    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
3334    matches null strings.
3335    
3336    4. Fix perltest so that it doesn't do unwanted things when fed an empty
3337    pattern. Perl treats empty patterns specially - it reuses the most recent
3338    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
3339    effect.
3340    
3341    5. The POSIX interface was broken in that it was just handing over the POSIX
3342    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
3343    required a bigger vector, with some working space on the end. This means that
3344    the POSIX wrapper now has to get and free some memory, and copy the results.
3345    
3346    6. Added some simple autoconf support, placing the test data and the
3347    documentation in separate directories, re-organizing some of the
3348    information files, and making it build pcre-config (a GNU standard). Also added
3349    libtool support for building PCRE as a shared library, which is now the
3350    default.
3351    
3352    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
3353    09 are not valid octal constants. Single digits will be used for minor values
3354    less than 10.
3355    
3356    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
3357    existing programs that set these in the POSIX interface can use PCRE without
3358    modification.
3359    
3360    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
3361    return all that pcre_info() returns, plus additional data. The pcre_info()
3362    function is retained for compatibility, but is considered to be obsolete.
3363    
3364    10. Added experimental recursion feature (?R) to handle one common case that
3365    Perl 5.6 will be able to do with (?p{...}).
3366    
3367    11. Added support for POSIX character classes like [:alpha:], which Perl is
3368    adopting.
3369    
3370    
3371    Version 2.08 31-Aug-99
3372    ----------------------
3373    
3374    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
3375    trying to match at the startoffset position, but instead was moving forward to
3376    the next newline as if a previous match had failed.
3377    
3378    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
3379    and could get into a loop if a null string was matched other than at the start
3380    of the subject.
3381    
3382    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
3383    be distinguished at compile time, and for completeness also added PCRE_DATE.
3384    
3385    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
3386    in GnuWin32 environments.
3387    
3388    
3389    Version 2.07 29-Jul-99
3390    ----------------------
3391    
3392    1. The documentation is now supplied in plain text form and HTML as well as in
3393    the form of man page sources.
3394    
3395    2. C++ compilers don't like assigning (void *) values to other pointer types.
3396    In particular this affects malloc(). Although there is no problem in Standard
3397    C, I've put in casts to keep C++ compilers happy.
3398    
3399    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
3400    should be (const char *).
3401    
3402    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
3403    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
3404    However, I haven't made this a standard facility. The documentation doesn't
3405    mention it, and the Makefile doesn't support it.
3406    
3407    5. The Makefile now contains an "install" target, with editable destinations at
3408    the top of the file. The pcretest program is not installed.
3409    
3410    6. pgrep -V now gives the PCRE version number and date.
3411    
3412    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
3413    causing the entire string to be ignored, instead of just the last character.
3414    
3415    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
3416    non-matching string, it can take a very, very long time, even for strings of
3417    quite modest length, because of the nested recursion. PCRE now does better in
3418    some of these cases. It does this by remembering the last required literal
3419    character in the pattern, and pre-searching the subject to ensure it is present
3420    before running the real match. In other words, it applies a heuristic to detect
3421    some types of certain failure quickly, and in the above example, if presented
3422    with a string that has no trailing " it gives "no match" very quickly.
3423    
3424    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
3425    other alternatives are tried instead.
3426    
3427    
3428    Version 2.06 09-Jun-99
3429    ----------------------
3430    
3431    1. Change pcretest's output for amount of store used to show just the code
3432    space, because the remainder (the data block) varies in size between 32-bit and
3433    64-bit systems.
3434    
3435    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
3436    start matching at. This allows lookbehinds to work when searching for multiple
3437    occurrences in a string.
3438    
3439    3. Added additional options to pcretest for testing multiple occurrences:
3440    
3441       /+   outputs the rest of the string that follows a match
3442       /g   loops for multiple occurrences, using the new startoffset argument
3443       /G   loops for multiple occurrences by passing an incremented pointer
3444    
3445    4. PCRE wasn't doing the "first character" optimization for patterns starting
3446    with \b or \B, though it was doing it for other lookbehind assertions. That is,
3447    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
3448    the letter 'x'. On long subject strings, this gives a significant speed-up.
3449    
3450    
3451    Version 2.05 21-Apr-99
3452    ----------------------
3453    
3454    1. Changed the type of magic_number from int to long int so that it works
3455    properly on 16-bit systems.
3456    
3457    2. Fixed a bug which caused patterns starting with .* not to work correctly
3458    when the subject string contained newline characters. PCRE was assuming
3459    anchoring for such patterns in all cases, which is not correct because .* will
3460    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
3461    DOTALL is set at top level; otherwise it knows that patterns starting with .*
3462    must be retried after every newline in the subject.
3463    
3464    
3465    Version 2.04 18-Feb-99
3466    ----------------------
3467    
3468    1. For parenthesized subpatterns with repeats whose minimum was zero, the
3469    computation of the store needed to hold the pattern was incorrect (too large).
3470    If such patterns were nested a few deep, this could multiply and become a real
3471    problem.
3472    
3473    2. Added /M option to pcretest to show the memory requirement of a specific
3474    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
3475    
3476    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
3477    compiled in such a way that the backtracking after subsequent failure was
3478    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
3479    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
3480    
3481    
3482  Version 2.03 02-Feb-99  Version 2.03 02-Feb-99
3483  ----------------------  ----------------------
3484    
3485  1. Fixed typo and small mistake in man page.  1. Fixed typo and small mistake in man page.
3486    
3487  2. Added 4th condition (GPL supersedes) and created separate LICENCE file.  2. Added 4th condition (GPL supersedes if conflict) and created separate
3488    LICENCE file containing the conditions.
3489    
3490  3. Updated pcretest so that patterns such as /abc\/def/ work like they do in  3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
3491  Perl, that is the internal \ allows the delimiter to be included in the  Perl, that is the internal \ allows the delimiter to be included in the

Legend:
Removed from v.29  
changed lines
  Added in v.514

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12