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

Diff of /code/trunk/ChangeLog

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

revision 25 by nigel, Sat Feb 24 21:38:45 2007 UTC revision 91 by nigel, Sat Feb 24 21:41:34 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 6.7 04-Jul-06
5    ---------------------
6    
7     1. In order to handle tests when input lines are enormously long, pcretest has
8        been re-factored so that it automatically extends its buffers when
9        necessary. The code is crude, but this _is_ just a test program. The
10        default size has been increased from 32K to 50K.
11    
12     2. The code in pcre_study() was using the value of the re argument before
13        testing it for NULL. (Of course, in any sensible call of the function, it
14        won't be NULL.)
15    
16     3. The memmove() emulation function in pcre_internal.h, which is used on
17        systems that lack both memmove() and bcopy() - that is, hardly ever -
18        was missing a "static" storage class specifier.
19    
20     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
21        containing an extended class (one that cannot be represented by a bitmap
22        because it contains high-valued characters or Unicode property items, e.g.
23        [\pZ]). Almost always one would set UTF-8 mode when processing such a
24        pattern, but PCRE should not loop if you do not (it no longer does).
25        [Detail: two cases were found: (a) a repeated subpattern containing an
26        extended class; (b) a recursive reference to a subpattern that followed a
27        previous extended class. It wasn't skipping over the extended class
28        correctly when UTF-8 mode was not set.]
29    
30     5. A negated single-character class was not being recognized as fixed-length
31        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
32        compile error "lookbehind assertion is not fixed length".
33    
34     6. The RunPerlTest auxiliary script was showing an unexpected difference
35        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
36        write a Perl script that can interpret lines of an input file either as
37        byte characters or as UTF-8, which is what "perltest" was being required to
38        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
39        can't do is switch easily at run time between having the "use utf8;" pragma
40        or not. In the end, I fudged it by using the RunPerlTest script to insert
41        "use utf8;" explicitly for the UTF-8 tests.
42    
43     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
44        the end of the subject string, contrary to the documentation and to what
45        Perl does. This was true of both matching functions. Now it matches only at
46        the start of the subject and immediately after *internal* newlines.
47    
48     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
49        a pointer to an int instead of a pointer to an unsigned long int. This
50        caused problems on 64-bit systems.
51    
52     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
53        instance of the 'standard' template library not being so standard".
54    
55    10. There was no check on the number of named subpatterns nor the maximum
56        length of a subpattern name. The product of these values is used to compute
57        the size of the memory block for a compiled pattern. By supplying a very
58        long subpattern name and a large number of named subpatterns, the size
59        computation could be caused to overflow. This is now prevented by limiting
60        the length of names to 32 characters, and the number of named subpatterns
61        to 10,000.
62    
63    11. Subpatterns that are repeated with specific counts have to be replicated in
64        the compiled pattern. The size of memory for this was computed from the
65        length of the subpattern and the repeat count. The latter is limited to
66        65535, but there was no limit on the former, meaning that integer overflow
67        could in principle occur. The compiled length of a repeated subpattern is
68        now limited to 30,000 bytes in order to prevent this.
69    
70    12. Added the optional facility to have named substrings with the same name.
71    
72    13. Added the ability to use a named substring as a condition, using the
73        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
74        are numbers (not recommended). Forward references are permitted.
75    
76    14. Added forward references in named backreferences (if you see what I mean).
77    
78    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
79        pattern could run off the end of the subject. For example, the pattern
80        "(?s)(.{1,5})"8 did this with the subject "ab".
81    
82    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
83        PCRE_CASELESS was set when matching characters that were quantified with ?
84        or *.
85    
86    17. A character class other than a single negated character that had a minimum
87        but no maximum quantifier - for example [ab]{6,} - was not handled
88        correctly by pce_dfa_exec(). It would match only one character.
89    
90    18. A valid (though odd) pattern that looked like a POSIX character
91        class but used an invalid character after [ (for example [[,abc,]]) caused
92        pcre_compile() to give the error "Failed: internal error: code overflow" or
93        in some cases to crash with a glibc free() error. This could even happen if
94        the pattern terminated after [[ but there just happened to be a sequence of
95        letters, a binary zero, and a closing ] in the memory that followed.
96    
97    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
98        over the years. Originally (before any Unicode support), just the bottom 8
99        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
100        output from "man perlunicode" includes this:
101    
102          The regular expression compiler produces polymorphic opcodes.  That
103          is, the pattern adapts to the data and automatically switches to
104          the Unicode character scheme when presented with Unicode data--or
105          instead uses a traditional byte scheme when presented with byte
106          data.
107    
108        Sadly, a wide octal escape does not cause a switch, and in a string with
109        no other multibyte characters, these octal escapes are treated as before.
110        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
111        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
112        Unicode string.
113    
114        I have not perpetrated such confusion in PCRE. Up till now, it took just
115        the bottom 8 bits, as in old Perl. I have now made octal escapes with
116        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
117        translate to the appropriate multibyte character.
118    
119    29. Applied some refactoring to reduce the number of warnings from Microsoft
120        and Borland compilers. This has included removing the fudge introduced
121        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
122        a warning about an unused variable.
123    
124    21. PCRE has not included VT (character 0x0b) in the set of whitespace
125        characters since release 4.0, because Perl (from release 5.004) does not.
126        [Or at least, is documented not to: some releases seem to be in conflict
127        with the documentation.] However, when a pattern was studied with
128        pcre_study() and all its branches started with \s, PCRE still included VT
129        as a possible starting character. Of course, this did no harm; it just
130        caused an unnecessary match attempt.
131    
132    22. Removed a now-redundant internal flag bit that recorded the fact that case
133        dependency changed within the pattern. This was once needed for "required
134        byte" processing, but is no longer used. This recovers a now-scarce options
135        bit. Also moved the least significant internal flag bit to the most-
136        significant bit of the word, which was not previously used (hangover from
137        the days when it was an int rather than a uint) to free up another bit for
138        the future.
139    
140    23. Added support for CRLF line endings as well as CR and LF. As well as the
141        default being selectable at build time, it can now be changed at runtime
142        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
143        specify that it is scanning data with non-default line endings.
144    
145    24. Changed the definition of CXXLINK to make it agree with the definition of
146        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
147    
148    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
149        recursions. This makes a big different to stack usage for some patterns.
150    
151    26. If a subpattern containing a named recursion or subroutine reference such
152        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
153        the space required for the compiled pattern went wrong and gave too small a
154        value. Depending on the environment, this could lead to "Failed: internal
155        error: code overflow at offset 49" or "glibc detected double free or
156        corruption" errors.
157    
158    27. Applied patches from Google (a) to support the new newline modes and (b) to
159        advance over multibyte UTF-8 characters in GlobalReplace.
160    
161    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
162        difference for some implementation of PCRE in some Windows version.
163    
164    29. Added some extra testing facilities to pcretest:
165    
166        \q<number>   in a data line sets the "match limit" value
167        \Q<number>   in a data line sets the "match recursion limt" value
168        -S <number>  sets the stack size, where <number> is in megabytes
169    
170        The -S option isn't available for Windows.
171    
172    
173    Version 6.6 06-Feb-06
174    ---------------------
175    
176     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
177        in pcreposix.h. I have copied the definition from pcre.h.
178    
179     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
180        because pcre.h is no longer a built file.
181    
182     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
183        not normally included in the compiled code.
184    
185    
186    Version 6.5 01-Feb-06
187    ---------------------
188    
189     1. When using the partial match feature with pcre_dfa_exec(), it was not
190        anchoring the second and subsequent partial matches at the new starting
191        point. This could lead to incorrect results. For example, with the pattern
192        /1234/, partially matching against "123" and then "a4" gave a match.
193    
194     2. Changes to pcregrep:
195    
196        (a) All non-match returns from pcre_exec() were being treated as failures
197            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
198            error message is output. Some extra information is given for the
199            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
200            probably the only errors that are likely to be caused by users (by
201            specifying a regex that has nested indefinite repeats, for instance).
202            If there are more than 20 of these errors, pcregrep is abandoned.
203    
204        (b) A binary zero was treated as data while matching, but terminated the
205            output line if it was written out. This has been fixed: binary zeroes
206            are now no different to any other data bytes.
207    
208        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
209            used to set a locale for matching. The --locale=xxxx long option has
210            been added (no short equivalent) to specify a locale explicitly on the
211            pcregrep command, overriding the environment variables.
212    
213        (d) When -B was used with -n, some line numbers in the output were one less
214            than they should have been.
215    
216        (e) Added the -o (--only-matching) option.
217    
218        (f) If -A or -C was used with -c (count only), some lines of context were
219            accidentally printed for the final match.
220    
221        (g) Added the -H (--with-filename) option.
222    
223        (h) The combination of options -rh failed to suppress file names for files
224            that were found from directory arguments.
225    
226        (i) Added the -D (--devices) and -d (--directories) options.
227    
228        (j) Added the -F (--fixed-strings) option.
229    
230        (k) Allow "-" to be used as a file name for -f as well as for a data file.
231    
232        (l) Added the --colo(u)r option.
233    
234        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
235            is not present by default.
236    
237     3. A nasty bug was discovered in the handling of recursive patterns, that is,
238        items such as (?R) or (?1), when the recursion could match a number of
239        alternatives. If it matched one of the alternatives, but subsequently,
240        outside the recursion, there was a failure, the code tried to back up into
241        the recursion. However, because of the way PCRE is implemented, this is not
242        possible, and the result was an incorrect result from the match.
243    
244        In order to prevent this happening, the specification of recursion has
245        been changed so that all such subpatterns are automatically treated as
246        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
247    
248     4. I had overlooked the fact that, in some locales, there are characters for
249        which isalpha() is true but neither isupper() nor islower() are true. In
250        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
251        and ordfeminine) are like this. This affected the treatment of \w and \W
252        when they appeared in character classes, but not when they appeared outside
253        a character class. The bit map for "word" characters is now created
254        separately from the results of isalnum() instead of just taking it from the
255        upper, lower, and digit maps. (Plus the underscore character, of course.)
256    
257     5. The above bug also affected the handling of POSIX character classes such as
258        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
259        permanent tables. Instead, the bit maps for such a class were previously
260        created as the appropriate unions of the upper, lower, and digit bitmaps.
261        Now they are created by subtraction from the [[:word:]] class, which has
262        its own bitmap.
263    
264     6. The [[:blank:]] character class matches horizontal, but not vertical space.
265        It is created by subtracting the vertical space characters (\x09, \x0a,
266        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
267        subtraction was done in the overall bitmap for a character class, meaning
268        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
269        be recognized. This bug has been fixed.
270    
271     7. Patches from the folks at Google:
272    
273          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
274          real life, but is still worth protecting against".
275    
276          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
277          regular expressions".
278    
279          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
280          have it.
281    
282          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
283          "configure" and the latter not, in order to fix a problem somebody had
284          with compiling the Arg class on HP-UX.
285    
286          (e) Improve the error-handling of the C++ wrapper a little bit.
287    
288          (f) New tests for checking recursion limiting.
289    
290     8. The pcre_memmove() function, which is used only if the environment does not
291        have a standard memmove() function (and is therefore rarely compiled),
292        contained two bugs: (a) use of int instead of size_t, and (b) it was not
293        returning a result (though PCRE never actually uses the result).
294    
295     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
296        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
297        returned instead of calling malloc() with an overflowing number that would
298        most likely cause subsequent chaos.
299    
300    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
301    
302    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
303        with this option is matched, the nmatch and pmatch options of regexec() are
304        ignored.
305    
306    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
307        provided in case anyone wants to the the POSIX interface with UTF-8
308        strings.
309    
310    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
311        C++ linking (needed for some HP-UX environments).
312    
313    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
314        (unused parameter) and in the pcre_printint() function (omitted "default"
315        switch label when the default is to do nothing).
316    
317    15. Added some code to make it possible, when PCRE is compiled as a C++
318        library, to replace subject pointers for pcre_exec() with a smart pointer
319        class, thus making it possible to process discontinuous strings.
320    
321    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
322        much the same function. They were added by different people who were trying
323        to make PCRE easy to compile on non-Unix systems. It has been suggested
324        that PCRE_EXPORT be abolished now that there is more automatic apparatus
325        for compiling on Windows systems. I have therefore replaced it with
326        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
327        defaults to "extern" for C or "extern C" for C++, which works fine on
328        Unix-like systems. It is now possible to override the value of PCRE_DATA_
329        SCOPE with something explicit in config.h. In addition:
330    
331        (a) pcreposix.h still had just "extern" instead of either of these macros;
332            I have replaced it with PCRE_DATA_SCOPE.
333    
334        (b) Functions such as _pcre_xclass(), which are internal to the library,
335            but external in the C sense, all had PCRE_EXPORT in their definitions.
336            This is apparently wrong for the Windows case, so I have removed it.
337            (It makes no difference on Unix-like systems.)
338    
339    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
340        of recursive calls to match(). This is different to MATCH_LIMIT because
341        that limits the total number of calls to match(), not all of which increase
342        the depth of recursion. Limiting the recursion depth limits the amount of
343        stack (or heap if NO_RECURSE is set) that is used. The default can be set
344        when PCRE is compiled, and changed at run time. A patch from Google adds
345        this functionality to the C++ interface.
346    
347    18. Changes to the handling of Unicode character properties:
348    
349        (a) Updated the table to Unicode 4.1.0.
350    
351        (b) Recognize characters that are not in the table as "Cn" (undefined).
352    
353        (c) I revised the way the table is implemented to a much improved format
354            which includes recognition of ranges. It now supports the ranges that
355            are defined in UnicodeData.txt, and it also amalgamates other
356            characters into ranges. This has reduced the number of entries in the
357            table from around 16,000 to around 3,000, thus reducing its size
358            considerably. I realized I did not need to use a tree structure after
359            all - a binary chop search is just as efficient. Having reduced the
360            number of entries, I extended their size from 6 bytes to 8 bytes to
361            allow for more data.
362    
363        (d) Added support for Unicode script names via properties such as \p{Han}.
364    
365    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
366        matching that character.
367    
368    20. When matching a repeated Unicode property with a minimum greater than zero,
369        (for example \pL{2,}), PCRE could look past the end of the subject if it
370        reached it while seeking the minimum number of characters. This could
371        happen only if some of the characters were more than one byte long, because
372        there is a check for at least the minimum number of bytes.
373    
374    21. Refactored the implementation of \p and \P so as to be more general, to
375        allow for more different types of property in future. This has changed the
376        compiled form incompatibly. Anybody with saved compiled patterns that use
377        \p or \P will have to recompile them.
378    
379    22. Added "Any" and "L&" to the supported property types.
380    
381    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
382        but give a compile time error if the value is greater than 0xff.
383    
384    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
385        accidentally not being installed or uninstalled.
386    
387    25. The pcre.h file was built from pcre.h.in, but the only changes that were
388        made were to insert the current release number. This seemed silly, because
389        it made things harder for people building PCRE on systems that don't run
390        "configure". I have turned pcre.h into a distributed file, no longer built
391        by "configure", with the version identification directly included. There is
392        no longer a pcre.h.in file.
393    
394        However, this change necessitated a change to the pcre-config script as
395        well. It is built from pcre-config.in, and one of the substitutions was the
396        release number. I have updated configure.ac so that ./configure now finds
397        the release number by grepping pcre.h.
398    
399    26. Added the ability to run the tests under valgrind.
400    
401    
402    Version 6.4 05-Sep-05
403    ---------------------
404    
405     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
406        "--" to be printed when multiple files were scanned, even when none of the
407        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
408        consider it to be a bug, and have restored the previous behaviour.
409    
410     2. A couple of code tidies to get rid of compiler warnings.
411    
412     3. The pcretest program used to cheat by referring to symbols in the library
413        whose names begin with _pcre_. These are internal symbols that are not
414        really supposed to be visible externally, and in some environments it is
415        possible to suppress them. The cheating is now confined to including
416        certain files from the library's source, which is a bit cleaner.
417    
418     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
419        file's purpose clearer.
420    
421     5. Reorganized pcre_ucp_findchar().
422    
423    
424    Version 6.3 15-Aug-05
425    ---------------------
426    
427     1. The file libpcre.pc.in did not have general read permission in the tarball.
428    
429     2. There were some problems when building without C++ support:
430    
431        (a) If C++ support was not built, "make install" and "make test" still
432            tried to test it.
433    
434        (b) There were problems when the value of CXX was explicitly set. Some
435            changes have been made to try to fix these, and ...
436    
437        (c) --disable-cpp can now be used to explicitly disable C++ support.
438    
439        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
440            backslash in a target when C++ was disabled. This confuses some
441            versions of "make", apparently. Using an intermediate variable solves
442            this. (Same for CPP_LOBJ.)
443    
444     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
445        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
446        necessary on certain architectures.
447    
448     4. Added a setting of -export-symbols-regex to the link command to remove
449        those symbols that are exported in the C sense, but actually are local
450        within the library, and not documented. Their names all begin with
451        "_pcre_". This is not a perfect job, because (a) we have to except some
452        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
453        available (and never for static libraries). I have made a note to try to
454        find a way round (a) in the future.
455    
456    
457    Version 6.2 01-Aug-05
458    ---------------------
459    
460     1. There was no test for integer overflow of quantifier values. A construction
461        such as {1111111111111111} would give undefined results. What is worse, if
462        a minimum quantifier for a parenthesized subpattern overflowed and became
463        negative, the calculation of the memory size went wrong. This could have
464        led to memory overwriting.
465    
466     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
467    
468     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
469        operating environments where this matters.
470    
471     4. Applied Giuseppe Maxia's patch to add additional features for controlling
472        PCRE options from within the C++ wrapper.
473    
474     5. Named capturing subpatterns were not being correctly counted when a pattern
475        was compiled. This caused two problems: (a) If there were more than 100
476        such subpatterns, the calculation of the memory needed for the whole
477        compiled pattern went wrong, leading to an overflow error. (b) Numerical
478        back references of the form \12, where the number was greater than 9, were
479        not recognized as back references, even though there were sufficient
480        previous subpatterns.
481    
482     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
483        versions of gcc, e.g. 2.95.4.
484    
485    
486    Version 6.1 21-Jun-05
487    ---------------------
488    
489     1. There was one reference to the variable "posix" in pcretest.c that was not
490        surrounded by "#if !defined NOPOSIX".
491    
492     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
493        the cross-check on the old pcre_info() function, for the benefit of the
494        cut-down version of PCRE that is currently imported into Exim.
495    
496     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
497        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
498        patterns that start (?i) but I don't think that matters. The use of (?i) is
499        just an example; this all applies to the other options as well.
500    
501     4. Since libtool seems to echo the compile commands it is issuing, the output
502        from "make" can be reduced a bit by putting "@" in front of each libtool
503        compile command.
504    
505     5. Patch from the folks at Google for configure.in to be a bit more thorough
506        in checking for a suitable C++ installation before trying to compile the
507        C++ stuff. This should fix a reported problem when a compiler was present,
508        but no suitable headers.
509    
510     6. The man pages all had just "PCRE" as their title. I have changed them to
511        be the relevant file name. I have also arranged that these names are
512        retained in the file doc/pcre.txt, which is a concatenation in text format
513        of all the man pages except the little individual ones for each function.
514    
515     7. The NON-UNIX-USE file had not been updated for the different set of source
516        files that come with release 6. I also added a few comments about the C++
517        wrapper.
518    
519    
520    Version 6.0 07-Jun-05
521    ---------------------
522    
523     1. Some minor internal re-organization to help with my DFA experiments.
524    
525     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
526        didn't matter for the library itself when fully configured, but did matter
527        when compiling without UCP support, or within Exim, where the ucp files are
528        not imported.
529    
530     3. Refactoring of the library code to split up the various functions into
531        different source modules. The addition of the new DFA matching code (see
532        below) to a single monolithic source would have made it really too
533        unwieldy, quite apart from causing all the code to be include in a
534        statically linked application, when only some functions are used. This is
535        relevant even without the DFA addition now that patterns can be compiled in
536        one application and matched in another.
537    
538        The downside of splitting up is that there have to be some external
539        functions and data tables that are used internally in different modules of
540        the library but which are not part of the API. These have all had their
541        names changed to start with "_pcre_" so that they are unlikely to clash
542        with other external names.
543    
544     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
545        a different (DFA) algorithm. Although it is slower than the original
546        function, it does have some advantages for certain types of matching
547        problem.
548    
549     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
550        including restarting after a partial match.
551    
552     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
553        defined when compiling for Windows was sent to me. I have put it into the
554        code, though I have no means of testing or verifying it.
555    
556     7. Added the pcre_refcount() auxiliary function.
557    
558     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
559        match before or at the first newline in the subject string. In pcretest,
560        the /f option on a pattern can be used to set this.
561    
562     9. A repeated \w when used in UTF-8 mode with characters greater than 256
563        would behave wrongly. This has been present in PCRE since release 4.0.
564    
565    10. A number of changes to the pcregrep command:
566    
567        (a) Refactored how -x works; insert ^(...)$ instead of setting
568            PCRE_ANCHORED and checking the length, in preparation for adding
569            something similar for -w.
570    
571        (b) Added the -w (match as a word) option.
572    
573        (c) Refactored the way lines are read and buffered so as to have more
574            than one at a time available.
575    
576        (d) Implemented a pcregrep test script.
577    
578        (e) Added the -M (multiline match) option. This allows patterns to match
579            over several lines of the subject. The buffering ensures that at least
580            8K, or the rest of the document (whichever is the shorter) is available
581            for matching (and similarly the previous 8K for lookbehind assertions).
582    
583        (f) Changed the --help output so that it now says
584    
585              -w, --word-regex(p)
586    
587            instead of two lines, one with "regex" and the other with "regexp"
588            because that confused at least one person since the short forms are the
589            same. (This required a bit of code, as the output is generated
590            automatically from a table. It wasn't just a text change.)
591    
592        (g) -- can be used to terminate pcregrep options if the next thing isn't an
593            option but starts with a hyphen. Could be a pattern or a path name
594            starting with a hyphen, for instance.
595    
596        (h) "-" can be given as a file name to represent stdin.
597    
598        (i) When file names are being printed, "(standard input)" is used for
599            the standard input, for compatibility with GNU grep. Previously
600            "<stdin>" was used.
601    
602        (j) The option --label=xxx can be used to supply a name to be used for
603            stdin when file names are being printed. There is no short form.
604    
605        (k) Re-factored the options decoding logic because we are going to add
606            two more options that take data. Such options can now be given in four
607            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
608    
609        (l) Added the -A, -B, and -C options for requesting that lines of context
610            around matches be printed.
611    
612        (m) Added the -L option to print the names of files that do not contain
613            any matching lines, that is, the complement of -l.
614    
615        (n) The return code is 2 if any file cannot be opened, but pcregrep does
616            continue to scan other files.
617    
618        (o) The -s option was incorrectly implemented. For compatibility with other
619            greps, it now suppresses the error message for a non-existent or non-
620            accessible file (but not the return code). There is a new option called
621            -q that suppresses the output of matching lines, which was what -s was
622            previously doing.
623    
624        (p) Added --include and --exclude options to specify files for inclusion
625            and exclusion when recursing.
626    
627    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
628        Hopefully, it now does.
629    
630    12. Missing cast in pcre_study().
631    
632    13. Added an "uninstall" target to the makefile.
633    
634    14. Replaced "extern" in the function prototypes in Makefile.in with
635        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
636        world, but is set differently for Windows.
637    
638    15. Added a second compiling function called pcre_compile2(). The only
639        difference is that it has an extra argument, which is a pointer to an
640        integer error code. When there is a compile-time failure, this is set
641        non-zero, in addition to the error test pointer being set to point to an
642        error message. The new argument may be NULL if no error number is required
643        (but then you may as well call pcre_compile(), which is now just a
644        wrapper). This facility is provided because some applications need a
645        numeric error indication, but it has also enabled me to tidy up the way
646        compile-time errors are handled in the POSIX wrapper.
647    
648    16. Added VPATH=.libs to the makefile; this should help when building with one
649        prefix path and installing with another. (Or so I'm told by someone who
650        knows more about this stuff than I do.)
651    
652    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
653        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
654        match everything, including newlines. This is not POSIX-compatible, but
655        somebody wanted the feature. From pcretest it can be activated by using
656        both the P and the s flags.
657    
658    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
659    
660    19. libpcre.pc was being incorrectly installed as executable.
661    
662    20. A couple of places in pcretest check for end-of-line by looking for '\n';
663        it now also looks for '\r' so that it will work unmodified on Windows.
664    
665    21. Added Google's contributed C++ wrapper to the distribution.
666    
667    22. Added some untidy missing memory free() calls in pcretest, to keep
668        Electric Fence happy when testing.
669    
670    
671    
672    Version 5.0 13-Sep-04
673    ---------------------
674    
675     1. Internal change: literal characters are no longer packed up into items
676        containing multiple characters in a single byte-string. Each character
677        is now matched using a separate opcode. However, there may be more than one
678        byte in the character in UTF-8 mode.
679    
680     2. The pcre_callout_block structure has two new fields: pattern_position and
681        next_item_length. These contain the offset in the pattern to the next match
682        item, and its length, respectively.
683    
684     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
685        insertion of callouts before each pattern item. Added the /C option to
686        pcretest to make use of this.
687    
688     4. On the advice of a Windows user, the lines
689    
690          #if defined(_WIN32) || defined(WIN32)
691          _setmode( _fileno( stdout ), 0x8000 );
692          #endif  /* defined(_WIN32) || defined(WIN32) */
693    
694        have been added to the source of pcretest. This apparently does useful
695        magic in relation to line terminators.
696    
697     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
698        for the benefit of those environments where the "b" makes a difference.
699    
700     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
701        to know about it. I have put a hack into configure.in that adds in code
702        to set GCC=yes if CC=icc. This seems to end up at a point in the
703        generated configure script that is early enough to affect the setting of
704        compiler options, which is what is needed, but I have no means of testing
705        whether it really works. (The user who reported this had patched the
706        generated configure script, which of course I cannot do.)
707    
708        LATER: After change 22 below (new libtool files), the configure script
709        seems to know about icc (and also ecc). Therefore, I have commented out
710        this hack in configure.in.
711    
712     7. Added support for pkg-config (2 patches were sent in).
713    
714     8. Negated POSIX character classes that used a combination of internal tables
715        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
716        [[:^ascii]]. Typically, they would match almost any characters. The other
717        POSIX classes were not broken in this way.
718    
719     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
720        to find the match, as PCRE was deluded into thinking that the match had to
721        start at the start point or following a newline. The same bug applied to
722        patterns with negative forward assertions or any backward assertions
723        preceding ".*" at the start, unless the pattern required a fixed first
724        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
725    
726    10. In UTF-8 mode, when moving forwards in the subject after a failed match
727        starting at the last subject character, bytes beyond the end of the subject
728        string were read.
729    
730    11. Renamed the variable "class" as "classbits" to make life easier for C++
731        users. (Previously there was a macro definition, but it apparently wasn't
732        enough.)
733    
734    12. Added the new field "tables" to the extra data so that tables can be passed
735        in at exec time, or the internal tables can be re-selected. This allows
736        a compiled regex to be saved and re-used at a later time by a different
737        program that might have everything at different addresses.
738    
739    13. Modified the pcre-config script so that, when run on Solaris, it shows a
740        -R library as well as a -L library.
741    
742    14. The debugging options of pcretest (-d on the command line or D on a
743        pattern) showed incorrect output for anything following an extended class
744        that contained multibyte characters and which was followed by a quantifier.
745    
746    15. Added optional support for general category Unicode character properties
747        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
748        support. It adds about 90K to the size of the library. The meanings of the
749        inbuilt class escapes such as \d and \s have NOT been changed.
750    
751    16. Updated pcredemo.c to include calls to free() to release the memory for the
752        compiled pattern.
753    
754    17. The generated file chartables.c was being created in the source directory
755        instead of in the building directory. This caused the build to fail if the
756        source directory was different from the building directory, and was
757        read-only.
758    
759    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
760        file. No doubt somebody will tell me if they don't make sense... Also added
761        Dan Mooney's comments about building on OpenVMS.
762    
763    19. Added support for partial matching via the PCRE_PARTIAL option for
764        pcre_exec() and the \P data escape in pcretest.
765    
766    20. Extended pcretest with 3 new pattern features:
767    
768        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
769              write the compiled pattern to the file whose name is "rest-of-line".
770              This is a straight binary dump of the data, with the saved pointer to
771              the character tables forced to be NULL. The study data, if any, is
772              written too. After writing, pcretest reads a new pattern.
773    
774        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
775              compiled pattern from the given file. There must not be any
776              occurrences of "<" in the file name (pretty unlikely); if there are,
777              pcretest will instead treat the initial "<" as a pattern delimiter.
778              After reading in the pattern, pcretest goes on to read data lines as
779              usual.
780    
781        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
782              and 16-bit fields in a compiled pattern, to simulate a pattern that
783              was compiled on a host of opposite endianness.
784    
785    21. The pcre-exec() function can now cope with patterns that were compiled on
786        hosts of opposite endianness, with this restriction:
787    
788          As for any compiled expression that is saved and used later, the tables
789          pointer field cannot be preserved; the extra_data field in the arguments
790          to pcre_exec() should be used to pass in a tables address if a value
791          other than the default internal tables were used at compile time.
792    
793    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
794        now diagnosed as an error. Previously, most of the time, a negative number
795        would have been treated as zero, but if in addition "ovector" was passed as
796        NULL, a crash could occur.
797    
798    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
799        new versions from the libtool 1.5 distribution (the last one is a copy of
800        a file called libtool.m4). This seems to have fixed the need to patch
801        "configure" to support Darwin 1.3 (which I used to do). However, I still
802        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
803        workstation).
804    
805    24. Changed the PCRE licence to be the more standard "BSD" licence.
806    
807    
808    Version 4.5 01-Dec-03
809    ---------------------
810    
811     1. There has been some re-arrangement of the code for the match() function so
812        that it can be compiled in a version that does not call itself recursively.
813        Instead, it keeps those local variables that need separate instances for
814        each "recursion" in a frame on the heap, and gets/frees frames whenever it
815        needs to "recurse". Keeping track of where control must go is done by means
816        of setjmp/longjmp. The whole thing is implemented by a set of macros that
817        hide most of the details from the main code, and operates only if
818        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
819        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
820        operating.
821    
822        To make it easier for callers to provide specially tailored get/free
823        functions for this usage, two new functions, pcre_stack_malloc, and
824        pcre_stack_free, are used. They are always called in strict stacking order,
825        and the size of block requested is always the same.
826    
827        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
828        PCRE has been compiled to use the stack or the heap for recursion. The
829        -C option of pcretest uses this to show which version is compiled.
830    
831        A new data escape \S, is added to pcretest; it causes the amounts of store
832        obtained and freed by both kinds of malloc/free at match time to be added
833        to the output.
834    
835     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
836        what's available on my current Linux desktop machine.
837    
838     3. When matching a UTF-8 string, the test for a valid string at the start has
839        been extended. If start_offset is not zero, PCRE now checks that it points
840        to a byte that is the start of a UTF-8 character. If not, it returns
841        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
842        this is necessary because there may be backward assertions in the pattern.
843        When matching the same subject several times, it may save resources to use
844        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
845    
846     4. The code for checking the validity of UTF-8 strings has been tightened so
847        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
848        containing "overlong sequences".
849    
850     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
851        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
852        should have been "&", but it just so happened that all the cases this let
853        through by mistake were picked up later in the function.
854    
855     6. I had used a variable called "isblank" - this is a C99 function, causing
856        some compilers to warn. To avoid this, I renamed it (as "blankclass").
857    
858     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
859        prompting; (b) run "./pcretest /dev/null" at the start of the test script
860        so the version is shown; (c) stop "make test" echoing "./RunTest".
861    
862     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
863    
864     9. The prototype for memmove() for systems that don't have it was using
865        size_t, but the inclusion of the header that defines size_t was later. I've
866        moved the #includes for the C headers earlier to avoid this.
867    
868    10. Added some adjustments to the code to make it easier to compiler on certain
869        special systems:
870    
871          (a) Some "const" qualifiers were missing.
872          (b) Added the macro EXPORT before all exported functions; by default this
873              is defined to be empty.
874          (c) Changed the dftables auxiliary program (that builds chartables.c) so
875              that it reads its output file name as an argument instead of writing
876              to the standard output and assuming this can be redirected.
877    
878    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
879        class containing characters with values greater than 255, PCRE compilation
880        went into a loop.
881    
882    12. A recursive reference to a subpattern that was within another subpattern
883        that had a minimum quantifier of zero caused PCRE to crash. For example,
884        (x(y(?2))z)? provoked this bug with a subject that got as far as the
885        recursion. If the recursively-called subpattern itself had a zero repeat,
886        that was OK.
887    
888    13. In pcretest, the buffer for reading a data line was set at 30K, but the
889        buffer into which it was copied (for escape processing) was still set at
890        1024, so long lines caused crashes.
891    
892    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
893        "internal error: code overflow...". This applied to any character class
894        that was followed by a possessive quantifier.
895    
896    15. Modified the Makefile to add libpcre.la as a prerequisite for
897        libpcreposix.la because I was told this is needed for a parallel build to
898        work.
899    
900    16. If a pattern that contained .* following optional items at the start was
901        studied, the wrong optimizing data was generated, leading to matching
902        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
903        matching string must start with a or b or c. The correct conclusion for
904        this pattern is that a match can start with any character.
905    
906    
907    Version 4.4 13-Aug-03
908    ---------------------
909    
910     1. In UTF-8 mode, a character class containing characters with values between
911        127 and 255 was not handled correctly if the compiled pattern was studied.
912        In fixing this, I have also improved the studying algorithm for such
913        classes (slightly).
914    
915     2. Three internal functions had redundant arguments passed to them. Removal
916        might give a very teeny performance improvement.
917    
918     3. Documentation bug: the value of the capture_top field in a callout is *one
919        more than* the number of the hightest numbered captured substring.
920    
921     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
922        in incorrectly linking with a previously installed version. They now link
923        explicitly with libpcre.la.
924    
925     5. configure.in no longer needs to recognize Cygwin specially.
926    
927     6. A problem in pcre.in for Windows platforms is fixed.
928    
929     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
930        pcretest, it used to include the size of the study block as part of its
931        output. Unfortunately, the structure contains a field that has a different
932        size on different hardware architectures. This meant that the tests that
933        showed this size failed. As the block is currently always of a fixed size,
934        this information isn't actually particularly useful in pcretest output, so
935        I have just removed it.
936    
937     8. Three pre-processor statements accidentally did not start in column 1.
938        Sadly, there are *still* compilers around that complain, even though
939        standard C has not required this for well over a decade. Sigh.
940    
941     9. In pcretest, the code for checking callouts passed small integers in the
942        callout_data field, which is a void * field. However, some picky compilers
943        complained about the casts involved for this on 64-bit systems. Now
944        pcretest passes the address of the small integer instead, which should get
945        rid of the warnings.
946    
947    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
948        both compile and run time, and gives an error if an invalid UTF-8 sequence
949        is found. There is a option for disabling this check in cases where the
950        string is known to be correct and/or the maximum performance is wanted.
951    
952    11. In response to a bug report, I changed one line in Makefile.in from
953    
954            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
955        to
956            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
957    
958        to look similar to other lines, but I have no way of telling whether this
959        is the right thing to do, as I do not use Windows. No doubt I'll get told
960        if it's wrong...
961    
962    
963    Version 4.3 21-May-03
964    ---------------------
965    
966    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
967       Makefile.
968    
969    2. Some refactoring to improve the quality of the code:
970    
971       (i)   The utf8_table... variables are now declared "const".
972    
973       (ii)  The code for \cx, which used the "case flipping" table to upper case
974             lower case letters, now just substracts 32. This is ASCII-specific,
975             but the whole concept of \cx is ASCII-specific, so it seems
976             reasonable.
977    
978       (iii) PCRE was using its character types table to recognize decimal and
979             hexadecimal digits in the pattern. This is silly, because it handles
980             only 0-9, a-f, and A-F, but the character types table is locale-
981             specific, which means strange things might happen. A private
982             table is now used for this - though it costs 256 bytes, a table is
983             much faster than multiple explicit tests. Of course, the standard
984             character types table is still used for matching digits in subject
985             strings against \d.
986    
987       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
988             ending in _t are). So I've renamed it as ESC_tee.
989    
990    3. The first argument for regexec() in the POSIX wrapper should have been
991       defined as "const".
992    
993    4. Changed pcretest to use malloc() for its buffers so that they can be
994       Electric Fenced for debugging.
995    
996    5. There were several places in the code where, in UTF-8 mode, PCRE would try
997       to read one or more bytes before the start of the subject string. Often this
998       had no effect on PCRE's behaviour, but in some circumstances it could
999       provoke a segmentation fault.
1000    
1001    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1002       to try to read one or more bytes before the start of the subject string.
1003    
1004    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1005       UTF-8 support could misbehave in various ways if the subject string
1006       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1007       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1008       back over UTF-8 characters.)
1009    
1010    
1011    Version 4.2 14-Apr-03
1012    ---------------------
1013    
1014    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1015    
1016    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1017         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1018         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1019         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1020         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1021           and BUILD_EXEEXT
1022         Note: automatic setting of the BUILD variables is not yet working
1023         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1024           compile-time but not at link-time
1025         [LINK]: use for linking executables only
1026         make different versions for Windows and non-Windows
1027         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1028           libraries
1029         [LINK_FOR_BUILD]: new variable
1030         [OBJEXT]: use throughout
1031         [EXEEXT]: use throughout
1032         <winshared>: new target
1033         <wininstall>: new target
1034         <dftables.o>: use native compiler
1035         <dftables>: use native linker
1036         <install>: handle Windows platform correctly
1037         <clean>: ditto
1038         <check>: ditto
1039         copy DLL to top builddir before testing
1040    
1041       As part of these changes, -no-undefined was removed again. This was reported
1042       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1043       in any case.
1044    
1045    3. Some tidies to get rid of compiler warnings:
1046    
1047       . In the match_data structure, match_limit was an unsigned long int, whereas
1048         match_call_count was an int. I've made them both unsigned long ints.
1049    
1050       . In pcretest the fact that a const uschar * doesn't automatically cast to
1051         a void * provoked a warning.
1052    
1053       . Turning on some more compiler warnings threw up some "shadow" variables
1054         and a few more missing casts.
1055    
1056    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1057       option, a class that contained a single character with a value between 128
1058       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1059    
1060    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1061       option, a class that contained several characters, but with at least one
1062       whose value was between 128 and 255 caused PCRE to crash.
1063    
1064    
1065    Version 4.1 12-Mar-03
1066    ---------------------
1067    
1068    1. Compiling with gcc -pedantic found a couple of places where casts were
1069    needed, and a string in dftables.c that was longer than standard compilers are
1070    required to support.
1071    
1072    2. Compiling with Sun's compiler found a few more places where the code could
1073    be tidied up in order to avoid warnings.
1074    
1075    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1076    first of these names is deprecated in the latest Autoconf in favour of the name
1077    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1078    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1079    analogy I have changed it to CFLAGS_FOR_BUILD.
1080    
1081    4. Added -no-undefined to the linking command in the Makefile, because this is
1082    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1083    linking step for the pcreposix library.
1084    
1085    5. PCRE was failing to diagnose the case of two named groups with the same
1086    name.
1087    
1088    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1089    literal character that is needed in the subject for a match, and scans along to
1090    ensure that it is present before embarking on the full matching process. This
1091    saves time in cases of nested unlimited repeats that are never going to match.
1092    Problem: the scan can take a lot of time if the subject is very long (e.g.
1093    megabytes), thus penalizing straightforward matches. It is now done only if the
1094    amount of subject to be scanned is less than 1000 bytes.
1095    
1096    7. A lesser problem with the same optimization is that it was recording the
1097    first character of an anchored pattern as "needed", thus provoking a search
1098    right along the subject, even when the first match of the pattern was going to
1099    fail. The "needed" character is now not set for anchored patterns, unless it
1100    follows something in the pattern that is of non-fixed length. Thus, it still
1101    fulfils its original purpose of finding quick non-matches in cases of nested
1102    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1103    
1104    
1105    Version 4.0 17-Feb-03
1106    ---------------------
1107    
1108    1. If a comment in an extended regex that started immediately after a meta-item
1109    extended to the end of string, PCRE compiled incorrect data. This could lead to
1110    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1111    
1112    2. Moved to autoconf 2.53 and libtool 1.4.2.
1113    
1114    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1115    the special perltest8 script is no longer needed - all the tests can be run
1116    from a single perltest script.
1117    
1118    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1119    by \s. It has now been removed in PCRE. This means it isn't recognized as
1120    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1121    class [:space:] *does* include VT, thereby creating a mess.
1122    
1123    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1124    space and tab.
1125    
1126    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1127    its new features into the main test script, reducing the number of scripts.
1128    
1129    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1130    were backward compatible, and made the (?i) apply to the whole pattern, as if
1131    /i were given. Now it behaves more logically, and applies the option setting
1132    only to what follows. PCRE has been changed to follow suit. However, if it
1133    finds options settings right at the start of the pattern, it extracts them into
1134    the global options, as before. Thus, they show up in the info data.
1135    
1136    8. Added support for the \Q...\E escape sequence. Characters in between are
1137    treated as literals. This is slightly different from Perl in that $ and @ are
1138    also handled as literals inside the quotes. In Perl, they will cause variable
1139    interpolation. Note the following examples:
1140    
1141        Pattern            PCRE matches      Perl matches
1142    
1143        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
1144        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
1145        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
1146    
1147    For compatibility with Perl, \Q...\E sequences are recognized inside character
1148    classes as well as outside them.
1149    
1150    9. Re-organized 3 code statements in pcretest to avoid "overflow in
1151    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1152    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1153    signed/unsigned warnings.
1154    
1155    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1156    option for pcretest, so I've replaced it by a simple function that does just
1157    that job.
1158    
1159    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1160    "pcregrep -".
1161    
1162    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1163    Java package. This provides some syntactic sugar for simple cases of what my
1164    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1165    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1166    item, you can use this simplified notation. Note that only makes sense with
1167    greedy quantifiers. Consequently, the use of the possessive quantifier forces
1168    greediness, whatever the setting of the PCRE_UNGREEDY option.
1169    
1170    13. A change of greediness default within a pattern was not taking effect at
1171    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1172    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1173    was abstracted outside.
1174    
1175    14. PCRE now supports the \G assertion. It is true when the current matching
1176    position is at the start point of the match. This differs from \A when the
1177    starting offset is non-zero. Used with the /g option of pcretest (or similar
1178    code), it works in the same way as it does for Perl's /g option. If all
1179    alternatives of a regex begin with \G, the expression is anchored to the start
1180    match position, and the "anchored" flag is set in the compiled expression.
1181    
1182    15. Some bugs concerning the handling of certain option changes within patterns
1183    have been fixed. These applied to options other than (?ims). For example,
1184    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1185    been the other way round. Some of this was related to change 7 above.
1186    
1187    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1188    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1189    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1190    POSIX classes only within a class (e.g. /[[:alpha:]]/).
1191    
1192    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1193    mode. Unlike ".", it always matches newline, whatever the setting of
1194    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1195    assertions. Perl allows it, but it doesn't (in general) work because it can't
1196    calculate the length of the lookbehind. At least, that's the case for Perl
1197    5.8.0 - I've been told they are going to document that it doesn't work in
1198    future.
1199    
1200    18. Added an error diagnosis for escapes that PCRE does not support: these are
1201    \L, \l, \N, \P, \p, \U, \u, and \X.
1202    
1203    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1204    reading past the end of the pattern in cases such as /[abcd/.
1205    
1206    20. PCRE was getting more memory than necessary for patterns with classes that
1207    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1208    
1209    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1210    compiling PCRE for use with Virtual Pascal.
1211    
1212    22. Small fix to the Makefile to make it work properly if the build is done
1213    outside the source tree.
1214    
1215    23. Added a new extension: a condition to go with recursion. If a conditional
1216    subpattern starts with (?(R) the "true" branch is used if recursion has
1217    happened, whereas the "false" branch is used only at the top level.
1218    
1219    24. When there was a very long string of literal characters (over 255 bytes
1220    without UTF support, over 250 bytes with UTF support), the computation of how
1221    much memory was required could be incorrect, leading to segfaults or other
1222    strange effects.
1223    
1224    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1225    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1226    there was a subsequent back reference to those brackets. This meant that, for
1227    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1228    possible to check for precisely this case. All we can do is abandon the
1229    optimization if .* occurs inside capturing brackets when there are any back
1230    references whatsoever. (See below for a better fix that came later.)
1231    
1232    26. The handling of the optimization for finding the first character of a
1233    non-anchored pattern, and for finding a character that is required later in the
1234    match were failing in some cases. This didn't break the matching; it just
1235    failed to optimize when it could. The way this is done has been re-implemented.
1236    
1237    27. Fixed typo in error message for invalid (?R item (it said "(?p").
1238    
1239    28. Added a new feature that provides some of the functionality that Perl
1240    provides with (?{...}). The facility is termed a "callout". The way it is done
1241    in PCRE is for the caller to provide an optional function, by setting
1242    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1243    global variable. By default it is unset, which disables all calling out. To get
1244    the function called, the regex must include (?C) at appropriate points. This
1245    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1246    This provides a means of identifying different callout points. When PCRE
1247    reaches such a point in the regex, if pcre_callout has been set, the external
1248    function is called. It is provided with data in a structure called
1249    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1250    matching continues; if it returns a non-zero value, the match at the current
1251    point fails. However, backtracking will occur if possible. [This was changed
1252    later and other features added - see item 49 below.]
1253    
1254    29. pcretest is upgraded to test the callout functionality. It provides a
1255    callout function that displays information. By default, it shows the start of
1256    the match and the current position in the text. There are some new data escapes
1257    to vary what happens:
1258    
1259        \C+         in addition, show current contents of captured substrings
1260        \C-         do not supply a callout function
1261        \C!n        return 1 when callout number n is reached
1262        \C!n!m      return 1 when callout number n is reached for the mth time
1263    
1264    30. If pcregrep was called with the -l option and just a single file name, it
1265    output "<stdin>" if a match was found, instead of the file name.
1266    
1267    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1268    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1269    pcre_exec(). This saves a malloc/free per call. The default value of
1270    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1271    when configuring.
1272    
1273    32. The default maximum size of a compiled pattern is 64K. There have been a
1274    few cases of people hitting this limit. The code now uses macros to handle the
1275    storing of links as offsets within the compiled pattern. It defaults to 2-byte
1276    links, but this can be changed to 3 or 4 bytes by --with-link-size when
1277    configuring. Tests 2 and 5 work only with 2-byte links because they output
1278    debugging information about compiled patterns.
1279    
1280    33. Internal code re-arrangements:
1281    
1282    (a) Moved the debugging function for printing out a compiled regex into
1283        its own source file (printint.c) and used #include to pull it into
1284        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1285        separate copies.
1286    
1287    (b) Defined the list of op-code names for debugging as a macro in
1288        internal.h so that it is next to the definition of the opcodes.
1289    
1290    (c) Defined a table of op-code lengths for simpler skipping along compiled
1291        code. This is again a macro in internal.h so that it is next to the
1292        definition of the opcodes.
1293    
1294    34. Added support for recursive calls to individual subpatterns, along the
1295    lines of Robin Houston's patch (but implemented somewhat differently).
1296    
1297    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1298    allow it to read and process whole directories in Win32. This code was
1299    contributed by Lionel Fourquaux; it has not been tested by me.
1300    
1301    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1302    used to name a group. Names consist of alphanumerics and underscores, and must
1303    be unique. Back references use the syntax (?P=name) and recursive calls use
1304    (?P>name) which is a PCRE extension to the Python extension. Groups still have
1305    numbers. The function pcre_fullinfo() can be used after compilation to extract
1306    a name/number map. There are three relevant calls:
1307    
1308      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
1309      PCRE_INFO_NAMECOUNT            yields the number of entries
1310      PCRE_INFO_NAMETABLE            yields a pointer to the map.
1311    
1312    The map is a vector of fixed-size entries. The size of each entry depends on
1313    the length of the longest name used. The first two bytes of each entry are the
1314    group number, most significant byte first. There follows the corresponding
1315    name, zero terminated. The names are in alphabetical order.
1316    
1317    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1318    case instead of 255. Making it the same both with and without UTF-8 support
1319    means that the same test output works with both.
1320    
1321    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1322    calling malloc() with a zero argument.
1323    
1324    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1325    optimization. I've improved things by keeping a bitmap of backreferences with
1326    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1327    fact referenced, the optimization can be applied. It is unlikely that a
1328    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1329    the match to follow \n) will appear inside brackets with a number greater than
1330    31, but if it does, any back reference > 31 suppresses the optimization.
1331    
1332    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1333    of disabling numbered capturing parentheses. Any opening parenthesis that is
1334    not followed by ? behaves as if it were followed by ?: but named parentheses
1335    can still be used for capturing (and they will acquire numbers in the usual
1336    way).
1337    
1338    41. Redesigned the return codes from the match() function into yes/no/error so
1339    that errors can be passed back from deep inside the nested calls. A malloc
1340    failure while inside a recursive subpattern call now causes the
1341    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1342    
1343    42. It is now possible to set a limit on the number of times the match()
1344    function is called in a call to pcre_exec(). This facility makes it possible to
1345    limit the amount of recursion and backtracking, though not in a directly
1346    obvious way, because the match() function is used in a number of different
1347    circumstances. The count starts from zero for each position in the subject
1348    string (for non-anchored patterns). The default limit is, for compatibility, a
1349    large number, namely 10 000 000. You can change this in two ways:
1350    
1351    (a) When configuring PCRE before making, you can use --with-match-limit=n
1352        to set a default value for the compiled library.
1353    
1354    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1355        a different value is set. See 45 below.
1356    
1357    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1358    
1359    43. Added a new function pcre_config(int, void *) to enable run-time extraction
1360    of things that can be changed at compile time. The first argument specifies
1361    what is wanted and the second points to where the information is to be placed.
1362    The current list of available information is:
1363    
1364      PCRE_CONFIG_UTF8
1365    
1366    The output is an integer that is set to one if UTF-8 support is available;
1367    otherwise it is set to zero.
1368    
1369      PCRE_CONFIG_NEWLINE
1370    
1371    The output is an integer that it set to the value of the code that is used for
1372    newline. It is either LF (10) or CR (13).
1373    
1374      PCRE_CONFIG_LINK_SIZE
1375    
1376    The output is an integer that contains the number of bytes used for internal
1377    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1378    
1379      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1380    
1381    The output is an integer that contains the threshold above which the POSIX
1382    interface uses malloc() for output vectors. See item 31 above.
1383    
1384      PCRE_CONFIG_MATCH_LIMIT
1385    
1386    The output is an unsigned integer that contains the default limit of the number
1387    of match() calls in a pcre_exec() execution. See 42 above.
1388    
1389    44. pcretest has been upgraded by the addition of the -C option. This causes it
1390    to extract all the available output from the new pcre_config() function, and to
1391    output it. The program then exits immediately.
1392    
1393    45. A need has arisen to pass over additional data with calls to pcre_exec() in
1394    order to support additional features. One way would have been to define
1395    pcre_exec2() (for example) with extra arguments, but this would not have been
1396    extensible, and would also have required all calls to the original function to
1397    be mapped to the new one. Instead, I have chosen to extend the mechanism that
1398    is used for passing in "extra" data from pcre_study().
1399    
1400    The pcre_extra structure is now exposed and defined in pcre.h. It currently
1401    contains the following fields:
1402    
1403      flags         a bitmap indicating which of the following fields are set
1404      study_data    opaque data from pcre_study()
1405      match_limit   a way of specifying a limit on match() calls for a specific
1406                      call to pcre_exec()
1407      callout_data  data for callouts (see 49 below)
1408    
1409    The flag bits are also defined in pcre.h, and are
1410    
1411      PCRE_EXTRA_STUDY_DATA
1412      PCRE_EXTRA_MATCH_LIMIT
1413      PCRE_EXTRA_CALLOUT_DATA
1414    
1415    The pcre_study() function now returns one of these new pcre_extra blocks, with
1416    the actual study data pointed to by the study_data field, and the
1417    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1418    before. That is, this change is entirely upwards-compatible and requires no
1419    change to existing code.
1420    
1421    If you want to pass in additional data to pcre_exec(), you can either place it
1422    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1423    block.
1424    
1425    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1426    data string contains the escape sequence \M, pcretest calls pcre_exec() several
1427    times with different match limits, until it finds the minimum value needed for
1428    pcre_exec() to complete. The value is then output. This can be instructive; for
1429    most simple matches the number is quite small, but for pathological cases it
1430    gets very large very quickly.
1431    
1432    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1433    returns the size of the data block pointed to by the study_data field in a
1434    pcre_extra block, that is, the value that was passed as the argument to
1435    pcre_malloc() when PCRE was getting memory in which to place the information
1436    created by pcre_study(). The fourth argument should point to a size_t variable.
1437    pcretest has been extended so that this information is shown after a successful
1438    pcre_study() call when information about the compiled regex is being displayed.
1439    
1440    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1441    because what follows is always an absolute path. (Later: it turns out that this
1442    is more than cosmetic for MinGW, because it doesn't like empty path
1443    components.)
1444    
1445    49. Some changes have been made to the callout feature (see 28 above):
1446    
1447    (i)  A callout function now has three choices for what it returns:
1448    
1449           0  =>  success, carry on matching
1450         > 0  =>  failure at this point, but backtrack if possible
1451         < 0  =>  serious error, return this value from pcre_exec()
1452    
1453         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1454         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1455         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1456         use by callout functions. It will never be used by PCRE itself.
1457    
1458    (ii) The pcre_extra structure (see 45 above) has a void * field called
1459         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1460         pcre_callout_block structure has a field of the same name. The contents of
1461         the field passed in the pcre_extra structure are passed to the callout
1462         function in the corresponding field in the callout block. This makes it
1463         easier to use the same callout-containing regex from multiple threads. For
1464         testing, the pcretest program has a new data escape
1465    
1466           \C*n        pass the number n (may be negative) as callout_data
1467    
1468         If the callout function in pcretest receives a non-zero value as
1469         callout_data, it returns that value.
1470    
1471    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1472    there were some redundant $(CFLAGS) in commands that are now specified as
1473    $(LINK), which already includes $(CFLAGS).
1474    
1475    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1476    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1477    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1478    one-byte characters throughout. Note that case-insensitive matching applies
1479    only to characters whose values are less than 256. PCRE doesn't support the
1480    notion of cases for higher-valued characters.
1481    
1482    (i)   A character class whose characters are all within 0-255 is handled as
1483          a bit map, and the map is inverted for negative classes. Previously, a
1484          character > 255 always failed to match such a class; however it should
1485          match if the class was a negative one (e.g. [^ab]). This has been fixed.
1486    
1487    (ii)  A negated character class with a single character < 255 is coded as
1488          "not this character" (OP_NOT). This wasn't working properly when the test
1489          character was multibyte, either singly or repeated.
1490    
1491    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1492          mode, for example: \x{100}{2,3}.
1493    
1494    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1495          singly or repeated) now correctly test multibyte characters. However,
1496          PCRE doesn't recognize any characters with values greater than 255 as
1497          digits, spaces, or word characters. Such characters always match \D, \S,
1498          and \W, and never match \d, \s, or \w.
1499    
1500    (v)   Classes may now contain characters and character ranges with values
1501          greater than 255. For example: [ab\x{100}-\x{400}].
1502    
1503    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
1504          PCRE in UTF-8 mode.
1505    
1506    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1507    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1508    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1509    value.)
1510    
1511    53. The single man page has become too large. I have therefore split it up into
1512    a number of separate man pages. These also give rise to individual HTML pages;
1513    these are now put in a separate directory, and there is an index.html page that
1514    lists them all. Some hyperlinking between the pages has been installed.
1515    
1516    54. Added convenience functions for handling named capturing parentheses.
1517    
1518    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1519    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1520    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1521    are faulted.
1522    
1523    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1524    calling configure. These values are used when compiling the dftables.c program
1525    which is run to generate the source of the default character tables. They
1526    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1527    you will need to set these values.
1528    
1529    57. Updated the building process for Windows DLL, as provided by Fred Cox.
1530    
1531    
1532    Version 3.9 02-Jan-02
1533    ---------------------
1534    
1535    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
1536    
1537    2. If --disable-static was given, the building process failed when trying to
1538    build pcretest and pcregrep. (For some reason it was using libtool to compile
1539    them, which is not right, as they aren't part of the library.)
1540    
1541    
1542    Version 3.8 18-Dec-01
1543    ---------------------
1544    
1545    1. The experimental UTF-8 code was completely screwed up. It was packing the
1546    bytes in the wrong order. How dumb can you get?
1547    
1548    
1549    Version 3.7 29-Oct-01
1550    ---------------------
1551    
1552    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
1553    This caused pcretest, when used on the test data, to segfault. Unfortunately,
1554    this didn't happen under Solaris 8, where I normally test things.
1555    
1556    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
1557    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
1558    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
1559    this fix an hour or so after the initial 3.7 release.)
1560    
1561    
1562    Version 3.6 23-Oct-01
1563    ---------------------
1564    
1565    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
1566    offsets passed as NULL with zero offset count.
1567    
1568    2. The config.guess and config.sub files had not been updated when I moved to
1569    the latest autoconf.
1570    
1571    
1572    Version 3.5 15-Aug-01
1573    ---------------------
1574    
1575    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
1576    had been forgotten.
1577    
1578    2. By using declared but undefined structures, we can avoid using "void"
1579    definitions in pcre.h while keeping the internal definitions of the structures
1580    private.
1581    
1582    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
1583    user point of view, this means that both static and shared libraries are built
1584    by default, but this can be individually controlled. More of the work of
1585    handling this static/shared cases is now inside libtool instead of PCRE's make
1586    file.
1587    
1588    4. The pcretest utility is now installed along with pcregrep because it is
1589    useful for users (to test regexs) and by doing this, it automatically gets
1590    relinked by libtool. The documentation has been turned into a man page, so
1591    there are now .1, .txt, and .html versions in /doc.
1592    
1593    5. Upgrades to pcregrep:
1594       (i)   Added long-form option names like gnu grep.
1595       (ii)  Added --help to list all options with an explanatory phrase.
1596       (iii) Added -r, --recursive to recurse into sub-directories.
1597       (iv)  Added -f, --file to read patterns from a file.
1598    
1599    6. pcre_exec() was referring to its "code" argument before testing that
1600    argument for NULL (and giving an error if it was NULL).
1601    
1602    7. Upgraded Makefile.in to allow for compiling in a different directory from
1603    the source directory.
1604    
1605    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
1606    options bits, the pointer it was passed was to an int instead of to an unsigned
1607    long int. This mattered only on 64-bit systems.
1608    
1609    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
1610    generated) instead of pcre.in, which it its source. Also made the same change
1611    in several of the .c files.
1612    
1613    10. A new release of gcc defines printf() as a macro, which broke pcretest
1614    because it had an ifdef in the middle of a string argument for printf(). Fixed
1615    by using separate calls to printf().
1616    
1617    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
1618    script, to force use of CR or LF instead of \n in the source. On non-Unix
1619    systems, the value can be set in config.h.
1620    
1621    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
1622    absolute limit. Changed the text of the error message to make this clear, and
1623    likewise updated the man page.
1624    
1625    13. The limit of 99 on the number of capturing subpatterns has been removed.
1626    The new limit is 65535, which I hope will not be a "real" limit.
1627    
1628    
1629    Version 3.4 22-Aug-00
1630    ---------------------
1631    
1632    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
1633    
1634    2. Diagnose condition (?(0) as an error instead of crashing on matching.
1635    
1636    
1637    Version 3.3 01-Aug-00
1638    ---------------------
1639    
1640    1. If an octal character was given, but the value was greater than \377, it
1641    was not getting masked to the least significant bits, as documented. This could
1642    lead to crashes in some systems.
1643    
1644    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
1645    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
1646    
1647    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
1648    These just pass their arguments on to (pcre_free)(), but they are provided
1649    because some uses of PCRE bind it to non-C systems that can call its functions,
1650    but cannot call free() or pcre_free() directly.
1651    
1652    4. Add "make test" as a synonym for "make check". Corrected some comments in
1653    the Makefile.
1654    
1655    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
1656    Makefile.
1657    
1658    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
1659    command called pgrep for grepping around the active processes.
1660    
1661    7. Added the beginnings of support for UTF-8 character strings.
1662    
1663    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
1664    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
1665    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
1666    out for the ar command.)
1667    
1668    
1669    Version 3.2 12-May-00
1670    ---------------------
1671    
1672    This is purely a bug fixing release.
1673    
1674    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
1675    of ZA. This was just one example of several cases that could provoke this bug,
1676    which was introduced by change 9 of version 2.00. The code for breaking
1677    infinite loops after an iteration that matches an empty string was't working
1678    correctly.
1679    
1680    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
1681    when matched against abbab (for example). After matching an empty string, it
1682    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
1683    caused it to match further down the string than it should.
1684    
1685    3. The code contained an inclusion of sys/types.h. It isn't clear why this
1686    was there because it doesn't seem to be needed, and it causes trouble on some
1687    systems, as it is not a Standard C header. It has been removed.
1688    
1689    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
1690    were reported on the Macintosh. The changes were from
1691    
1692      while ((c = *(++ptr)) != 0 && c != '\n');
1693    to
1694      while ((c = *(++ptr)) != 0 && c != '\n') ;
1695    
1696    Totally extraordinary, but if that's what it takes...
1697    
1698    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
1699    available. Added HAVE_BCOPY and an autoconf test for it; if neither
1700    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
1701    assumes the way PCRE uses memmove() (always moving upwards).
1702    
1703    6. PCRE is being used in one environment where strchr() is not available. There
1704    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
1705    faster code anyway.
1706    
1707    
1708    Version 3.1 09-Feb-00
1709    ---------------------
1710    
1711    The only change in this release is the fixing of some bugs in Makefile.in for
1712    the "install" target:
1713    
1714    (1) It was failing to install pcreposix.h.
1715    
1716    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
1717    
1718    
1719    Version 3.0 01-Feb-00
1720    ---------------------
1721    
1722    1. Add support for the /+ modifier to perltest (to output $` like it does in
1723    pcretest).
1724    
1725    2. Add support for the /g modifier to perltest.
1726    
1727    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
1728    matches null strings.
1729    
1730    4. Fix perltest so that it doesn't do unwanted things when fed an empty
1731    pattern. Perl treats empty patterns specially - it reuses the most recent
1732    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
1733    effect.
1734    
1735    5. The POSIX interface was broken in that it was just handing over the POSIX
1736    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
1737    required a bigger vector, with some working space on the end. This means that
1738    the POSIX wrapper now has to get and free some memory, and copy the results.
1739    
1740    6. Added some simple autoconf support, placing the test data and the
1741    documentation in separate directories, re-organizing some of the
1742    information files, and making it build pcre-config (a GNU standard). Also added
1743    libtool support for building PCRE as a shared library, which is now the
1744    default.
1745    
1746    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
1747    09 are not valid octal constants. Single digits will be used for minor values
1748    less than 10.
1749    
1750    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
1751    existing programs that set these in the POSIX interface can use PCRE without
1752    modification.
1753    
1754    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
1755    return all that pcre_info() returns, plus additional data. The pcre_info()
1756    function is retained for compatibility, but is considered to be obsolete.
1757    
1758    10. Added experimental recursion feature (?R) to handle one common case that
1759    Perl 5.6 will be able to do with (?p{...}).
1760    
1761    11. Added support for POSIX character classes like [:alpha:], which Perl is
1762    adopting.
1763    
1764    
1765    Version 2.08 31-Aug-99
1766    ----------------------
1767    
1768    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
1769    trying to match at the startoffset position, but instead was moving forward to
1770    the next newline as if a previous match had failed.
1771    
1772    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
1773    and could get into a loop if a null string was matched other than at the start
1774    of the subject.
1775    
1776    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
1777    be distinguished at compile time, and for completeness also added PCRE_DATE.
1778    
1779    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
1780    in GnuWin32 environments.
1781    
1782    
1783    Version 2.07 29-Jul-99
1784    ----------------------
1785    
1786    1. The documentation is now supplied in plain text form and HTML as well as in
1787    the form of man page sources.
1788    
1789    2. C++ compilers don't like assigning (void *) values to other pointer types.
1790    In particular this affects malloc(). Although there is no problem in Standard
1791    C, I've put in casts to keep C++ compilers happy.
1792    
1793    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
1794    should be (const char *).
1795    
1796    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
1797    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
1798    However, I haven't made this a standard facility. The documentation doesn't
1799    mention it, and the Makefile doesn't support it.
1800    
1801    5. The Makefile now contains an "install" target, with editable destinations at
1802    the top of the file. The pcretest program is not installed.
1803    
1804    6. pgrep -V now gives the PCRE version number and date.
1805    
1806    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1807    causing the entire string to be ignored, instead of just the last character.
1808    
1809    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1810    non-matching string, it can take a very, very long time, even for strings of
1811    quite modest length, because of the nested recursion. PCRE now does better in
1812    some of these cases. It does this by remembering the last required literal
1813    character in the pattern, and pre-searching the subject to ensure it is present
1814    before running the real match. In other words, it applies a heuristic to detect
1815    some types of certain failure quickly, and in the above example, if presented
1816    with a string that has no trailing " it gives "no match" very quickly.
1817    
1818    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1819    other alternatives are tried instead.
1820    
1821    
1822    Version 2.06 09-Jun-99
1823    ----------------------
1824    
1825    1. Change pcretest's output for amount of store used to show just the code
1826    space, because the remainder (the data block) varies in size between 32-bit and
1827    64-bit systems.
1828    
1829    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1830    start matching at. This allows lookbehinds to work when searching for multiple
1831    occurrences in a string.
1832    
1833    3. Added additional options to pcretest for testing multiple occurrences:
1834    
1835       /+   outputs the rest of the string that follows a match
1836       /g   loops for multiple occurrences, using the new startoffset argument
1837       /G   loops for multiple occurrences by passing an incremented pointer
1838    
1839    4. PCRE wasn't doing the "first character" optimization for patterns starting
1840    with \b or \B, though it was doing it for other lookbehind assertions. That is,
1841    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1842    the letter 'x'. On long subject strings, this gives a significant speed-up.
1843    
1844    
1845    Version 2.05 21-Apr-99
1846    ----------------------
1847    
1848    1. Changed the type of magic_number from int to long int so that it works
1849    properly on 16-bit systems.
1850    
1851    2. Fixed a bug which caused patterns starting with .* not to work correctly
1852    when the subject string contained newline characters. PCRE was assuming
1853    anchoring for such patterns in all cases, which is not correct because .* will
1854    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1855    DOTALL is set at top level; otherwise it knows that patterns starting with .*
1856    must be retried after every newline in the subject.
1857    
1858    
1859    Version 2.04 18-Feb-99
1860    ----------------------
1861    
1862    1. For parenthesized subpatterns with repeats whose minimum was zero, the
1863    computation of the store needed to hold the pattern was incorrect (too large).
1864    If such patterns were nested a few deep, this could multiply and become a real
1865    problem.
1866    
1867    2. Added /M option to pcretest to show the memory requirement of a specific
1868    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1869    
1870    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1871    compiled in such a way that the backtracking after subsequent failure was
1872    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1873    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1874    
1875    
1876    Version 2.03 02-Feb-99
1877    ----------------------
1878    
1879    1. Fixed typo and small mistake in man page.
1880    
1881    2. Added 4th condition (GPL supersedes if conflict) and created separate
1882    LICENCE file containing the conditions.
1883    
1884    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1885    Perl, that is the internal \ allows the delimiter to be included in the
1886    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1887    the final delimiter, add \ to the end of the pattern (to test the error).
1888    
1889    4. Added the convenience functions for extracting substrings after a successful
1890    match. Updated pcretest to make it able to test these functions.
1891    
1892    
1893    Version 2.02 14-Jan-99
1894    ----------------------
1895    
1896    1. Initialized the working variables associated with each extraction so that
1897    their saving and restoring doesn't refer to uninitialized store.
1898    
1899    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1900    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1901    fix the problem.
1902    
1903    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1904    calls, and wasn't printing the correct value for compiling calls. Increased the
1905    default value of LOOPREPEAT, and the number of significant figures in the
1906    times.
1907    
1908    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1909    
1910    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1911    a building problem on Windows NT with a FAT file system.
1912    
1913    
1914  Version 2.01 21-Oct-98  Version 2.01 21-Oct-98
1915  ----------------------  ----------------------

Legend:
Removed from v.25  
changed lines
  Added in v.91

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12