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

Diff of /code/trunk/ChangeLog

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

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8        was out-of-date, and there was no check on the pcre_compile() error code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

Legend:
Removed from v.29  
changed lines
  Added in v.423

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12