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

Diff of /code/trunk/ChangeLog

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

revision 13 by nigel, Sat Feb 24 21:38:21 2007 UTC revision 274 by ph10, Tue Nov 20 10:05:23 2007 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4    Version 7.5 12-Nov-07
5    ---------------------
6    
7    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
8        values in parens when parsing an RE using the C++ wrapper."
9    
10    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
11        Characters greater than 255 were excluded from the class instead of being
12        included.
13    
14    3.  The same bug as (2) above applied to negated POSIX classes such as
15        [:^space:].
16    
17    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
18        defined or documented. It seems to have been a typo for PCRE_STATIC, so
19        I have changed it.
20    
21    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
22        first named subpattern) and a construct such as (?&a) would reference the
23        first named subpattern whose name started with "a" (in other words, the
24        length check was missing). Both these problems are fixed. "Subpattern name
25        expected" is now given for (?&) (a zero-length name), and this patch also
26        makes it give the same error for \k'' (previously it complained that that
27        was a reference to a non-existent subpattern).
28    
29    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
30        this is right because (?- can be followed by option settings as well as by
31        digits. I have, however, made the messages clearer.
32    
33    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
34        than the number used in the conditional) now cause a compile-time error.
35        This is actually not compatible with Perl, which accepts such patterns, but
36        treats the conditional as always being FALSE (as PCRE used to), but it
37        seems to me that giving a diagnostic is better.
38    
39    8.  Change "alphameric" to the more common word "alphanumeric" in comments
40        and messages.
41    
42    
43    Version 7.4 21-Sep-07
44    ---------------------
45    
46    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
47        means that a class such as [\s] counted as "explicit reference to CR or
48        LF". That isn't really right - the whole point of the change was to try to
49        help when there was an actual mention of one of the two characters. So now
50        the change happens only if \r or \n (or a literal CR or LF) character is
51        encountered.
52    
53    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
54        of both had grown to the point where there were only 3 bits left.
55        Fortunately, there was spare space in the data structure, and so I have
56        moved the internal flags into a new 16-bit field to free up more option
57        bits.
58    
59    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
60        but did not set the internal JCHANGED flag - either of these is enough to
61        control the way the "get" function works - but the PCRE_INFO_JCHANGED
62        facility is supposed to tell if (?J) was ever used, so now (?J) at the
63        start sets both bits.
64    
65    4.  Added options (at build time, compile time, exec time) to change \R from
66        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
67    
68    5.  doc/pcresyntax.html was missing from the distribution.
69    
70    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
71        compatibility, even though it is no longer used.
72    
73    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
74        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
75        windows.h file is present (where different names are used). [This was
76        reversed later after testing - see 16 below.]
77    
78    8.  Changed all #include <config.h> to #include "config.h". There were also
79        some further <pcre.h> cases that I changed to "pcre.h".
80    
81    9.  When pcregrep was used with the --colour option, it missed the line ending
82        sequence off the lines that it output.
83    
84    10. It was pointed out to me that arrays of string pointers cause lots of
85        relocations when a shared library is dynamically loaded. A technique of
86        using a single long string with a table of offsets can drastically reduce
87        these. I have refactored PCRE in four places to do this. The result is
88        dramatic:
89    
90          Originally:                          290
91          After changing UCP table:            187
92          After changing error message table:   43
93          After changing table of "verbs"       36
94          After changing table of Posix names   22
95    
96        Thanks to the folks working on Gregex for glib for this insight.
97    
98    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
99        unicode-properties was also set.
100    
101    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
102    
103    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
104        checked only for CRLF.
105    
106    14. Added casts to pcretest.c to avoid compiler warnings.
107    
108    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
109    
110    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
111        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
112        entirely. This removes changes made in 7 above.
113    
114    17. The CMake files have been updated, and there is now more information about
115        building with CMake in the NON-UNIX-USE document.
116    
117    
118    Version 7.3 28-Aug-07
119    ---------------------
120    
121     1. In the rejigging of the build system that eventually resulted in 7.1, the
122        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
123        brackets there is not right, since it causes compilers to look for an
124        installed pcre.h, not the version that is in the source that is being
125        compiled (which of course may be different). I have changed it back to:
126    
127          #include "pcre.h"
128    
129        I have a vague recollection that the change was concerned with compiling in
130        different directories, but in the new build system, that is taken care of
131        by the VPATH setting the Makefile.
132    
133     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
134        when the subject happened to end in the byte 0x85 (e.g. if the last
135        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
136        characters but of course it shouldn't be taken as a newline when it is part
137        of another character. The bug was that, for an unlimited repeat of . in
138        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
139        characters when looking for a newline.
140    
141     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
142    
143     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
144        in debug output.
145    
146     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
147        long printing in the pcrecpp unittest when running under MinGW.
148    
149     6. ESC_K was left out of the EBCDIC table.
150    
151     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
152        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
153        limit also applies to "virtual nesting" when a pattern is recursive, and in
154        this case 1000 isn't so big. I have been able to remove this limit at the
155        expense of backing off one optimization in certain circumstances. Normally,
156        when pcre_exec() would call its internal match() function recursively and
157        immediately return the result unconditionally, it uses a "tail recursion"
158        feature to save stack. However, when a subpattern that can match an empty
159        string has an unlimited repetition quantifier, it no longer makes this
160        optimization. That gives it a stack frame in which to save the data for
161        checking that an empty string has been matched. Previously this was taken
162        from the 1000-entry workspace that had been reserved. So now there is no
163        explicit limit, but more stack is used.
164    
165     8. Applied Daniel's patches to solve problems with the import/export magic
166        syntax that is required for Windows, and which was going wrong for the
167        pcreposix and pcrecpp parts of the library. These were overlooked when this
168        problem was solved for the main library.
169    
170     9. There were some crude static tests to avoid integer overflow when computing
171        the size of patterns that contain repeated groups with explicit upper
172        limits. As the maximum quantifier is 65535, the maximum group length was
173        set at 30,000 so that the product of these two numbers did not overflow a
174        32-bit integer. However, it turns out that people want to use groups that
175        are longer than 30,000 bytes (though not repeat them that many times).
176        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
177        made it possible to implement the integer overflow checks in a much more
178        dynamic way, which I have now done. The artificial limitation on group
179        length has been removed - we now have only the limit on the total length of
180        the compiled pattern, which depends on the LINK_SIZE setting.
181    
182    10. Fixed a bug in the documentation for get/copy named substring when
183        duplicate names are permitted. If none of the named substrings are set, the
184        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
185        empty string.
186    
187    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
188        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
189        because the ] is interpreted as the first data character and the
190        terminating ] is not found. PCRE has been made compatible with Perl in this
191        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
192        cause memory overwriting.
193    
194    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
195        string has been matched (to stop an infinite loop). It was not recognizing
196        a conditional subpattern that could match an empty string if that
197        subpattern was within another subpattern. For example, it looped when
198        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
199        condition was not nested. This bug has been fixed.
200    
201    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
202        past the start of the subject in the presence of bytes with the top bit
203        set, for example "\x8aBCD".
204    
205    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
206        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
207    
208    14. Optimized (?!) to (*FAIL).
209    
210    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
211        This restricts code points to be within the range 0 to 0x10FFFF, excluding
212        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
213        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
214        does: it's just the validity check that is more restrictive.
215    
216    16. Inserted checks for integer overflows during escape sequence (backslash)
217        processing, and also fixed erroneous offset values for syntax errors during
218        backslash processing.
219    
220    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
221        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
222    
223    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
224        caused an overrun.
225    
226    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
227        something other than just ASCII characters) inside a group that had an
228        unlimited repeat caused a loop at compile time (while checking to see
229        whether the group could match an empty string).
230    
231    20. Debugging a pattern containing \p or \P could cause a crash. For example,
232        [\P{Any}] did so. (Error in the code for printing property names.)
233    
234    21. An orphan \E inside a character class could cause a crash.
235    
236    22. A repeated capturing bracket such as (A)? could cause a wild memory
237        reference during compilation.
238    
239    23. There are several functions in pcre_compile() that scan along a compiled
240        expression for various reasons (e.g. to see if it's fixed length for look
241        behind). There were bugs in these functions when a repeated \p or \P was
242        present in the pattern. These operators have additional parameters compared
243        with \d, etc, and these were not being taken into account when moving along
244        the compiled data. Specifically:
245    
246        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
247            length.
248    
249        (b) An item such as \pL+ within a repeated group could cause crashes or
250            loops.
251    
252        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
253            "reference to non-existent subpattern" error.
254    
255        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
256    
257    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
258        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
259    
260    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
261    
262    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
263        character were causing crashes (broken optimization).
264    
265    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
266        \p or \P) caused a compile-time loop.
267    
268    28. More problems have arisen in unanchored patterns when CRLF is a valid line
269        break. For example, the unstudied pattern [\r\n]A does not match the string
270        "\r\nA" because change 7.0/46 below moves the current point on by two
271        characters after failing to match at the start. However, the pattern \nA
272        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
273        the same is true. There doesn't seem any very clean way out of this, but
274        what I have chosen to do makes the common cases work: PCRE now takes note
275        of whether there can be an explicit match for \r or \n anywhere in the
276        pattern, and if so, 7.0/46 no longer applies. As part of this change,
277        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
278        pattern has explicit CR or LF references.
279    
280    29. Added (*CR) etc for changing newline setting at start of pattern.
281    
282    
283    Version 7.2 19-Jun-07
284    ---------------------
285    
286     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
287        which is apparently normally available under Windows.
288    
289     2. Re-jig the pcregrep tests with different newline settings in an attempt
290        to make them independent of the local environment's newline setting.
291    
292     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
293    
294     4. Some of the "internals" tests were previously cut out when the link size
295        was not 2, because the output contained actual offsets. The recent new
296        "Z" feature of pcretest means that these can be cut out, making the tests
297        usable with all link sizes.
298    
299     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
300        stack recursion. This gives a massive performance boost under BSD, but just
301        a small improvement under Linux. However, it saves one field in the frame
302        in all cases.
303    
304     6. Added more features from the forthcoming Perl 5.10:
305    
306        (a) (?-n) (where n is a string of digits) is a relative subroutine or
307            recursion call. It refers to the nth most recently opened parentheses.
308    
309        (b) (?+n) is also a relative subroutine call; it refers to the nth next
310            to be opened parentheses.
311    
312        (c) Conditions that refer to capturing parentheses can be specified
313            relatively, for example, (?(-2)... or (?(+3)...
314    
315        (d) \K resets the start of the current match so that everything before
316            is not part of it.
317    
318        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
319    
320        (f) \g{name} is another synonym - part of Perl 5.10's unification of
321            reference syntax.
322    
323        (g) (?| introduces a group in which the numbering of parentheses in each
324            alternative starts with the same number.
325    
326        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
327    
328     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
329        PCRE_INFO_JCHANGED.
330    
331     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
332        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
333        for detecting groups that can match an empty string.
334    
335     9. A pattern with a very large number of alternatives (more than several
336        hundred) was running out of internal workspace during the pre-compile
337        phase, where pcre_compile() figures out how much memory will be needed. A
338        bit of new cunning has reduced the workspace needed for groups with
339        alternatives. The 1000-alternative test pattern now uses 12 bytes of
340        workspace instead of running out of the 4096 that are available.
341    
342    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
343    
344    11. Applied patch from Google to remove an optimization that didn't quite work.
345        The report of the bug said:
346    
347          pcrecpp::RE("a*").FullMatch("aaa") matches, while
348          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
349          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
350    
351    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
352        it matched the wrong number of bytes.
353    
354    
355    Version 7.1 24-Apr-07
356    ---------------------
357    
358     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
359        that is more "standard", making use of automake and other Autotools. There
360        is some re-arrangement of the files and adjustment of comments consequent
361        on this.
362    
363     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
364        for recursive directory scanning broke on some systems because the files
365        are not scanned in any specific order and on different systems the order
366        was different. A call to "sort" has been inserted into RunGrepTest for the
367        approprate test as a short-term fix. In the longer term there may be an
368        alternative.
369    
370     3. I had an email from Eric Raymond about problems translating some of PCRE's
371        man pages to HTML (despite the fact that I distribute HTML pages, some
372        people do their own conversions for various reasons). The problems
373        concerned the use of low-level troff macros .br and .in. I have therefore
374        removed all such uses from the man pages (some were redundant, some could
375        be replaced by .nf/.fi pairs). The 132html script that I use to generate
376        HTML has been updated to handle .nf/.fi and to complain if it encounters
377        .br or .in.
378    
379     4. Updated comments in configure.ac that get placed in config.h.in and also
380        arranged for config.h to be included in the distribution, with the name
381        config.h.generic, for the benefit of those who have to compile without
382        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
383    
384     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
385        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
386        makevp.bat for the current PCRE, using the additional files
387        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
388    
389     6. A Windows user reported a minor discrepancy with test 2, which turned out
390        to be caused by a trailing space on an input line that had got lost in his
391        copy. The trailing space was an accident, so I've just removed it.
392    
393     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
394        that is needed.
395    
396     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
397        as "const" (a) because they are and (b) because it helps the PHP
398        maintainers who have recently made a script to detect big data structures
399        in the php code that should be moved to the .rodata section. I remembered
400        to update Builducptable as well, so it won't revert if ucptable.h is ever
401        re-created.
402    
403     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
404        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
405        order to be able to cut out the UTF-8 tables in the latter when UTF-8
406        support is not required. This saves 1.5-2K of code, which is important in
407        some applications.
408    
409        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
410        so as not to refer to the tables, even though these functions will never be
411        called when UTF-8 support is disabled. Otherwise there are problems with a
412        shared library.
413    
414    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
415    
416        (a) It was defining its arguments as char * instead of void *.
417    
418        (b) It was assuming that all moves were upwards in memory; this was true
419            a long time ago when I wrote it, but is no longer the case.
420    
421        The emulated memove() is provided for those environments that have neither
422        memmove() nor bcopy(). I didn't think anyone used it these days, but that
423        is clearly not the case, as these two bugs were recently reported.
424    
425    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
426        and Detrail to create the HTML documentation, the .txt form of the man
427        pages, and it removes trailing spaces from listed files. It also creates
428        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
429        case, it wraps all the #defines with #ifndefs. This script should be run
430        before "make dist".
431    
432    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
433        with Unicode property support.
434    
435        (a) For a maximizing quantifier, if the two different cases of the
436            character were of different lengths in their UTF-8 codings (there are
437            some cases like this - I found 11), and the matching function had to
438            back up over a mixture of the two cases, it incorrectly assumed they
439            were both the same length.
440    
441        (b) When PCRE was configured to use the heap rather than the stack for
442            recursion during matching, it was not correctly preserving the data for
443            the other case of a UTF-8 character when checking ahead for a match
444            while processing a minimizing repeat. If the check also involved
445            matching a wide character, but failed, corruption could cause an
446            erroneous result when trying to check for a repeat of the original
447            character.
448    
449    13. Some tidying changes to the testing mechanism:
450    
451        (a) The RunTest script now detects the internal link size and whether there
452            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
453            values substituted by "configure". (The RunGrepTest script already did
454            this for UTF-8.) The configure.ac script no longer substitutes the
455            relevant variables.
456    
457        (b) The debugging options /B and /D in pcretest show the compiled bytecode
458            with length and offset values. This means that the output is different
459            for different internal link sizes. Test 2 is skipped for link sizes
460            other than 2 because of this, bypassing the problem. Unfortunately,
461            there was also a test in test 3 (the locale tests) that used /B and
462            failed for link sizes other than 2. Rather than cut the whole test out,
463            I have added a new /Z option to pcretest that replaces the length and
464            offset values with spaces. This is now used to make test 3 independent
465            of link size. (Test 2 will be tidied up later.)
466    
467    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
468        segmentation fault instead of returning the appropriate error message.
469    
470    15. In multiline mode when the newline sequence was set to "any", the pattern
471        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
472        This doesn't seem right; it now treats the CRLF combination as the line
473        ending, and so does not match in that case. It's only a pattern such as ^$
474        that would hit this one: something like ^ABC$ would have failed after \r
475        and then tried again after \r\n.
476    
477    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
478        in an attempt to make files that differ only in their line terminators
479        compare equal. This works on Linux.
480    
481    17. Under certain error circumstances pcregrep might try to free random memory
482        as it exited. This is now fixed, thanks to valgrind.
483    
484    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
485        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
486        was because its rules for how to advance for /g after matching an empty
487        string at the end of a line did not allow for this case. They now check for
488        it specially.
489    
490    20. pcretest is supposed to handle patterns and data of any length, by
491        extending its buffers when necessary. It was getting this wrong when the
492        buffer for a data line had to be extended.
493    
494    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
495        CRLF as a newline sequence.
496    
497    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
498        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
499        I have nevertheless tidied it up.
500    
501    23. Added some casts to kill warnings from HP-UX ia64 compiler.
502    
503    24. Added a man page for pcre-config.
504    
505    
506    Version 7.0 19-Dec-06
507    ---------------------
508    
509     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
510        moving to gcc 4.1.1.
511    
512     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
513        sys/time.h, which is documented as needed for this function. It doesn't
514        seem to matter on Linux, but it showed up on some releases of OS X.
515    
516     3. It seems that there are systems where bytes whose values are greater than
517        127 match isprint() in the "C" locale. The "C" locale should be the
518        default when a C program starts up. In most systems, only ASCII printing
519        characters match isprint(). This difference caused the output from pcretest
520        to vary, making some of the tests fail. I have changed pcretest so that:
521    
522        (a) When it is outputting text in the compiled version of a pattern, bytes
523            other than 32-126 are always shown as hex escapes.
524    
525        (b) When it is outputting text that is a matched part of a subject string,
526            it does the same, unless a different locale has been set for the match
527            (using the /L modifier). In this case, it uses isprint() to decide.
528    
529     4. Fixed a major bug that caused incorrect computation of the amount of memory
530        required for a compiled pattern when options that changed within the
531        pattern affected the logic of the preliminary scan that determines the
532        length. The relevant options are -x, and -i in UTF-8 mode. The result was
533        that the computed length was too small. The symptoms of this bug were
534        either the PCRE error "internal error: code overflow" from pcre_compile(),
535        or a glibc crash with a message such as "pcretest: free(): invalid next
536        size (fast)". Examples of patterns that provoked this bug (shown in
537        pcretest format) are:
538    
539          /(?-x: )/x
540          /(?x)(?-x: \s*#\s*)/
541          /((?i)[\x{c0}])/8
542          /(?i:[\x{c0}])/8
543    
544        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
545        is now done differently.
546    
547     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
548        wrapper classes; (b) implement a new function in the C++ scanner that is
549        more efficient than the old way of doing things because it avoids levels of
550        recursion in the regex matching; (c) add a paragraph to the documentation
551        for the FullMatch() function.
552    
553     6. The escape sequence \n was being treated as whatever was defined as
554        "newline". Not only was this contrary to the documentation, which states
555        that \n is character 10 (hex 0A), but it also went horribly wrong when
556        "newline" was defined as CRLF. This has been fixed.
557    
558     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
559        was being set to -1 for the "end of line" case (supposedly a value that no
560        character can have). Though this value is never used (the check for end of
561        line is "zero bytes in current character"), it caused compiler complaints.
562        I've changed it to 0xffffffff.
563    
564     8. In pcre_version.c, the version string was being built by a sequence of
565        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
566        string (as it is for production releases) called a macro with an empty
567        argument. The C standard says the result of this is undefined. The gcc
568        compiler treats it as an empty string (which was what was wanted) but it is
569        reported that Visual C gives an error. The source has been hacked around to
570        avoid this problem.
571    
572     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
573        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
574        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
575        of them did).
576    
577    10. Originally, pcretest opened its input and output without "b"; then I was
578        told that "b" was needed in some environments, so it was added for release
579        5.0 to both the input and output. (It makes no difference on Unix-like
580        systems.) Later I was told that it is wrong for the input on Windows. I've
581        now abstracted the modes into two macros, to make it easier to fiddle with
582        them, and removed "b" from the input mode under Windows.
583    
584    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
585    
586    12. Added -help and --help to pcretest as an official way of being reminded
587        of the options.
588    
589    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
590        and pcrecpp.cc because they annoy compilers at high warning levels.
591    
592    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
593    
594    15. Fixed an occurrence of == in configure.ac that should have been = (shell
595        scripts are not C programs :-) and which was not noticed because it works
596        on Linux.
597    
598    16. pcretest is supposed to handle any length of pattern and data line (as one
599        line or as a continued sequence of lines) by extending its input buffer if
600        necessary. This feature was broken for very long pattern lines, leading to
601        a string of junk being passed to pcre_compile() if the pattern was longer
602        than about 50K.
603    
604    17. I have done a major re-factoring of the way pcre_compile() computes the
605        amount of memory needed for a compiled pattern. Previously, there was code
606        that made a preliminary scan of the pattern in order to do this. That was
607        OK when PCRE was new, but as the facilities have expanded, it has become
608        harder and harder to keep it in step with the real compile phase, and there
609        have been a number of bugs (see for example, 4 above). I have now found a
610        cunning way of running the real compile function in a "fake" mode that
611        enables it to compute how much memory it would need, while actually only
612        ever using a few hundred bytes of working memory and without too many
613        tests of the mode. This should make future maintenance and development
614        easier. A side effect of this work is that the limit of 200 on the nesting
615        depth of parentheses has been removed (though this was never a serious
616        limitation, I suspect). However, there is a downside: pcre_compile() now
617        runs more slowly than before (30% or more, depending on the pattern). I
618        hope this isn't a big issue. There is no effect on runtime performance.
619    
620    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
621        newline (only possible for the last line of a file) and it was a
622        pattern that set a locale (followed by /Lsomething), pcretest crashed.
623    
624    19. Added additional timing features to pcretest. (1) The -tm option now times
625        matching only, not compiling. (2) Both -t and -tm can be followed, as a
626        separate command line item, by a number that specifies the number of
627        repeats to use when timing. The default is 50000; this gives better
628        precision, but takes uncomfortably long for very large patterns.
629    
630    20. Extended pcre_study() to be more clever in cases where a branch of a
631        subpattern has no definite first character. For example, (a*|b*)[cd] would
632        previously give no result from pcre_study(). Now it recognizes that the
633        first character must be a, b, c, or d.
634    
635    21. There was an incorrect error "recursive call could loop indefinitely" if
636        a subpattern (or the entire pattern) that was being tested for matching an
637        empty string contained only one non-empty item after a nested subpattern.
638        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
639        incorrectly, because the \d was being skipped in the check.
640    
641    22. The pcretest program now has a new pattern option /B and a command line
642        option -b, which is equivalent to adding /B to every pattern. This causes
643        it to show the compiled bytecode, without the additional information that
644        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
645        is the same as /B/I).
646    
647    23. A new optimization is now able automatically to treat some sequences such
648        as a*b as a*+b. More specifically, if something simple (such as a character
649        or a simple class like \d) has an unlimited quantifier, and is followed by
650        something that cannot possibly match the quantified thing, the quantifier
651        is automatically "possessified".
652    
653    24. A recursive reference to a subpattern whose number was greater than 39
654        went wrong under certain circumstances in UTF-8 mode. This bug could also
655        have affected the operation of pcre_study().
656    
657    25. Realized that a little bit of performance could be had by replacing
658        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
659    
660    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
661    
662    27. Possessive quantifiers such as a++ were previously implemented by turning
663        them into atomic groups such as ($>a+). Now they have their own opcodes,
664        which improves performance. This includes the automatically created ones
665        from 23 above.
666    
667    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
668        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
669        the first matched character to be a colon. This applied both to named and
670        numbered groups.
671    
672    29. The ucpinternal.h header file was missing its idempotency #ifdef.
673    
674    30. I was sent a "project" file called libpcre.a.dev which I understand makes
675        building PCRE on Windows easier, so I have included it in the distribution.
676    
677    31. There is now a check in pcretest against a ridiculously large number being
678        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
679        loop, the loop is abandoned.
680    
681    32. Forward references to subpatterns in conditions such as (?(2)...) where
682        subpattern 2 is defined later cause pcre_compile() to search forwards in
683        the pattern for the relevant set of parentheses. This search went wrong
684        when there were unescaped parentheses in a character class, parentheses
685        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
686    
687    33. "Subroutine" calls and backreferences were previously restricted to
688        referencing subpatterns earlier in the regex. This restriction has now
689        been removed.
690    
691    34. Added a number of extra features that are going to be in Perl 5.10. On the
692        whole, these are just syntactic alternatives for features that PCRE had
693        previously implemented using the Python syntax or my own invention. The
694        other formats are all retained for compatibility.
695    
696        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
697            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
698            also .NET compatible.
699    
700        (b) A recursion or subroutine call to a named group can now be defined as
701            (?&name) as well as (?P>name).
702    
703        (c) A backreference to a named group can now be defined as \k<name> or
704            \k'name' as well as (?P=name). The new forms, as well as being in Perl
705            5.10, are also .NET compatible.
706    
707        (d) A conditional reference to a named group can now use the syntax
708            (?(<name>) or (?('name') as well as (?(name).
709    
710        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
711            groups (named and numbered) that are never evaluated inline, but can be
712            called as "subroutines" from elsewhere. In effect, the DEFINE condition
713            is always false. There may be only one alternative in such a group.
714    
715        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
716            as the simple (?(R). The condition is true only if the most recent
717            recursion is that of the given number or name. It does not search out
718            through the entire recursion stack.
719    
720        (g) The escape \gN or \g{N} has been added, where N is a positive or
721            negative number, specifying an absolute or relative reference.
722    
723    35. Tidied to get rid of some further signed/unsigned compiler warnings and
724        some "unreachable code" warnings.
725    
726    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
727        things, this adds five new scripts.
728    
729    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
730        There were also incompatibilities regarding the handling of \Q..\E inside
731        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
732        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
733    
734    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
735        matches an empty string, and forcibly breaks the loop. There were bugs in
736        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
737        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
738        separate and independent bugs (that affected different cases) have been
739        fixed.
740    
741    39. Refactored the code to abolish the use of different opcodes for small
742        capturing bracket numbers. This is a tidy that I avoided doing when I
743        removed the limit on the number of capturing brackets for 3.5 back in 2001.
744        The new approach is not only tidier, it makes it possible to reduce the
745        memory needed to fix the previous bug (38).
746    
747    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
748        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
749        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
750        mode.
751    
752    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
753        report.
754    
755    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
756        copy construction and assignment in the C++ wrapper.
757    
758    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
759        couple of bugs that could have given wrong results in the "--newline=crlf"
760        case.
761    
762    44. Added a number of casts and did some reorganization of signed/unsigned int
763        variables following suggestions from Dair Grant. Also renamed the variable
764        "this" as "item" because it is a C++ keyword.
765    
766    45. Arranged for dftables to add
767    
768          #include "pcre_internal.h"
769    
770        to pcre_chartables.c because without it, gcc 4.x may remove the array
771        definition from the final binary if PCRE is built into a static library and
772        dead code stripping is activated.
773    
774    46. For an unanchored pattern, if a match attempt fails at the start of a
775        newline sequence, and the newline setting is CRLF or ANY, and the next two
776        characters are CRLF, advance by two characters instead of one.
777    
778    
779    Version 6.7 04-Jul-06
780    ---------------------
781    
782     1. In order to handle tests when input lines are enormously long, pcretest has
783        been re-factored so that it automatically extends its buffers when
784        necessary. The code is crude, but this _is_ just a test program. The
785        default size has been increased from 32K to 50K.
786    
787     2. The code in pcre_study() was using the value of the re argument before
788        testing it for NULL. (Of course, in any sensible call of the function, it
789        won't be NULL.)
790    
791     3. The memmove() emulation function in pcre_internal.h, which is used on
792        systems that lack both memmove() and bcopy() - that is, hardly ever -
793        was missing a "static" storage class specifier.
794    
795     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
796        containing an extended class (one that cannot be represented by a bitmap
797        because it contains high-valued characters or Unicode property items, e.g.
798        [\pZ]). Almost always one would set UTF-8 mode when processing such a
799        pattern, but PCRE should not loop if you do not (it no longer does).
800        [Detail: two cases were found: (a) a repeated subpattern containing an
801        extended class; (b) a recursive reference to a subpattern that followed a
802        previous extended class. It wasn't skipping over the extended class
803        correctly when UTF-8 mode was not set.]
804    
805     5. A negated single-character class was not being recognized as fixed-length
806        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
807        compile error "lookbehind assertion is not fixed length".
808    
809     6. The RunPerlTest auxiliary script was showing an unexpected difference
810        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
811        write a Perl script that can interpret lines of an input file either as
812        byte characters or as UTF-8, which is what "perltest" was being required to
813        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
814        can't do is switch easily at run time between having the "use utf8;" pragma
815        or not. In the end, I fudged it by using the RunPerlTest script to insert
816        "use utf8;" explicitly for the UTF-8 tests.
817    
818     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
819        the end of the subject string, contrary to the documentation and to what
820        Perl does. This was true of both matching functions. Now it matches only at
821        the start of the subject and immediately after *internal* newlines.
822    
823     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
824        a pointer to an int instead of a pointer to an unsigned long int. This
825        caused problems on 64-bit systems.
826    
827     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
828        instance of the 'standard' template library not being so standard".
829    
830    10. There was no check on the number of named subpatterns nor the maximum
831        length of a subpattern name. The product of these values is used to compute
832        the size of the memory block for a compiled pattern. By supplying a very
833        long subpattern name and a large number of named subpatterns, the size
834        computation could be caused to overflow. This is now prevented by limiting
835        the length of names to 32 characters, and the number of named subpatterns
836        to 10,000.
837    
838    11. Subpatterns that are repeated with specific counts have to be replicated in
839        the compiled pattern. The size of memory for this was computed from the
840        length of the subpattern and the repeat count. The latter is limited to
841        65535, but there was no limit on the former, meaning that integer overflow
842        could in principle occur. The compiled length of a repeated subpattern is
843        now limited to 30,000 bytes in order to prevent this.
844    
845    12. Added the optional facility to have named substrings with the same name.
846    
847    13. Added the ability to use a named substring as a condition, using the
848        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
849        are numbers (not recommended). Forward references are permitted.
850    
851    14. Added forward references in named backreferences (if you see what I mean).
852    
853    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
854        pattern could run off the end of the subject. For example, the pattern
855        "(?s)(.{1,5})"8 did this with the subject "ab".
856    
857    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
858        PCRE_CASELESS was set when matching characters that were quantified with ?
859        or *.
860    
861    17. A character class other than a single negated character that had a minimum
862        but no maximum quantifier - for example [ab]{6,} - was not handled
863        correctly by pce_dfa_exec(). It would match only one character.
864    
865    18. A valid (though odd) pattern that looked like a POSIX character
866        class but used an invalid character after [ (for example [[,abc,]]) caused
867        pcre_compile() to give the error "Failed: internal error: code overflow" or
868        in some cases to crash with a glibc free() error. This could even happen if
869        the pattern terminated after [[ but there just happened to be a sequence of
870        letters, a binary zero, and a closing ] in the memory that followed.
871    
872    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
873        over the years. Originally (before any Unicode support), just the bottom 8
874        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
875        output from "man perlunicode" includes this:
876    
877          The regular expression compiler produces polymorphic opcodes.  That
878          is, the pattern adapts to the data and automatically switches to
879          the Unicode character scheme when presented with Unicode data--or
880          instead uses a traditional byte scheme when presented with byte
881          data.
882    
883        Sadly, a wide octal escape does not cause a switch, and in a string with
884        no other multibyte characters, these octal escapes are treated as before.
885        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
886        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
887        Unicode string.
888    
889        I have not perpetrated such confusion in PCRE. Up till now, it took just
890        the bottom 8 bits, as in old Perl. I have now made octal escapes with
891        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
892        translate to the appropriate multibyte character.
893    
894    29. Applied some refactoring to reduce the number of warnings from Microsoft
895        and Borland compilers. This has included removing the fudge introduced
896        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
897        a warning about an unused variable.
898    
899    21. PCRE has not included VT (character 0x0b) in the set of whitespace
900        characters since release 4.0, because Perl (from release 5.004) does not.
901        [Or at least, is documented not to: some releases seem to be in conflict
902        with the documentation.] However, when a pattern was studied with
903        pcre_study() and all its branches started with \s, PCRE still included VT
904        as a possible starting character. Of course, this did no harm; it just
905        caused an unnecessary match attempt.
906    
907    22. Removed a now-redundant internal flag bit that recorded the fact that case
908        dependency changed within the pattern. This was once needed for "required
909        byte" processing, but is no longer used. This recovers a now-scarce options
910        bit. Also moved the least significant internal flag bit to the most-
911        significant bit of the word, which was not previously used (hangover from
912        the days when it was an int rather than a uint) to free up another bit for
913        the future.
914    
915    23. Added support for CRLF line endings as well as CR and LF. As well as the
916        default being selectable at build time, it can now be changed at runtime
917        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
918        specify that it is scanning data with non-default line endings.
919    
920    24. Changed the definition of CXXLINK to make it agree with the definition of
921        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
922    
923    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
924        recursions. This makes a big different to stack usage for some patterns.
925    
926    26. If a subpattern containing a named recursion or subroutine reference such
927        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
928        the space required for the compiled pattern went wrong and gave too small a
929        value. Depending on the environment, this could lead to "Failed: internal
930        error: code overflow at offset 49" or "glibc detected double free or
931        corruption" errors.
932    
933    27. Applied patches from Google (a) to support the new newline modes and (b) to
934        advance over multibyte UTF-8 characters in GlobalReplace.
935    
936    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
937        difference for some implementation of PCRE in some Windows version.
938    
939    29. Added some extra testing facilities to pcretest:
940    
941        \q<number>   in a data line sets the "match limit" value
942        \Q<number>   in a data line sets the "match recursion limt" value
943        -S <number>  sets the stack size, where <number> is in megabytes
944    
945        The -S option isn't available for Windows.
946    
947    
948    Version 6.6 06-Feb-06
949    ---------------------
950    
951     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
952        in pcreposix.h. I have copied the definition from pcre.h.
953    
954     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
955        because pcre.h is no longer a built file.
956    
957     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
958        not normally included in the compiled code.
959    
960    
961    Version 6.5 01-Feb-06
962    ---------------------
963    
964     1. When using the partial match feature with pcre_dfa_exec(), it was not
965        anchoring the second and subsequent partial matches at the new starting
966        point. This could lead to incorrect results. For example, with the pattern
967        /1234/, partially matching against "123" and then "a4" gave a match.
968    
969     2. Changes to pcregrep:
970    
971        (a) All non-match returns from pcre_exec() were being treated as failures
972            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
973            error message is output. Some extra information is given for the
974            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
975            probably the only errors that are likely to be caused by users (by
976            specifying a regex that has nested indefinite repeats, for instance).
977            If there are more than 20 of these errors, pcregrep is abandoned.
978    
979        (b) A binary zero was treated as data while matching, but terminated the
980            output line if it was written out. This has been fixed: binary zeroes
981            are now no different to any other data bytes.
982    
983        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
984            used to set a locale for matching. The --locale=xxxx long option has
985            been added (no short equivalent) to specify a locale explicitly on the
986            pcregrep command, overriding the environment variables.
987    
988        (d) When -B was used with -n, some line numbers in the output were one less
989            than they should have been.
990    
991        (e) Added the -o (--only-matching) option.
992    
993        (f) If -A or -C was used with -c (count only), some lines of context were
994            accidentally printed for the final match.
995    
996        (g) Added the -H (--with-filename) option.
997    
998        (h) The combination of options -rh failed to suppress file names for files
999            that were found from directory arguments.
1000    
1001        (i) Added the -D (--devices) and -d (--directories) options.
1002    
1003        (j) Added the -F (--fixed-strings) option.
1004    
1005        (k) Allow "-" to be used as a file name for -f as well as for a data file.
1006    
1007        (l) Added the --colo(u)r option.
1008    
1009        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
1010            is not present by default.
1011    
1012     3. A nasty bug was discovered in the handling of recursive patterns, that is,
1013        items such as (?R) or (?1), when the recursion could match a number of
1014        alternatives. If it matched one of the alternatives, but subsequently,
1015        outside the recursion, there was a failure, the code tried to back up into
1016        the recursion. However, because of the way PCRE is implemented, this is not
1017        possible, and the result was an incorrect result from the match.
1018    
1019        In order to prevent this happening, the specification of recursion has
1020        been changed so that all such subpatterns are automatically treated as
1021        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
1022    
1023     4. I had overlooked the fact that, in some locales, there are characters for
1024        which isalpha() is true but neither isupper() nor islower() are true. In
1025        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
1026        and ordfeminine) are like this. This affected the treatment of \w and \W
1027        when they appeared in character classes, but not when they appeared outside
1028        a character class. The bit map for "word" characters is now created
1029        separately from the results of isalnum() instead of just taking it from the
1030        upper, lower, and digit maps. (Plus the underscore character, of course.)
1031    
1032     5. The above bug also affected the handling of POSIX character classes such as
1033        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
1034        permanent tables. Instead, the bit maps for such a class were previously
1035        created as the appropriate unions of the upper, lower, and digit bitmaps.
1036        Now they are created by subtraction from the [[:word:]] class, which has
1037        its own bitmap.
1038    
1039     6. The [[:blank:]] character class matches horizontal, but not vertical space.
1040        It is created by subtracting the vertical space characters (\x09, \x0a,
1041        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
1042        subtraction was done in the overall bitmap for a character class, meaning
1043        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
1044        be recognized. This bug has been fixed.
1045    
1046     7. Patches from the folks at Google:
1047    
1048          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1049          real life, but is still worth protecting against".
1050    
1051          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1052          regular expressions".
1053    
1054          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1055          have it.
1056    
1057          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1058          "configure" and the latter not, in order to fix a problem somebody had
1059          with compiling the Arg class on HP-UX.
1060    
1061          (e) Improve the error-handling of the C++ wrapper a little bit.
1062    
1063          (f) New tests for checking recursion limiting.
1064    
1065     8. The pcre_memmove() function, which is used only if the environment does not
1066        have a standard memmove() function (and is therefore rarely compiled),
1067        contained two bugs: (a) use of int instead of size_t, and (b) it was not
1068        returning a result (though PCRE never actually uses the result).
1069    
1070     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1071        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1072        returned instead of calling malloc() with an overflowing number that would
1073        most likely cause subsequent chaos.
1074    
1075    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1076    
1077    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1078        with this option is matched, the nmatch and pmatch options of regexec() are
1079        ignored.
1080    
1081    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1082        provided in case anyone wants to the the POSIX interface with UTF-8
1083        strings.
1084    
1085    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1086        C++ linking (needed for some HP-UX environments).
1087    
1088    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1089        (unused parameter) and in the pcre_printint() function (omitted "default"
1090        switch label when the default is to do nothing).
1091    
1092    15. Added some code to make it possible, when PCRE is compiled as a C++
1093        library, to replace subject pointers for pcre_exec() with a smart pointer
1094        class, thus making it possible to process discontinuous strings.
1095    
1096    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1097        much the same function. They were added by different people who were trying
1098        to make PCRE easy to compile on non-Unix systems. It has been suggested
1099        that PCRE_EXPORT be abolished now that there is more automatic apparatus
1100        for compiling on Windows systems. I have therefore replaced it with
1101        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1102        defaults to "extern" for C or "extern C" for C++, which works fine on
1103        Unix-like systems. It is now possible to override the value of PCRE_DATA_
1104        SCOPE with something explicit in config.h. In addition:
1105    
1106        (a) pcreposix.h still had just "extern" instead of either of these macros;
1107            I have replaced it with PCRE_DATA_SCOPE.
1108    
1109        (b) Functions such as _pcre_xclass(), which are internal to the library,
1110            but external in the C sense, all had PCRE_EXPORT in their definitions.
1111            This is apparently wrong for the Windows case, so I have removed it.
1112            (It makes no difference on Unix-like systems.)
1113    
1114    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1115        of recursive calls to match(). This is different to MATCH_LIMIT because
1116        that limits the total number of calls to match(), not all of which increase
1117        the depth of recursion. Limiting the recursion depth limits the amount of
1118        stack (or heap if NO_RECURSE is set) that is used. The default can be set
1119        when PCRE is compiled, and changed at run time. A patch from Google adds
1120        this functionality to the C++ interface.
1121    
1122    18. Changes to the handling of Unicode character properties:
1123    
1124        (a) Updated the table to Unicode 4.1.0.
1125    
1126        (b) Recognize characters that are not in the table as "Cn" (undefined).
1127    
1128        (c) I revised the way the table is implemented to a much improved format
1129            which includes recognition of ranges. It now supports the ranges that
1130            are defined in UnicodeData.txt, and it also amalgamates other
1131            characters into ranges. This has reduced the number of entries in the
1132            table from around 16,000 to around 3,000, thus reducing its size
1133            considerably. I realized I did not need to use a tree structure after
1134            all - a binary chop search is just as efficient. Having reduced the
1135            number of entries, I extended their size from 6 bytes to 8 bytes to
1136            allow for more data.
1137    
1138        (d) Added support for Unicode script names via properties such as \p{Han}.
1139    
1140    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1141        matching that character.
1142    
1143    20. When matching a repeated Unicode property with a minimum greater than zero,
1144        (for example \pL{2,}), PCRE could look past the end of the subject if it
1145        reached it while seeking the minimum number of characters. This could
1146        happen only if some of the characters were more than one byte long, because
1147        there is a check for at least the minimum number of bytes.
1148    
1149    21. Refactored the implementation of \p and \P so as to be more general, to
1150        allow for more different types of property in future. This has changed the
1151        compiled form incompatibly. Anybody with saved compiled patterns that use
1152        \p or \P will have to recompile them.
1153    
1154    22. Added "Any" and "L&" to the supported property types.
1155    
1156    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1157        but give a compile time error if the value is greater than 0xff.
1158    
1159    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1160        accidentally not being installed or uninstalled.
1161    
1162    25. The pcre.h file was built from pcre.h.in, but the only changes that were
1163        made were to insert the current release number. This seemed silly, because
1164        it made things harder for people building PCRE on systems that don't run
1165        "configure". I have turned pcre.h into a distributed file, no longer built
1166        by "configure", with the version identification directly included. There is
1167        no longer a pcre.h.in file.
1168    
1169        However, this change necessitated a change to the pcre-config script as
1170        well. It is built from pcre-config.in, and one of the substitutions was the
1171        release number. I have updated configure.ac so that ./configure now finds
1172        the release number by grepping pcre.h.
1173    
1174    26. Added the ability to run the tests under valgrind.
1175    
1176    
1177    Version 6.4 05-Sep-05
1178    ---------------------
1179    
1180     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1181        "--" to be printed when multiple files were scanned, even when none of the
1182        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1183        consider it to be a bug, and have restored the previous behaviour.
1184    
1185     2. A couple of code tidies to get rid of compiler warnings.
1186    
1187     3. The pcretest program used to cheat by referring to symbols in the library
1188        whose names begin with _pcre_. These are internal symbols that are not
1189        really supposed to be visible externally, and in some environments it is
1190        possible to suppress them. The cheating is now confined to including
1191        certain files from the library's source, which is a bit cleaner.
1192    
1193     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1194        file's purpose clearer.
1195    
1196     5. Reorganized pcre_ucp_findchar().
1197    
1198    
1199    Version 6.3 15-Aug-05
1200    ---------------------
1201    
1202     1. The file libpcre.pc.in did not have general read permission in the tarball.
1203    
1204     2. There were some problems when building without C++ support:
1205    
1206        (a) If C++ support was not built, "make install" and "make test" still
1207            tried to test it.
1208    
1209        (b) There were problems when the value of CXX was explicitly set. Some
1210            changes have been made to try to fix these, and ...
1211    
1212        (c) --disable-cpp can now be used to explicitly disable C++ support.
1213    
1214        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1215            backslash in a target when C++ was disabled. This confuses some
1216            versions of "make", apparently. Using an intermediate variable solves
1217            this. (Same for CPP_LOBJ.)
1218    
1219     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1220        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1221        necessary on certain architectures.
1222    
1223     4. Added a setting of -export-symbols-regex to the link command to remove
1224        those symbols that are exported in the C sense, but actually are local
1225        within the library, and not documented. Their names all begin with
1226        "_pcre_". This is not a perfect job, because (a) we have to except some
1227        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1228        available (and never for static libraries). I have made a note to try to
1229        find a way round (a) in the future.
1230    
1231    
1232    Version 6.2 01-Aug-05
1233    ---------------------
1234    
1235     1. There was no test for integer overflow of quantifier values. A construction
1236        such as {1111111111111111} would give undefined results. What is worse, if
1237        a minimum quantifier for a parenthesized subpattern overflowed and became
1238        negative, the calculation of the memory size went wrong. This could have
1239        led to memory overwriting.
1240    
1241     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1242    
1243     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1244        operating environments where this matters.
1245    
1246     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1247        PCRE options from within the C++ wrapper.
1248    
1249     5. Named capturing subpatterns were not being correctly counted when a pattern
1250        was compiled. This caused two problems: (a) If there were more than 100
1251        such subpatterns, the calculation of the memory needed for the whole
1252        compiled pattern went wrong, leading to an overflow error. (b) Numerical
1253        back references of the form \12, where the number was greater than 9, were
1254        not recognized as back references, even though there were sufficient
1255        previous subpatterns.
1256    
1257     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1258        versions of gcc, e.g. 2.95.4.
1259    
1260    
1261    Version 6.1 21-Jun-05
1262    ---------------------
1263    
1264     1. There was one reference to the variable "posix" in pcretest.c that was not
1265        surrounded by "#if !defined NOPOSIX".
1266    
1267     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1268        the cross-check on the old pcre_info() function, for the benefit of the
1269        cut-down version of PCRE that is currently imported into Exim.
1270    
1271     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1272        allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1273        patterns that start (?i) but I don't think that matters. The use of (?i) is
1274        just an example; this all applies to the other options as well.
1275    
1276     4. Since libtool seems to echo the compile commands it is issuing, the output
1277        from "make" can be reduced a bit by putting "@" in front of each libtool
1278        compile command.
1279    
1280     5. Patch from the folks at Google for configure.in to be a bit more thorough
1281        in checking for a suitable C++ installation before trying to compile the
1282        C++ stuff. This should fix a reported problem when a compiler was present,
1283        but no suitable headers.
1284    
1285     6. The man pages all had just "PCRE" as their title. I have changed them to
1286        be the relevant file name. I have also arranged that these names are
1287        retained in the file doc/pcre.txt, which is a concatenation in text format
1288        of all the man pages except the little individual ones for each function.
1289    
1290     7. The NON-UNIX-USE file had not been updated for the different set of source
1291        files that come with release 6. I also added a few comments about the C++
1292        wrapper.
1293    
1294    
1295    Version 6.0 07-Jun-05
1296    ---------------------
1297    
1298     1. Some minor internal re-organization to help with my DFA experiments.
1299    
1300     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1301        didn't matter for the library itself when fully configured, but did matter
1302        when compiling without UCP support, or within Exim, where the ucp files are
1303        not imported.
1304    
1305     3. Refactoring of the library code to split up the various functions into
1306        different source modules. The addition of the new DFA matching code (see
1307        below) to a single monolithic source would have made it really too
1308        unwieldy, quite apart from causing all the code to be include in a
1309        statically linked application, when only some functions are used. This is
1310        relevant even without the DFA addition now that patterns can be compiled in
1311        one application and matched in another.
1312    
1313        The downside of splitting up is that there have to be some external
1314        functions and data tables that are used internally in different modules of
1315        the library but which are not part of the API. These have all had their
1316        names changed to start with "_pcre_" so that they are unlikely to clash
1317        with other external names.
1318    
1319     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1320        a different (DFA) algorithm. Although it is slower than the original
1321        function, it does have some advantages for certain types of matching
1322        problem.
1323    
1324     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1325        including restarting after a partial match.
1326    
1327     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1328        defined when compiling for Windows was sent to me. I have put it into the
1329        code, though I have no means of testing or verifying it.
1330    
1331     7. Added the pcre_refcount() auxiliary function.
1332    
1333     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1334        match before or at the first newline in the subject string. In pcretest,
1335        the /f option on a pattern can be used to set this.
1336    
1337     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1338        would behave wrongly. This has been present in PCRE since release 4.0.
1339    
1340    10. A number of changes to the pcregrep command:
1341    
1342        (a) Refactored how -x works; insert ^(...)$ instead of setting
1343            PCRE_ANCHORED and checking the length, in preparation for adding
1344            something similar for -w.
1345    
1346        (b) Added the -w (match as a word) option.
1347    
1348        (c) Refactored the way lines are read and buffered so as to have more
1349            than one at a time available.
1350    
1351        (d) Implemented a pcregrep test script.
1352    
1353        (e) Added the -M (multiline match) option. This allows patterns to match
1354            over several lines of the subject. The buffering ensures that at least
1355            8K, or the rest of the document (whichever is the shorter) is available
1356            for matching (and similarly the previous 8K for lookbehind assertions).
1357    
1358        (f) Changed the --help output so that it now says
1359    
1360              -w, --word-regex(p)
1361    
1362            instead of two lines, one with "regex" and the other with "regexp"
1363            because that confused at least one person since the short forms are the
1364            same. (This required a bit of code, as the output is generated
1365            automatically from a table. It wasn't just a text change.)
1366    
1367        (g) -- can be used to terminate pcregrep options if the next thing isn't an
1368            option but starts with a hyphen. Could be a pattern or a path name
1369            starting with a hyphen, for instance.
1370    
1371        (h) "-" can be given as a file name to represent stdin.
1372    
1373        (i) When file names are being printed, "(standard input)" is used for
1374            the standard input, for compatibility with GNU grep. Previously
1375            "<stdin>" was used.
1376    
1377        (j) The option --label=xxx can be used to supply a name to be used for
1378            stdin when file names are being printed. There is no short form.
1379    
1380        (k) Re-factored the options decoding logic because we are going to add
1381            two more options that take data. Such options can now be given in four
1382            different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1383    
1384        (l) Added the -A, -B, and -C options for requesting that lines of context
1385            around matches be printed.
1386    
1387        (m) Added the -L option to print the names of files that do not contain
1388            any matching lines, that is, the complement of -l.
1389    
1390        (n) The return code is 2 if any file cannot be opened, but pcregrep does
1391            continue to scan other files.
1392    
1393        (o) The -s option was incorrectly implemented. For compatibility with other
1394            greps, it now suppresses the error message for a non-existent or non-
1395            accessible file (but not the return code). There is a new option called
1396            -q that suppresses the output of matching lines, which was what -s was
1397            previously doing.
1398    
1399        (p) Added --include and --exclude options to specify files for inclusion
1400            and exclusion when recursing.
1401    
1402    11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1403        Hopefully, it now does.
1404    
1405    12. Missing cast in pcre_study().
1406    
1407    13. Added an "uninstall" target to the makefile.
1408    
1409    14. Replaced "extern" in the function prototypes in Makefile.in with
1410        "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1411        world, but is set differently for Windows.
1412    
1413    15. Added a second compiling function called pcre_compile2(). The only
1414        difference is that it has an extra argument, which is a pointer to an
1415        integer error code. When there is a compile-time failure, this is set
1416        non-zero, in addition to the error test pointer being set to point to an
1417        error message. The new argument may be NULL if no error number is required
1418        (but then you may as well call pcre_compile(), which is now just a
1419        wrapper). This facility is provided because some applications need a
1420        numeric error indication, but it has also enabled me to tidy up the way
1421        compile-time errors are handled in the POSIX wrapper.
1422    
1423    16. Added VPATH=.libs to the makefile; this should help when building with one
1424        prefix path and installing with another. (Or so I'm told by someone who
1425        knows more about this stuff than I do.)
1426    
1427    17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1428        passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1429        match everything, including newlines. This is not POSIX-compatible, but
1430        somebody wanted the feature. From pcretest it can be activated by using
1431        both the P and the s flags.
1432    
1433    18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1434    
1435    19. libpcre.pc was being incorrectly installed as executable.
1436    
1437    20. A couple of places in pcretest check for end-of-line by looking for '\n';
1438        it now also looks for '\r' so that it will work unmodified on Windows.
1439    
1440    21. Added Google's contributed C++ wrapper to the distribution.
1441    
1442    22. Added some untidy missing memory free() calls in pcretest, to keep
1443        Electric Fence happy when testing.
1444    
1445    
1446    
1447    Version 5.0 13-Sep-04
1448    ---------------------
1449    
1450     1. Internal change: literal characters are no longer packed up into items
1451        containing multiple characters in a single byte-string. Each character
1452        is now matched using a separate opcode. However, there may be more than one
1453        byte in the character in UTF-8 mode.
1454    
1455     2. The pcre_callout_block structure has two new fields: pattern_position and
1456        next_item_length. These contain the offset in the pattern to the next match
1457        item, and its length, respectively.
1458    
1459     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1460        insertion of callouts before each pattern item. Added the /C option to
1461        pcretest to make use of this.
1462    
1463     4. On the advice of a Windows user, the lines
1464    
1465          #if defined(_WIN32) || defined(WIN32)
1466          _setmode( _fileno( stdout ), 0x8000 );
1467          #endif  /* defined(_WIN32) || defined(WIN32) */
1468    
1469        have been added to the source of pcretest. This apparently does useful
1470        magic in relation to line terminators.
1471    
1472     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1473        for the benefit of those environments where the "b" makes a difference.
1474    
1475     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1476        to know about it. I have put a hack into configure.in that adds in code
1477        to set GCC=yes if CC=icc. This seems to end up at a point in the
1478        generated configure script that is early enough to affect the setting of
1479        compiler options, which is what is needed, but I have no means of testing
1480        whether it really works. (The user who reported this had patched the
1481        generated configure script, which of course I cannot do.)
1482    
1483        LATER: After change 22 below (new libtool files), the configure script
1484        seems to know about icc (and also ecc). Therefore, I have commented out
1485        this hack in configure.in.
1486    
1487     7. Added support for pkg-config (2 patches were sent in).
1488    
1489     8. Negated POSIX character classes that used a combination of internal tables
1490        were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1491        [[:^ascii]]. Typically, they would match almost any characters. The other
1492        POSIX classes were not broken in this way.
1493    
1494     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1495        to find the match, as PCRE was deluded into thinking that the match had to
1496        start at the start point or following a newline. The same bug applied to
1497        patterns with negative forward assertions or any backward assertions
1498        preceding ".*" at the start, unless the pattern required a fixed first
1499        character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1500    
1501    10. In UTF-8 mode, when moving forwards in the subject after a failed match
1502        starting at the last subject character, bytes beyond the end of the subject
1503        string were read.
1504    
1505    11. Renamed the variable "class" as "classbits" to make life easier for C++
1506        users. (Previously there was a macro definition, but it apparently wasn't
1507        enough.)
1508    
1509    12. Added the new field "tables" to the extra data so that tables can be passed
1510        in at exec time, or the internal tables can be re-selected. This allows
1511        a compiled regex to be saved and re-used at a later time by a different
1512        program that might have everything at different addresses.
1513    
1514    13. Modified the pcre-config script so that, when run on Solaris, it shows a
1515        -R library as well as a -L library.
1516    
1517    14. The debugging options of pcretest (-d on the command line or D on a
1518        pattern) showed incorrect output for anything following an extended class
1519        that contained multibyte characters and which was followed by a quantifier.
1520    
1521    15. Added optional support for general category Unicode character properties
1522        via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1523        support. It adds about 90K to the size of the library. The meanings of the
1524        inbuilt class escapes such as \d and \s have NOT been changed.
1525    
1526    16. Updated pcredemo.c to include calls to free() to release the memory for the
1527        compiled pattern.
1528    
1529    17. The generated file chartables.c was being created in the source directory
1530        instead of in the building directory. This caused the build to fail if the
1531        source directory was different from the building directory, and was
1532        read-only.
1533    
1534    18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1535        file. No doubt somebody will tell me if they don't make sense... Also added
1536        Dan Mooney's comments about building on OpenVMS.
1537    
1538    19. Added support for partial matching via the PCRE_PARTIAL option for
1539        pcre_exec() and the \P data escape in pcretest.
1540    
1541    20. Extended pcretest with 3 new pattern features:
1542    
1543        (i)   A pattern option of the form ">rest-of-line" causes pcretest to
1544              write the compiled pattern to the file whose name is "rest-of-line".
1545              This is a straight binary dump of the data, with the saved pointer to
1546              the character tables forced to be NULL. The study data, if any, is
1547              written too. After writing, pcretest reads a new pattern.
1548    
1549        (ii)  If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1550              compiled pattern from the given file. There must not be any
1551              occurrences of "<" in the file name (pretty unlikely); if there are,
1552              pcretest will instead treat the initial "<" as a pattern delimiter.
1553              After reading in the pattern, pcretest goes on to read data lines as
1554              usual.
1555    
1556        (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1557              and 16-bit fields in a compiled pattern, to simulate a pattern that
1558              was compiled on a host of opposite endianness.
1559    
1560    21. The pcre-exec() function can now cope with patterns that were compiled on
1561        hosts of opposite endianness, with this restriction:
1562    
1563          As for any compiled expression that is saved and used later, the tables
1564          pointer field cannot be preserved; the extra_data field in the arguments
1565          to pcre_exec() should be used to pass in a tables address if a value
1566          other than the default internal tables were used at compile time.
1567    
1568    22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1569        now diagnosed as an error. Previously, most of the time, a negative number
1570        would have been treated as zero, but if in addition "ovector" was passed as
1571        NULL, a crash could occur.
1572    
1573    23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1574        new versions from the libtool 1.5 distribution (the last one is a copy of
1575        a file called libtool.m4). This seems to have fixed the need to patch
1576        "configure" to support Darwin 1.3 (which I used to do). However, I still
1577        had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1578        workstation).
1579    
1580    24. Changed the PCRE licence to be the more standard "BSD" licence.
1581    
1582    
1583    Version 4.5 01-Dec-03
1584    ---------------------
1585    
1586     1. There has been some re-arrangement of the code for the match() function so
1587        that it can be compiled in a version that does not call itself recursively.
1588        Instead, it keeps those local variables that need separate instances for
1589        each "recursion" in a frame on the heap, and gets/frees frames whenever it
1590        needs to "recurse". Keeping track of where control must go is done by means
1591        of setjmp/longjmp. The whole thing is implemented by a set of macros that
1592        hide most of the details from the main code, and operates only if
1593        NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1594        "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1595        operating.
1596    
1597        To make it easier for callers to provide specially tailored get/free
1598        functions for this usage, two new functions, pcre_stack_malloc, and
1599        pcre_stack_free, are used. They are always called in strict stacking order,
1600        and the size of block requested is always the same.
1601    
1602        The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1603        PCRE has been compiled to use the stack or the heap for recursion. The
1604        -C option of pcretest uses this to show which version is compiled.
1605    
1606        A new data escape \S, is added to pcretest; it causes the amounts of store
1607        obtained and freed by both kinds of malloc/free at match time to be added
1608        to the output.
1609    
1610     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1611        what's available on my current Linux desktop machine.
1612    
1613     3. When matching a UTF-8 string, the test for a valid string at the start has
1614        been extended. If start_offset is not zero, PCRE now checks that it points
1615        to a byte that is the start of a UTF-8 character. If not, it returns
1616        PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1617        this is necessary because there may be backward assertions in the pattern.
1618        When matching the same subject several times, it may save resources to use
1619        PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1620    
1621     4. The code for checking the validity of UTF-8 strings has been tightened so
1622        that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1623        containing "overlong sequences".
1624    
1625     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1626        I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1627        should have been "&", but it just so happened that all the cases this let
1628        through by mistake were picked up later in the function.
1629    
1630     6. I had used a variable called "isblank" - this is a C99 function, causing
1631        some compilers to warn. To avoid this, I renamed it (as "blankclass").
1632    
1633     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1634        prompting; (b) run "./pcretest /dev/null" at the start of the test script
1635        so the version is shown; (c) stop "make test" echoing "./RunTest".
1636    
1637     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1638    
1639     9. The prototype for memmove() for systems that don't have it was using
1640        size_t, but the inclusion of the header that defines size_t was later. I've
1641        moved the #includes for the C headers earlier to avoid this.
1642    
1643    10. Added some adjustments to the code to make it easier to compiler on certain
1644        special systems:
1645    
1646          (a) Some "const" qualifiers were missing.
1647          (b) Added the macro EXPORT before all exported functions; by default this
1648              is defined to be empty.
1649          (c) Changed the dftables auxiliary program (that builds chartables.c) so
1650              that it reads its output file name as an argument instead of writing
1651              to the standard output and assuming this can be redirected.
1652    
1653    11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1654        class containing characters with values greater than 255, PCRE compilation
1655        went into a loop.
1656    
1657    12. A recursive reference to a subpattern that was within another subpattern
1658        that had a minimum quantifier of zero caused PCRE to crash. For example,
1659        (x(y(?2))z)? provoked this bug with a subject that got as far as the
1660        recursion. If the recursively-called subpattern itself had a zero repeat,
1661        that was OK.
1662    
1663    13. In pcretest, the buffer for reading a data line was set at 30K, but the
1664        buffer into which it was copied (for escape processing) was still set at
1665        1024, so long lines caused crashes.
1666    
1667    14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1668        "internal error: code overflow...". This applied to any character class
1669        that was followed by a possessive quantifier.
1670    
1671    15. Modified the Makefile to add libpcre.la as a prerequisite for
1672        libpcreposix.la because I was told this is needed for a parallel build to
1673        work.
1674    
1675    16. If a pattern that contained .* following optional items at the start was
1676        studied, the wrong optimizing data was generated, leading to matching
1677        errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1678        matching string must start with a or b or c. The correct conclusion for
1679        this pattern is that a match can start with any character.
1680    
1681    
1682    Version 4.4 13-Aug-03
1683    ---------------------
1684    
1685     1. In UTF-8 mode, a character class containing characters with values between
1686        127 and 255 was not handled correctly if the compiled pattern was studied.
1687        In fixing this, I have also improved the studying algorithm for such
1688        classes (slightly).
1689    
1690     2. Three internal functions had redundant arguments passed to them. Removal
1691        might give a very teeny performance improvement.
1692    
1693     3. Documentation bug: the value of the capture_top field in a callout is *one
1694        more than* the number of the hightest numbered captured substring.
1695    
1696     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1697        in incorrectly linking with a previously installed version. They now link
1698        explicitly with libpcre.la.
1699    
1700     5. configure.in no longer needs to recognize Cygwin specially.
1701    
1702     6. A problem in pcre.in for Windows platforms is fixed.
1703    
1704     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1705        pcretest, it used to include the size of the study block as part of its
1706        output. Unfortunately, the structure contains a field that has a different
1707        size on different hardware architectures. This meant that the tests that
1708        showed this size failed. As the block is currently always of a fixed size,
1709        this information isn't actually particularly useful in pcretest output, so
1710        I have just removed it.
1711    
1712     8. Three pre-processor statements accidentally did not start in column 1.
1713        Sadly, there are *still* compilers around that complain, even though
1714        standard C has not required this for well over a decade. Sigh.
1715    
1716     9. In pcretest, the code for checking callouts passed small integers in the
1717        callout_data field, which is a void * field. However, some picky compilers
1718        complained about the casts involved for this on 64-bit systems. Now
1719        pcretest passes the address of the small integer instead, which should get
1720        rid of the warnings.
1721    
1722    10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1723        both compile and run time, and gives an error if an invalid UTF-8 sequence
1724        is found. There is a option for disabling this check in cases where the
1725        string is known to be correct and/or the maximum performance is wanted.
1726    
1727    11. In response to a bug report, I changed one line in Makefile.in from
1728    
1729            -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1730        to
1731            -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1732    
1733        to look similar to other lines, but I have no way of telling whether this
1734        is the right thing to do, as I do not use Windows. No doubt I'll get told
1735        if it's wrong...
1736    
1737    
1738    Version 4.3 21-May-03
1739    ---------------------
1740    
1741    1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1742       Makefile.
1743    
1744    2. Some refactoring to improve the quality of the code:
1745    
1746       (i)   The utf8_table... variables are now declared "const".
1747    
1748       (ii)  The code for \cx, which used the "case flipping" table to upper case
1749             lower case letters, now just substracts 32. This is ASCII-specific,
1750             but the whole concept of \cx is ASCII-specific, so it seems
1751             reasonable.
1752    
1753       (iii) PCRE was using its character types table to recognize decimal and
1754             hexadecimal digits in the pattern. This is silly, because it handles
1755             only 0-9, a-f, and A-F, but the character types table is locale-
1756             specific, which means strange things might happen. A private
1757             table is now used for this - though it costs 256 bytes, a table is
1758             much faster than multiple explicit tests. Of course, the standard
1759             character types table is still used for matching digits in subject
1760             strings against \d.
1761    
1762       (iv)  Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1763             ending in _t are). So I've renamed it as ESC_tee.
1764    
1765    3. The first argument for regexec() in the POSIX wrapper should have been
1766       defined as "const".
1767    
1768    4. Changed pcretest to use malloc() for its buffers so that they can be
1769       Electric Fenced for debugging.
1770    
1771    5. There were several places in the code where, in UTF-8 mode, PCRE would try
1772       to read one or more bytes before the start of the subject string. Often this
1773       had no effect on PCRE's behaviour, but in some circumstances it could
1774       provoke a segmentation fault.
1775    
1776    6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1777       to try to read one or more bytes before the start of the subject string.
1778    
1779    7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1780       UTF-8 support could misbehave in various ways if the subject string
1781       contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1782       area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1783       back over UTF-8 characters.)
1784    
1785    
1786    Version 4.2 14-Apr-03
1787    ---------------------
1788    
1789    1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1790    
1791    2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1792         [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1793         [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1794         [WIN_PREFIX]: new variable, "cyg" for Cygwin
1795         * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1796           and BUILD_EXEEXT
1797         Note: automatic setting of the BUILD variables is not yet working
1798         set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1799           compile-time but not at link-time
1800         [LINK]: use for linking executables only
1801         make different versions for Windows and non-Windows
1802         [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1803           libraries
1804         [LINK_FOR_BUILD]: new variable
1805         [OBJEXT]: use throughout
1806         [EXEEXT]: use throughout
1807         <winshared>: new target
1808         <wininstall>: new target
1809         <dftables.o>: use native compiler
1810         <dftables>: use native linker
1811         <install>: handle Windows platform correctly
1812         <clean>: ditto
1813         <check>: ditto
1814         copy DLL to top builddir before testing
1815    
1816       As part of these changes, -no-undefined was removed again. This was reported
1817       to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1818       in any case.
1819    
1820    3. Some tidies to get rid of compiler warnings:
1821    
1822       . In the match_data structure, match_limit was an unsigned long int, whereas
1823         match_call_count was an int. I've made them both unsigned long ints.
1824    
1825       . In pcretest the fact that a const uschar * doesn't automatically cast to
1826         a void * provoked a warning.
1827    
1828       . Turning on some more compiler warnings threw up some "shadow" variables
1829         and a few more missing casts.
1830    
1831    4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1832       option, a class that contained a single character with a value between 128
1833       and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1834    
1835    5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1836       option, a class that contained several characters, but with at least one
1837       whose value was between 128 and 255 caused PCRE to crash.
1838    
1839    
1840    Version 4.1 12-Mar-03
1841    ---------------------
1842    
1843    1. Compiling with gcc -pedantic found a couple of places where casts were
1844    needed, and a string in dftables.c that was longer than standard compilers are
1845    required to support.
1846    
1847    2. Compiling with Sun's compiler found a few more places where the code could
1848    be tidied up in order to avoid warnings.
1849    
1850    3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1851    first of these names is deprecated in the latest Autoconf in favour of the name
1852    CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1853    compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1854    analogy I have changed it to CFLAGS_FOR_BUILD.
1855    
1856    4. Added -no-undefined to the linking command in the Makefile, because this is
1857    apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1858    linking step for the pcreposix library.
1859    
1860    5. PCRE was failing to diagnose the case of two named groups with the same
1861    name.
1862    
1863    6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1864    literal character that is needed in the subject for a match, and scans along to
1865    ensure that it is present before embarking on the full matching process. This
1866    saves time in cases of nested unlimited repeats that are never going to match.
1867    Problem: the scan can take a lot of time if the subject is very long (e.g.
1868    megabytes), thus penalizing straightforward matches. It is now done only if the
1869    amount of subject to be scanned is less than 1000 bytes.
1870    
1871    7. A lesser problem with the same optimization is that it was recording the
1872    first character of an anchored pattern as "needed", thus provoking a search
1873    right along the subject, even when the first match of the pattern was going to
1874    fail. The "needed" character is now not set for anchored patterns, unless it
1875    follows something in the pattern that is of non-fixed length. Thus, it still
1876    fulfils its original purpose of finding quick non-matches in cases of nested
1877    unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1878    
1879    
1880    Version 4.0 17-Feb-03
1881    ---------------------
1882    
1883    1. If a comment in an extended regex that started immediately after a meta-item
1884    extended to the end of string, PCRE compiled incorrect data. This could lead to
1885    all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1886    
1887    2. Moved to autoconf 2.53 and libtool 1.4.2.
1888    
1889    3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1890    the special perltest8 script is no longer needed - all the tests can be run
1891    from a single perltest script.
1892    
1893    4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1894    by \s. It has now been removed in PCRE. This means it isn't recognized as
1895    whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1896    class [:space:] *does* include VT, thereby creating a mess.
1897    
1898    5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1899    space and tab.
1900    
1901    6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1902    its new features into the main test script, reducing the number of scripts.
1903    
1904    7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1905    were backward compatible, and made the (?i) apply to the whole pattern, as if
1906    /i were given. Now it behaves more logically, and applies the option setting
1907    only to what follows. PCRE has been changed to follow suit. However, if it
1908    finds options settings right at the start of the pattern, it extracts them into
1909    the global options, as before. Thus, they show up in the info data.
1910    
1911    8. Added support for the \Q...\E escape sequence. Characters in between are
1912    treated as literals. This is slightly different from Perl in that $ and @ are
1913    also handled as literals inside the quotes. In Perl, they will cause variable
1914    interpolation. Note the following examples:
1915    
1916        Pattern            PCRE matches      Perl matches
1917    
1918        \Qabc$xyz\E        abc$xyz           abc followed by the contents of $xyz
1919        \Qabc\$xyz\E       abc\$xyz          abc\$xyz
1920        \Qabc\E\$\Qxyz\E   abc$xyz           abc$xyz
1921    
1922    For compatibility with Perl, \Q...\E sequences are recognized inside character
1923    classes as well as outside them.
1924    
1925    9. Re-organized 3 code statements in pcretest to avoid "overflow in
1926    floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1927    (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1928    signed/unsigned warnings.
1929    
1930    10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1931    option for pcretest, so I've replaced it by a simple function that does just
1932    that job.
1933    
1934    11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1935    "pcregrep -".
1936    
1937    12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1938    Java package. This provides some syntactic sugar for simple cases of what my
1939    documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1940    as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1941    item, you can use this simplified notation. Note that only makes sense with
1942    greedy quantifiers. Consequently, the use of the possessive quantifier forces
1943    greediness, whatever the setting of the PCRE_UNGREEDY option.
1944    
1945    13. A change of greediness default within a pattern was not taking effect at
1946    the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1947    subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1948    was abstracted outside.
1949    
1950    14. PCRE now supports the \G assertion. It is true when the current matching
1951    position is at the start point of the match. This differs from \A when the
1952    starting offset is non-zero. Used with the /g option of pcretest (or similar
1953    code), it works in the same way as it does for Perl's /g option. If all
1954    alternatives of a regex begin with \G, the expression is anchored to the start
1955    match position, and the "anchored" flag is set in the compiled expression.
1956    
1957    15. Some bugs concerning the handling of certain option changes within patterns
1958    have been fixed. These applied to options other than (?ims). For example,
1959    "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1960    been the other way round. Some of this was related to change 7 above.
1961    
1962    16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1963    features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1964    and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1965    POSIX classes only within a class (e.g. /[[:alpha:]]/).
1966    
1967    17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1968    mode. Unlike ".", it always matches newline, whatever the setting of
1969    PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1970    assertions. Perl allows it, but it doesn't (in general) work because it can't
1971    calculate the length of the lookbehind. At least, that's the case for Perl
1972    5.8.0 - I've been told they are going to document that it doesn't work in
1973    future.
1974    
1975    18. Added an error diagnosis for escapes that PCRE does not support: these are
1976    \L, \l, \N, \P, \p, \U, \u, and \X.
1977    
1978    19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1979    reading past the end of the pattern in cases such as /[abcd/.
1980    
1981    20. PCRE was getting more memory than necessary for patterns with classes that
1982    contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1983    
1984    21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1985    compiling PCRE for use with Virtual Pascal.
1986    
1987    22. Small fix to the Makefile to make it work properly if the build is done
1988    outside the source tree.
1989    
1990    23. Added a new extension: a condition to go with recursion. If a conditional
1991    subpattern starts with (?(R) the "true" branch is used if recursion has
1992    happened, whereas the "false" branch is used only at the top level.
1993    
1994    24. When there was a very long string of literal characters (over 255 bytes
1995    without UTF support, over 250 bytes with UTF support), the computation of how
1996    much memory was required could be incorrect, leading to segfaults or other
1997    strange effects.
1998    
1999    25. PCRE was incorrectly assuming anchoring (either to start of subject or to
2000    start of line for a non-DOTALL pattern) when a pattern started with (.*) and
2001    there was a subsequent back reference to those brackets. This meant that, for
2002    example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
2003    possible to check for precisely this case. All we can do is abandon the
2004    optimization if .* occurs inside capturing brackets when there are any back
2005    references whatsoever. (See below for a better fix that came later.)
2006    
2007    26. The handling of the optimization for finding the first character of a
2008    non-anchored pattern, and for finding a character that is required later in the
2009    match were failing in some cases. This didn't break the matching; it just
2010    failed to optimize when it could. The way this is done has been re-implemented.
2011    
2012    27. Fixed typo in error message for invalid (?R item (it said "(?p").
2013    
2014    28. Added a new feature that provides some of the functionality that Perl
2015    provides with (?{...}). The facility is termed a "callout". The way it is done
2016    in PCRE is for the caller to provide an optional function, by setting
2017    pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
2018    global variable. By default it is unset, which disables all calling out. To get
2019    the function called, the regex must include (?C) at appropriate points. This
2020    is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
2021    This provides a means of identifying different callout points. When PCRE
2022    reaches such a point in the regex, if pcre_callout has been set, the external
2023    function is called. It is provided with data in a structure called
2024    pcre_callout_block, which is defined in pcre.h. If the function returns 0,
2025    matching continues; if it returns a non-zero value, the match at the current
2026    point fails. However, backtracking will occur if possible. [This was changed
2027    later and other features added - see item 49 below.]
2028    
2029    29. pcretest is upgraded to test the callout functionality. It provides a
2030    callout function that displays information. By default, it shows the start of
2031    the match and the current position in the text. There are some new data escapes
2032    to vary what happens:
2033    
2034        \C+         in addition, show current contents of captured substrings
2035        \C-         do not supply a callout function
2036        \C!n        return 1 when callout number n is reached
2037        \C!n!m      return 1 when callout number n is reached for the mth time
2038    
2039    30. If pcregrep was called with the -l option and just a single file name, it
2040    output "<stdin>" if a match was found, instead of the file name.
2041    
2042    31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
2043    slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
2044    pcre_exec(). This saves a malloc/free per call. The default value of
2045    POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
2046    when configuring.
2047    
2048    32. The default maximum size of a compiled pattern is 64K. There have been a
2049    few cases of people hitting this limit. The code now uses macros to handle the
2050    storing of links as offsets within the compiled pattern. It defaults to 2-byte
2051    links, but this can be changed to 3 or 4 bytes by --with-link-size when
2052    configuring. Tests 2 and 5 work only with 2-byte links because they output
2053    debugging information about compiled patterns.
2054    
2055    33. Internal code re-arrangements:
2056    
2057    (a) Moved the debugging function for printing out a compiled regex into
2058        its own source file (printint.c) and used #include to pull it into
2059        pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2060        separate copies.
2061    
2062    (b) Defined the list of op-code names for debugging as a macro in
2063        internal.h so that it is next to the definition of the opcodes.
2064    
2065    (c) Defined a table of op-code lengths for simpler skipping along compiled
2066        code. This is again a macro in internal.h so that it is next to the
2067        definition of the opcodes.
2068    
2069    34. Added support for recursive calls to individual subpatterns, along the
2070    lines of Robin Houston's patch (but implemented somewhat differently).
2071    
2072    35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2073    allow it to read and process whole directories in Win32. This code was
2074    contributed by Lionel Fourquaux; it has not been tested by me.
2075    
2076    36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2077    used to name a group. Names consist of alphanumerics and underscores, and must
2078    be unique. Back references use the syntax (?P=name) and recursive calls use
2079    (?P>name) which is a PCRE extension to the Python extension. Groups still have
2080    numbers. The function pcre_fullinfo() can be used after compilation to extract
2081    a name/number map. There are three relevant calls:
2082    
2083      PCRE_INFO_NAMEENTRYSIZE        yields the size of each entry in the map
2084      PCRE_INFO_NAMECOUNT            yields the number of entries
2085      PCRE_INFO_NAMETABLE            yields a pointer to the map.
2086    
2087    The map is a vector of fixed-size entries. The size of each entry depends on
2088    the length of the longest name used. The first two bytes of each entry are the
2089    group number, most significant byte first. There follows the corresponding
2090    name, zero terminated. The names are in alphabetical order.
2091    
2092    37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2093    case instead of 255. Making it the same both with and without UTF-8 support
2094    means that the same test output works with both.
2095    
2096    38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2097    calling malloc() with a zero argument.
2098    
2099    39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2100    optimization. I've improved things by keeping a bitmap of backreferences with
2101    numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2102    fact referenced, the optimization can be applied. It is unlikely that a
2103    relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2104    the match to follow \n) will appear inside brackets with a number greater than
2105    31, but if it does, any back reference > 31 suppresses the optimization.
2106    
2107    40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2108    of disabling numbered capturing parentheses. Any opening parenthesis that is
2109    not followed by ? behaves as if it were followed by ?: but named parentheses
2110    can still be used for capturing (and they will acquire numbers in the usual
2111    way).
2112    
2113    41. Redesigned the return codes from the match() function into yes/no/error so
2114    that errors can be passed back from deep inside the nested calls. A malloc
2115    failure while inside a recursive subpattern call now causes the
2116    PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2117    
2118    42. It is now possible to set a limit on the number of times the match()
2119    function is called in a call to pcre_exec(). This facility makes it possible to
2120    limit the amount of recursion and backtracking, though not in a directly
2121    obvious way, because the match() function is used in a number of different
2122    circumstances. The count starts from zero for each position in the subject
2123    string (for non-anchored patterns). The default limit is, for compatibility, a
2124    large number, namely 10 000 000. You can change this in two ways:
2125    
2126    (a) When configuring PCRE before making, you can use --with-match-limit=n
2127        to set a default value for the compiled library.
2128    
2129    (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2130        a different value is set. See 45 below.
2131    
2132    If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2133    
2134    43. Added a new function pcre_config(int, void *) to enable run-time extraction
2135    of things that can be changed at compile time. The first argument specifies
2136    what is wanted and the second points to where the information is to be placed.
2137    The current list of available information is:
2138    
2139      PCRE_CONFIG_UTF8
2140    
2141    The output is an integer that is set to one if UTF-8 support is available;
2142    otherwise it is set to zero.
2143    
2144      PCRE_CONFIG_NEWLINE
2145    
2146    The output is an integer that it set to the value of the code that is used for
2147    newline. It is either LF (10) or CR (13).
2148    
2149      PCRE_CONFIG_LINK_SIZE
2150    
2151    The output is an integer that contains the number of bytes used for internal
2152    linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2153    
2154      PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2155    
2156    The output is an integer that contains the threshold above which the POSIX
2157    interface uses malloc() for output vectors. See item 31 above.
2158    
2159      PCRE_CONFIG_MATCH_LIMIT
2160    
2161    The output is an unsigned integer that contains the default limit of the number
2162    of match() calls in a pcre_exec() execution. See 42 above.
2163    
2164    44. pcretest has been upgraded by the addition of the -C option. This causes it
2165    to extract all the available output from the new pcre_config() function, and to
2166    output it. The program then exits immediately.
2167    
2168    45. A need has arisen to pass over additional data with calls to pcre_exec() in
2169    order to support additional features. One way would have been to define
2170    pcre_exec2() (for example) with extra arguments, but this would not have been
2171    extensible, and would also have required all calls to the original function to
2172    be mapped to the new one. Instead, I have chosen to extend the mechanism that
2173    is used for passing in "extra" data from pcre_study().
2174    
2175    The pcre_extra structure is now exposed and defined in pcre.h. It currently
2176    contains the following fields:
2177    
2178      flags         a bitmap indicating which of the following fields are set
2179      study_data    opaque data from pcre_study()
2180      match_limit   a way of specifying a limit on match() calls for a specific
2181                      call to pcre_exec()
2182      callout_data  data for callouts (see 49 below)
2183    
2184    The flag bits are also defined in pcre.h, and are
2185    
2186      PCRE_EXTRA_STUDY_DATA
2187      PCRE_EXTRA_MATCH_LIMIT
2188      PCRE_EXTRA_CALLOUT_DATA
2189    
2190    The pcre_study() function now returns one of these new pcre_extra blocks, with
2191    the actual study data pointed to by the study_data field, and the
2192    PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2193    before. That is, this change is entirely upwards-compatible and requires no
2194    change to existing code.
2195    
2196    If you want to pass in additional data to pcre_exec(), you can either place it
2197    in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2198    block.
2199    
2200    46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2201    data string contains the escape sequence \M, pcretest calls pcre_exec() several
2202    times with different match limits, until it finds the minimum value needed for
2203    pcre_exec() to complete. The value is then output. This can be instructive; for
2204    most simple matches the number is quite small, but for pathological cases it
2205    gets very large very quickly.
2206    
2207    47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2208    returns the size of the data block pointed to by the study_data field in a
2209    pcre_extra block, that is, the value that was passed as the argument to
2210    pcre_malloc() when PCRE was getting memory in which to place the information
2211    created by pcre_study(). The fourth argument should point to a size_t variable.
2212    pcretest has been extended so that this information is shown after a successful
2213    pcre_study() call when information about the compiled regex is being displayed.
2214    
2215    48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2216    because what follows is always an absolute path. (Later: it turns out that this
2217    is more than cosmetic for MinGW, because it doesn't like empty path
2218    components.)
2219    
2220    49. Some changes have been made to the callout feature (see 28 above):
2221    
2222    (i)  A callout function now has three choices for what it returns:
2223    
2224           0  =>  success, carry on matching
2225         > 0  =>  failure at this point, but backtrack if possible
2226         < 0  =>  serious error, return this value from pcre_exec()
2227    
2228         Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2229         values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2230         "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2231         use by callout functions. It will never be used by PCRE itself.
2232    
2233    (ii) The pcre_extra structure (see 45 above) has a void * field called
2234         callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2235         pcre_callout_block structure has a field of the same name. The contents of
2236         the field passed in the pcre_extra structure are passed to the callout
2237         function in the corresponding field in the callout block. This makes it
2238         easier to use the same callout-containing regex from multiple threads. For
2239         testing, the pcretest program has a new data escape
2240    
2241           \C*n        pass the number n (may be negative) as callout_data
2242    
2243         If the callout function in pcretest receives a non-zero value as
2244         callout_data, it returns that value.
2245    
2246    50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2247    there were some redundant $(CFLAGS) in commands that are now specified as
2248    $(LINK), which already includes $(CFLAGS).
2249    
2250    51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2251    has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2252    with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2253    one-byte characters throughout. Note that case-insensitive matching applies
2254    only to characters whose values are less than 256. PCRE doesn't support the
2255    notion of cases for higher-valued characters.
2256    
2257    (i)   A character class whose characters are all within 0-255 is handled as
2258          a bit map, and the map is inverted for negative classes. Previously, a
2259          character > 255 always failed to match such a class; however it should
2260          match if the class was a negative one (e.g. [^ab]). This has been fixed.
2261    
2262    (ii)  A negated character class with a single character < 255 is coded as
2263          "not this character" (OP_NOT). This wasn't working properly when the test
2264          character was multibyte, either singly or repeated.
2265    
2266    (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2267          mode, for example: \x{100}{2,3}.
2268    
2269    (iv)  The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2270          singly or repeated) now correctly test multibyte characters. However,
2271          PCRE doesn't recognize any characters with values greater than 255 as
2272          digits, spaces, or word characters. Such characters always match \D, \S,
2273          and \W, and never match \d, \s, or \w.
2274    
2275    (v)   Classes may now contain characters and character ranges with values
2276          greater than 255. For example: [ab\x{100}-\x{400}].
2277    
2278    (vi)  pcregrep now has a --utf-8 option (synonym -u) which makes it call
2279          PCRE in UTF-8 mode.
2280    
2281    52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2282    PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2283    retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2284    value.)
2285    
2286    53. The single man page has become too large. I have therefore split it up into
2287    a number of separate man pages. These also give rise to individual HTML pages;
2288    these are now put in a separate directory, and there is an index.html page that
2289    lists them all. Some hyperlinking between the pages has been installed.
2290    
2291    54. Added convenience functions for handling named capturing parentheses.
2292    
2293    55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2294    aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2295    true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2296    are faulted.
2297    
2298    56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2299    calling configure. These values are used when compiling the dftables.c program
2300    which is run to generate the source of the default character tables. They
2301    default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2302    you will need to set these values.
2303    
2304    57. Updated the building process for Windows DLL, as provided by Fred Cox.
2305    
2306    
2307    Version 3.9 02-Jan-02
2308    ---------------------
2309    
2310    1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2311    
2312    2. If --disable-static was given, the building process failed when trying to
2313    build pcretest and pcregrep. (For some reason it was using libtool to compile
2314    them, which is not right, as they aren't part of the library.)
2315    
2316    
2317    Version 3.8 18-Dec-01
2318    ---------------------
2319    
2320    1. The experimental UTF-8 code was completely screwed up. It was packing the
2321    bytes in the wrong order. How dumb can you get?
2322    
2323    
2324    Version 3.7 29-Oct-01
2325    ---------------------
2326    
2327    1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2328    This caused pcretest, when used on the test data, to segfault. Unfortunately,
2329    this didn't happen under Solaris 8, where I normally test things.
2330    
2331    2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2332    doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2333    isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2334    this fix an hour or so after the initial 3.7 release.)
2335    
2336    
2337    Version 3.6 23-Oct-01
2338    ---------------------
2339    
2340    1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2341    offsets passed as NULL with zero offset count.
2342    
2343    2. The config.guess and config.sub files had not been updated when I moved to
2344    the latest autoconf.
2345    
2346    
2347    Version 3.5 15-Aug-01
2348    ---------------------
2349    
2350    1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2351    had been forgotten.
2352    
2353    2. By using declared but undefined structures, we can avoid using "void"
2354    definitions in pcre.h while keeping the internal definitions of the structures
2355    private.
2356    
2357    3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2358    user point of view, this means that both static and shared libraries are built
2359    by default, but this can be individually controlled. More of the work of
2360    handling this static/shared cases is now inside libtool instead of PCRE's make
2361    file.
2362    
2363    4. The pcretest utility is now installed along with pcregrep because it is
2364    useful for users (to test regexs) and by doing this, it automatically gets
2365    relinked by libtool. The documentation has been turned into a man page, so
2366    there are now .1, .txt, and .html versions in /doc.
2367    
2368    5. Upgrades to pcregrep:
2369       (i)   Added long-form option names like gnu grep.
2370       (ii)  Added --help to list all options with an explanatory phrase.
2371       (iii) Added -r, --recursive to recurse into sub-directories.
2372       (iv)  Added -f, --file to read patterns from a file.
2373    
2374    6. pcre_exec() was referring to its "code" argument before testing that
2375    argument for NULL (and giving an error if it was NULL).
2376    
2377    7. Upgraded Makefile.in to allow for compiling in a different directory from
2378    the source directory.
2379    
2380    8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2381    options bits, the pointer it was passed was to an int instead of to an unsigned
2382    long int. This mattered only on 64-bit systems.
2383    
2384    9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2385    generated) instead of pcre.in, which it its source. Also made the same change
2386    in several of the .c files.
2387    
2388    10. A new release of gcc defines printf() as a macro, which broke pcretest
2389    because it had an ifdef in the middle of a string argument for printf(). Fixed
2390    by using separate calls to printf().
2391    
2392    11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2393    script, to force use of CR or LF instead of \n in the source. On non-Unix
2394    systems, the value can be set in config.h.
2395    
2396    12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2397    absolute limit. Changed the text of the error message to make this clear, and
2398    likewise updated the man page.
2399    
2400    13. The limit of 99 on the number of capturing subpatterns has been removed.
2401    The new limit is 65535, which I hope will not be a "real" limit.
2402    
2403    
2404    Version 3.4 22-Aug-00
2405    ---------------------
2406    
2407    1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2408    
2409    2. Diagnose condition (?(0) as an error instead of crashing on matching.
2410    
2411    
2412    Version 3.3 01-Aug-00
2413    ---------------------
2414    
2415    1. If an octal character was given, but the value was greater than \377, it
2416    was not getting masked to the least significant bits, as documented. This could
2417    lead to crashes in some systems.
2418    
2419    2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2420    the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2421    
2422    3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2423    These just pass their arguments on to (pcre_free)(), but they are provided
2424    because some uses of PCRE bind it to non-C systems that can call its functions,
2425    but cannot call free() or pcre_free() directly.
2426    
2427    4. Add "make test" as a synonym for "make check". Corrected some comments in
2428    the Makefile.
2429    
2430    5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2431    Makefile.
2432    
2433    6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2434    command called pgrep for grepping around the active processes.
2435    
2436    7. Added the beginnings of support for UTF-8 character strings.
2437    
2438    8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2439    RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2440    the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2441    out for the ar command.)
2442    
2443    
2444    Version 3.2 12-May-00
2445    ---------------------
2446    
2447    This is purely a bug fixing release.
2448    
2449    1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2450    of ZA. This was just one example of several cases that could provoke this bug,
2451    which was introduced by change 9 of version 2.00. The code for breaking
2452    infinite loops after an iteration that matches an empty string was't working
2453    correctly.
2454    
2455    2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2456    when matched against abbab (for example). After matching an empty string, it
2457    wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2458    caused it to match further down the string than it should.
2459    
2460    3. The code contained an inclusion of sys/types.h. It isn't clear why this
2461    was there because it doesn't seem to be needed, and it causes trouble on some
2462    systems, as it is not a Standard C header. It has been removed.
2463    
2464    4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2465    were reported on the Macintosh. The changes were from
2466    
2467      while ((c = *(++ptr)) != 0 && c != '\n');
2468    to
2469      while ((c = *(++ptr)) != 0 && c != '\n') ;
2470    
2471    Totally extraordinary, but if that's what it takes...
2472    
2473    5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2474    available. Added HAVE_BCOPY and an autoconf test for it; if neither
2475    HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2476    assumes the way PCRE uses memmove() (always moving upwards).
2477    
2478    6. PCRE is being used in one environment where strchr() is not available. There
2479    was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2480    faster code anyway.
2481    
2482    
2483    Version 3.1 09-Feb-00
2484    ---------------------
2485    
2486    The only change in this release is the fixing of some bugs in Makefile.in for
2487    the "install" target:
2488    
2489    (1) It was failing to install pcreposix.h.
2490    
2491    (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2492    
2493    
2494    Version 3.0 01-Feb-00
2495    ---------------------
2496    
2497    1. Add support for the /+ modifier to perltest (to output $` like it does in
2498    pcretest).
2499    
2500    2. Add support for the /g modifier to perltest.
2501    
2502    3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2503    matches null strings.
2504    
2505    4. Fix perltest so that it doesn't do unwanted things when fed an empty
2506    pattern. Perl treats empty patterns specially - it reuses the most recent
2507    pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2508    effect.
2509    
2510    5. The POSIX interface was broken in that it was just handing over the POSIX
2511    captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2512    required a bigger vector, with some working space on the end. This means that
2513    the POSIX wrapper now has to get and free some memory, and copy the results.
2514    
2515    6. Added some simple autoconf support, placing the test data and the
2516    documentation in separate directories, re-organizing some of the
2517    information files, and making it build pcre-config (a GNU standard). Also added
2518    libtool support for building PCRE as a shared library, which is now the
2519    default.
2520    
2521    7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2522    09 are not valid octal constants. Single digits will be used for minor values
2523    less than 10.
2524    
2525    8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2526    existing programs that set these in the POSIX interface can use PCRE without
2527    modification.
2528    
2529    9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2530    return all that pcre_info() returns, plus additional data. The pcre_info()
2531    function is retained for compatibility, but is considered to be obsolete.
2532    
2533    10. Added experimental recursion feature (?R) to handle one common case that
2534    Perl 5.6 will be able to do with (?p{...}).
2535    
2536    11. Added support for POSIX character classes like [:alpha:], which Perl is
2537    adopting.
2538    
2539    
2540    Version 2.08 31-Aug-99
2541    ----------------------
2542    
2543    1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2544    trying to match at the startoffset position, but instead was moving forward to
2545    the next newline as if a previous match had failed.
2546    
2547    2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2548    and could get into a loop if a null string was matched other than at the start
2549    of the subject.
2550    
2551    3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2552    be distinguished at compile time, and for completeness also added PCRE_DATE.
2553    
2554    5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2555    in GnuWin32 environments.
2556    
2557    
2558    Version 2.07 29-Jul-99
2559    ----------------------
2560    
2561    1. The documentation is now supplied in plain text form and HTML as well as in
2562    the form of man page sources.
2563    
2564    2. C++ compilers don't like assigning (void *) values to other pointer types.
2565    In particular this affects malloc(). Although there is no problem in Standard
2566    C, I've put in casts to keep C++ compilers happy.
2567    
2568    3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2569    should be (const char *).
2570    
2571    4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2572    be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2573    However, I haven't made this a standard facility. The documentation doesn't
2574    mention it, and the Makefile doesn't support it.
2575    
2576    5. The Makefile now contains an "install" target, with editable destinations at
2577    the top of the file. The pcretest program is not installed.
2578    
2579    6. pgrep -V now gives the PCRE version number and date.
2580    
2581    7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2582    causing the entire string to be ignored, instead of just the last character.
2583    
2584    8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2585    non-matching string, it can take a very, very long time, even for strings of
2586    quite modest length, because of the nested recursion. PCRE now does better in
2587    some of these cases. It does this by remembering the last required literal
2588    character in the pattern, and pre-searching the subject to ensure it is present
2589    before running the real match. In other words, it applies a heuristic to detect
2590    some types of certain failure quickly, and in the above example, if presented
2591    with a string that has no trailing " it gives "no match" very quickly.
2592    
2593    9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2594    other alternatives are tried instead.
2595    
2596    
2597    Version 2.06 09-Jun-99
2598    ----------------------
2599    
2600    1. Change pcretest's output for amount of store used to show just the code
2601    space, because the remainder (the data block) varies in size between 32-bit and
2602    64-bit systems.
2603    
2604    2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2605    start matching at. This allows lookbehinds to work when searching for multiple
2606    occurrences in a string.
2607    
2608    3. Added additional options to pcretest for testing multiple occurrences:
2609    
2610       /+   outputs the rest of the string that follows a match
2611       /g   loops for multiple occurrences, using the new startoffset argument
2612       /G   loops for multiple occurrences by passing an incremented pointer
2613    
2614    4. PCRE wasn't doing the "first character" optimization for patterns starting
2615    with \b or \B, though it was doing it for other lookbehind assertions. That is,
2616    it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2617    the letter 'x'. On long subject strings, this gives a significant speed-up.
2618    
2619    
2620    Version 2.05 21-Apr-99
2621    ----------------------
2622    
2623    1. Changed the type of magic_number from int to long int so that it works
2624    properly on 16-bit systems.
2625    
2626    2. Fixed a bug which caused patterns starting with .* not to work correctly
2627    when the subject string contained newline characters. PCRE was assuming
2628    anchoring for such patterns in all cases, which is not correct because .* will
2629    not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2630    DOTALL is set at top level; otherwise it knows that patterns starting with .*
2631    must be retried after every newline in the subject.
2632    
2633    
2634    Version 2.04 18-Feb-99
2635    ----------------------
2636    
2637    1. For parenthesized subpatterns with repeats whose minimum was zero, the
2638    computation of the store needed to hold the pattern was incorrect (too large).
2639    If such patterns were nested a few deep, this could multiply and become a real
2640    problem.
2641    
2642    2. Added /M option to pcretest to show the memory requirement of a specific
2643    pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2644    
2645    3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2646    compiled in such a way that the backtracking after subsequent failure was
2647    pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2648    ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2649    
2650    
2651    Version 2.03 02-Feb-99
2652    ----------------------
2653    
2654    1. Fixed typo and small mistake in man page.
2655    
2656    2. Added 4th condition (GPL supersedes if conflict) and created separate
2657    LICENCE file containing the conditions.
2658    
2659    3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2660    Perl, that is the internal \ allows the delimiter to be included in the
2661    pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2662    the final delimiter, add \ to the end of the pattern (to test the error).
2663    
2664    4. Added the convenience functions for extracting substrings after a successful
2665    match. Updated pcretest to make it able to test these functions.
2666    
2667    
2668    Version 2.02 14-Jan-99
2669    ----------------------
2670    
2671    1. Initialized the working variables associated with each extraction so that
2672    their saving and restoring doesn't refer to uninitialized store.
2673    
2674    2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2675    compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2676    fix the problem.
2677    
2678    3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2679    calls, and wasn't printing the correct value for compiling calls. Increased the
2680    default value of LOOPREPEAT, and the number of significant figures in the
2681    times.
2682    
2683    4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2684    
2685    5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2686    a building problem on Windows NT with a FAT file system.
2687    
2688    
2689    Version 2.01 21-Oct-98
2690    ----------------------
2691    
2692    1. Changed the API for pcre_compile() to allow for the provision of a pointer
2693    to character tables built by pcre_maketables() in the current locale. If NULL
2694    is passed, the default tables are used.
2695    
2696    
2697    Version 2.00 24-Sep-98
2698    ----------------------
2699    
2700    1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2701    it any more.
2702    
2703    2. Allow quantification of (?>) groups, and make it work correctly.
2704    
2705    3. The first character computation wasn't working for (?>) groups.
2706    
2707    4. Correct the implementation of \Z (it is permitted to match on the \n at the
2708    end of the subject) and add 5.005's \z, which really does match only at the
2709    very end of the subject.
2710    
2711    5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2712    
2713    6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2714    DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2715    localized options. All options to pcre_study() were also removed.
2716    
2717    7. Add other new features from 5.005:
2718    
2719       $(?<=           positive lookbehind
2720       $(?<!           negative lookbehind
2721       (?imsx-imsx)    added the unsetting capability
2722                       such a setting is global if at outer level; local otherwise
2723       (?imsx-imsx:)   non-capturing groups with option setting
2724       (?(cond)re|re)  conditional pattern matching
2725    
2726       A backreference to itself in a repeated group matches the previous
2727       captured string.
2728    
2729    8. General tidying up of studying (both automatic and via "study")
2730    consequential on the addition of new assertions.
2731    
2732    9. As in 5.005, unlimited repeated groups that could match an empty substring
2733    are no longer faulted at compile time. Instead, the loop is forcibly broken at
2734    runtime if any iteration does actually match an empty substring.
2735    
2736    10. Include the RunTest script in the distribution.
2737    
2738    11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2739    discrepancies, some of which were old and were also with respect to 5.004. They
2740    have now been fixed.
2741    
2742    
2743    Version 1.09 28-Apr-98
2744    ----------------------
2745    
2746    1. A negated single character class followed by a quantifier with a minimum
2747    value of one (e.g.  [^x]{1,6}  ) was not compiled correctly. This could lead to
2748    program crashes, or just wrong answers. This did not apply to negated classes
2749    containing more than one character, or to minima other than one.
2750    
2751    
2752    Version 1.08 27-Mar-98
2753    ----------------------
2754    
2755    1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2756    
2757    2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2758    latter must appear before anything that relies on it in the pattern.
2759    
2760    
2761    Version 1.07 16-Feb-98
2762    ----------------------
2763    
2764    1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2765    repeat of a potentially empty string).
2766    
2767    
2768    Version 1.06 23-Jan-98
2769    ----------------------
2770    
2771    1. Added Markus Oberhumer's little patches for C++.
2772    
2773    2. Literal strings longer than 255 characters were broken.
2774    
2775    
2776  Version 1.05 23-Dec-97  Version 1.05 23-Dec-97
2777  ----------------------  ----------------------

Legend:
Removed from v.13  
changed lines
  Added in v.274

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12