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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.59  
changed lines
  Added in v.270

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12