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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.35  
changed lines
  Added in v.267

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12