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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.25  
changed lines
  Added in v.230

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12