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

Diff of /code/trunk/ChangeLog

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

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

Legend:
Removed from v.31  
changed lines
  Added in v.240

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12