ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 169 - (hide annotations) (download)
Mon Jun 4 10:49:21 2007 UTC (7 years, 10 months ago) by ph10
File size: 128451 byte(s)

1 nigel 3 ChangeLog for PCRE
2     ------------------
4 ph10 157 Version 7.2 01-May-07
5     ---------------------
7     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
8     which is apparently normally available under Windows.
9 ph10 159
10     2. Re-jig the pcregrep tests with different newline settings in an attempt
11     to make them independent of the local environment's newline setting.
12 ph10 157
13 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
14 ph10 165
15 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
16     was not 2, because the output contained actual offsets. The recent new
17 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
18     usable with all link sizes.
20 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
21     stack recursion. This gives a massive performance boost under BSD, but just
22 ph10 165 a small improvement under Linux. However, it saves one field in the frame
23 ph10 164 in all cases.
24 ph10 166
25     6. Added more features from the forthcoming Perl 5.10:
27     (a) (?-n) (where n is a string of digits) is a relative subroutine or
28     recursion call. It refers to the nth most recently opened parentheses.
30     (b) (?+n) is also a relative subroutine call; it refers to the nth next
31     to be opened parentheses.
32 ph10 167
33     (c) Conditions that refer to capturing parentheses can be specified
34     relatively, for example, (?(-2)... or (?(+3)...
35 ph10 168
36     (d) \K resets the start of the current match so that everything before
37     is not part of it.
38 ph10 169
39     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
41 ph10 159
42 ph10 160
43 ph10 155 Version 7.1 24-Apr-07
44 ph10 98 ---------------------
46 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
47 ph10 122 that is more "standard", making use of automake and other Autotools. There
48 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
49     on this.
50 ph10 111
51     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
52     for recursive directory scanning broke on some systems because the files
53     are not scanned in any specific order and on different systems the order
54     was different. A call to "sort" has been inserted into RunGrepTest for the
55     approprate test as a short-term fix. In the longer term there may be an
56 ph10 100 alternative.
57 ph10 111
58 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
59 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
60     people do their own conversions for various reasons). The problems
61     concerned the use of low-level troff macros .br and .in. I have therefore
62     removed all such uses from the man pages (some were redundant, some could
63 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
64     HTML has been updated to handle .nf/.fi and to complain if it encounters
65     .br or .in.
66 ph10 111
67 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
68 ph10 123 arranged for config.h to be included in the distribution, with the name
69 ph10 111 config.h.generic, for the benefit of those who have to compile without
70     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
72     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
73     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
74 ph10 127 makevp.bat for the current PCRE, using the additional files
75 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
76 ph10 111
77     6. A Windows user reported a minor discrepancy with test 2, which turned out
78     to be caused by a trailing space on an input line that had got lost in his
79 ph10 102 copy. The trailing space was an accident, so I've just removed it.
80 ph10 111
81 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
82 ph10 111 that is needed.
84 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
85     as "const" (a) because they are and (b) because it helps the PHP
86     maintainers who have recently made a script to detect big data structures
87 ph10 111 in the php code that should be moved to the .rodata section. I remembered
88     to update Builducptable as well, so it won't revert if ucptable.h is ever
89 ph10 105 re-created.
90 ph10 111
91     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
92 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
93 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
94     support is not required. This saves 1.5-2K of code, which is important in
95 ph10 107 some applications.
96 ph10 111
97 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
98 ph10 111 so as not to refer to the tables, even though these functions will never be
99     called when UTF-8 support is disabled. Otherwise there are problems with a
100     shared library.
102 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
104     (a) It was defining its arguments as char * instead of void *.
106 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
107     a long time ago when I wrote it, but is no longer the case.
109 ph10 118 The emulated memove() is provided for those environments that have neither
110 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
111 ph10 118 is clearly not the case, as these two bugs were recently reported.
112 ph10 123
113 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
114 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
115     pages, and it removes trailing spaces from listed files. It also creates
116     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
117     case, it wraps all the #defines with #ifndefs. This script should be run
118 ph10 111 before "make dist".
119 ph10 123
120 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
121     with Unicode property support.
122 ph10 123
123     (a) For a maximizing quantifier, if the two different cases of the
124     character were of different lengths in their UTF-8 codings (there are
125     some cases like this - I found 11), and the matching function had to
126 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
127     were both the same length.
128 ph10 123
129     (b) When PCRE was configured to use the heap rather than the stack for
130     recursion during matching, it was not correctly preserving the data for
131     the other case of a UTF-8 character when checking ahead for a match
132     while processing a minimizing repeat. If the check also involved
133 ph10 115 matching a wide character, but failed, corruption could cause an
134     erroneous result when trying to check for a repeat of the original
135     character.
136 ph10 123
137 ph10 116 13. Some tidying changes to the testing mechanism:
138 ph10 98
139 ph10 116 (a) The RunTest script now detects the internal link size and whether there
140     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
141 ph10 123 values substituted by "configure". (The RunGrepTest script already did
142     this for UTF-8.) The configure.ac script no longer substitutes the
143     relevant variables.
145 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
146     with length and offset values. This means that the output is different
147     for different internal link sizes. Test 2 is skipped for link sizes
148     other than 2 because of this, bypassing the problem. Unfortunately,
149     there was also a test in test 3 (the locale tests) that used /B and
150 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
151     I have added a new /Z option to pcretest that replaces the length and
152     offset values with spaces. This is now used to make test 3 independent
153 ph10 122 of link size. (Test 2 will be tidied up later.)
154 ph10 123
155     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
156 ph10 122 segmentation fault instead of returning the appropriate error message.
157 ph10 134
158 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
159 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
160     This doesn't seem right; it now treats the CRLF combination as the line
161     ending, and so does not match in that case. It's only a pattern such as ^$
162     that would hit this one: something like ^ABC$ would have failed after \r
163     and then tried again after \r\n.
165 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
166 ph10 134 in an attempt to make files that differ only in their line terminators
167     compare equal. This works on Linux.
168 ph10 142
169 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
170     as it exited. This is now fixed, thanks to valgrind.
171 ph10 142
172 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
173 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
174     was because its rules for how to advance for /g after matching an empty
175 ph10 143 string at the end of a line did not allow for this case. They now check for
176     it specially.
177 ph10 150
178     20. pcretest is supposed to handle patterns and data of any length, by
179     extending its buffers when necessary. It was getting this wrong when the
180 ph10 147 buffer for a data line had to be extended.
181 ph10 150
182 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
183 ph10 150 CRLF as a newline sequence.
184 ph10 152
185 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
186 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
187     I have nevertheless tidied it up.
188 ph10 111
189 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
190 ph10 123
191 ph10 153 24. Added a man page for pcre-config.
192 ph10 152
193 ph10 153
194 nigel 93 Version 7.0 19-Dec-06
195     ---------------------
197     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
198     moving to gcc 4.1.1.
200     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
201     sys/time.h, which is documented as needed for this function. It doesn't
202     seem to matter on Linux, but it showed up on some releases of OS X.
204     3. It seems that there are systems where bytes whose values are greater than
205     127 match isprint() in the "C" locale. The "C" locale should be the
206     default when a C program starts up. In most systems, only ASCII printing
207     characters match isprint(). This difference caused the output from pcretest
208     to vary, making some of the tests fail. I have changed pcretest so that:
210     (a) When it is outputting text in the compiled version of a pattern, bytes
211     other than 32-126 are always shown as hex escapes.
213     (b) When it is outputting text that is a matched part of a subject string,
214     it does the same, unless a different locale has been set for the match
215     (using the /L modifier). In this case, it uses isprint() to decide.
217     4. Fixed a major bug that caused incorrect computation of the amount of memory
218     required for a compiled pattern when options that changed within the
219     pattern affected the logic of the preliminary scan that determines the
220     length. The relevant options are -x, and -i in UTF-8 mode. The result was
221     that the computed length was too small. The symptoms of this bug were
222     either the PCRE error "internal error: code overflow" from pcre_compile(),
223     or a glibc crash with a message such as "pcretest: free(): invalid next
224     size (fast)". Examples of patterns that provoked this bug (shown in
225     pcretest format) are:
227     /(?-x: )/x
228     /(?x)(?-x: \s*#\s*)/
229     /((?i)[\x{c0}])/8
230     /(?i:[\x{c0}])/8
232     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
233     is now done differently.
235     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
236     wrapper classes; (b) implement a new function in the C++ scanner that is
237     more efficient than the old way of doing things because it avoids levels of
238     recursion in the regex matching; (c) add a paragraph to the documentation
239     for the FullMatch() function.
241     6. The escape sequence \n was being treated as whatever was defined as
242     "newline". Not only was this contrary to the documentation, which states
243     that \n is character 10 (hex 0A), but it also went horribly wrong when
244     "newline" was defined as CRLF. This has been fixed.
246     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
247     was being set to -1 for the "end of line" case (supposedly a value that no
248     character can have). Though this value is never used (the check for end of
249     line is "zero bytes in current character"), it caused compiler complaints.
250     I've changed it to 0xffffffff.
252     8. In pcre_version.c, the version string was being built by a sequence of
253     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
254     string (as it is for production releases) called a macro with an empty
255     argument. The C standard says the result of this is undefined. The gcc
256     compiler treats it as an empty string (which was what was wanted) but it is
257     reported that Visual C gives an error. The source has been hacked around to
258     avoid this problem.
260     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
261     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
262     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
263     of them did).
265     10. Originally, pcretest opened its input and output without "b"; then I was
266     told that "b" was needed in some environments, so it was added for release
267     5.0 to both the input and output. (It makes no difference on Unix-like
268     systems.) Later I was told that it is wrong for the input on Windows. I've
269     now abstracted the modes into two macros, to make it easier to fiddle with
270     them, and removed "b" from the input mode under Windows.
272     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
274     12. Added -help and --help to pcretest as an official way of being reminded
275     of the options.
277     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
278     and pcrecpp.cc because they annoy compilers at high warning levels.
280     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
282     15. Fixed an occurrence of == in configure.ac that should have been = (shell
283     scripts are not C programs :-) and which was not noticed because it works
284     on Linux.
286     16. pcretest is supposed to handle any length of pattern and data line (as one
287     line or as a continued sequence of lines) by extending its input buffer if
288     necessary. This feature was broken for very long pattern lines, leading to
289     a string of junk being passed to pcre_compile() if the pattern was longer
290     than about 50K.
292     17. I have done a major re-factoring of the way pcre_compile() computes the
293     amount of memory needed for a compiled pattern. Previously, there was code
294     that made a preliminary scan of the pattern in order to do this. That was
295     OK when PCRE was new, but as the facilities have expanded, it has become
296     harder and harder to keep it in step with the real compile phase, and there
297     have been a number of bugs (see for example, 4 above). I have now found a
298     cunning way of running the real compile function in a "fake" mode that
299     enables it to compute how much memory it would need, while actually only
300     ever using a few hundred bytes of working memory and without too many
301     tests of the mode. This should make future maintenance and development
302     easier. A side effect of this work is that the limit of 200 on the nesting
303     depth of parentheses has been removed (though this was never a serious
304     limitation, I suspect). However, there is a downside: pcre_compile() now
305     runs more slowly than before (30% or more, depending on the pattern). I
306     hope this isn't a big issue. There is no effect on runtime performance.
308     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
309     newline (only possible for the last line of a file) and it was a
310     pattern that set a locale (followed by /Lsomething), pcretest crashed.
312     19. Added additional timing features to pcretest. (1) The -tm option now times
313     matching only, not compiling. (2) Both -t and -tm can be followed, as a
314     separate command line item, by a number that specifies the number of
315     repeats to use when timing. The default is 50000; this gives better
316     precision, but takes uncomfortably long for very large patterns.
318     20. Extended pcre_study() to be more clever in cases where a branch of a
319     subpattern has no definite first character. For example, (a*|b*)[cd] would
320     previously give no result from pcre_study(). Now it recognizes that the
321     first character must be a, b, c, or d.
323     21. There was an incorrect error "recursive call could loop indefinitely" if
324     a subpattern (or the entire pattern) that was being tested for matching an
325     empty string contained only one non-empty item after a nested subpattern.
326     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
327     incorrectly, because the \d was being skipped in the check.
329     22. The pcretest program now has a new pattern option /B and a command line
330     option -b, which is equivalent to adding /B to every pattern. This causes
331     it to show the compiled bytecode, without the additional information that
332     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
333     is the same as /B/I).
335     23. A new optimization is now able automatically to treat some sequences such
336     as a*b as a*+b. More specifically, if something simple (such as a character
337     or a simple class like \d) has an unlimited quantifier, and is followed by
338     something that cannot possibly match the quantified thing, the quantifier
339     is automatically "possessified".
341     24. A recursive reference to a subpattern whose number was greater than 39
342     went wrong under certain circumstances in UTF-8 mode. This bug could also
343     have affected the operation of pcre_study().
345     25. Realized that a little bit of performance could be had by replacing
346     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
348     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
350     27. Possessive quantifiers such as a++ were previously implemented by turning
351     them into atomic groups such as ($>a+). Now they have their own opcodes,
352     which improves performance. This includes the automatically created ones
353     from 23 above.
355     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
356     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
357     the first matched character to be a colon. This applied both to named and
358     numbered groups.
360     29. The ucpinternal.h header file was missing its idempotency #ifdef.
362     30. I was sent a "project" file called libpcre.a.dev which I understand makes
363     building PCRE on Windows easier, so I have included it in the distribution.
365     31. There is now a check in pcretest against a ridiculously large number being
366     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
367     loop, the loop is abandoned.
369     32. Forward references to subpatterns in conditions such as (?(2)...) where
370     subpattern 2 is defined later cause pcre_compile() to search forwards in
371     the pattern for the relevant set of parentheses. This search went wrong
372     when there were unescaped parentheses in a character class, parentheses
373     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
375     33. "Subroutine" calls and backreferences were previously restricted to
376     referencing subpatterns earlier in the regex. This restriction has now
377     been removed.
379     34. Added a number of extra features that are going to be in Perl 5.10. On the
380     whole, these are just syntactic alternatives for features that PCRE had
381     previously implemented using the Python syntax or my own invention. The
382     other formats are all retained for compatibility.
384     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
385     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
386     also .NET compatible.
388     (b) A recursion or subroutine call to a named group can now be defined as
389     (?&name) as well as (?P>name).
391     (c) A backreference to a named group can now be defined as \k<name> or
392     \k'name' as well as (?P=name). The new forms, as well as being in Perl
393     5.10, are also .NET compatible.
395     (d) A conditional reference to a named group can now use the syntax
396     (?(<name>) or (?('name') as well as (?(name).
398     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
399     groups (named and numbered) that are never evaluated inline, but can be
400     called as "subroutines" from elsewhere. In effect, the DEFINE condition
401     is always false. There may be only one alternative in such a group.
403     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
404     as the simple (?(R). The condition is true only if the most recent
405     recursion is that of the given number or name. It does not search out
406     through the entire recursion stack.
408     (g) The escape \gN or \g{N} has been added, where N is a positive or
409     negative number, specifying an absolute or relative reference.
411     35. Tidied to get rid of some further signed/unsigned compiler warnings and
412     some "unreachable code" warnings.
414     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
415     things, this adds five new scripts.
417     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
418     There were also incompatibilities regarding the handling of \Q..\E inside
419     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
420     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
422     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
423     matches an empty string, and forcibly breaks the loop. There were bugs in
424     this code in non-simple cases. For a pattern such as ^(a()*)* matched
425     against aaaa the result was just "a" rather than "aaaa", for example. Two
426     separate and independent bugs (that affected different cases) have been
427     fixed.
429     39. Refactored the code to abolish the use of different opcodes for small
430     capturing bracket numbers. This is a tidy that I avoided doing when I
431     removed the limit on the number of capturing brackets for 3.5 back in 2001.
432     The new approach is not only tidier, it makes it possible to reduce the
433     memory needed to fix the previous bug (38).
435     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
436     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
437     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
438     mode.
440     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
441     report.
443     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
444     copy construction and assignment in the C++ wrapper.
446     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
447     couple of bugs that could have given wrong results in the "--newline=crlf"
448     case.
450     44. Added a number of casts and did some reorganization of signed/unsigned int
451     variables following suggestions from Dair Grant. Also renamed the variable
452     "this" as "item" because it is a C++ keyword.
454     45. Arranged for dftables to add
456     #include "pcre_internal.h"
458     to pcre_chartables.c because without it, gcc 4.x may remove the array
459     definition from the final binary if PCRE is built into a static library and
460     dead code stripping is activated.
462     46. For an unanchored pattern, if a match attempt fails at the start of a
463     newline sequence, and the newline setting is CRLF or ANY, and the next two
464     characters are CRLF, advance by two characters instead of one.
467 nigel 91 Version 6.7 04-Jul-06
468     ---------------------
470     1. In order to handle tests when input lines are enormously long, pcretest has
471     been re-factored so that it automatically extends its buffers when
472     necessary. The code is crude, but this _is_ just a test program. The
473     default size has been increased from 32K to 50K.
475     2. The code in pcre_study() was using the value of the re argument before
476     testing it for NULL. (Of course, in any sensible call of the function, it
477     won't be NULL.)
479     3. The memmove() emulation function in pcre_internal.h, which is used on
480     systems that lack both memmove() and bcopy() - that is, hardly ever -
481     was missing a "static" storage class specifier.
483     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
484     containing an extended class (one that cannot be represented by a bitmap
485     because it contains high-valued characters or Unicode property items, e.g.
486     [\pZ]). Almost always one would set UTF-8 mode when processing such a
487     pattern, but PCRE should not loop if you do not (it no longer does).
488     [Detail: two cases were found: (a) a repeated subpattern containing an
489     extended class; (b) a recursive reference to a subpattern that followed a
490     previous extended class. It wasn't skipping over the extended class
491     correctly when UTF-8 mode was not set.]
493     5. A negated single-character class was not being recognized as fixed-length
494     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
495     compile error "lookbehind assertion is not fixed length".
497     6. The RunPerlTest auxiliary script was showing an unexpected difference
498     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
499     write a Perl script that can interpret lines of an input file either as
500     byte characters or as UTF-8, which is what "perltest" was being required to
501     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
502     can't do is switch easily at run time between having the "use utf8;" pragma
503     or not. In the end, I fudged it by using the RunPerlTest script to insert
504     "use utf8;" explicitly for the UTF-8 tests.
506     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
507     the end of the subject string, contrary to the documentation and to what
508     Perl does. This was true of both matching functions. Now it matches only at
509     the start of the subject and immediately after *internal* newlines.
511     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
512     a pointer to an int instead of a pointer to an unsigned long int. This
513     caused problems on 64-bit systems.
515     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
516     instance of the 'standard' template library not being so standard".
518     10. There was no check on the number of named subpatterns nor the maximum
519     length of a subpattern name. The product of these values is used to compute
520     the size of the memory block for a compiled pattern. By supplying a very
521     long subpattern name and a large number of named subpatterns, the size
522     computation could be caused to overflow. This is now prevented by limiting
523     the length of names to 32 characters, and the number of named subpatterns
524     to 10,000.
526     11. Subpatterns that are repeated with specific counts have to be replicated in
527     the compiled pattern. The size of memory for this was computed from the
528     length of the subpattern and the repeat count. The latter is limited to
529     65535, but there was no limit on the former, meaning that integer overflow
530     could in principle occur. The compiled length of a repeated subpattern is
531     now limited to 30,000 bytes in order to prevent this.
533     12. Added the optional facility to have named substrings with the same name.
535     13. Added the ability to use a named substring as a condition, using the
536     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
537     are numbers (not recommended). Forward references are permitted.
539     14. Added forward references in named backreferences (if you see what I mean).
541     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
542     pattern could run off the end of the subject. For example, the pattern
543     "(?s)(.{1,5})"8 did this with the subject "ab".
545     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
546     PCRE_CASELESS was set when matching characters that were quantified with ?
547     or *.
549     17. A character class other than a single negated character that had a minimum
550     but no maximum quantifier - for example [ab]{6,} - was not handled
551     correctly by pce_dfa_exec(). It would match only one character.
553     18. A valid (though odd) pattern that looked like a POSIX character
554     class but used an invalid character after [ (for example [[,abc,]]) caused
555     pcre_compile() to give the error "Failed: internal error: code overflow" or
556     in some cases to crash with a glibc free() error. This could even happen if
557     the pattern terminated after [[ but there just happened to be a sequence of
558     letters, a binary zero, and a closing ] in the memory that followed.
560     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
561     over the years. Originally (before any Unicode support), just the bottom 8
562     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
563     output from "man perlunicode" includes this:
565     The regular expression compiler produces polymorphic opcodes. That
566     is, the pattern adapts to the data and automatically switches to
567     the Unicode character scheme when presented with Unicode data--or
568     instead uses a traditional byte scheme when presented with byte
569     data.
571     Sadly, a wide octal escape does not cause a switch, and in a string with
572     no other multibyte characters, these octal escapes are treated as before.
573     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
574     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
575     Unicode string.
577     I have not perpetrated such confusion in PCRE. Up till now, it took just
578     the bottom 8 bits, as in old Perl. I have now made octal escapes with
579     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
580     translate to the appropriate multibyte character.
582     29. Applied some refactoring to reduce the number of warnings from Microsoft
583     and Borland compilers. This has included removing the fudge introduced
584     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
585     a warning about an unused variable.
587     21. PCRE has not included VT (character 0x0b) in the set of whitespace
588     characters since release 4.0, because Perl (from release 5.004) does not.
589     [Or at least, is documented not to: some releases seem to be in conflict
590     with the documentation.] However, when a pattern was studied with
591     pcre_study() and all its branches started with \s, PCRE still included VT
592     as a possible starting character. Of course, this did no harm; it just
593     caused an unnecessary match attempt.
595     22. Removed a now-redundant internal flag bit that recorded the fact that case
596     dependency changed within the pattern. This was once needed for "required
597     byte" processing, but is no longer used. This recovers a now-scarce options
598     bit. Also moved the least significant internal flag bit to the most-
599     significant bit of the word, which was not previously used (hangover from
600     the days when it was an int rather than a uint) to free up another bit for
601     the future.
603     23. Added support for CRLF line endings as well as CR and LF. As well as the
604     default being selectable at build time, it can now be changed at runtime
605     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
606     specify that it is scanning data with non-default line endings.
608     24. Changed the definition of CXXLINK to make it agree with the definition of
609     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
611     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
612     recursions. This makes a big different to stack usage for some patterns.
614     26. If a subpattern containing a named recursion or subroutine reference such
615     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
616     the space required for the compiled pattern went wrong and gave too small a
617     value. Depending on the environment, this could lead to "Failed: internal
618     error: code overflow at offset 49" or "glibc detected double free or
619     corruption" errors.
621     27. Applied patches from Google (a) to support the new newline modes and (b) to
622     advance over multibyte UTF-8 characters in GlobalReplace.
624     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
625     difference for some implementation of PCRE in some Windows version.
627     29. Added some extra testing facilities to pcretest:
629     \q<number> in a data line sets the "match limit" value
630     \Q<number> in a data line sets the "match recursion limt" value
631     -S <number> sets the stack size, where <number> is in megabytes
633     The -S option isn't available for Windows.
636 nigel 89 Version 6.6 06-Feb-06
637     ---------------------
639     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
640     in pcreposix.h. I have copied the definition from pcre.h.
642     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
643     because pcre.h is no longer a built file.
645     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
646     not normally included in the compiled code.
649 nigel 87 Version 6.5 01-Feb-06
650     ---------------------
652     1. When using the partial match feature with pcre_dfa_exec(), it was not
653     anchoring the second and subsequent partial matches at the new starting
654     point. This could lead to incorrect results. For example, with the pattern
655     /1234/, partially matching against "123" and then "a4" gave a match.
657     2. Changes to pcregrep:
659     (a) All non-match returns from pcre_exec() were being treated as failures
660     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
661     error message is output. Some extra information is given for the
663     probably the only errors that are likely to be caused by users (by
664     specifying a regex that has nested indefinite repeats, for instance).
665     If there are more than 20 of these errors, pcregrep is abandoned.
667     (b) A binary zero was treated as data while matching, but terminated the
668     output line if it was written out. This has been fixed: binary zeroes
669     are now no different to any other data bytes.
671     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
672     used to set a locale for matching. The --locale=xxxx long option has
673     been added (no short equivalent) to specify a locale explicitly on the
674     pcregrep command, overriding the environment variables.
676     (d) When -B was used with -n, some line numbers in the output were one less
677     than they should have been.
679     (e) Added the -o (--only-matching) option.
681     (f) If -A or -C was used with -c (count only), some lines of context were
682     accidentally printed for the final match.
684     (g) Added the -H (--with-filename) option.
686     (h) The combination of options -rh failed to suppress file names for files
687     that were found from directory arguments.
689     (i) Added the -D (--devices) and -d (--directories) options.
691     (j) Added the -F (--fixed-strings) option.
693     (k) Allow "-" to be used as a file name for -f as well as for a data file.
695     (l) Added the --colo(u)r option.
697     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
698     is not present by default.
700     3. A nasty bug was discovered in the handling of recursive patterns, that is,
701     items such as (?R) or (?1), when the recursion could match a number of
702     alternatives. If it matched one of the alternatives, but subsequently,
703     outside the recursion, there was a failure, the code tried to back up into
704     the recursion. However, because of the way PCRE is implemented, this is not
705     possible, and the result was an incorrect result from the match.
707     In order to prevent this happening, the specification of recursion has
708     been changed so that all such subpatterns are automatically treated as
709     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
711     4. I had overlooked the fact that, in some locales, there are characters for
712     which isalpha() is true but neither isupper() nor islower() are true. In
713     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
714     and ordfeminine) are like this. This affected the treatment of \w and \W
715     when they appeared in character classes, but not when they appeared outside
716     a character class. The bit map for "word" characters is now created
717     separately from the results of isalnum() instead of just taking it from the
718     upper, lower, and digit maps. (Plus the underscore character, of course.)
720     5. The above bug also affected the handling of POSIX character classes such as
721     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
722     permanent tables. Instead, the bit maps for such a class were previously
723     created as the appropriate unions of the upper, lower, and digit bitmaps.
724     Now they are created by subtraction from the [[:word:]] class, which has
725     its own bitmap.
727     6. The [[:blank:]] character class matches horizontal, but not vertical space.
728     It is created by subtracting the vertical space characters (\x09, \x0a,
729     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
730     subtraction was done in the overall bitmap for a character class, meaning
731     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
732     be recognized. This bug has been fixed.
734     7. Patches from the folks at Google:
736     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
737     real life, but is still worth protecting against".
739     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
740     regular expressions".
742     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
743     have it.
745     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
746     "configure" and the latter not, in order to fix a problem somebody had
747     with compiling the Arg class on HP-UX.
749     (e) Improve the error-handling of the C++ wrapper a little bit.
751     (f) New tests for checking recursion limiting.
753     8. The pcre_memmove() function, which is used only if the environment does not
754     have a standard memmove() function (and is therefore rarely compiled),
755     contained two bugs: (a) use of int instead of size_t, and (b) it was not
756     returning a result (though PCRE never actually uses the result).
758     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
759     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
760     returned instead of calling malloc() with an overflowing number that would
761     most likely cause subsequent chaos.
763     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
765     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
766     with this option is matched, the nmatch and pmatch options of regexec() are
767     ignored.
769     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
770     provided in case anyone wants to the the POSIX interface with UTF-8
771     strings.
773     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
774     C++ linking (needed for some HP-UX environments).
776     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
777     (unused parameter) and in the pcre_printint() function (omitted "default"
778     switch label when the default is to do nothing).
780     15. Added some code to make it possible, when PCRE is compiled as a C++
781     library, to replace subject pointers for pcre_exec() with a smart pointer
782     class, thus making it possible to process discontinuous strings.
784     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
785     much the same function. They were added by different people who were trying
786     to make PCRE easy to compile on non-Unix systems. It has been suggested
787     that PCRE_EXPORT be abolished now that there is more automatic apparatus
788     for compiling on Windows systems. I have therefore replaced it with
789     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
790     defaults to "extern" for C or "extern C" for C++, which works fine on
791     Unix-like systems. It is now possible to override the value of PCRE_DATA_
792     SCOPE with something explicit in config.h. In addition:
794     (a) pcreposix.h still had just "extern" instead of either of these macros;
795     I have replaced it with PCRE_DATA_SCOPE.
797     (b) Functions such as _pcre_xclass(), which are internal to the library,
798     but external in the C sense, all had PCRE_EXPORT in their definitions.
799     This is apparently wrong for the Windows case, so I have removed it.
800     (It makes no difference on Unix-like systems.)
802     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
803     of recursive calls to match(). This is different to MATCH_LIMIT because
804     that limits the total number of calls to match(), not all of which increase
805     the depth of recursion. Limiting the recursion depth limits the amount of
806     stack (or heap if NO_RECURSE is set) that is used. The default can be set
807     when PCRE is compiled, and changed at run time. A patch from Google adds
808     this functionality to the C++ interface.
810     18. Changes to the handling of Unicode character properties:
812     (a) Updated the table to Unicode 4.1.0.
814     (b) Recognize characters that are not in the table as "Cn" (undefined).
816     (c) I revised the way the table is implemented to a much improved format
817     which includes recognition of ranges. It now supports the ranges that
818     are defined in UnicodeData.txt, and it also amalgamates other
819     characters into ranges. This has reduced the number of entries in the
820     table from around 16,000 to around 3,000, thus reducing its size
821     considerably. I realized I did not need to use a tree structure after
822     all - a binary chop search is just as efficient. Having reduced the
823     number of entries, I extended their size from 6 bytes to 8 bytes to
824     allow for more data.
826     (d) Added support for Unicode script names via properties such as \p{Han}.
828     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
829     matching that character.
831     20. When matching a repeated Unicode property with a minimum greater than zero,
832     (for example \pL{2,}), PCRE could look past the end of the subject if it
833     reached it while seeking the minimum number of characters. This could
834     happen only if some of the characters were more than one byte long, because
835     there is a check for at least the minimum number of bytes.
837     21. Refactored the implementation of \p and \P so as to be more general, to
838     allow for more different types of property in future. This has changed the
839     compiled form incompatibly. Anybody with saved compiled patterns that use
840     \p or \P will have to recompile them.
842     22. Added "Any" and "L&" to the supported property types.
844     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
845     but give a compile time error if the value is greater than 0xff.
847     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
848     accidentally not being installed or uninstalled.
850     25. The pcre.h file was built from pcre.h.in, but the only changes that were
851     made were to insert the current release number. This seemed silly, because
852     it made things harder for people building PCRE on systems that don't run
853     "configure". I have turned pcre.h into a distributed file, no longer built
854     by "configure", with the version identification directly included. There is
855     no longer a pcre.h.in file.
857     However, this change necessitated a change to the pcre-config script as
858     well. It is built from pcre-config.in, and one of the substitutions was the
859     release number. I have updated configure.ac so that ./configure now finds
860     the release number by grepping pcre.h.
862     26. Added the ability to run the tests under valgrind.
865 nigel 85 Version 6.4 05-Sep-05
866     ---------------------
868     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
869     "--" to be printed when multiple files were scanned, even when none of the
870     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
871     consider it to be a bug, and have restored the previous behaviour.
873     2. A couple of code tidies to get rid of compiler warnings.
875     3. The pcretest program used to cheat by referring to symbols in the library
876     whose names begin with _pcre_. These are internal symbols that are not
877     really supposed to be visible externally, and in some environments it is
878     possible to suppress them. The cheating is now confined to including
879     certain files from the library's source, which is a bit cleaner.
881     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
882     file's purpose clearer.
884     5. Reorganized pcre_ucp_findchar().
887 nigel 83 Version 6.3 15-Aug-05
888     ---------------------
890     1. The file libpcre.pc.in did not have general read permission in the tarball.
892     2. There were some problems when building without C++ support:
894     (a) If C++ support was not built, "make install" and "make test" still
895     tried to test it.
897     (b) There were problems when the value of CXX was explicitly set. Some
898     changes have been made to try to fix these, and ...
900     (c) --disable-cpp can now be used to explicitly disable C++ support.
902     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
903     backslash in a target when C++ was disabled. This confuses some
904     versions of "make", apparently. Using an intermediate variable solves
905     this. (Same for CPP_LOBJ.)
907     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
908     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
909     necessary on certain architectures.
911     4. Added a setting of -export-symbols-regex to the link command to remove
912     those symbols that are exported in the C sense, but actually are local
913     within the library, and not documented. Their names all begin with
914     "_pcre_". This is not a perfect job, because (a) we have to except some
915     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
916     available (and never for static libraries). I have made a note to try to
917     find a way round (a) in the future.
920 nigel 81 Version 6.2 01-Aug-05
921     ---------------------
923     1. There was no test for integer overflow of quantifier values. A construction
924     such as {1111111111111111} would give undefined results. What is worse, if
925     a minimum quantifier for a parenthesized subpattern overflowed and became
926     negative, the calculation of the memory size went wrong. This could have
927     led to memory overwriting.
929     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
931     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
932     operating environments where this matters.
934     4. Applied Giuseppe Maxia's patch to add additional features for controlling
935     PCRE options from within the C++ wrapper.
937     5. Named capturing subpatterns were not being correctly counted when a pattern
938     was compiled. This caused two problems: (a) If there were more than 100
939     such subpatterns, the calculation of the memory needed for the whole
940     compiled pattern went wrong, leading to an overflow error. (b) Numerical
941     back references of the form \12, where the number was greater than 9, were
942     not recognized as back references, even though there were sufficient
943     previous subpatterns.
945     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
946     versions of gcc, e.g. 2.95.4.
949 nigel 79 Version 6.1 21-Jun-05
950     ---------------------
952     1. There was one reference to the variable "posix" in pcretest.c that was not
953     surrounded by "#if !defined NOPOSIX".
955     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
956     the cross-check on the old pcre_info() function, for the benefit of the
957     cut-down version of PCRE that is currently imported into Exim.
959     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
960     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
961     patterns that start (?i) but I don't think that matters. The use of (?i) is
962     just an example; this all applies to the other options as well.
964     4. Since libtool seems to echo the compile commands it is issuing, the output
965     from "make" can be reduced a bit by putting "@" in front of each libtool
966     compile command.
968     5. Patch from the folks at Google for configure.in to be a bit more thorough
969     in checking for a suitable C++ installation before trying to compile the
970     C++ stuff. This should fix a reported problem when a compiler was present,
971     but no suitable headers.
973     6. The man pages all had just "PCRE" as their title. I have changed them to
974     be the relevant file name. I have also arranged that these names are
975     retained in the file doc/pcre.txt, which is a concatenation in text format
976     of all the man pages except the little individual ones for each function.
978     7. The NON-UNIX-USE file had not been updated for the different set of source
979     files that come with release 6. I also added a few comments about the C++
980     wrapper.
983 nigel 77 Version 6.0 07-Jun-05
984     ---------------------
986     1. Some minor internal re-organization to help with my DFA experiments.
988     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
989     didn't matter for the library itself when fully configured, but did matter
990     when compiling without UCP support, or within Exim, where the ucp files are
991     not imported.
993     3. Refactoring of the library code to split up the various functions into
994     different source modules. The addition of the new DFA matching code (see
995     below) to a single monolithic source would have made it really too
996     unwieldy, quite apart from causing all the code to be include in a
997     statically linked application, when only some functions are used. This is
998     relevant even without the DFA addition now that patterns can be compiled in
999     one application and matched in another.
1001     The downside of splitting up is that there have to be some external
1002     functions and data tables that are used internally in different modules of
1003     the library but which are not part of the API. These have all had their
1004     names changed to start with "_pcre_" so that they are unlikely to clash
1005     with other external names.
1007     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1008     a different (DFA) algorithm. Although it is slower than the original
1009     function, it does have some advantages for certain types of matching
1010     problem.
1012     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1013     including restarting after a partial match.
1015     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1016     defined when compiling for Windows was sent to me. I have put it into the
1017     code, though I have no means of testing or verifying it.
1019     7. Added the pcre_refcount() auxiliary function.
1021     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1022     match before or at the first newline in the subject string. In pcretest,
1023     the /f option on a pattern can be used to set this.
1025     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1026     would behave wrongly. This has been present in PCRE since release 4.0.
1028     10. A number of changes to the pcregrep command:
1030     (a) Refactored how -x works; insert ^(...)$ instead of setting
1031     PCRE_ANCHORED and checking the length, in preparation for adding
1032     something similar for -w.
1034     (b) Added the -w (match as a word) option.
1036     (c) Refactored the way lines are read and buffered so as to have more
1037     than one at a time available.
1039     (d) Implemented a pcregrep test script.
1041     (e) Added the -M (multiline match) option. This allows patterns to match
1042     over several lines of the subject. The buffering ensures that at least
1043     8K, or the rest of the document (whichever is the shorter) is available
1044     for matching (and similarly the previous 8K for lookbehind assertions).
1046     (f) Changed the --help output so that it now says
1048     -w, --word-regex(p)
1050     instead of two lines, one with "regex" and the other with "regexp"
1051     because that confused at least one person since the short forms are the
1052     same. (This required a bit of code, as the output is generated
1053     automatically from a table. It wasn't just a text change.)
1055     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1056     option but starts with a hyphen. Could be a pattern or a path name
1057     starting with a hyphen, for instance.
1059     (h) "-" can be given as a file name to represent stdin.
1061     (i) When file names are being printed, "(standard input)" is used for
1062     the standard input, for compatibility with GNU grep. Previously
1063     "<stdin>" was used.
1065     (j) The option --label=xxx can be used to supply a name to be used for
1066     stdin when file names are being printed. There is no short form.
1068     (k) Re-factored the options decoding logic because we are going to add
1069     two more options that take data. Such options can now be given in four
1070     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1072     (l) Added the -A, -B, and -C options for requesting that lines of context
1073     around matches be printed.
1075     (m) Added the -L option to print the names of files that do not contain
1076     any matching lines, that is, the complement of -l.
1078     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1079     continue to scan other files.
1081     (o) The -s option was incorrectly implemented. For compatibility with other
1082     greps, it now suppresses the error message for a non-existent or non-
1083     accessible file (but not the return code). There is a new option called
1084     -q that suppresses the output of matching lines, which was what -s was
1085     previously doing.
1087     (p) Added --include and --exclude options to specify files for inclusion
1088     and exclusion when recursing.
1090     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1091     Hopefully, it now does.
1093     12. Missing cast in pcre_study().
1095     13. Added an "uninstall" target to the makefile.
1097     14. Replaced "extern" in the function prototypes in Makefile.in with
1098     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1099     world, but is set differently for Windows.
1101     15. Added a second compiling function called pcre_compile2(). The only
1102     difference is that it has an extra argument, which is a pointer to an
1103     integer error code. When there is a compile-time failure, this is set
1104     non-zero, in addition to the error test pointer being set to point to an
1105     error message. The new argument may be NULL if no error number is required
1106     (but then you may as well call pcre_compile(), which is now just a
1107     wrapper). This facility is provided because some applications need a
1108     numeric error indication, but it has also enabled me to tidy up the way
1109     compile-time errors are handled in the POSIX wrapper.
1111     16. Added VPATH=.libs to the makefile; this should help when building with one
1112     prefix path and installing with another. (Or so I'm told by someone who
1113     knows more about this stuff than I do.)
1115     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1116     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1117     match everything, including newlines. This is not POSIX-compatible, but
1118     somebody wanted the feature. From pcretest it can be activated by using
1119     both the P and the s flags.
1121     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1123     19. libpcre.pc was being incorrectly installed as executable.
1125     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1126     it now also looks for '\r' so that it will work unmodified on Windows.
1128     21. Added Google's contributed C++ wrapper to the distribution.
1130     22. Added some untidy missing memory free() calls in pcretest, to keep
1131     Electric Fence happy when testing.
1135 nigel 75 Version 5.0 13-Sep-04
1136     ---------------------
1138     1. Internal change: literal characters are no longer packed up into items
1139     containing multiple characters in a single byte-string. Each character
1140     is now matched using a separate opcode. However, there may be more than one
1141     byte in the character in UTF-8 mode.
1143     2. The pcre_callout_block structure has two new fields: pattern_position and
1144     next_item_length. These contain the offset in the pattern to the next match
1145     item, and its length, respectively.
1147     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1148     insertion of callouts before each pattern item. Added the /C option to
1149     pcretest to make use of this.
1151     4. On the advice of a Windows user, the lines
1153     #if defined(_WIN32) || defined(WIN32)
1154     _setmode( _fileno( stdout ), 0x8000 );
1155     #endif /* defined(_WIN32) || defined(WIN32) */
1157     have been added to the source of pcretest. This apparently does useful
1158     magic in relation to line terminators.
1160     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1161     for the benefit of those environments where the "b" makes a difference.
1163     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1164     to know about it. I have put a hack into configure.in that adds in code
1165     to set GCC=yes if CC=icc. This seems to end up at a point in the
1166     generated configure script that is early enough to affect the setting of
1167     compiler options, which is what is needed, but I have no means of testing
1168     whether it really works. (The user who reported this had patched the
1169     generated configure script, which of course I cannot do.)
1171     LATER: After change 22 below (new libtool files), the configure script
1172     seems to know about icc (and also ecc). Therefore, I have commented out
1173     this hack in configure.in.
1175     7. Added support for pkg-config (2 patches were sent in).
1177     8. Negated POSIX character classes that used a combination of internal tables
1178     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1179     [[:^ascii]]. Typically, they would match almost any characters. The other
1180     POSIX classes were not broken in this way.
1182     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1183     to find the match, as PCRE was deluded into thinking that the match had to
1184     start at the start point or following a newline. The same bug applied to
1185     patterns with negative forward assertions or any backward assertions
1186     preceding ".*" at the start, unless the pattern required a fixed first
1187     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1189     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1190     starting at the last subject character, bytes beyond the end of the subject
1191     string were read.
1193     11. Renamed the variable "class" as "classbits" to make life easier for C++
1194     users. (Previously there was a macro definition, but it apparently wasn't
1195     enough.)
1197     12. Added the new field "tables" to the extra data so that tables can be passed
1198     in at exec time, or the internal tables can be re-selected. This allows
1199     a compiled regex to be saved and re-used at a later time by a different
1200     program that might have everything at different addresses.
1202     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1203     -R library as well as a -L library.
1205     14. The debugging options of pcretest (-d on the command line or D on a
1206     pattern) showed incorrect output for anything following an extended class
1207     that contained multibyte characters and which was followed by a quantifier.
1209     15. Added optional support for general category Unicode character properties
1210     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1211     support. It adds about 90K to the size of the library. The meanings of the
1212     inbuilt class escapes such as \d and \s have NOT been changed.
1214     16. Updated pcredemo.c to include calls to free() to release the memory for the
1215     compiled pattern.
1217     17. The generated file chartables.c was being created in the source directory
1218     instead of in the building directory. This caused the build to fail if the
1219     source directory was different from the building directory, and was
1220     read-only.
1222     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1223     file. No doubt somebody will tell me if they don't make sense... Also added
1224     Dan Mooney's comments about building on OpenVMS.
1226     19. Added support for partial matching via the PCRE_PARTIAL option for
1227     pcre_exec() and the \P data escape in pcretest.
1229     20. Extended pcretest with 3 new pattern features:
1231     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1232     write the compiled pattern to the file whose name is "rest-of-line".
1233     This is a straight binary dump of the data, with the saved pointer to
1234     the character tables forced to be NULL. The study data, if any, is
1235     written too. After writing, pcretest reads a new pattern.
1237     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1238     compiled pattern from the given file. There must not be any
1239     occurrences of "<" in the file name (pretty unlikely); if there are,
1240     pcretest will instead treat the initial "<" as a pattern delimiter.
1241     After reading in the pattern, pcretest goes on to read data lines as
1242     usual.
1244     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1245     and 16-bit fields in a compiled pattern, to simulate a pattern that
1246     was compiled on a host of opposite endianness.
1248     21. The pcre-exec() function can now cope with patterns that were compiled on
1249     hosts of opposite endianness, with this restriction:
1251     As for any compiled expression that is saved and used later, the tables
1252     pointer field cannot be preserved; the extra_data field in the arguments
1253     to pcre_exec() should be used to pass in a tables address if a value
1254     other than the default internal tables were used at compile time.
1256     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1257     now diagnosed as an error. Previously, most of the time, a negative number
1258     would have been treated as zero, but if in addition "ovector" was passed as
1259     NULL, a crash could occur.
1261     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1262     new versions from the libtool 1.5 distribution (the last one is a copy of
1263     a file called libtool.m4). This seems to have fixed the need to patch
1264     "configure" to support Darwin 1.3 (which I used to do). However, I still
1265     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1266     workstation).
1268     24. Changed the PCRE licence to be the more standard "BSD" licence.
1271 nigel 73 Version 4.5 01-Dec-03
1272     ---------------------
1274     1. There has been some re-arrangement of the code for the match() function so
1275     that it can be compiled in a version that does not call itself recursively.
1276     Instead, it keeps those local variables that need separate instances for
1277     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1278     needs to "recurse". Keeping track of where control must go is done by means
1279     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1280     hide most of the details from the main code, and operates only if
1281     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1282     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1283     operating.
1285     To make it easier for callers to provide specially tailored get/free
1286     functions for this usage, two new functions, pcre_stack_malloc, and
1287     pcre_stack_free, are used. They are always called in strict stacking order,
1288     and the size of block requested is always the same.
1290     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1291     PCRE has been compiled to use the stack or the heap for recursion. The
1292     -C option of pcretest uses this to show which version is compiled.
1294     A new data escape \S, is added to pcretest; it causes the amounts of store
1295     obtained and freed by both kinds of malloc/free at match time to be added
1296     to the output.
1298     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1299     what's available on my current Linux desktop machine.
1301     3. When matching a UTF-8 string, the test for a valid string at the start has
1302     been extended. If start_offset is not zero, PCRE now checks that it points
1303     to a byte that is the start of a UTF-8 character. If not, it returns
1304     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1305     this is necessary because there may be backward assertions in the pattern.
1306     When matching the same subject several times, it may save resources to use
1307     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1309     4. The code for checking the validity of UTF-8 strings has been tightened so
1310     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1311     containing "overlong sequences".
1313     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1314     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1315     should have been "&", but it just so happened that all the cases this let
1316     through by mistake were picked up later in the function.
1318     6. I had used a variable called "isblank" - this is a C99 function, causing
1319     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1321     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1322     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1323     so the version is shown; (c) stop "make test" echoing "./RunTest".
1325     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1327     9. The prototype for memmove() for systems that don't have it was using
1328     size_t, but the inclusion of the header that defines size_t was later. I've
1329     moved the #includes for the C headers earlier to avoid this.
1331     10. Added some adjustments to the code to make it easier to compiler on certain
1332     special systems:
1334     (a) Some "const" qualifiers were missing.
1335     (b) Added the macro EXPORT before all exported functions; by default this
1336     is defined to be empty.
1337     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1338     that it reads its output file name as an argument instead of writing
1339     to the standard output and assuming this can be redirected.
1341     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1342     class containing characters with values greater than 255, PCRE compilation
1343     went into a loop.
1345     12. A recursive reference to a subpattern that was within another subpattern
1346     that had a minimum quantifier of zero caused PCRE to crash. For example,
1347     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1348     recursion. If the recursively-called subpattern itself had a zero repeat,
1349     that was OK.
1351     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1352     buffer into which it was copied (for escape processing) was still set at
1353     1024, so long lines caused crashes.
1355     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1356     "internal error: code overflow...". This applied to any character class
1357     that was followed by a possessive quantifier.
1359     15. Modified the Makefile to add libpcre.la as a prerequisite for
1360     libpcreposix.la because I was told this is needed for a parallel build to
1361     work.
1363     16. If a pattern that contained .* following optional items at the start was
1364     studied, the wrong optimizing data was generated, leading to matching
1365     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1366     matching string must start with a or b or c. The correct conclusion for
1367     this pattern is that a match can start with any character.
1370 nigel 71 Version 4.4 13-Aug-03
1371     ---------------------
1373     1. In UTF-8 mode, a character class containing characters with values between
1374     127 and 255 was not handled correctly if the compiled pattern was studied.
1375     In fixing this, I have also improved the studying algorithm for such
1376     classes (slightly).
1378     2. Three internal functions had redundant arguments passed to them. Removal
1379     might give a very teeny performance improvement.
1381     3. Documentation bug: the value of the capture_top field in a callout is *one
1382     more than* the number of the hightest numbered captured substring.
1384     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1385     in incorrectly linking with a previously installed version. They now link
1386     explicitly with libpcre.la.
1388     5. configure.in no longer needs to recognize Cygwin specially.
1390     6. A problem in pcre.in for Windows platforms is fixed.
1392     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1393     pcretest, it used to include the size of the study block as part of its
1394     output. Unfortunately, the structure contains a field that has a different
1395     size on different hardware architectures. This meant that the tests that
1396     showed this size failed. As the block is currently always of a fixed size,
1397     this information isn't actually particularly useful in pcretest output, so
1398     I have just removed it.
1400     8. Three pre-processor statements accidentally did not start in column 1.
1401     Sadly, there are *still* compilers around that complain, even though
1402     standard C has not required this for well over a decade. Sigh.
1404     9. In pcretest, the code for checking callouts passed small integers in the
1405     callout_data field, which is a void * field. However, some picky compilers
1406     complained about the casts involved for this on 64-bit systems. Now
1407     pcretest passes the address of the small integer instead, which should get
1408     rid of the warnings.
1410     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1411     both compile and run time, and gives an error if an invalid UTF-8 sequence
1412     is found. There is a option for disabling this check in cases where the
1413     string is known to be correct and/or the maximum performance is wanted.
1415     11. In response to a bug report, I changed one line in Makefile.in from
1417     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1418     to
1419     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1421     to look similar to other lines, but I have no way of telling whether this
1422     is the right thing to do, as I do not use Windows. No doubt I'll get told
1423     if it's wrong...
1426 nigel 69 Version 4.3 21-May-03
1427     ---------------------
1429     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1430     Makefile.
1432     2. Some refactoring to improve the quality of the code:
1434     (i) The utf8_table... variables are now declared "const".
1436     (ii) The code for \cx, which used the "case flipping" table to upper case
1437     lower case letters, now just substracts 32. This is ASCII-specific,
1438     but the whole concept of \cx is ASCII-specific, so it seems
1439     reasonable.
1441     (iii) PCRE was using its character types table to recognize decimal and
1442     hexadecimal digits in the pattern. This is silly, because it handles
1443     only 0-9, a-f, and A-F, but the character types table is locale-
1444     specific, which means strange things might happen. A private
1445     table is now used for this - though it costs 256 bytes, a table is
1446     much faster than multiple explicit tests. Of course, the standard
1447     character types table is still used for matching digits in subject
1448     strings against \d.
1450     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1451     ending in _t are). So I've renamed it as ESC_tee.
1453     3. The first argument for regexec() in the POSIX wrapper should have been
1454     defined as "const".
1456     4. Changed pcretest to use malloc() for its buffers so that they can be
1457     Electric Fenced for debugging.
1459     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1460     to read one or more bytes before the start of the subject string. Often this
1461     had no effect on PCRE's behaviour, but in some circumstances it could
1462     provoke a segmentation fault.
1464     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1465     to try to read one or more bytes before the start of the subject string.
1467     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1468     UTF-8 support could misbehave in various ways if the subject string
1469     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1470     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1471     back over UTF-8 characters.)
1474 nigel 67 Version 4.2 14-Apr-03
1475     ---------------------
1477     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1479     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1480     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1481     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1482     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1483     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1484     and BUILD_EXEEXT
1485     Note: automatic setting of the BUILD variables is not yet working
1486     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1487     compile-time but not at link-time
1488     [LINK]: use for linking executables only
1489     make different versions for Windows and non-Windows
1490     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1491     libraries
1492     [LINK_FOR_BUILD]: new variable
1493     [OBJEXT]: use throughout
1494     [EXEEXT]: use throughout
1495     <winshared>: new target
1496     <wininstall>: new target
1497     <dftables.o>: use native compiler
1498     <dftables>: use native linker
1499     <install>: handle Windows platform correctly
1500     <clean>: ditto
1501     <check>: ditto
1502     copy DLL to top builddir before testing
1504     As part of these changes, -no-undefined was removed again. This was reported
1505     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1506     in any case.
1508     3. Some tidies to get rid of compiler warnings:
1510     . In the match_data structure, match_limit was an unsigned long int, whereas
1511     match_call_count was an int. I've made them both unsigned long ints.
1513     . In pcretest the fact that a const uschar * doesn't automatically cast to
1514     a void * provoked a warning.
1516     . Turning on some more compiler warnings threw up some "shadow" variables
1517     and a few more missing casts.
1519     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1520     option, a class that contained a single character with a value between 128
1521     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1523     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1524     option, a class that contained several characters, but with at least one
1525     whose value was between 128 and 255 caused PCRE to crash.
1528 nigel 65 Version 4.1 12-Mar-03
1529     ---------------------
1530 nigel 63
1531 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1532     needed, and a string in dftables.c that was longer than standard compilers are
1533     required to support.
1535     2. Compiling with Sun's compiler found a few more places where the code could
1536     be tidied up in order to avoid warnings.
1538     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1539     first of these names is deprecated in the latest Autoconf in favour of the name
1540     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1541     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1542     analogy I have changed it to CFLAGS_FOR_BUILD.
1544     4. Added -no-undefined to the linking command in the Makefile, because this is
1545     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1546     linking step for the pcreposix library.
1548     5. PCRE was failing to diagnose the case of two named groups with the same
1549     name.
1551     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1552     literal character that is needed in the subject for a match, and scans along to
1553     ensure that it is present before embarking on the full matching process. This
1554     saves time in cases of nested unlimited repeats that are never going to match.
1555     Problem: the scan can take a lot of time if the subject is very long (e.g.
1556     megabytes), thus penalizing straightforward matches. It is now done only if the
1557     amount of subject to be scanned is less than 1000 bytes.
1559     7. A lesser problem with the same optimization is that it was recording the
1560     first character of an anchored pattern as "needed", thus provoking a search
1561     right along the subject, even when the first match of the pattern was going to
1562     fail. The "needed" character is now not set for anchored patterns, unless it
1563     follows something in the pattern that is of non-fixed length. Thus, it still
1564     fulfils its original purpose of finding quick non-matches in cases of nested
1565     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1568     Version 4.0 17-Feb-03
1569     ---------------------
1571 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1572     extended to the end of string, PCRE compiled incorrect data. This could lead to
1573     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1575     2. Moved to autoconf 2.53 and libtool 1.4.2.
1577     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1578     the special perltest8 script is no longer needed - all the tests can be run
1579     from a single perltest script.
1581     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1582     by \s. It has now been removed in PCRE. This means it isn't recognized as
1583     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1584     class [:space:] *does* include VT, thereby creating a mess.
1586     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1587     space and tab.
1589     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1590     its new features into the main test script, reducing the number of scripts.
1592     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1593     were backward compatible, and made the (?i) apply to the whole pattern, as if
1594     /i were given. Now it behaves more logically, and applies the option setting
1595     only to what follows. PCRE has been changed to follow suit. However, if it
1596     finds options settings right at the start of the pattern, it extracts them into
1597     the global options, as before. Thus, they show up in the info data.
1599     8. Added support for the \Q...\E escape sequence. Characters in between are
1600     treated as literals. This is slightly different from Perl in that $ and @ are
1601     also handled as literals inside the quotes. In Perl, they will cause variable
1602     interpolation. Note the following examples:
1604     Pattern PCRE matches Perl matches
1606     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1607     \Qabc\$xyz\E abc\$xyz abc\$xyz
1608     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1610     For compatibility with Perl, \Q...\E sequences are recognized inside character
1611     classes as well as outside them.
1613     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1614     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1615     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1616     signed/unsigned warnings.
1618     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1619     option for pcretest, so I've replaced it by a simple function that does just
1620     that job.
1622     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1623     "pcregrep -".
1625     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1626     Java package. This provides some syntactic sugar for simple cases of what my
1627     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1628     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1629     item, you can use this simplified notation. Note that only makes sense with
1630     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1631     greediness, whatever the setting of the PCRE_UNGREEDY option.
1633     13. A change of greediness default within a pattern was not taking effect at
1634     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1635     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1636     was abstracted outside.
1638     14. PCRE now supports the \G assertion. It is true when the current matching
1639     position is at the start point of the match. This differs from \A when the
1640     starting offset is non-zero. Used with the /g option of pcretest (or similar
1641     code), it works in the same way as it does for Perl's /g option. If all
1642     alternatives of a regex begin with \G, the expression is anchored to the start
1643     match position, and the "anchored" flag is set in the compiled expression.
1645     15. Some bugs concerning the handling of certain option changes within patterns
1646     have been fixed. These applied to options other than (?ims). For example,
1647     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1648     been the other way round. Some of this was related to change 7 above.
1650     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1651     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1652     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1653     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1655     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1656     mode. Unlike ".", it always matches newline, whatever the setting of
1657     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1658     assertions. Perl allows it, but it doesn't (in general) work because it can't
1659     calculate the length of the lookbehind. At least, that's the case for Perl
1660     5.8.0 - I've been told they are going to document that it doesn't work in
1661     future.
1663     18. Added an error diagnosis for escapes that PCRE does not support: these are
1664     \L, \l, \N, \P, \p, \U, \u, and \X.
1666     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1667     reading past the end of the pattern in cases such as /[abcd/.
1669     20. PCRE was getting more memory than necessary for patterns with classes that
1670     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1672     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1673     compiling PCRE for use with Virtual Pascal.
1675     22. Small fix to the Makefile to make it work properly if the build is done
1676     outside the source tree.
1678     23. Added a new extension: a condition to go with recursion. If a conditional
1679     subpattern starts with (?(R) the "true" branch is used if recursion has
1680     happened, whereas the "false" branch is used only at the top level.
1682     24. When there was a very long string of literal characters (over 255 bytes
1683     without UTF support, over 250 bytes with UTF support), the computation of how
1684     much memory was required could be incorrect, leading to segfaults or other
1685     strange effects.
1687     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1688     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1689     there was a subsequent back reference to those brackets. This meant that, for
1690     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1691     possible to check for precisely this case. All we can do is abandon the
1692     optimization if .* occurs inside capturing brackets when there are any back
1693     references whatsoever. (See below for a better fix that came later.)
1695     26. The handling of the optimization for finding the first character of a
1696     non-anchored pattern, and for finding a character that is required later in the
1697     match were failing in some cases. This didn't break the matching; it just
1698     failed to optimize when it could. The way this is done has been re-implemented.
1700     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1702     28. Added a new feature that provides some of the functionality that Perl
1703     provides with (?{...}). The facility is termed a "callout". The way it is done
1704     in PCRE is for the caller to provide an optional function, by setting
1705     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1706     global variable. By default it is unset, which disables all calling out. To get
1707     the function called, the regex must include (?C) at appropriate points. This
1708     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1709     This provides a means of identifying different callout points. When PCRE
1710     reaches such a point in the regex, if pcre_callout has been set, the external
1711     function is called. It is provided with data in a structure called
1712     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1713     matching continues; if it returns a non-zero value, the match at the current
1714     point fails. However, backtracking will occur if possible. [This was changed
1715     later and other features added - see item 49 below.]
1717     29. pcretest is upgraded to test the callout functionality. It provides a
1718     callout function that displays information. By default, it shows the start of
1719     the match and the current position in the text. There are some new data escapes
1720     to vary what happens:
1722     \C+ in addition, show current contents of captured substrings
1723     \C- do not supply a callout function
1724     \C!n return 1 when callout number n is reached
1725     \C!n!m return 1 when callout number n is reached for the mth time
1727     30. If pcregrep was called with the -l option and just a single file name, it
1728     output "<stdin>" if a match was found, instead of the file name.
1730     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1731     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1732     pcre_exec(). This saves a malloc/free per call. The default value of
1733     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1734     when configuring.
1736     32. The default maximum size of a compiled pattern is 64K. There have been a
1737     few cases of people hitting this limit. The code now uses macros to handle the
1738     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1739     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1740     configuring. Tests 2 and 5 work only with 2-byte links because they output
1741     debugging information about compiled patterns.
1743     33. Internal code re-arrangements:
1745     (a) Moved the debugging function for printing out a compiled regex into
1746     its own source file (printint.c) and used #include to pull it into
1747     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1748     separate copies.
1750     (b) Defined the list of op-code names for debugging as a macro in
1751     internal.h so that it is next to the definition of the opcodes.
1753     (c) Defined a table of op-code lengths for simpler skipping along compiled
1754     code. This is again a macro in internal.h so that it is next to the
1755     definition of the opcodes.
1757     34. Added support for recursive calls to individual subpatterns, along the
1758     lines of Robin Houston's patch (but implemented somewhat differently).
1760     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1761     allow it to read and process whole directories in Win32. This code was
1762     contributed by Lionel Fourquaux; it has not been tested by me.
1764     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1765     used to name a group. Names consist of alphanumerics and underscores, and must
1766     be unique. Back references use the syntax (?P=name) and recursive calls use
1767     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1768     numbers. The function pcre_fullinfo() can be used after compilation to extract
1769     a name/number map. There are three relevant calls:
1771     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1772     PCRE_INFO_NAMECOUNT yields the number of entries
1773     PCRE_INFO_NAMETABLE yields a pointer to the map.
1775     The map is a vector of fixed-size entries. The size of each entry depends on
1776     the length of the longest name used. The first two bytes of each entry are the
1777     group number, most significant byte first. There follows the corresponding
1778     name, zero terminated. The names are in alphabetical order.
1780     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1781     case instead of 255. Making it the same both with and without UTF-8 support
1782     means that the same test output works with both.
1784     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1785     calling malloc() with a zero argument.
1787     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1788     optimization. I've improved things by keeping a bitmap of backreferences with
1789     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1790     fact referenced, the optimization can be applied. It is unlikely that a
1791     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1792     the match to follow \n) will appear inside brackets with a number greater than
1793     31, but if it does, any back reference > 31 suppresses the optimization.
1795     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1796     of disabling numbered capturing parentheses. Any opening parenthesis that is
1797     not followed by ? behaves as if it were followed by ?: but named parentheses
1798     can still be used for capturing (and they will acquire numbers in the usual
1799     way).
1801     41. Redesigned the return codes from the match() function into yes/no/error so
1802     that errors can be passed back from deep inside the nested calls. A malloc
1803     failure while inside a recursive subpattern call now causes the
1804     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1806     42. It is now possible to set a limit on the number of times the match()
1807     function is called in a call to pcre_exec(). This facility makes it possible to
1808     limit the amount of recursion and backtracking, though not in a directly
1809     obvious way, because the match() function is used in a number of different
1810     circumstances. The count starts from zero for each position in the subject
1811     string (for non-anchored patterns). The default limit is, for compatibility, a
1812     large number, namely 10 000 000. You can change this in two ways:
1814     (a) When configuring PCRE before making, you can use --with-match-limit=n
1815     to set a default value for the compiled library.
1817     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1818     a different value is set. See 45 below.
1820     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1822     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1823     of things that can be changed at compile time. The first argument specifies
1824     what is wanted and the second points to where the information is to be placed.
1825     The current list of available information is:
1829     The output is an integer that is set to one if UTF-8 support is available;
1830     otherwise it is set to zero.
1834     The output is an integer that it set to the value of the code that is used for
1835     newline. It is either LF (10) or CR (13).
1839     The output is an integer that contains the number of bytes used for internal
1840     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1844     The output is an integer that contains the threshold above which the POSIX
1845     interface uses malloc() for output vectors. See item 31 above.
1849     The output is an unsigned integer that contains the default limit of the number
1850     of match() calls in a pcre_exec() execution. See 42 above.
1852     44. pcretest has been upgraded by the addition of the -C option. This causes it
1853     to extract all the available output from the new pcre_config() function, and to
1854     output it. The program then exits immediately.
1856     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1857     order to support additional features. One way would have been to define
1858     pcre_exec2() (for example) with extra arguments, but this would not have been
1859     extensible, and would also have required all calls to the original function to
1860     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1861     is used for passing in "extra" data from pcre_study().
1863     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1864     contains the following fields:
1866     flags a bitmap indicating which of the following fields are set
1867     study_data opaque data from pcre_study()
1868     match_limit a way of specifying a limit on match() calls for a specific
1869     call to pcre_exec()
1870     callout_data data for callouts (see 49 below)
1872     The flag bits are also defined in pcre.h, and are
1878     The pcre_study() function now returns one of these new pcre_extra blocks, with
1879     the actual study data pointed to by the study_data field, and the
1880     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
1881     before. That is, this change is entirely upwards-compatible and requires no
1882     change to existing code.
1884     If you want to pass in additional data to pcre_exec(), you can either place it
1885     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
1886     block.
1888     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
1889     data string contains the escape sequence \M, pcretest calls pcre_exec() several
1890     times with different match limits, until it finds the minimum value needed for
1891     pcre_exec() to complete. The value is then output. This can be instructive; for
1892     most simple matches the number is quite small, but for pathological cases it
1893     gets very large very quickly.
1895     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
1896     returns the size of the data block pointed to by the study_data field in a
1897     pcre_extra block, that is, the value that was passed as the argument to
1898     pcre_malloc() when PCRE was getting memory in which to place the information
1899     created by pcre_study(). The fourth argument should point to a size_t variable.
1900     pcretest has been extended so that this information is shown after a successful
1901     pcre_study() call when information about the compiled regex is being displayed.
1903     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
1904     because what follows is always an absolute path. (Later: it turns out that this
1905     is more than cosmetic for MinGW, because it doesn't like empty path
1906     components.)
1908     49. Some changes have been made to the callout feature (see 28 above):
1910     (i) A callout function now has three choices for what it returns:
1912     0 => success, carry on matching
1913     > 0 => failure at this point, but backtrack if possible
1914     < 0 => serious error, return this value from pcre_exec()
1916     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
1917     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
1918     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
1919     use by callout functions. It will never be used by PCRE itself.
1921     (ii) The pcre_extra structure (see 45 above) has a void * field called
1922     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
1923     pcre_callout_block structure has a field of the same name. The contents of
1924     the field passed in the pcre_extra structure are passed to the callout
1925     function in the corresponding field in the callout block. This makes it
1926     easier to use the same callout-containing regex from multiple threads. For
1927     testing, the pcretest program has a new data escape
1929     \C*n pass the number n (may be negative) as callout_data
1931     If the callout function in pcretest receives a non-zero value as
1932     callout_data, it returns that value.
1934     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
1935     there were some redundant $(CFLAGS) in commands that are now specified as
1936     $(LINK), which already includes $(CFLAGS).
1938     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
1939     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
1940     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
1941     one-byte characters throughout. Note that case-insensitive matching applies
1942     only to characters whose values are less than 256. PCRE doesn't support the
1943     notion of cases for higher-valued characters.
1945     (i) A character class whose characters are all within 0-255 is handled as
1946     a bit map, and the map is inverted for negative classes. Previously, a
1947     character > 255 always failed to match such a class; however it should
1948     match if the class was a negative one (e.g. [^ab]). This has been fixed.
1950     (ii) A negated character class with a single character < 255 is coded as
1951     "not this character" (OP_NOT). This wasn't working properly when the test
1952     character was multibyte, either singly or repeated.
1954     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
1955     mode, for example: \x{100}{2,3}.
1957     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
1958     singly or repeated) now correctly test multibyte characters. However,
1959     PCRE doesn't recognize any characters with values greater than 255 as
1960     digits, spaces, or word characters. Such characters always match \D, \S,
1961     and \W, and never match \d, \s, or \w.
1963     (v) Classes may now contain characters and character ranges with values
1964     greater than 255. For example: [ab\x{100}-\x{400}].
1966     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
1967     PCRE in UTF-8 mode.
1969     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
1970     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
1971     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
1972     value.)
1974     53. The single man page has become too large. I have therefore split it up into
1975     a number of separate man pages. These also give rise to individual HTML pages;
1976     these are now put in a separate directory, and there is an index.html page that
1977     lists them all. Some hyperlinking between the pages has been installed.
1979     54. Added convenience functions for handling named capturing parentheses.
1981     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
1982     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
1983     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
1984     are faulted.
1986     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
1987     calling configure. These values are used when compiling the dftables.c program
1988     which is run to generate the source of the default character tables. They
1989     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
1990     you will need to set these values.
1992     57. Updated the building process for Windows DLL, as provided by Fred Cox.
1995     Version 3.9 02-Jan-02
1996 nigel 61 ---------------------
1998     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2000     2. If --disable-static was given, the building process failed when trying to
2001     build pcretest and pcregrep. (For some reason it was using libtool to compile
2002     them, which is not right, as they aren't part of the library.)
2005 nigel 59 Version 3.8 18-Dec-01
2006     ---------------------
2008     1. The experimental UTF-8 code was completely screwed up. It was packing the
2009     bytes in the wrong order. How dumb can you get?
2012 nigel 57 Version 3.7 29-Oct-01
2013     ---------------------
2015     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2016     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2017     this didn't happen under Solaris 8, where I normally test things.
2019 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2020     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2021     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2022     this fix an hour or so after the initial 3.7 release.)
2023 nigel 57
2024 nigel 59
2025 nigel 55 Version 3.6 23-Oct-01
2026     ---------------------
2028     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2029     offsets passed as NULL with zero offset count.
2031     2. The config.guess and config.sub files had not been updated when I moved to
2032     the latest autoconf.
2035 nigel 53 Version 3.5 15-Aug-01
2036     ---------------------
2037 nigel 5
2038 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2039     had been forgotten.
2041     2. By using declared but undefined structures, we can avoid using "void"
2042     definitions in pcre.h while keeping the internal definitions of the structures
2043     private.
2045     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2046     user point of view, this means that both static and shared libraries are built
2047     by default, but this can be individually controlled. More of the work of
2048     handling this static/shared cases is now inside libtool instead of PCRE's make
2049     file.
2051     4. The pcretest utility is now installed along with pcregrep because it is
2052     useful for users (to test regexs) and by doing this, it automatically gets
2053     relinked by libtool. The documentation has been turned into a man page, so
2054     there are now .1, .txt, and .html versions in /doc.
2056     5. Upgrades to pcregrep:
2057     (i) Added long-form option names like gnu grep.
2058     (ii) Added --help to list all options with an explanatory phrase.
2059     (iii) Added -r, --recursive to recurse into sub-directories.
2060     (iv) Added -f, --file to read patterns from a file.
2062     6. pcre_exec() was referring to its "code" argument before testing that
2063     argument for NULL (and giving an error if it was NULL).
2065     7. Upgraded Makefile.in to allow for compiling in a different directory from
2066     the source directory.
2068     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2069     options bits, the pointer it was passed was to an int instead of to an unsigned
2070     long int. This mattered only on 64-bit systems.
2072     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2073     generated) instead of pcre.in, which it its source. Also made the same change
2074     in several of the .c files.
2076     10. A new release of gcc defines printf() as a macro, which broke pcretest
2077     because it had an ifdef in the middle of a string argument for printf(). Fixed
2078     by using separate calls to printf().
2080     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2081     script, to force use of CR or LF instead of \n in the source. On non-Unix
2082     systems, the value can be set in config.h.
2084     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2085     absolute limit. Changed the text of the error message to make this clear, and
2086     likewise updated the man page.
2088     13. The limit of 99 on the number of capturing subpatterns has been removed.
2089     The new limit is 65535, which I hope will not be a "real" limit.
2092 nigel 51 Version 3.4 22-Aug-00
2093     ---------------------
2095     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2097     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2100 nigel 49 Version 3.3 01-Aug-00
2101     ---------------------
2103     1. If an octal character was given, but the value was greater than \377, it
2104     was not getting masked to the least significant bits, as documented. This could
2105     lead to crashes in some systems.
2107     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2108     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2110     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2111     These just pass their arguments on to (pcre_free)(), but they are provided
2112     because some uses of PCRE bind it to non-C systems that can call its functions,
2113     but cannot call free() or pcre_free() directly.
2115     4. Add "make test" as a synonym for "make check". Corrected some comments in
2116     the Makefile.
2118     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2119     Makefile.
2121     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2122     command called pgrep for grepping around the active processes.
2124     7. Added the beginnings of support for UTF-8 character strings.
2126     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2127     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2128     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2129     out for the ar command.)
2132 nigel 47 Version 3.2 12-May-00
2133     ---------------------
2135     This is purely a bug fixing release.
2137     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2138     of ZA. This was just one example of several cases that could provoke this bug,
2139     which was introduced by change 9 of version 2.00. The code for breaking
2140     infinite loops after an iteration that matches an empty string was't working
2141     correctly.
2143     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2144     when matched against abbab (for example). After matching an empty string, it
2145     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2146     caused it to match further down the string than it should.
2148     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2149     was there because it doesn't seem to be needed, and it causes trouble on some
2150     systems, as it is not a Standard C header. It has been removed.
2152     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2153     were reported on the Macintosh. The changes were from
2155     while ((c = *(++ptr)) != 0 && c != '\n');
2156     to
2157     while ((c = *(++ptr)) != 0 && c != '\n') ;
2159     Totally extraordinary, but if that's what it takes...
2161     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2162     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2163     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2164     assumes the way PCRE uses memmove() (always moving upwards).
2166     6. PCRE is being used in one environment where strchr() is not available. There
2167     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2168     faster code anyway.
2171 nigel 45 Version 3.1 09-Feb-00
2172     ---------------------
2174     The only change in this release is the fixing of some bugs in Makefile.in for
2175     the "install" target:
2177     (1) It was failing to install pcreposix.h.
2179     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2182 nigel 43 Version 3.0 01-Feb-00
2183     ---------------------
2184 nigel 41
2185     1. Add support for the /+ modifier to perltest (to output $` like it does in
2186     pcretest).
2188     2. Add support for the /g modifier to perltest.
2190     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2191     matches null strings.
2193     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2194     pattern. Perl treats empty patterns specially - it reuses the most recent
2195     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2196     effect.
2198     5. The POSIX interface was broken in that it was just handing over the POSIX
2199     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2200     required a bigger vector, with some working space on the end. This means that
2201     the POSIX wrapper now has to get and free some memory, and copy the results.
2203 nigel 43 6. Added some simple autoconf support, placing the test data and the
2204     documentation in separate directories, re-organizing some of the
2205     information files, and making it build pcre-config (a GNU standard). Also added
2206     libtool support for building PCRE as a shared library, which is now the
2207     default.
2208 nigel 41
2209 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2210     09 are not valid octal constants. Single digits will be used for minor values
2211     less than 10.
2213     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2214     existing programs that set these in the POSIX interface can use PCRE without
2215     modification.
2217     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2218     return all that pcre_info() returns, plus additional data. The pcre_info()
2219     function is retained for compatibility, but is considered to be obsolete.
2221     10. Added experimental recursion feature (?R) to handle one common case that
2222     Perl 5.6 will be able to do with (?p{...}).
2224     11. Added support for POSIX character classes like [:alpha:], which Perl is
2225     adopting.
2228 nigel 39 Version 2.08 31-Aug-99
2229     ----------------------
2231     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2232     trying to match at the startoffset position, but instead was moving forward to
2233     the next newline as if a previous match had failed.
2235     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2236     and could get into a loop if a null string was matched other than at the start
2237     of the subject.
2239     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2240     be distinguished at compile time, and for completeness also added PCRE_DATE.
2242     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2243     in GnuWin32 environments.
2246 nigel 37 Version 2.07 29-Jul-99
2247     ----------------------
2249     1. The documentation is now supplied in plain text form and HTML as well as in
2250     the form of man page sources.
2252     2. C++ compilers don't like assigning (void *) values to other pointer types.
2253     In particular this affects malloc(). Although there is no problem in Standard
2254     C, I've put in casts to keep C++ compilers happy.
2256     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2257     should be (const char *).
2259     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2260     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2261     However, I haven't made this a standard facility. The documentation doesn't
2262     mention it, and the Makefile doesn't support it.
2264     5. The Makefile now contains an "install" target, with editable destinations at
2265     the top of the file. The pcretest program is not installed.
2267     6. pgrep -V now gives the PCRE version number and date.
2269     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2270     causing the entire string to be ignored, instead of just the last character.
2272     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2273     non-matching string, it can take a very, very long time, even for strings of
2274     quite modest length, because of the nested recursion. PCRE now does better in
2275     some of these cases. It does this by remembering the last required literal
2276     character in the pattern, and pre-searching the subject to ensure it is present
2277     before running the real match. In other words, it applies a heuristic to detect
2278     some types of certain failure quickly, and in the above example, if presented
2279     with a string that has no trailing " it gives "no match" very quickly.
2281     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2282     other alternatives are tried instead.
2285 nigel 35 Version 2.06 09-Jun-99
2286     ----------------------
2288     1. Change pcretest's output for amount of store used to show just the code
2289     space, because the remainder (the data block) varies in size between 32-bit and
2290     64-bit systems.
2292     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2293     start matching at. This allows lookbehinds to work when searching for multiple
2294     occurrences in a string.
2296     3. Added additional options to pcretest for testing multiple occurrences:
2298     /+ outputs the rest of the string that follows a match
2299     /g loops for multiple occurrences, using the new startoffset argument
2300     /G loops for multiple occurrences by passing an incremented pointer
2302     4. PCRE wasn't doing the "first character" optimization for patterns starting
2303     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2304     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2305     the letter 'x'. On long subject strings, this gives a significant speed-up.
2308 nigel 33 Version 2.05 21-Apr-99
2309     ----------------------
2311     1. Changed the type of magic_number from int to long int so that it works
2312     properly on 16-bit systems.
2314     2. Fixed a bug which caused patterns starting with .* not to work correctly
2315     when the subject string contained newline characters. PCRE was assuming
2316     anchoring for such patterns in all cases, which is not correct because .* will
2317     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2318     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2319     must be retried after every newline in the subject.
2322 nigel 31 Version 2.04 18-Feb-99
2323     ----------------------
2325     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2326     computation of the store needed to hold the pattern was incorrect (too large).
2327     If such patterns were nested a few deep, this could multiply and become a real
2328     problem.
2330     2. Added /M option to pcretest to show the memory requirement of a specific
2331     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2333     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2334     compiled in such a way that the backtracking after subsequent failure was
2335     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2336     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2339 nigel 29 Version 2.03 02-Feb-99
2340     ----------------------
2342     1. Fixed typo and small mistake in man page.
2344 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2345     LICENCE file containing the conditions.
2346 nigel 29
2347     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2348     Perl, that is the internal \ allows the delimiter to be included in the
2349     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2350     the final delimiter, add \ to the end of the pattern (to test the error).
2352     4. Added the convenience functions for extracting substrings after a successful
2353     match. Updated pcretest to make it able to test these functions.
2356 nigel 27 Version 2.02 14-Jan-99
2357     ----------------------
2359     1. Initialized the working variables associated with each extraction so that
2360     their saving and restoring doesn't refer to uninitialized store.
2362     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2363     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2364     fix the problem.
2366     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2367     calls, and wasn't printing the correct value for compiling calls. Increased the
2368     default value of LOOPREPEAT, and the number of significant figures in the
2369     times.
2371     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2373     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2374     a building problem on Windows NT with a FAT file system.
2377 nigel 25 Version 2.01 21-Oct-98
2378     ----------------------
2380     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2381     to character tables built by pcre_maketables() in the current locale. If NULL
2382     is passed, the default tables are used.
2385 nigel 23 Version 2.00 24-Sep-98
2386 nigel 21 ----------------------
2388 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2389     it any more.
2391     2. Allow quantification of (?>) groups, and make it work correctly.
2393     3. The first character computation wasn't working for (?>) groups.
2395     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2396     end of the subject) and add 5.005's \z, which really does match only at the
2397     very end of the subject.
2399     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2401     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2402     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2403     localized options. All options to pcre_study() were also removed.
2405     7. Add other new features from 5.005:
2407     $(?<= positive lookbehind
2408     $(?<! negative lookbehind
2409     (?imsx-imsx) added the unsetting capability
2410     such a setting is global if at outer level; local otherwise
2411     (?imsx-imsx:) non-capturing groups with option setting
2412     (?(cond)re|re) conditional pattern matching
2414     A backreference to itself in a repeated group matches the previous
2415     captured string.
2417     8. General tidying up of studying (both automatic and via "study")
2418     consequential on the addition of new assertions.
2420     9. As in 5.005, unlimited repeated groups that could match an empty substring
2421     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2422     runtime if any iteration does actually match an empty substring.
2424     10. Include the RunTest script in the distribution.
2426     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2427     discrepancies, some of which were old and were also with respect to 5.004. They
2428     have now been fixed.
2431     Version 1.09 28-Apr-98
2432     ----------------------
2434 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2435     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2436     program crashes, or just wrong answers. This did not apply to negated classes
2437     containing more than one character, or to minima other than one.
2440 nigel 19 Version 1.08 27-Mar-98
2441     ----------------------
2443     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2445     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2446     latter must appear before anything that relies on it in the pattern.
2449 nigel 17 Version 1.07 16-Feb-98
2450     ----------------------
2452     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2453     repeat of a potentially empty string).
2456 nigel 15 Version 1.06 23-Jan-98
2457     ----------------------
2459     1. Added Markus Oberhumer's little patches for C++.
2461     2. Literal strings longer than 255 characters were broken.
2464 nigel 13 Version 1.05 23-Dec-97
2465     ----------------------
2467     1. Negated character classes containing more than one character were failing if
2468     PCRE_CASELESS was set at run time.
2471 nigel 11 Version 1.04 19-Dec-97
2472     ----------------------
2474     1. Corrected the man page, where some "const" qualifiers had been omitted.
2476     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2477     input syntax.
2479     3. Fixed memory leak which occurred when a regex with back references was
2480     matched with an offsets vector that wasn't big enough. The temporary memory
2481     that is used in this case wasn't being freed if the match failed.
2483     4. Tidied pcretest to ensure it frees memory that it gets.
2485     5. Temporary memory was being obtained in the case where the passed offsets
2486     vector was exactly big enough.
2488     6. Corrected definition of offsetof() from change 5 below.
2490     7. I had screwed up change 6 below and broken the rules for the use of
2491     setjmp(). Now fixed.
2494 nigel 9 Version 1.03 18-Dec-97
2495     ----------------------
2497     1. A erroneous regex with a missing opening parenthesis was correctly
2498     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2499     on some systems.
2501     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2502     it was reported that one broken compiler failed on the former because "code" is
2503     also an independent variable.
2505     3. The erroneous regex a[]b caused an array overrun reference.
2507     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2508     fail on data ending with that character. (It was going on too far, and checking
2509     the next character, typically a binary zero.) This was specific to the
2510     optimized code for single-character negative classes.
2512     5. Added a contributed patch from the TIN world which does the following:
2514     + Add an undef for memmove, in case the the system defines a macro for it.
2516     + Add a definition of offsetof(), in case there isn't one. (I don't know
2517     the reason behind this - offsetof() is part of the ANSI standard - but
2518     it does no harm).
2520     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2521     most of the places where whitespace preceded '#'. I have given up and
2522     allowed the remaining 2 cases to be at the margin.
2524     + Rename some variables in pcre to eliminate shadowing. This seems very
2525     pedantic, but does no harm, of course.
2527     6. Moved the call to setjmp() into its own function, to get rid of warnings
2528     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2530     7. Constructs such as \d{8,} were compiling into the equivalent of
2531 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2532 nigel 9 outcome, but in this particular case used more store than had been allocated,
2533     which caused the bug to be discovered because it threw up an internal error.
2535     8. The debugging code in both pcre and pcretest for outputting the compiled
2536     form of a regex was going wrong in the case of back references followed by
2537     curly-bracketed repeats.
2540 nigel 7 Version 1.02 12-Dec-97
2541     ----------------------
2543     1. Typos in pcre.3 and comments in the source fixed.
2545     2. Applied a contributed patch to get rid of places where it used to remove
2546     'const' from variables, and fixed some signed/unsigned and uninitialized
2547     variable warnings.
2549     3. Added the "runtest" target to Makefile.
2551     4. Set default compiler flag to -O2 rather than just -O.
2554 nigel 5 Version 1.01 19-Nov-97
2555     ----------------------
2557     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2558     like /([ab]*)*/, that is, for classes with more than one character in them.
2560     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2561     as /((?>a*))*/ (a PCRE_EXTRA facility).
2564     Version 1.00 18-Nov-97
2565     ----------------------
2567     1. Added compile-time macros to support systems such as SunOS4 which don't have
2568     memmove() or strerror() but have other things that can be used instead.
2570     2. Arranged that "make clean" removes the executables.
2573 nigel 3 Version 0.99 27-Oct-97
2574     ----------------------
2576     1. Fixed bug in code for optimizing classes with only one character. It was
2577     initializing a 32-byte map regardless, which could cause it to run off the end
2578     of the memory it had got.
2580     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2583     Version 0.98 22-Oct-97
2584     ----------------------
2586     1. Fixed bug in code for handling temporary memory usage when there are more
2587     back references than supplied space in the ovector. This could cause segfaults.
2590     Version 0.97 21-Oct-97
2591     ----------------------
2593     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2595     2. Optimized negated single characters not to use a bit map.
2597     3. Brought error texts together as macro definitions; clarified some of them;
2598     fixed one that was wrong - it said "range out of order" when it meant "invalid
2599     escape sequence".
2601     4. Changed some char * arguments to const char *.
2603     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2605     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2606     pcretest.
2609     Version 0.96 16-Oct-97
2610     ----------------------
2612     1. Added a simple "pgrep" utility to the distribution.
2614     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2615     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2616     where "ddd" means "one or more decimal digits".
2618     3. Fixed serious bug. If a pattern had a back reference, but the call to
2619     pcre_exec() didn't supply a large enough ovector to record the related
2620     identifying subpattern, the match always failed. PCRE now remembers the number
2621     of the largest back reference, and gets some temporary memory in which to save
2622     the offsets during matching if necessary, in order to ensure that
2623     backreferences always work.
2625     4. Increased the compatibility with Perl in a number of ways:
2627     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2628     to request this handling. The option can be set at compile or exec time.
2630     (b) $ matches before a terminating newline by default; an option
2631     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2632     mode). The option can be set at compile or exec time.
2634     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2635     the same as Perl's. If the decimal number it represents is less than 10
2636     or there aren't that many previous left capturing parentheses, an octal
2637     escape is read. Inside a character class, it's always an octal escape,
2638     even if it is a single digit.
2640     (d) An escaped but undefined alphabetic character is taken as a literal,
2641     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2642     escapes.
2644     (e) {0} is now permitted. (The previous item is removed from the compiled
2645     pattern).
2647     5. Changed all the names of code files so that the basic parts are no longer
2648     than 10 characters, and abolished the teeny "globals.c" file.
2650     6. Changed the handling of character classes; they are now done with a 32-byte
2651     bit map always.
2653     7. Added the -d and /D options to pcretest to make it possible to look at the
2654     internals of compilation without having to recompile pcre.
2657     Version 0.95 23-Sep-97
2658     ----------------------
2660     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2661     \x20 at the start of a run of normal characters. These were being treated as
2662     real characters, instead of the source characters being re-checked.
2665     Version 0.94 18-Sep-97
2666     ----------------------
2668     1. The functions are now thread-safe, with the caveat that the global variables
2669     containing pointers to malloc() and free() or alternative functions are the
2670     same for all threads.
2672     2. Get pcre_study() to generate a bitmap of initial characters for non-
2673     anchored patterns when this is possible, and use it if passed to pcre_exec().
2676     Version 0.93 15-Sep-97
2677     ----------------------
2679     1. /(b)|(:+)/ was computing an incorrect first character.
2681     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2682     but not actually doing anything yet.
2684     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2685     as Perl does (e.g. [-az] or [az-]).
2687     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2688     all possible positions.
2690     5. Split up into different modules to avoid including unneeded functions in a
2691     compiled binary. However, compile and exec are still in one module. The "study"
2692     function is split off.
2694     6. The character tables are now in a separate module whose source is generated
2695     by an auxiliary program - but can then be edited by hand if required. There are
2696     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2697     toupper() in the code.
2699     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2700     make them global. Abolish the function for setting them, as the caller can now
2701     set them directly.
2704     Version 0.92 11-Sep-97
2705     ----------------------
2707     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2708     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2710     2. Caseless matching was not working in character classes if the characters in
2711     the pattern were in upper case.
2713     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2715     4. Make PCRE_ANCHORED public and accept as a compile option.
2717     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2718     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2719     pass them.
2721     6. Give an error if bad option bits passed at compile or run time.
2723     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2724     pcretest to cause it to pass that flag.
2726     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2727     options, and the first character, if set.
2729     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2732     Version 0.91 10-Sep-97
2733     ----------------------
2735     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2736     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2738     2. PCRE was looping on encountering an indefinitely repeated back reference to
2739     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2740     Perl does - treats the match as successful.
2742     ****


Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

ViewVC Help
Powered by ViewVC 1.1.12