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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 210 - (hide annotations) (download)
Wed Aug 8 14:24:50 2007 UTC (7 years, 1 month ago) by ph10
File size: 134972 byte(s)
Add Perl 5.10's backtracking verbs.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 189 Version 7.3 05-Jul-07
5 ph10 157 ---------------------
6    
7 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
8     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
9     brackets there is not right, since it causes compilers to look for an
10     installed pcre.h, not the version that is in the source that is being
11     compiled (which of course may be different). I have changed it back to:
12 ph10 197
13 ph10 189 #include "pcre.h"
14 ph10 197
15     I have a vague recollection that the change was concerned with compiling in
16     different directories, but in the new build system, that is taken care of
17     by the VPATH setting the Makefile.
18    
19 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
20     when the subject happened to end in the byte 0x85 (e.g. if the last
21     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
22     characters but of course it shouldn't be taken as a newline when it is part
23     of another character. The bug was that, for an unlimited repeat of . in
24     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
25     characters when looking for a newline.
26 ph10 197
27     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
28    
29     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
30     in debug output.
31    
32 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
33     long printing in the pcrecpp unittest when running under MinGW.
34 ph10 189
35 ph10 197 6. ESC_K was left out of the EBCDIC table.
36 ph10 189
37 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
38     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
39     limit also applies to "virtual nesting" when a pattern is recursive, and in
40     this case 1000 isn't so big. I have been able to remove this limit at the
41     expense of backing off one optimization in certain circumstances. Normally,
42     when pcre_exec() would call its internal match() function recursively and
43     immediately return the result unconditionally, it uses a "tail recursion"
44     feature to save stack. However, when a subpattern that can match an empty
45     string has an unlimited repetition quantifier, it no longer makes this
46     optimization. That gives it a stack frame in which to save the data for
47     checking that an empty string has been matched. Previously this was taken
48     from the 1000-entry workspace that had been reserved. So now there is no
49     explicit limit, but more stack is used.
50 ph10 189
51 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
52     syntax that is required for Windows, and which was going wrong for the
53     pcreposix and pcrecpp parts of the library. These were overlooked when this
54     problem was solved for the main library.
55 ph10 197
56 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
57     the size of patterns that contain repeated groups with explicit upper
58     limits. As the maximum quantifier is 65535, the maximum group length was
59     set at 30,000 so that the product of these two numbers did not overflow a
60     32-bit integer. However, it turns out that people want to use groups that
61     are longer than 30,000 bytes (though not repeat them that many times).
62     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
63     made it possible to implement the integer overflow checks in a much more
64     dynamic way, which I have now done. The artificial limitation on group
65     length has been removed - we now have only the limit on the total length of
66     the compiled pattern, which depends on the LINK_SIZE setting.
67 ph10 208
68     10. Fixed a bug in the documentation for get/copy named substring when
69     duplicate names are permitted. If none of the named substrings are set, the
70     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
71     empty string.
72    
73     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
74     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
75     because the ] is interpreted as the first data character and the
76     terminating ] is not found. PCRE has been made compatible with Perl in this
77     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
78     cause memory overwriting.
79    
80 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
81     string has been matched (to stop an infinite loop). It was not recognizing
82 ph10 208 a conditional subpattern that could match an empty string if that
83 ph10 206 subpattern was within another subpattern. For example, it looped when
84 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
85 ph10 206 condition was not nested. This bug has been fixed.
86 ph10 208
87 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
88     past the start of the subject in the presence of bytes with the top bit
89     set, for example "\x8aBCD".
90 ph10 210
91     13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
92     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
93 ph10 200
94 ph10 210 14. Optimized (?!) to (*FAIL).
95 ph10 202
96 ph10 210
97 ph10 189 Version 7.2 19-Jun-07
98     ---------------------
99    
100 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
101     which is apparently normally available under Windows.
102 ph10 159
103     2. Re-jig the pcregrep tests with different newline settings in an attempt
104     to make them independent of the local environment's newline setting.
105 ph10 157
106 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
107 ph10 165
108 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
109     was not 2, because the output contained actual offsets. The recent new
110 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
111     usable with all link sizes.
112    
113 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
114     stack recursion. This gives a massive performance boost under BSD, but just
115 ph10 165 a small improvement under Linux. However, it saves one field in the frame
116 ph10 164 in all cases.
117 ph10 172
118 ph10 166 6. Added more features from the forthcoming Perl 5.10:
119 ph10 172
120 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
121     recursion call. It refers to the nth most recently opened parentheses.
122 ph10 172
123 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
124 ph10 172 to be opened parentheses.
125    
126     (c) Conditions that refer to capturing parentheses can be specified
127 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
128 ph10 172
129 ph10 168 (d) \K resets the start of the current match so that everything before
130 ph10 172 is not part of it.
131    
132 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
133 ph10 172
134 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
135 ph10 172 reference syntax.
136 ph10 182
137 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
138 ph10 182 alternative starts with the same number.
139 ph10 172
140 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
141    
142 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
143     PCRE_INFO_JCHANGED.
144    
145     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
146     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
147 ph10 170 for detecting groups that can match an empty string.
148 ph10 159
149 ph10 172 9. A pattern with a very large number of alternatives (more than several
150     hundred) was running out of internal workspace during the pre-compile
151     phase, where pcre_compile() figures out how much memory will be needed. A
152     bit of new cunning has reduced the workspace needed for groups with
153     alternatives. The 1000-alternative test pattern now uses 12 bytes of
154     workspace instead of running out of the 4096 that are available.
155 ph10 182
156 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
157 ph10 172
158 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
159     The report of the bug said:
160 ph10 182
161 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
162     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
163 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
164 ph10 185
165 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
166 ph10 185 it matched the wrong number of bytes.
167 ph10 172
168 ph10 179
169 ph10 155 Version 7.1 24-Apr-07
170 ph10 98 ---------------------
171    
172 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
173 ph10 122 that is more "standard", making use of automake and other Autotools. There
174 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
175     on this.
176 ph10 111
177     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
178     for recursive directory scanning broke on some systems because the files
179     are not scanned in any specific order and on different systems the order
180     was different. A call to "sort" has been inserted into RunGrepTest for the
181     approprate test as a short-term fix. In the longer term there may be an
182 ph10 100 alternative.
183 ph10 111
184 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
185 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
186     people do their own conversions for various reasons). The problems
187     concerned the use of low-level troff macros .br and .in. I have therefore
188     removed all such uses from the man pages (some were redundant, some could
189 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
190     HTML has been updated to handle .nf/.fi and to complain if it encounters
191     .br or .in.
192 ph10 111
193 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
194 ph10 123 arranged for config.h to be included in the distribution, with the name
195 ph10 111 config.h.generic, for the benefit of those who have to compile without
196     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
197    
198     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
199     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
200 ph10 127 makevp.bat for the current PCRE, using the additional files
201 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
202 ph10 111
203     6. A Windows user reported a minor discrepancy with test 2, which turned out
204     to be caused by a trailing space on an input line that had got lost in his
205 ph10 102 copy. The trailing space was an accident, so I've just removed it.
206 ph10 111
207 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
208 ph10 111 that is needed.
209    
210 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
211     as "const" (a) because they are and (b) because it helps the PHP
212     maintainers who have recently made a script to detect big data structures
213 ph10 111 in the php code that should be moved to the .rodata section. I remembered
214     to update Builducptable as well, so it won't revert if ucptable.h is ever
215 ph10 105 re-created.
216 ph10 111
217     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
218 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
219 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
220     support is not required. This saves 1.5-2K of code, which is important in
221 ph10 107 some applications.
222 ph10 111
223 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
224 ph10 111 so as not to refer to the tables, even though these functions will never be
225     called when UTF-8 support is disabled. Otherwise there are problems with a
226     shared library.
227    
228 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
229    
230     (a) It was defining its arguments as char * instead of void *.
231    
232 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
233     a long time ago when I wrote it, but is no longer the case.
234    
235 ph10 118 The emulated memove() is provided for those environments that have neither
236 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
237 ph10 118 is clearly not the case, as these two bugs were recently reported.
238 ph10 123
239 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
240 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
241     pages, and it removes trailing spaces from listed files. It also creates
242     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
243     case, it wraps all the #defines with #ifndefs. This script should be run
244 ph10 111 before "make dist".
245 ph10 123
246 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
247     with Unicode property support.
248 ph10 123
249     (a) For a maximizing quantifier, if the two different cases of the
250     character were of different lengths in their UTF-8 codings (there are
251     some cases like this - I found 11), and the matching function had to
252 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
253     were both the same length.
254 ph10 123
255     (b) When PCRE was configured to use the heap rather than the stack for
256     recursion during matching, it was not correctly preserving the data for
257     the other case of a UTF-8 character when checking ahead for a match
258     while processing a minimizing repeat. If the check also involved
259 ph10 115 matching a wide character, but failed, corruption could cause an
260     erroneous result when trying to check for a repeat of the original
261     character.
262 ph10 123
263 ph10 116 13. Some tidying changes to the testing mechanism:
264 ph10 98
265 ph10 116 (a) The RunTest script now detects the internal link size and whether there
266     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
267 ph10 123 values substituted by "configure". (The RunGrepTest script already did
268     this for UTF-8.) The configure.ac script no longer substitutes the
269     relevant variables.
270    
271 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
272     with length and offset values. This means that the output is different
273     for different internal link sizes. Test 2 is skipped for link sizes
274     other than 2 because of this, bypassing the problem. Unfortunately,
275     there was also a test in test 3 (the locale tests) that used /B and
276 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
277     I have added a new /Z option to pcretest that replaces the length and
278     offset values with spaces. This is now used to make test 3 independent
279 ph10 122 of link size. (Test 2 will be tidied up later.)
280 ph10 123
281     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
282 ph10 122 segmentation fault instead of returning the appropriate error message.
283 ph10 134
284 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
285 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
286     This doesn't seem right; it now treats the CRLF combination as the line
287     ending, and so does not match in that case. It's only a pattern such as ^$
288     that would hit this one: something like ^ABC$ would have failed after \r
289     and then tried again after \r\n.
290    
291 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
292 ph10 134 in an attempt to make files that differ only in their line terminators
293     compare equal. This works on Linux.
294 ph10 142
295 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
296     as it exited. This is now fixed, thanks to valgrind.
297 ph10 142
298 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
299 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
300     was because its rules for how to advance for /g after matching an empty
301 ph10 143 string at the end of a line did not allow for this case. They now check for
302     it specially.
303 ph10 150
304     20. pcretest is supposed to handle patterns and data of any length, by
305     extending its buffers when necessary. It was getting this wrong when the
306 ph10 147 buffer for a data line had to be extended.
307 ph10 150
308 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
309 ph10 150 CRLF as a newline sequence.
310 ph10 152
311 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
312 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
313     I have nevertheless tidied it up.
314 ph10 111
315 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
316 ph10 123
317 ph10 153 24. Added a man page for pcre-config.
318 ph10 152
319 ph10 153
320 nigel 93 Version 7.0 19-Dec-06
321     ---------------------
322    
323     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
324     moving to gcc 4.1.1.
325    
326     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
327     sys/time.h, which is documented as needed for this function. It doesn't
328     seem to matter on Linux, but it showed up on some releases of OS X.
329    
330     3. It seems that there are systems where bytes whose values are greater than
331     127 match isprint() in the "C" locale. The "C" locale should be the
332     default when a C program starts up. In most systems, only ASCII printing
333     characters match isprint(). This difference caused the output from pcretest
334     to vary, making some of the tests fail. I have changed pcretest so that:
335    
336     (a) When it is outputting text in the compiled version of a pattern, bytes
337     other than 32-126 are always shown as hex escapes.
338    
339     (b) When it is outputting text that is a matched part of a subject string,
340     it does the same, unless a different locale has been set for the match
341     (using the /L modifier). In this case, it uses isprint() to decide.
342    
343     4. Fixed a major bug that caused incorrect computation of the amount of memory
344     required for a compiled pattern when options that changed within the
345     pattern affected the logic of the preliminary scan that determines the
346     length. The relevant options are -x, and -i in UTF-8 mode. The result was
347     that the computed length was too small. The symptoms of this bug were
348     either the PCRE error "internal error: code overflow" from pcre_compile(),
349     or a glibc crash with a message such as "pcretest: free(): invalid next
350     size (fast)". Examples of patterns that provoked this bug (shown in
351     pcretest format) are:
352    
353     /(?-x: )/x
354     /(?x)(?-x: \s*#\s*)/
355     /((?i)[\x{c0}])/8
356     /(?i:[\x{c0}])/8
357    
358     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
359     is now done differently.
360    
361     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
362     wrapper classes; (b) implement a new function in the C++ scanner that is
363     more efficient than the old way of doing things because it avoids levels of
364     recursion in the regex matching; (c) add a paragraph to the documentation
365     for the FullMatch() function.
366    
367     6. The escape sequence \n was being treated as whatever was defined as
368     "newline". Not only was this contrary to the documentation, which states
369     that \n is character 10 (hex 0A), but it also went horribly wrong when
370     "newline" was defined as CRLF. This has been fixed.
371    
372     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
373     was being set to -1 for the "end of line" case (supposedly a value that no
374     character can have). Though this value is never used (the check for end of
375     line is "zero bytes in current character"), it caused compiler complaints.
376     I've changed it to 0xffffffff.
377    
378     8. In pcre_version.c, the version string was being built by a sequence of
379     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
380     string (as it is for production releases) called a macro with an empty
381     argument. The C standard says the result of this is undefined. The gcc
382     compiler treats it as an empty string (which was what was wanted) but it is
383     reported that Visual C gives an error. The source has been hacked around to
384     avoid this problem.
385    
386     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
387     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
388     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
389     of them did).
390    
391     10. Originally, pcretest opened its input and output without "b"; then I was
392     told that "b" was needed in some environments, so it was added for release
393     5.0 to both the input and output. (It makes no difference on Unix-like
394     systems.) Later I was told that it is wrong for the input on Windows. I've
395     now abstracted the modes into two macros, to make it easier to fiddle with
396     them, and removed "b" from the input mode under Windows.
397    
398     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
399    
400     12. Added -help and --help to pcretest as an official way of being reminded
401     of the options.
402    
403     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
404     and pcrecpp.cc because they annoy compilers at high warning levels.
405    
406     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
407    
408     15. Fixed an occurrence of == in configure.ac that should have been = (shell
409     scripts are not C programs :-) and which was not noticed because it works
410     on Linux.
411    
412     16. pcretest is supposed to handle any length of pattern and data line (as one
413     line or as a continued sequence of lines) by extending its input buffer if
414     necessary. This feature was broken for very long pattern lines, leading to
415     a string of junk being passed to pcre_compile() if the pattern was longer
416     than about 50K.
417    
418     17. I have done a major re-factoring of the way pcre_compile() computes the
419     amount of memory needed for a compiled pattern. Previously, there was code
420     that made a preliminary scan of the pattern in order to do this. That was
421     OK when PCRE was new, but as the facilities have expanded, it has become
422     harder and harder to keep it in step with the real compile phase, and there
423     have been a number of bugs (see for example, 4 above). I have now found a
424     cunning way of running the real compile function in a "fake" mode that
425     enables it to compute how much memory it would need, while actually only
426     ever using a few hundred bytes of working memory and without too many
427     tests of the mode. This should make future maintenance and development
428     easier. A side effect of this work is that the limit of 200 on the nesting
429     depth of parentheses has been removed (though this was never a serious
430     limitation, I suspect). However, there is a downside: pcre_compile() now
431     runs more slowly than before (30% or more, depending on the pattern). I
432     hope this isn't a big issue. There is no effect on runtime performance.
433    
434     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
435     newline (only possible for the last line of a file) and it was a
436     pattern that set a locale (followed by /Lsomething), pcretest crashed.
437    
438     19. Added additional timing features to pcretest. (1) The -tm option now times
439     matching only, not compiling. (2) Both -t and -tm can be followed, as a
440     separate command line item, by a number that specifies the number of
441     repeats to use when timing. The default is 50000; this gives better
442     precision, but takes uncomfortably long for very large patterns.
443    
444     20. Extended pcre_study() to be more clever in cases where a branch of a
445     subpattern has no definite first character. For example, (a*|b*)[cd] would
446     previously give no result from pcre_study(). Now it recognizes that the
447     first character must be a, b, c, or d.
448    
449     21. There was an incorrect error "recursive call could loop indefinitely" if
450     a subpattern (or the entire pattern) that was being tested for matching an
451     empty string contained only one non-empty item after a nested subpattern.
452     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
453     incorrectly, because the \d was being skipped in the check.
454    
455     22. The pcretest program now has a new pattern option /B and a command line
456     option -b, which is equivalent to adding /B to every pattern. This causes
457     it to show the compiled bytecode, without the additional information that
458     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
459     is the same as /B/I).
460    
461     23. A new optimization is now able automatically to treat some sequences such
462     as a*b as a*+b. More specifically, if something simple (such as a character
463     or a simple class like \d) has an unlimited quantifier, and is followed by
464     something that cannot possibly match the quantified thing, the quantifier
465     is automatically "possessified".
466    
467     24. A recursive reference to a subpattern whose number was greater than 39
468     went wrong under certain circumstances in UTF-8 mode. This bug could also
469     have affected the operation of pcre_study().
470    
471     25. Realized that a little bit of performance could be had by replacing
472     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
473    
474     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
475    
476     27. Possessive quantifiers such as a++ were previously implemented by turning
477     them into atomic groups such as ($>a+). Now they have their own opcodes,
478     which improves performance. This includes the automatically created ones
479     from 23 above.
480    
481     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
482     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
483     the first matched character to be a colon. This applied both to named and
484     numbered groups.
485    
486     29. The ucpinternal.h header file was missing its idempotency #ifdef.
487    
488     30. I was sent a "project" file called libpcre.a.dev which I understand makes
489     building PCRE on Windows easier, so I have included it in the distribution.
490    
491     31. There is now a check in pcretest against a ridiculously large number being
492     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
493     loop, the loop is abandoned.
494    
495     32. Forward references to subpatterns in conditions such as (?(2)...) where
496     subpattern 2 is defined later cause pcre_compile() to search forwards in
497     the pattern for the relevant set of parentheses. This search went wrong
498     when there were unescaped parentheses in a character class, parentheses
499     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
500    
501     33. "Subroutine" calls and backreferences were previously restricted to
502     referencing subpatterns earlier in the regex. This restriction has now
503     been removed.
504    
505     34. Added a number of extra features that are going to be in Perl 5.10. On the
506     whole, these are just syntactic alternatives for features that PCRE had
507     previously implemented using the Python syntax or my own invention. The
508     other formats are all retained for compatibility.
509    
510     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
511     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
512     also .NET compatible.
513    
514     (b) A recursion or subroutine call to a named group can now be defined as
515     (?&name) as well as (?P>name).
516    
517     (c) A backreference to a named group can now be defined as \k<name> or
518     \k'name' as well as (?P=name). The new forms, as well as being in Perl
519     5.10, are also .NET compatible.
520    
521     (d) A conditional reference to a named group can now use the syntax
522     (?(<name>) or (?('name') as well as (?(name).
523    
524     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
525     groups (named and numbered) that are never evaluated inline, but can be
526     called as "subroutines" from elsewhere. In effect, the DEFINE condition
527     is always false. There may be only one alternative in such a group.
528    
529     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
530     as the simple (?(R). The condition is true only if the most recent
531     recursion is that of the given number or name. It does not search out
532     through the entire recursion stack.
533    
534     (g) The escape \gN or \g{N} has been added, where N is a positive or
535     negative number, specifying an absolute or relative reference.
536    
537     35. Tidied to get rid of some further signed/unsigned compiler warnings and
538     some "unreachable code" warnings.
539    
540     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
541     things, this adds five new scripts.
542    
543     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
544     There were also incompatibilities regarding the handling of \Q..\E inside
545     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
546     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
547    
548     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
549     matches an empty string, and forcibly breaks the loop. There were bugs in
550     this code in non-simple cases. For a pattern such as ^(a()*)* matched
551     against aaaa the result was just "a" rather than "aaaa", for example. Two
552     separate and independent bugs (that affected different cases) have been
553     fixed.
554    
555     39. Refactored the code to abolish the use of different opcodes for small
556     capturing bracket numbers. This is a tidy that I avoided doing when I
557     removed the limit on the number of capturing brackets for 3.5 back in 2001.
558     The new approach is not only tidier, it makes it possible to reduce the
559     memory needed to fix the previous bug (38).
560    
561     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
562     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
563     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
564     mode.
565    
566     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
567     report.
568    
569     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
570     copy construction and assignment in the C++ wrapper.
571    
572     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
573     couple of bugs that could have given wrong results in the "--newline=crlf"
574     case.
575    
576     44. Added a number of casts and did some reorganization of signed/unsigned int
577     variables following suggestions from Dair Grant. Also renamed the variable
578     "this" as "item" because it is a C++ keyword.
579    
580     45. Arranged for dftables to add
581    
582     #include "pcre_internal.h"
583    
584     to pcre_chartables.c because without it, gcc 4.x may remove the array
585     definition from the final binary if PCRE is built into a static library and
586     dead code stripping is activated.
587    
588     46. For an unanchored pattern, if a match attempt fails at the start of a
589     newline sequence, and the newline setting is CRLF or ANY, and the next two
590     characters are CRLF, advance by two characters instead of one.
591    
592    
593 nigel 91 Version 6.7 04-Jul-06
594     ---------------------
595    
596     1. In order to handle tests when input lines are enormously long, pcretest has
597     been re-factored so that it automatically extends its buffers when
598     necessary. The code is crude, but this _is_ just a test program. The
599     default size has been increased from 32K to 50K.
600    
601     2. The code in pcre_study() was using the value of the re argument before
602     testing it for NULL. (Of course, in any sensible call of the function, it
603     won't be NULL.)
604    
605     3. The memmove() emulation function in pcre_internal.h, which is used on
606     systems that lack both memmove() and bcopy() - that is, hardly ever -
607     was missing a "static" storage class specifier.
608    
609     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
610     containing an extended class (one that cannot be represented by a bitmap
611     because it contains high-valued characters or Unicode property items, e.g.
612     [\pZ]). Almost always one would set UTF-8 mode when processing such a
613     pattern, but PCRE should not loop if you do not (it no longer does).
614     [Detail: two cases were found: (a) a repeated subpattern containing an
615     extended class; (b) a recursive reference to a subpattern that followed a
616     previous extended class. It wasn't skipping over the extended class
617     correctly when UTF-8 mode was not set.]
618    
619     5. A negated single-character class was not being recognized as fixed-length
620     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
621     compile error "lookbehind assertion is not fixed length".
622    
623     6. The RunPerlTest auxiliary script was showing an unexpected difference
624     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
625     write a Perl script that can interpret lines of an input file either as
626     byte characters or as UTF-8, which is what "perltest" was being required to
627     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
628     can't do is switch easily at run time between having the "use utf8;" pragma
629     or not. In the end, I fudged it by using the RunPerlTest script to insert
630     "use utf8;" explicitly for the UTF-8 tests.
631    
632     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
633     the end of the subject string, contrary to the documentation and to what
634     Perl does. This was true of both matching functions. Now it matches only at
635     the start of the subject and immediately after *internal* newlines.
636    
637     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
638     a pointer to an int instead of a pointer to an unsigned long int. This
639     caused problems on 64-bit systems.
640    
641     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
642     instance of the 'standard' template library not being so standard".
643    
644     10. There was no check on the number of named subpatterns nor the maximum
645     length of a subpattern name. The product of these values is used to compute
646     the size of the memory block for a compiled pattern. By supplying a very
647     long subpattern name and a large number of named subpatterns, the size
648     computation could be caused to overflow. This is now prevented by limiting
649     the length of names to 32 characters, and the number of named subpatterns
650     to 10,000.
651    
652     11. Subpatterns that are repeated with specific counts have to be replicated in
653     the compiled pattern. The size of memory for this was computed from the
654     length of the subpattern and the repeat count. The latter is limited to
655     65535, but there was no limit on the former, meaning that integer overflow
656     could in principle occur. The compiled length of a repeated subpattern is
657     now limited to 30,000 bytes in order to prevent this.
658    
659     12. Added the optional facility to have named substrings with the same name.
660    
661     13. Added the ability to use a named substring as a condition, using the
662     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
663     are numbers (not recommended). Forward references are permitted.
664    
665     14. Added forward references in named backreferences (if you see what I mean).
666    
667     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
668     pattern could run off the end of the subject. For example, the pattern
669     "(?s)(.{1,5})"8 did this with the subject "ab".
670    
671     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
672     PCRE_CASELESS was set when matching characters that were quantified with ?
673     or *.
674    
675     17. A character class other than a single negated character that had a minimum
676     but no maximum quantifier - for example [ab]{6,} - was not handled
677     correctly by pce_dfa_exec(). It would match only one character.
678    
679     18. A valid (though odd) pattern that looked like a POSIX character
680     class but used an invalid character after [ (for example [[,abc,]]) caused
681     pcre_compile() to give the error "Failed: internal error: code overflow" or
682     in some cases to crash with a glibc free() error. This could even happen if
683     the pattern terminated after [[ but there just happened to be a sequence of
684     letters, a binary zero, and a closing ] in the memory that followed.
685    
686     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
687     over the years. Originally (before any Unicode support), just the bottom 8
688     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
689     output from "man perlunicode" includes this:
690    
691     The regular expression compiler produces polymorphic opcodes. That
692     is, the pattern adapts to the data and automatically switches to
693     the Unicode character scheme when presented with Unicode data--or
694     instead uses a traditional byte scheme when presented with byte
695     data.
696    
697     Sadly, a wide octal escape does not cause a switch, and in a string with
698     no other multibyte characters, these octal escapes are treated as before.
699     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
700     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
701     Unicode string.
702    
703     I have not perpetrated such confusion in PCRE. Up till now, it took just
704     the bottom 8 bits, as in old Perl. I have now made octal escapes with
705     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
706     translate to the appropriate multibyte character.
707    
708     29. Applied some refactoring to reduce the number of warnings from Microsoft
709     and Borland compilers. This has included removing the fudge introduced
710     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
711     a warning about an unused variable.
712    
713     21. PCRE has not included VT (character 0x0b) in the set of whitespace
714     characters since release 4.0, because Perl (from release 5.004) does not.
715     [Or at least, is documented not to: some releases seem to be in conflict
716     with the documentation.] However, when a pattern was studied with
717     pcre_study() and all its branches started with \s, PCRE still included VT
718     as a possible starting character. Of course, this did no harm; it just
719     caused an unnecessary match attempt.
720    
721     22. Removed a now-redundant internal flag bit that recorded the fact that case
722     dependency changed within the pattern. This was once needed for "required
723     byte" processing, but is no longer used. This recovers a now-scarce options
724     bit. Also moved the least significant internal flag bit to the most-
725     significant bit of the word, which was not previously used (hangover from
726     the days when it was an int rather than a uint) to free up another bit for
727     the future.
728    
729     23. Added support for CRLF line endings as well as CR and LF. As well as the
730     default being selectable at build time, it can now be changed at runtime
731     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
732     specify that it is scanning data with non-default line endings.
733    
734     24. Changed the definition of CXXLINK to make it agree with the definition of
735     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
736    
737     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
738     recursions. This makes a big different to stack usage for some patterns.
739    
740     26. If a subpattern containing a named recursion or subroutine reference such
741     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
742     the space required for the compiled pattern went wrong and gave too small a
743     value. Depending on the environment, this could lead to "Failed: internal
744     error: code overflow at offset 49" or "glibc detected double free or
745     corruption" errors.
746    
747     27. Applied patches from Google (a) to support the new newline modes and (b) to
748     advance over multibyte UTF-8 characters in GlobalReplace.
749    
750     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
751     difference for some implementation of PCRE in some Windows version.
752    
753     29. Added some extra testing facilities to pcretest:
754    
755     \q<number> in a data line sets the "match limit" value
756     \Q<number> in a data line sets the "match recursion limt" value
757     -S <number> sets the stack size, where <number> is in megabytes
758    
759     The -S option isn't available for Windows.
760    
761    
762 nigel 89 Version 6.6 06-Feb-06
763     ---------------------
764    
765     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
766     in pcreposix.h. I have copied the definition from pcre.h.
767    
768     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
769     because pcre.h is no longer a built file.
770    
771     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
772     not normally included in the compiled code.
773    
774    
775 nigel 87 Version 6.5 01-Feb-06
776     ---------------------
777    
778     1. When using the partial match feature with pcre_dfa_exec(), it was not
779     anchoring the second and subsequent partial matches at the new starting
780     point. This could lead to incorrect results. For example, with the pattern
781     /1234/, partially matching against "123" and then "a4" gave a match.
782    
783     2. Changes to pcregrep:
784    
785     (a) All non-match returns from pcre_exec() were being treated as failures
786     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
787     error message is output. Some extra information is given for the
788     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
789     probably the only errors that are likely to be caused by users (by
790     specifying a regex that has nested indefinite repeats, for instance).
791     If there are more than 20 of these errors, pcregrep is abandoned.
792    
793     (b) A binary zero was treated as data while matching, but terminated the
794     output line if it was written out. This has been fixed: binary zeroes
795     are now no different to any other data bytes.
796    
797     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
798     used to set a locale for matching. The --locale=xxxx long option has
799     been added (no short equivalent) to specify a locale explicitly on the
800     pcregrep command, overriding the environment variables.
801    
802     (d) When -B was used with -n, some line numbers in the output were one less
803     than they should have been.
804    
805     (e) Added the -o (--only-matching) option.
806    
807     (f) If -A or -C was used with -c (count only), some lines of context were
808     accidentally printed for the final match.
809    
810     (g) Added the -H (--with-filename) option.
811    
812     (h) The combination of options -rh failed to suppress file names for files
813     that were found from directory arguments.
814    
815     (i) Added the -D (--devices) and -d (--directories) options.
816    
817     (j) Added the -F (--fixed-strings) option.
818    
819     (k) Allow "-" to be used as a file name for -f as well as for a data file.
820    
821     (l) Added the --colo(u)r option.
822    
823     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
824     is not present by default.
825    
826     3. A nasty bug was discovered in the handling of recursive patterns, that is,
827     items such as (?R) or (?1), when the recursion could match a number of
828     alternatives. If it matched one of the alternatives, but subsequently,
829     outside the recursion, there was a failure, the code tried to back up into
830     the recursion. However, because of the way PCRE is implemented, this is not
831     possible, and the result was an incorrect result from the match.
832    
833     In order to prevent this happening, the specification of recursion has
834     been changed so that all such subpatterns are automatically treated as
835     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
836    
837     4. I had overlooked the fact that, in some locales, there are characters for
838     which isalpha() is true but neither isupper() nor islower() are true. In
839     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
840     and ordfeminine) are like this. This affected the treatment of \w and \W
841     when they appeared in character classes, but not when they appeared outside
842     a character class. The bit map for "word" characters is now created
843     separately from the results of isalnum() instead of just taking it from the
844     upper, lower, and digit maps. (Plus the underscore character, of course.)
845    
846     5. The above bug also affected the handling of POSIX character classes such as
847     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
848     permanent tables. Instead, the bit maps for such a class were previously
849     created as the appropriate unions of the upper, lower, and digit bitmaps.
850     Now they are created by subtraction from the [[:word:]] class, which has
851     its own bitmap.
852    
853     6. The [[:blank:]] character class matches horizontal, but not vertical space.
854     It is created by subtracting the vertical space characters (\x09, \x0a,
855     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
856     subtraction was done in the overall bitmap for a character class, meaning
857     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
858     be recognized. This bug has been fixed.
859    
860     7. Patches from the folks at Google:
861    
862     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
863     real life, but is still worth protecting against".
864    
865     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
866     regular expressions".
867    
868     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
869     have it.
870    
871     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
872     "configure" and the latter not, in order to fix a problem somebody had
873     with compiling the Arg class on HP-UX.
874    
875     (e) Improve the error-handling of the C++ wrapper a little bit.
876    
877     (f) New tests for checking recursion limiting.
878    
879     8. The pcre_memmove() function, which is used only if the environment does not
880     have a standard memmove() function (and is therefore rarely compiled),
881     contained two bugs: (a) use of int instead of size_t, and (b) it was not
882     returning a result (though PCRE never actually uses the result).
883    
884     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
885     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
886     returned instead of calling malloc() with an overflowing number that would
887     most likely cause subsequent chaos.
888    
889     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
890    
891     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
892     with this option is matched, the nmatch and pmatch options of regexec() are
893     ignored.
894    
895     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
896     provided in case anyone wants to the the POSIX interface with UTF-8
897     strings.
898    
899     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
900     C++ linking (needed for some HP-UX environments).
901    
902     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
903     (unused parameter) and in the pcre_printint() function (omitted "default"
904     switch label when the default is to do nothing).
905    
906     15. Added some code to make it possible, when PCRE is compiled as a C++
907     library, to replace subject pointers for pcre_exec() with a smart pointer
908     class, thus making it possible to process discontinuous strings.
909    
910     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
911     much the same function. They were added by different people who were trying
912     to make PCRE easy to compile on non-Unix systems. It has been suggested
913     that PCRE_EXPORT be abolished now that there is more automatic apparatus
914     for compiling on Windows systems. I have therefore replaced it with
915     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
916     defaults to "extern" for C or "extern C" for C++, which works fine on
917     Unix-like systems. It is now possible to override the value of PCRE_DATA_
918     SCOPE with something explicit in config.h. In addition:
919    
920     (a) pcreposix.h still had just "extern" instead of either of these macros;
921     I have replaced it with PCRE_DATA_SCOPE.
922    
923     (b) Functions such as _pcre_xclass(), which are internal to the library,
924     but external in the C sense, all had PCRE_EXPORT in their definitions.
925     This is apparently wrong for the Windows case, so I have removed it.
926     (It makes no difference on Unix-like systems.)
927    
928     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
929     of recursive calls to match(). This is different to MATCH_LIMIT because
930     that limits the total number of calls to match(), not all of which increase
931     the depth of recursion. Limiting the recursion depth limits the amount of
932     stack (or heap if NO_RECURSE is set) that is used. The default can be set
933     when PCRE is compiled, and changed at run time. A patch from Google adds
934     this functionality to the C++ interface.
935    
936     18. Changes to the handling of Unicode character properties:
937    
938     (a) Updated the table to Unicode 4.1.0.
939    
940     (b) Recognize characters that are not in the table as "Cn" (undefined).
941    
942     (c) I revised the way the table is implemented to a much improved format
943     which includes recognition of ranges. It now supports the ranges that
944     are defined in UnicodeData.txt, and it also amalgamates other
945     characters into ranges. This has reduced the number of entries in the
946     table from around 16,000 to around 3,000, thus reducing its size
947     considerably. I realized I did not need to use a tree structure after
948     all - a binary chop search is just as efficient. Having reduced the
949     number of entries, I extended their size from 6 bytes to 8 bytes to
950     allow for more data.
951    
952     (d) Added support for Unicode script names via properties such as \p{Han}.
953    
954     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
955     matching that character.
956    
957     20. When matching a repeated Unicode property with a minimum greater than zero,
958     (for example \pL{2,}), PCRE could look past the end of the subject if it
959     reached it while seeking the minimum number of characters. This could
960     happen only if some of the characters were more than one byte long, because
961     there is a check for at least the minimum number of bytes.
962    
963     21. Refactored the implementation of \p and \P so as to be more general, to
964     allow for more different types of property in future. This has changed the
965     compiled form incompatibly. Anybody with saved compiled patterns that use
966     \p or \P will have to recompile them.
967    
968     22. Added "Any" and "L&" to the supported property types.
969    
970     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
971     but give a compile time error if the value is greater than 0xff.
972    
973     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
974     accidentally not being installed or uninstalled.
975    
976     25. The pcre.h file was built from pcre.h.in, but the only changes that were
977     made were to insert the current release number. This seemed silly, because
978     it made things harder for people building PCRE on systems that don't run
979     "configure". I have turned pcre.h into a distributed file, no longer built
980     by "configure", with the version identification directly included. There is
981     no longer a pcre.h.in file.
982    
983     However, this change necessitated a change to the pcre-config script as
984     well. It is built from pcre-config.in, and one of the substitutions was the
985     release number. I have updated configure.ac so that ./configure now finds
986     the release number by grepping pcre.h.
987    
988     26. Added the ability to run the tests under valgrind.
989    
990    
991 nigel 85 Version 6.4 05-Sep-05
992     ---------------------
993    
994     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
995     "--" to be printed when multiple files were scanned, even when none of the
996     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
997     consider it to be a bug, and have restored the previous behaviour.
998    
999     2. A couple of code tidies to get rid of compiler warnings.
1000    
1001     3. The pcretest program used to cheat by referring to symbols in the library
1002     whose names begin with _pcre_. These are internal symbols that are not
1003     really supposed to be visible externally, and in some environments it is
1004     possible to suppress them. The cheating is now confined to including
1005     certain files from the library's source, which is a bit cleaner.
1006    
1007     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1008     file's purpose clearer.
1009    
1010     5. Reorganized pcre_ucp_findchar().
1011    
1012    
1013 nigel 83 Version 6.3 15-Aug-05
1014     ---------------------
1015    
1016     1. The file libpcre.pc.in did not have general read permission in the tarball.
1017    
1018     2. There were some problems when building without C++ support:
1019    
1020     (a) If C++ support was not built, "make install" and "make test" still
1021     tried to test it.
1022    
1023     (b) There were problems when the value of CXX was explicitly set. Some
1024     changes have been made to try to fix these, and ...
1025    
1026     (c) --disable-cpp can now be used to explicitly disable C++ support.
1027    
1028     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1029     backslash in a target when C++ was disabled. This confuses some
1030     versions of "make", apparently. Using an intermediate variable solves
1031     this. (Same for CPP_LOBJ.)
1032    
1033     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1034     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1035     necessary on certain architectures.
1036    
1037     4. Added a setting of -export-symbols-regex to the link command to remove
1038     those symbols that are exported in the C sense, but actually are local
1039     within the library, and not documented. Their names all begin with
1040     "_pcre_". This is not a perfect job, because (a) we have to except some
1041     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1042     available (and never for static libraries). I have made a note to try to
1043     find a way round (a) in the future.
1044    
1045    
1046 nigel 81 Version 6.2 01-Aug-05
1047     ---------------------
1048    
1049     1. There was no test for integer overflow of quantifier values. A construction
1050     such as {1111111111111111} would give undefined results. What is worse, if
1051     a minimum quantifier for a parenthesized subpattern overflowed and became
1052     negative, the calculation of the memory size went wrong. This could have
1053     led to memory overwriting.
1054    
1055     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1056    
1057     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1058     operating environments where this matters.
1059    
1060     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1061     PCRE options from within the C++ wrapper.
1062    
1063     5. Named capturing subpatterns were not being correctly counted when a pattern
1064     was compiled. This caused two problems: (a) If there were more than 100
1065     such subpatterns, the calculation of the memory needed for the whole
1066     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1067     back references of the form \12, where the number was greater than 9, were
1068     not recognized as back references, even though there were sufficient
1069     previous subpatterns.
1070    
1071     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1072     versions of gcc, e.g. 2.95.4.
1073    
1074    
1075 nigel 79 Version 6.1 21-Jun-05
1076     ---------------------
1077    
1078     1. There was one reference to the variable "posix" in pcretest.c that was not
1079     surrounded by "#if !defined NOPOSIX".
1080    
1081     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1082     the cross-check on the old pcre_info() function, for the benefit of the
1083     cut-down version of PCRE that is currently imported into Exim.
1084    
1085     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1086     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1087     patterns that start (?i) but I don't think that matters. The use of (?i) is
1088     just an example; this all applies to the other options as well.
1089    
1090     4. Since libtool seems to echo the compile commands it is issuing, the output
1091     from "make" can be reduced a bit by putting "@" in front of each libtool
1092     compile command.
1093    
1094     5. Patch from the folks at Google for configure.in to be a bit more thorough
1095     in checking for a suitable C++ installation before trying to compile the
1096     C++ stuff. This should fix a reported problem when a compiler was present,
1097     but no suitable headers.
1098    
1099     6. The man pages all had just "PCRE" as their title. I have changed them to
1100     be the relevant file name. I have also arranged that these names are
1101     retained in the file doc/pcre.txt, which is a concatenation in text format
1102     of all the man pages except the little individual ones for each function.
1103    
1104     7. The NON-UNIX-USE file had not been updated for the different set of source
1105     files that come with release 6. I also added a few comments about the C++
1106     wrapper.
1107    
1108    
1109 nigel 77 Version 6.0 07-Jun-05
1110     ---------------------
1111    
1112     1. Some minor internal re-organization to help with my DFA experiments.
1113    
1114     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1115     didn't matter for the library itself when fully configured, but did matter
1116     when compiling without UCP support, or within Exim, where the ucp files are
1117     not imported.
1118    
1119     3. Refactoring of the library code to split up the various functions into
1120     different source modules. The addition of the new DFA matching code (see
1121     below) to a single monolithic source would have made it really too
1122     unwieldy, quite apart from causing all the code to be include in a
1123     statically linked application, when only some functions are used. This is
1124     relevant even without the DFA addition now that patterns can be compiled in
1125     one application and matched in another.
1126    
1127     The downside of splitting up is that there have to be some external
1128     functions and data tables that are used internally in different modules of
1129     the library but which are not part of the API. These have all had their
1130     names changed to start with "_pcre_" so that they are unlikely to clash
1131     with other external names.
1132    
1133     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1134     a different (DFA) algorithm. Although it is slower than the original
1135     function, it does have some advantages for certain types of matching
1136     problem.
1137    
1138     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1139     including restarting after a partial match.
1140    
1141     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1142     defined when compiling for Windows was sent to me. I have put it into the
1143     code, though I have no means of testing or verifying it.
1144    
1145     7. Added the pcre_refcount() auxiliary function.
1146    
1147     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1148     match before or at the first newline in the subject string. In pcretest,
1149     the /f option on a pattern can be used to set this.
1150    
1151     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1152     would behave wrongly. This has been present in PCRE since release 4.0.
1153    
1154     10. A number of changes to the pcregrep command:
1155    
1156     (a) Refactored how -x works; insert ^(...)$ instead of setting
1157     PCRE_ANCHORED and checking the length, in preparation for adding
1158     something similar for -w.
1159    
1160     (b) Added the -w (match as a word) option.
1161    
1162     (c) Refactored the way lines are read and buffered so as to have more
1163     than one at a time available.
1164    
1165     (d) Implemented a pcregrep test script.
1166    
1167     (e) Added the -M (multiline match) option. This allows patterns to match
1168     over several lines of the subject. The buffering ensures that at least
1169     8K, or the rest of the document (whichever is the shorter) is available
1170     for matching (and similarly the previous 8K for lookbehind assertions).
1171    
1172     (f) Changed the --help output so that it now says
1173    
1174     -w, --word-regex(p)
1175    
1176     instead of two lines, one with "regex" and the other with "regexp"
1177     because that confused at least one person since the short forms are the
1178     same. (This required a bit of code, as the output is generated
1179     automatically from a table. It wasn't just a text change.)
1180    
1181     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1182     option but starts with a hyphen. Could be a pattern or a path name
1183     starting with a hyphen, for instance.
1184    
1185     (h) "-" can be given as a file name to represent stdin.
1186    
1187     (i) When file names are being printed, "(standard input)" is used for
1188     the standard input, for compatibility with GNU grep. Previously
1189     "<stdin>" was used.
1190    
1191     (j) The option --label=xxx can be used to supply a name to be used for
1192     stdin when file names are being printed. There is no short form.
1193    
1194     (k) Re-factored the options decoding logic because we are going to add
1195     two more options that take data. Such options can now be given in four
1196     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1197    
1198     (l) Added the -A, -B, and -C options for requesting that lines of context
1199     around matches be printed.
1200    
1201     (m) Added the -L option to print the names of files that do not contain
1202     any matching lines, that is, the complement of -l.
1203    
1204     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1205     continue to scan other files.
1206    
1207     (o) The -s option was incorrectly implemented. For compatibility with other
1208     greps, it now suppresses the error message for a non-existent or non-
1209     accessible file (but not the return code). There is a new option called
1210     -q that suppresses the output of matching lines, which was what -s was
1211     previously doing.
1212    
1213     (p) Added --include and --exclude options to specify files for inclusion
1214     and exclusion when recursing.
1215    
1216     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1217     Hopefully, it now does.
1218    
1219     12. Missing cast in pcre_study().
1220    
1221     13. Added an "uninstall" target to the makefile.
1222    
1223     14. Replaced "extern" in the function prototypes in Makefile.in with
1224     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1225     world, but is set differently for Windows.
1226    
1227     15. Added a second compiling function called pcre_compile2(). The only
1228     difference is that it has an extra argument, which is a pointer to an
1229     integer error code. When there is a compile-time failure, this is set
1230     non-zero, in addition to the error test pointer being set to point to an
1231     error message. The new argument may be NULL if no error number is required
1232     (but then you may as well call pcre_compile(), which is now just a
1233     wrapper). This facility is provided because some applications need a
1234     numeric error indication, but it has also enabled me to tidy up the way
1235     compile-time errors are handled in the POSIX wrapper.
1236    
1237     16. Added VPATH=.libs to the makefile; this should help when building with one
1238     prefix path and installing with another. (Or so I'm told by someone who
1239     knows more about this stuff than I do.)
1240    
1241     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1242     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1243     match everything, including newlines. This is not POSIX-compatible, but
1244     somebody wanted the feature. From pcretest it can be activated by using
1245     both the P and the s flags.
1246    
1247     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1248    
1249     19. libpcre.pc was being incorrectly installed as executable.
1250    
1251     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1252     it now also looks for '\r' so that it will work unmodified on Windows.
1253    
1254     21. Added Google's contributed C++ wrapper to the distribution.
1255    
1256     22. Added some untidy missing memory free() calls in pcretest, to keep
1257     Electric Fence happy when testing.
1258    
1259    
1260    
1261 nigel 75 Version 5.0 13-Sep-04
1262     ---------------------
1263    
1264     1. Internal change: literal characters are no longer packed up into items
1265     containing multiple characters in a single byte-string. Each character
1266     is now matched using a separate opcode. However, there may be more than one
1267     byte in the character in UTF-8 mode.
1268    
1269     2. The pcre_callout_block structure has two new fields: pattern_position and
1270     next_item_length. These contain the offset in the pattern to the next match
1271     item, and its length, respectively.
1272    
1273     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1274     insertion of callouts before each pattern item. Added the /C option to
1275     pcretest to make use of this.
1276    
1277     4. On the advice of a Windows user, the lines
1278    
1279     #if defined(_WIN32) || defined(WIN32)
1280     _setmode( _fileno( stdout ), 0x8000 );
1281     #endif /* defined(_WIN32) || defined(WIN32) */
1282    
1283     have been added to the source of pcretest. This apparently does useful
1284     magic in relation to line terminators.
1285    
1286     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1287     for the benefit of those environments where the "b" makes a difference.
1288    
1289     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1290     to know about it. I have put a hack into configure.in that adds in code
1291     to set GCC=yes if CC=icc. This seems to end up at a point in the
1292     generated configure script that is early enough to affect the setting of
1293     compiler options, which is what is needed, but I have no means of testing
1294     whether it really works. (The user who reported this had patched the
1295     generated configure script, which of course I cannot do.)
1296    
1297     LATER: After change 22 below (new libtool files), the configure script
1298     seems to know about icc (and also ecc). Therefore, I have commented out
1299     this hack in configure.in.
1300    
1301     7. Added support for pkg-config (2 patches were sent in).
1302    
1303     8. Negated POSIX character classes that used a combination of internal tables
1304     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1305     [[:^ascii]]. Typically, they would match almost any characters. The other
1306     POSIX classes were not broken in this way.
1307    
1308     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1309     to find the match, as PCRE was deluded into thinking that the match had to
1310     start at the start point or following a newline. The same bug applied to
1311     patterns with negative forward assertions or any backward assertions
1312     preceding ".*" at the start, unless the pattern required a fixed first
1313     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1314    
1315     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1316     starting at the last subject character, bytes beyond the end of the subject
1317     string were read.
1318    
1319     11. Renamed the variable "class" as "classbits" to make life easier for C++
1320     users. (Previously there was a macro definition, but it apparently wasn't
1321     enough.)
1322    
1323     12. Added the new field "tables" to the extra data so that tables can be passed
1324     in at exec time, or the internal tables can be re-selected. This allows
1325     a compiled regex to be saved and re-used at a later time by a different
1326     program that might have everything at different addresses.
1327    
1328     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1329     -R library as well as a -L library.
1330    
1331     14. The debugging options of pcretest (-d on the command line or D on a
1332     pattern) showed incorrect output for anything following an extended class
1333     that contained multibyte characters and which was followed by a quantifier.
1334    
1335     15. Added optional support for general category Unicode character properties
1336     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1337     support. It adds about 90K to the size of the library. The meanings of the
1338     inbuilt class escapes such as \d and \s have NOT been changed.
1339    
1340     16. Updated pcredemo.c to include calls to free() to release the memory for the
1341     compiled pattern.
1342    
1343     17. The generated file chartables.c was being created in the source directory
1344     instead of in the building directory. This caused the build to fail if the
1345     source directory was different from the building directory, and was
1346     read-only.
1347    
1348     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1349     file. No doubt somebody will tell me if they don't make sense... Also added
1350     Dan Mooney's comments about building on OpenVMS.
1351    
1352     19. Added support for partial matching via the PCRE_PARTIAL option for
1353     pcre_exec() and the \P data escape in pcretest.
1354    
1355     20. Extended pcretest with 3 new pattern features:
1356    
1357     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1358     write the compiled pattern to the file whose name is "rest-of-line".
1359     This is a straight binary dump of the data, with the saved pointer to
1360     the character tables forced to be NULL. The study data, if any, is
1361     written too. After writing, pcretest reads a new pattern.
1362    
1363     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1364     compiled pattern from the given file. There must not be any
1365     occurrences of "<" in the file name (pretty unlikely); if there are,
1366     pcretest will instead treat the initial "<" as a pattern delimiter.
1367     After reading in the pattern, pcretest goes on to read data lines as
1368     usual.
1369    
1370     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1371     and 16-bit fields in a compiled pattern, to simulate a pattern that
1372     was compiled on a host of opposite endianness.
1373    
1374     21. The pcre-exec() function can now cope with patterns that were compiled on
1375     hosts of opposite endianness, with this restriction:
1376    
1377     As for any compiled expression that is saved and used later, the tables
1378     pointer field cannot be preserved; the extra_data field in the arguments
1379     to pcre_exec() should be used to pass in a tables address if a value
1380     other than the default internal tables were used at compile time.
1381    
1382     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1383     now diagnosed as an error. Previously, most of the time, a negative number
1384     would have been treated as zero, but if in addition "ovector" was passed as
1385     NULL, a crash could occur.
1386    
1387     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1388     new versions from the libtool 1.5 distribution (the last one is a copy of
1389     a file called libtool.m4). This seems to have fixed the need to patch
1390     "configure" to support Darwin 1.3 (which I used to do). However, I still
1391     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1392     workstation).
1393    
1394     24. Changed the PCRE licence to be the more standard "BSD" licence.
1395    
1396    
1397 nigel 73 Version 4.5 01-Dec-03
1398     ---------------------
1399    
1400     1. There has been some re-arrangement of the code for the match() function so
1401     that it can be compiled in a version that does not call itself recursively.
1402     Instead, it keeps those local variables that need separate instances for
1403     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1404     needs to "recurse". Keeping track of where control must go is done by means
1405     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1406     hide most of the details from the main code, and operates only if
1407     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1408     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1409     operating.
1410    
1411     To make it easier for callers to provide specially tailored get/free
1412     functions for this usage, two new functions, pcre_stack_malloc, and
1413     pcre_stack_free, are used. They are always called in strict stacking order,
1414     and the size of block requested is always the same.
1415    
1416     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1417     PCRE has been compiled to use the stack or the heap for recursion. The
1418     -C option of pcretest uses this to show which version is compiled.
1419    
1420     A new data escape \S, is added to pcretest; it causes the amounts of store
1421     obtained and freed by both kinds of malloc/free at match time to be added
1422     to the output.
1423    
1424     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1425     what's available on my current Linux desktop machine.
1426    
1427     3. When matching a UTF-8 string, the test for a valid string at the start has
1428     been extended. If start_offset is not zero, PCRE now checks that it points
1429     to a byte that is the start of a UTF-8 character. If not, it returns
1430     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1431     this is necessary because there may be backward assertions in the pattern.
1432     When matching the same subject several times, it may save resources to use
1433     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1434    
1435     4. The code for checking the validity of UTF-8 strings has been tightened so
1436     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1437     containing "overlong sequences".
1438    
1439     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1440     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1441     should have been "&", but it just so happened that all the cases this let
1442     through by mistake were picked up later in the function.
1443    
1444     6. I had used a variable called "isblank" - this is a C99 function, causing
1445     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1446    
1447     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1448     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1449     so the version is shown; (c) stop "make test" echoing "./RunTest".
1450    
1451     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1452    
1453     9. The prototype for memmove() for systems that don't have it was using
1454     size_t, but the inclusion of the header that defines size_t was later. I've
1455     moved the #includes for the C headers earlier to avoid this.
1456    
1457     10. Added some adjustments to the code to make it easier to compiler on certain
1458     special systems:
1459    
1460     (a) Some "const" qualifiers were missing.
1461     (b) Added the macro EXPORT before all exported functions; by default this
1462     is defined to be empty.
1463     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1464     that it reads its output file name as an argument instead of writing
1465     to the standard output and assuming this can be redirected.
1466    
1467     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1468     class containing characters with values greater than 255, PCRE compilation
1469     went into a loop.
1470    
1471     12. A recursive reference to a subpattern that was within another subpattern
1472     that had a minimum quantifier of zero caused PCRE to crash. For example,
1473     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1474     recursion. If the recursively-called subpattern itself had a zero repeat,
1475     that was OK.
1476    
1477     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1478     buffer into which it was copied (for escape processing) was still set at
1479     1024, so long lines caused crashes.
1480    
1481     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1482     "internal error: code overflow...". This applied to any character class
1483     that was followed by a possessive quantifier.
1484    
1485     15. Modified the Makefile to add libpcre.la as a prerequisite for
1486     libpcreposix.la because I was told this is needed for a parallel build to
1487     work.
1488    
1489     16. If a pattern that contained .* following optional items at the start was
1490     studied, the wrong optimizing data was generated, leading to matching
1491     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1492     matching string must start with a or b or c. The correct conclusion for
1493     this pattern is that a match can start with any character.
1494    
1495    
1496 nigel 71 Version 4.4 13-Aug-03
1497     ---------------------
1498    
1499     1. In UTF-8 mode, a character class containing characters with values between
1500     127 and 255 was not handled correctly if the compiled pattern was studied.
1501     In fixing this, I have also improved the studying algorithm for such
1502     classes (slightly).
1503    
1504     2. Three internal functions had redundant arguments passed to them. Removal
1505     might give a very teeny performance improvement.
1506    
1507     3. Documentation bug: the value of the capture_top field in a callout is *one
1508     more than* the number of the hightest numbered captured substring.
1509    
1510     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1511     in incorrectly linking with a previously installed version. They now link
1512     explicitly with libpcre.la.
1513    
1514     5. configure.in no longer needs to recognize Cygwin specially.
1515    
1516     6. A problem in pcre.in for Windows platforms is fixed.
1517    
1518     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1519     pcretest, it used to include the size of the study block as part of its
1520     output. Unfortunately, the structure contains a field that has a different
1521     size on different hardware architectures. This meant that the tests that
1522     showed this size failed. As the block is currently always of a fixed size,
1523     this information isn't actually particularly useful in pcretest output, so
1524     I have just removed it.
1525    
1526     8. Three pre-processor statements accidentally did not start in column 1.
1527     Sadly, there are *still* compilers around that complain, even though
1528     standard C has not required this for well over a decade. Sigh.
1529    
1530     9. In pcretest, the code for checking callouts passed small integers in the
1531     callout_data field, which is a void * field. However, some picky compilers
1532     complained about the casts involved for this on 64-bit systems. Now
1533     pcretest passes the address of the small integer instead, which should get
1534     rid of the warnings.
1535    
1536     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1537     both compile and run time, and gives an error if an invalid UTF-8 sequence
1538     is found. There is a option for disabling this check in cases where the
1539     string is known to be correct and/or the maximum performance is wanted.
1540    
1541     11. In response to a bug report, I changed one line in Makefile.in from
1542    
1543     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1544     to
1545     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1546    
1547     to look similar to other lines, but I have no way of telling whether this
1548     is the right thing to do, as I do not use Windows. No doubt I'll get told
1549     if it's wrong...
1550    
1551    
1552 nigel 69 Version 4.3 21-May-03
1553     ---------------------
1554    
1555     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1556     Makefile.
1557    
1558     2. Some refactoring to improve the quality of the code:
1559    
1560     (i) The utf8_table... variables are now declared "const".
1561    
1562     (ii) The code for \cx, which used the "case flipping" table to upper case
1563     lower case letters, now just substracts 32. This is ASCII-specific,
1564     but the whole concept of \cx is ASCII-specific, so it seems
1565     reasonable.
1566    
1567     (iii) PCRE was using its character types table to recognize decimal and
1568     hexadecimal digits in the pattern. This is silly, because it handles
1569     only 0-9, a-f, and A-F, but the character types table is locale-
1570     specific, which means strange things might happen. A private
1571     table is now used for this - though it costs 256 bytes, a table is
1572     much faster than multiple explicit tests. Of course, the standard
1573     character types table is still used for matching digits in subject
1574     strings against \d.
1575    
1576     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1577     ending in _t are). So I've renamed it as ESC_tee.
1578    
1579     3. The first argument for regexec() in the POSIX wrapper should have been
1580     defined as "const".
1581    
1582     4. Changed pcretest to use malloc() for its buffers so that they can be
1583     Electric Fenced for debugging.
1584    
1585     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1586     to read one or more bytes before the start of the subject string. Often this
1587     had no effect on PCRE's behaviour, but in some circumstances it could
1588     provoke a segmentation fault.
1589    
1590     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1591     to try to read one or more bytes before the start of the subject string.
1592    
1593     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1594     UTF-8 support could misbehave in various ways if the subject string
1595     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1596     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1597     back over UTF-8 characters.)
1598    
1599    
1600 nigel 67 Version 4.2 14-Apr-03
1601     ---------------------
1602    
1603     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1604    
1605     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1606     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1607     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1608     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1609     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1610     and BUILD_EXEEXT
1611     Note: automatic setting of the BUILD variables is not yet working
1612     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1613     compile-time but not at link-time
1614     [LINK]: use for linking executables only
1615     make different versions for Windows and non-Windows
1616     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1617     libraries
1618     [LINK_FOR_BUILD]: new variable
1619     [OBJEXT]: use throughout
1620     [EXEEXT]: use throughout
1621     <winshared>: new target
1622     <wininstall>: new target
1623     <dftables.o>: use native compiler
1624     <dftables>: use native linker
1625     <install>: handle Windows platform correctly
1626     <clean>: ditto
1627     <check>: ditto
1628     copy DLL to top builddir before testing
1629    
1630     As part of these changes, -no-undefined was removed again. This was reported
1631     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1632     in any case.
1633    
1634     3. Some tidies to get rid of compiler warnings:
1635    
1636     . In the match_data structure, match_limit was an unsigned long int, whereas
1637     match_call_count was an int. I've made them both unsigned long ints.
1638    
1639     . In pcretest the fact that a const uschar * doesn't automatically cast to
1640     a void * provoked a warning.
1641    
1642     . Turning on some more compiler warnings threw up some "shadow" variables
1643     and a few more missing casts.
1644    
1645     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1646     option, a class that contained a single character with a value between 128
1647     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1648    
1649     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1650     option, a class that contained several characters, but with at least one
1651     whose value was between 128 and 255 caused PCRE to crash.
1652    
1653    
1654 nigel 65 Version 4.1 12-Mar-03
1655     ---------------------
1656 nigel 63
1657 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1658     needed, and a string in dftables.c that was longer than standard compilers are
1659     required to support.
1660    
1661     2. Compiling with Sun's compiler found a few more places where the code could
1662     be tidied up in order to avoid warnings.
1663    
1664     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1665     first of these names is deprecated in the latest Autoconf in favour of the name
1666     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1667     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1668     analogy I have changed it to CFLAGS_FOR_BUILD.
1669    
1670     4. Added -no-undefined to the linking command in the Makefile, because this is
1671     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1672     linking step for the pcreposix library.
1673    
1674     5. PCRE was failing to diagnose the case of two named groups with the same
1675     name.
1676    
1677     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1678     literal character that is needed in the subject for a match, and scans along to
1679     ensure that it is present before embarking on the full matching process. This
1680     saves time in cases of nested unlimited repeats that are never going to match.
1681     Problem: the scan can take a lot of time if the subject is very long (e.g.
1682     megabytes), thus penalizing straightforward matches. It is now done only if the
1683     amount of subject to be scanned is less than 1000 bytes.
1684    
1685     7. A lesser problem with the same optimization is that it was recording the
1686     first character of an anchored pattern as "needed", thus provoking a search
1687     right along the subject, even when the first match of the pattern was going to
1688     fail. The "needed" character is now not set for anchored patterns, unless it
1689     follows something in the pattern that is of non-fixed length. Thus, it still
1690     fulfils its original purpose of finding quick non-matches in cases of nested
1691     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1692    
1693    
1694     Version 4.0 17-Feb-03
1695     ---------------------
1696    
1697 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1698     extended to the end of string, PCRE compiled incorrect data. This could lead to
1699     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1700    
1701     2. Moved to autoconf 2.53 and libtool 1.4.2.
1702    
1703     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1704     the special perltest8 script is no longer needed - all the tests can be run
1705     from a single perltest script.
1706    
1707     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1708     by \s. It has now been removed in PCRE. This means it isn't recognized as
1709     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1710     class [:space:] *does* include VT, thereby creating a mess.
1711    
1712     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1713     space and tab.
1714    
1715     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1716     its new features into the main test script, reducing the number of scripts.
1717    
1718     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1719     were backward compatible, and made the (?i) apply to the whole pattern, as if
1720     /i were given. Now it behaves more logically, and applies the option setting
1721     only to what follows. PCRE has been changed to follow suit. However, if it
1722     finds options settings right at the start of the pattern, it extracts them into
1723     the global options, as before. Thus, they show up in the info data.
1724    
1725     8. Added support for the \Q...\E escape sequence. Characters in between are
1726     treated as literals. This is slightly different from Perl in that $ and @ are
1727     also handled as literals inside the quotes. In Perl, they will cause variable
1728     interpolation. Note the following examples:
1729    
1730     Pattern PCRE matches Perl matches
1731    
1732     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1733     \Qabc\$xyz\E abc\$xyz abc\$xyz
1734     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1735    
1736     For compatibility with Perl, \Q...\E sequences are recognized inside character
1737     classes as well as outside them.
1738    
1739     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1740     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1741     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1742     signed/unsigned warnings.
1743    
1744     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1745     option for pcretest, so I've replaced it by a simple function that does just
1746     that job.
1747    
1748     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1749     "pcregrep -".
1750    
1751     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1752     Java package. This provides some syntactic sugar for simple cases of what my
1753     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1754     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1755     item, you can use this simplified notation. Note that only makes sense with
1756     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1757     greediness, whatever the setting of the PCRE_UNGREEDY option.
1758    
1759     13. A change of greediness default within a pattern was not taking effect at
1760     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1761     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1762     was abstracted outside.
1763    
1764     14. PCRE now supports the \G assertion. It is true when the current matching
1765     position is at the start point of the match. This differs from \A when the
1766     starting offset is non-zero. Used with the /g option of pcretest (or similar
1767     code), it works in the same way as it does for Perl's /g option. If all
1768     alternatives of a regex begin with \G, the expression is anchored to the start
1769     match position, and the "anchored" flag is set in the compiled expression.
1770    
1771     15. Some bugs concerning the handling of certain option changes within patterns
1772     have been fixed. These applied to options other than (?ims). For example,
1773     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1774     been the other way round. Some of this was related to change 7 above.
1775    
1776     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1777     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1778     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1779     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1780    
1781     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1782     mode. Unlike ".", it always matches newline, whatever the setting of
1783     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1784     assertions. Perl allows it, but it doesn't (in general) work because it can't
1785     calculate the length of the lookbehind. At least, that's the case for Perl
1786     5.8.0 - I've been told they are going to document that it doesn't work in
1787     future.
1788    
1789     18. Added an error diagnosis for escapes that PCRE does not support: these are
1790     \L, \l, \N, \P, \p, \U, \u, and \X.
1791    
1792     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1793     reading past the end of the pattern in cases such as /[abcd/.
1794    
1795     20. PCRE was getting more memory than necessary for patterns with classes that
1796     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1797    
1798     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1799     compiling PCRE for use with Virtual Pascal.
1800    
1801     22. Small fix to the Makefile to make it work properly if the build is done
1802     outside the source tree.
1803    
1804     23. Added a new extension: a condition to go with recursion. If a conditional
1805     subpattern starts with (?(R) the "true" branch is used if recursion has
1806     happened, whereas the "false" branch is used only at the top level.
1807    
1808     24. When there was a very long string of literal characters (over 255 bytes
1809     without UTF support, over 250 bytes with UTF support), the computation of how
1810     much memory was required could be incorrect, leading to segfaults or other
1811     strange effects.
1812    
1813     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1814     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1815     there was a subsequent back reference to those brackets. This meant that, for
1816     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1817     possible to check for precisely this case. All we can do is abandon the
1818     optimization if .* occurs inside capturing brackets when there are any back
1819     references whatsoever. (See below for a better fix that came later.)
1820    
1821     26. The handling of the optimization for finding the first character of a
1822     non-anchored pattern, and for finding a character that is required later in the
1823     match were failing in some cases. This didn't break the matching; it just
1824     failed to optimize when it could. The way this is done has been re-implemented.
1825    
1826     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1827    
1828     28. Added a new feature that provides some of the functionality that Perl
1829     provides with (?{...}). The facility is termed a "callout". The way it is done
1830     in PCRE is for the caller to provide an optional function, by setting
1831     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1832     global variable. By default it is unset, which disables all calling out. To get
1833     the function called, the regex must include (?C) at appropriate points. This
1834     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1835     This provides a means of identifying different callout points. When PCRE
1836     reaches such a point in the regex, if pcre_callout has been set, the external
1837     function is called. It is provided with data in a structure called
1838     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1839     matching continues; if it returns a non-zero value, the match at the current
1840     point fails. However, backtracking will occur if possible. [This was changed
1841     later and other features added - see item 49 below.]
1842    
1843     29. pcretest is upgraded to test the callout functionality. It provides a
1844     callout function that displays information. By default, it shows the start of
1845     the match and the current position in the text. There are some new data escapes
1846     to vary what happens:
1847    
1848     \C+ in addition, show current contents of captured substrings
1849     \C- do not supply a callout function
1850     \C!n return 1 when callout number n is reached
1851     \C!n!m return 1 when callout number n is reached for the mth time
1852    
1853     30. If pcregrep was called with the -l option and just a single file name, it
1854     output "<stdin>" if a match was found, instead of the file name.
1855    
1856     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1857     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1858     pcre_exec(). This saves a malloc/free per call. The default value of
1859     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1860     when configuring.
1861    
1862     32. The default maximum size of a compiled pattern is 64K. There have been a
1863     few cases of people hitting this limit. The code now uses macros to handle the
1864     storing of links as offsets within the compiled pattern. It defaults to 2-byte
1865     links, but this can be changed to 3 or 4 bytes by --with-link-size when
1866     configuring. Tests 2 and 5 work only with 2-byte links because they output
1867     debugging information about compiled patterns.
1868    
1869     33. Internal code re-arrangements:
1870    
1871     (a) Moved the debugging function for printing out a compiled regex into
1872     its own source file (printint.c) and used #include to pull it into
1873     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
1874     separate copies.
1875    
1876     (b) Defined the list of op-code names for debugging as a macro in
1877     internal.h so that it is next to the definition of the opcodes.
1878    
1879     (c) Defined a table of op-code lengths for simpler skipping along compiled
1880     code. This is again a macro in internal.h so that it is next to the
1881     definition of the opcodes.
1882    
1883     34. Added support for recursive calls to individual subpatterns, along the
1884     lines of Robin Houston's patch (but implemented somewhat differently).
1885    
1886     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
1887     allow it to read and process whole directories in Win32. This code was
1888     contributed by Lionel Fourquaux; it has not been tested by me.
1889    
1890     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
1891     used to name a group. Names consist of alphanumerics and underscores, and must
1892     be unique. Back references use the syntax (?P=name) and recursive calls use
1893     (?P>name) which is a PCRE extension to the Python extension. Groups still have
1894     numbers. The function pcre_fullinfo() can be used after compilation to extract
1895     a name/number map. There are three relevant calls:
1896    
1897     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
1898     PCRE_INFO_NAMECOUNT yields the number of entries
1899     PCRE_INFO_NAMETABLE yields a pointer to the map.
1900    
1901     The map is a vector of fixed-size entries. The size of each entry depends on
1902     the length of the longest name used. The first two bytes of each entry are the
1903     group number, most significant byte first. There follows the corresponding
1904     name, zero terminated. The names are in alphabetical order.
1905    
1906     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
1907     case instead of 255. Making it the same both with and without UTF-8 support
1908     means that the same test output works with both.
1909    
1910     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
1911     calling malloc() with a zero argument.
1912    
1913     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
1914     optimization. I've improved things by keeping a bitmap of backreferences with
1915     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
1916     fact referenced, the optimization can be applied. It is unlikely that a
1917     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
1918     the match to follow \n) will appear inside brackets with a number greater than
1919     31, but if it does, any back reference > 31 suppresses the optimization.
1920    
1921     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
1922     of disabling numbered capturing parentheses. Any opening parenthesis that is
1923     not followed by ? behaves as if it were followed by ?: but named parentheses
1924     can still be used for capturing (and they will acquire numbers in the usual
1925     way).
1926    
1927     41. Redesigned the return codes from the match() function into yes/no/error so
1928     that errors can be passed back from deep inside the nested calls. A malloc
1929     failure while inside a recursive subpattern call now causes the
1930     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
1931    
1932     42. It is now possible to set a limit on the number of times the match()
1933     function is called in a call to pcre_exec(). This facility makes it possible to
1934     limit the amount of recursion and backtracking, though not in a directly
1935     obvious way, because the match() function is used in a number of different
1936     circumstances. The count starts from zero for each position in the subject
1937     string (for non-anchored patterns). The default limit is, for compatibility, a
1938     large number, namely 10 000 000. You can change this in two ways:
1939    
1940     (a) When configuring PCRE before making, you can use --with-match-limit=n
1941     to set a default value for the compiled library.
1942    
1943     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
1944     a different value is set. See 45 below.
1945    
1946     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1947    
1948     43. Added a new function pcre_config(int, void *) to enable run-time extraction
1949     of things that can be changed at compile time. The first argument specifies
1950     what is wanted and the second points to where the information is to be placed.
1951     The current list of available information is:
1952    
1953     PCRE_CONFIG_UTF8
1954    
1955     The output is an integer that is set to one if UTF-8 support is available;
1956     otherwise it is set to zero.
1957    
1958     PCRE_CONFIG_NEWLINE
1959    
1960     The output is an integer that it set to the value of the code that is used for
1961     newline. It is either LF (10) or CR (13).
1962    
1963     PCRE_CONFIG_LINK_SIZE
1964    
1965     The output is an integer that contains the number of bytes used for internal
1966     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
1967    
1968     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1969    
1970     The output is an integer that contains the threshold above which the POSIX
1971     interface uses malloc() for output vectors. See item 31 above.
1972    
1973     PCRE_CONFIG_MATCH_LIMIT
1974    
1975     The output is an unsigned integer that contains the default limit of the number
1976     of match() calls in a pcre_exec() execution. See 42 above.
1977    
1978     44. pcretest has been upgraded by the addition of the -C option. This causes it
1979     to extract all the available output from the new pcre_config() function, and to
1980     output it. The program then exits immediately.
1981    
1982     45. A need has arisen to pass over additional data with calls to pcre_exec() in
1983     order to support additional features. One way would have been to define
1984     pcre_exec2() (for example) with extra arguments, but this would not have been
1985     extensible, and would also have required all calls to the original function to
1986     be mapped to the new one. Instead, I have chosen to extend the mechanism that
1987     is used for passing in "extra" data from pcre_study().
1988    
1989     The pcre_extra structure is now exposed and defined in pcre.h. It currently
1990     contains the following fields:
1991    
1992     flags a bitmap indicating which of the following fields are set
1993     study_data opaque data from pcre_study()
1994     match_limit a way of specifying a limit on match() calls for a specific
1995     call to pcre_exec()
1996     callout_data data for callouts (see 49 below)
1997    
1998     The flag bits are also defined in pcre.h, and are
1999    
2000     PCRE_EXTRA_STUDY_DATA
2001     PCRE_EXTRA_MATCH_LIMIT
2002     PCRE_EXTRA_CALLOUT_DATA
2003    
2004     The pcre_study() function now returns one of these new pcre_extra blocks, with
2005     the actual study data pointed to by the study_data field, and the
2006     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2007     before. That is, this change is entirely upwards-compatible and requires no
2008     change to existing code.
2009    
2010     If you want to pass in additional data to pcre_exec(), you can either place it
2011     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2012     block.
2013    
2014     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2015     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2016     times with different match limits, until it finds the minimum value needed for
2017     pcre_exec() to complete. The value is then output. This can be instructive; for
2018     most simple matches the number is quite small, but for pathological cases it
2019     gets very large very quickly.
2020    
2021     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2022     returns the size of the data block pointed to by the study_data field in a
2023     pcre_extra block, that is, the value that was passed as the argument to
2024     pcre_malloc() when PCRE was getting memory in which to place the information
2025     created by pcre_study(). The fourth argument should point to a size_t variable.
2026     pcretest has been extended so that this information is shown after a successful
2027     pcre_study() call when information about the compiled regex is being displayed.
2028    
2029     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2030     because what follows is always an absolute path. (Later: it turns out that this
2031     is more than cosmetic for MinGW, because it doesn't like empty path
2032     components.)
2033    
2034     49. Some changes have been made to the callout feature (see 28 above):
2035    
2036     (i) A callout function now has three choices for what it returns:
2037    
2038     0 => success, carry on matching
2039     > 0 => failure at this point, but backtrack if possible
2040     < 0 => serious error, return this value from pcre_exec()
2041    
2042     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2043     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2044     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2045     use by callout functions. It will never be used by PCRE itself.
2046    
2047     (ii) The pcre_extra structure (see 45 above) has a void * field called
2048     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2049     pcre_callout_block structure has a field of the same name. The contents of
2050     the field passed in the pcre_extra structure are passed to the callout
2051     function in the corresponding field in the callout block. This makes it
2052     easier to use the same callout-containing regex from multiple threads. For
2053     testing, the pcretest program has a new data escape
2054    
2055     \C*n pass the number n (may be negative) as callout_data
2056    
2057     If the callout function in pcretest receives a non-zero value as
2058     callout_data, it returns that value.
2059    
2060     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2061     there were some redundant $(CFLAGS) in commands that are now specified as
2062     $(LINK), which already includes $(CFLAGS).
2063    
2064     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2065     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2066     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2067     one-byte characters throughout. Note that case-insensitive matching applies
2068     only to characters whose values are less than 256. PCRE doesn't support the
2069     notion of cases for higher-valued characters.
2070    
2071     (i) A character class whose characters are all within 0-255 is handled as
2072     a bit map, and the map is inverted for negative classes. Previously, a
2073     character > 255 always failed to match such a class; however it should
2074     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2075    
2076     (ii) A negated character class with a single character < 255 is coded as
2077     "not this character" (OP_NOT). This wasn't working properly when the test
2078     character was multibyte, either singly or repeated.
2079    
2080     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2081     mode, for example: \x{100}{2,3}.
2082    
2083     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2084     singly or repeated) now correctly test multibyte characters. However,
2085     PCRE doesn't recognize any characters with values greater than 255 as
2086     digits, spaces, or word characters. Such characters always match \D, \S,
2087     and \W, and never match \d, \s, or \w.
2088    
2089     (v) Classes may now contain characters and character ranges with values
2090     greater than 255. For example: [ab\x{100}-\x{400}].
2091    
2092     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2093     PCRE in UTF-8 mode.
2094    
2095     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2096     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2097     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2098     value.)
2099    
2100     53. The single man page has become too large. I have therefore split it up into
2101     a number of separate man pages. These also give rise to individual HTML pages;
2102     these are now put in a separate directory, and there is an index.html page that
2103     lists them all. Some hyperlinking between the pages has been installed.
2104    
2105     54. Added convenience functions for handling named capturing parentheses.
2106    
2107     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2108     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2109     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2110     are faulted.
2111    
2112     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2113     calling configure. These values are used when compiling the dftables.c program
2114     which is run to generate the source of the default character tables. They
2115     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2116     you will need to set these values.
2117    
2118     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2119    
2120    
2121     Version 3.9 02-Jan-02
2122 nigel 61 ---------------------
2123    
2124     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2125    
2126     2. If --disable-static was given, the building process failed when trying to
2127     build pcretest and pcregrep. (For some reason it was using libtool to compile
2128     them, which is not right, as they aren't part of the library.)
2129    
2130    
2131 nigel 59 Version 3.8 18-Dec-01
2132     ---------------------
2133    
2134     1. The experimental UTF-8 code was completely screwed up. It was packing the
2135     bytes in the wrong order. How dumb can you get?
2136    
2137    
2138 nigel 57 Version 3.7 29-Oct-01
2139     ---------------------
2140    
2141     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2142     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2143     this didn't happen under Solaris 8, where I normally test things.
2144    
2145 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2146     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2147     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2148     this fix an hour or so after the initial 3.7 release.)
2149 nigel 57
2150 nigel 59
2151 nigel 55 Version 3.6 23-Oct-01
2152     ---------------------
2153    
2154     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2155     offsets passed as NULL with zero offset count.
2156    
2157     2. The config.guess and config.sub files had not been updated when I moved to
2158     the latest autoconf.
2159    
2160    
2161 nigel 53 Version 3.5 15-Aug-01
2162     ---------------------
2163 nigel 5
2164 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2165     had been forgotten.
2166    
2167     2. By using declared but undefined structures, we can avoid using "void"
2168     definitions in pcre.h while keeping the internal definitions of the structures
2169     private.
2170    
2171     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2172     user point of view, this means that both static and shared libraries are built
2173     by default, but this can be individually controlled. More of the work of
2174     handling this static/shared cases is now inside libtool instead of PCRE's make
2175     file.
2176    
2177     4. The pcretest utility is now installed along with pcregrep because it is
2178     useful for users (to test regexs) and by doing this, it automatically gets
2179     relinked by libtool. The documentation has been turned into a man page, so
2180     there are now .1, .txt, and .html versions in /doc.
2181    
2182     5. Upgrades to pcregrep:
2183     (i) Added long-form option names like gnu grep.
2184     (ii) Added --help to list all options with an explanatory phrase.
2185     (iii) Added -r, --recursive to recurse into sub-directories.
2186     (iv) Added -f, --file to read patterns from a file.
2187    
2188     6. pcre_exec() was referring to its "code" argument before testing that
2189     argument for NULL (and giving an error if it was NULL).
2190    
2191     7. Upgraded Makefile.in to allow for compiling in a different directory from
2192     the source directory.
2193    
2194     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2195     options bits, the pointer it was passed was to an int instead of to an unsigned
2196     long int. This mattered only on 64-bit systems.
2197    
2198     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2199     generated) instead of pcre.in, which it its source. Also made the same change
2200     in several of the .c files.
2201    
2202     10. A new release of gcc defines printf() as a macro, which broke pcretest
2203     because it had an ifdef in the middle of a string argument for printf(). Fixed
2204     by using separate calls to printf().
2205    
2206     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2207     script, to force use of CR or LF instead of \n in the source. On non-Unix
2208     systems, the value can be set in config.h.
2209    
2210     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2211     absolute limit. Changed the text of the error message to make this clear, and
2212     likewise updated the man page.
2213    
2214     13. The limit of 99 on the number of capturing subpatterns has been removed.
2215     The new limit is 65535, which I hope will not be a "real" limit.
2216    
2217    
2218 nigel 51 Version 3.4 22-Aug-00
2219     ---------------------
2220    
2221     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2222    
2223     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2224    
2225    
2226 nigel 49 Version 3.3 01-Aug-00
2227     ---------------------
2228    
2229     1. If an octal character was given, but the value was greater than \377, it
2230     was not getting masked to the least significant bits, as documented. This could
2231     lead to crashes in some systems.
2232    
2233     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2234     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2235    
2236     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2237     These just pass their arguments on to (pcre_free)(), but they are provided
2238     because some uses of PCRE bind it to non-C systems that can call its functions,
2239     but cannot call free() or pcre_free() directly.
2240    
2241     4. Add "make test" as a synonym for "make check". Corrected some comments in
2242     the Makefile.
2243    
2244     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2245     Makefile.
2246    
2247     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2248     command called pgrep for grepping around the active processes.
2249    
2250     7. Added the beginnings of support for UTF-8 character strings.
2251    
2252     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2253     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2254     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2255     out for the ar command.)
2256    
2257    
2258 nigel 47 Version 3.2 12-May-00
2259     ---------------------
2260    
2261     This is purely a bug fixing release.
2262    
2263     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2264     of ZA. This was just one example of several cases that could provoke this bug,
2265     which was introduced by change 9 of version 2.00. The code for breaking
2266     infinite loops after an iteration that matches an empty string was't working
2267     correctly.
2268    
2269     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2270     when matched against abbab (for example). After matching an empty string, it
2271     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2272     caused it to match further down the string than it should.
2273    
2274     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2275     was there because it doesn't seem to be needed, and it causes trouble on some
2276     systems, as it is not a Standard C header. It has been removed.
2277    
2278     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2279     were reported on the Macintosh. The changes were from
2280    
2281     while ((c = *(++ptr)) != 0 && c != '\n');
2282     to
2283     while ((c = *(++ptr)) != 0 && c != '\n') ;
2284    
2285     Totally extraordinary, but if that's what it takes...
2286    
2287     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2288     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2289     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2290     assumes the way PCRE uses memmove() (always moving upwards).
2291    
2292     6. PCRE is being used in one environment where strchr() is not available. There
2293     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2294     faster code anyway.
2295    
2296    
2297 nigel 45 Version 3.1 09-Feb-00
2298     ---------------------
2299    
2300     The only change in this release is the fixing of some bugs in Makefile.in for
2301     the "install" target:
2302    
2303     (1) It was failing to install pcreposix.h.
2304    
2305     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2306    
2307    
2308 nigel 43 Version 3.0 01-Feb-00
2309     ---------------------
2310 nigel 41
2311     1. Add support for the /+ modifier to perltest (to output $` like it does in
2312     pcretest).
2313    
2314     2. Add support for the /g modifier to perltest.
2315    
2316     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2317     matches null strings.
2318    
2319     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2320     pattern. Perl treats empty patterns specially - it reuses the most recent
2321     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2322     effect.
2323    
2324     5. The POSIX interface was broken in that it was just handing over the POSIX
2325     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2326     required a bigger vector, with some working space on the end. This means that
2327     the POSIX wrapper now has to get and free some memory, and copy the results.
2328    
2329 nigel 43 6. Added some simple autoconf support, placing the test data and the
2330     documentation in separate directories, re-organizing some of the
2331     information files, and making it build pcre-config (a GNU standard). Also added
2332     libtool support for building PCRE as a shared library, which is now the
2333     default.
2334 nigel 41
2335 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2336     09 are not valid octal constants. Single digits will be used for minor values
2337     less than 10.
2338    
2339     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2340     existing programs that set these in the POSIX interface can use PCRE without
2341     modification.
2342    
2343     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2344     return all that pcre_info() returns, plus additional data. The pcre_info()
2345     function is retained for compatibility, but is considered to be obsolete.
2346    
2347     10. Added experimental recursion feature (?R) to handle one common case that
2348     Perl 5.6 will be able to do with (?p{...}).
2349    
2350     11. Added support for POSIX character classes like [:alpha:], which Perl is
2351     adopting.
2352    
2353    
2354 nigel 39 Version 2.08 31-Aug-99
2355     ----------------------
2356    
2357     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2358     trying to match at the startoffset position, but instead was moving forward to
2359     the next newline as if a previous match had failed.
2360    
2361     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2362     and could get into a loop if a null string was matched other than at the start
2363     of the subject.
2364    
2365     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2366     be distinguished at compile time, and for completeness also added PCRE_DATE.
2367    
2368     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2369     in GnuWin32 environments.
2370    
2371    
2372 nigel 37 Version 2.07 29-Jul-99
2373     ----------------------
2374    
2375     1. The documentation is now supplied in plain text form and HTML as well as in
2376     the form of man page sources.
2377    
2378     2. C++ compilers don't like assigning (void *) values to other pointer types.
2379     In particular this affects malloc(). Although there is no problem in Standard
2380     C, I've put in casts to keep C++ compilers happy.
2381    
2382     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2383     should be (const char *).
2384    
2385     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2386     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2387     However, I haven't made this a standard facility. The documentation doesn't
2388     mention it, and the Makefile doesn't support it.
2389    
2390     5. The Makefile now contains an "install" target, with editable destinations at
2391     the top of the file. The pcretest program is not installed.
2392    
2393     6. pgrep -V now gives the PCRE version number and date.
2394    
2395     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2396     causing the entire string to be ignored, instead of just the last character.
2397    
2398     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2399     non-matching string, it can take a very, very long time, even for strings of
2400     quite modest length, because of the nested recursion. PCRE now does better in
2401     some of these cases. It does this by remembering the last required literal
2402     character in the pattern, and pre-searching the subject to ensure it is present
2403     before running the real match. In other words, it applies a heuristic to detect
2404     some types of certain failure quickly, and in the above example, if presented
2405     with a string that has no trailing " it gives "no match" very quickly.
2406    
2407     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2408     other alternatives are tried instead.
2409    
2410    
2411 nigel 35 Version 2.06 09-Jun-99
2412     ----------------------
2413    
2414     1. Change pcretest's output for amount of store used to show just the code
2415     space, because the remainder (the data block) varies in size between 32-bit and
2416     64-bit systems.
2417    
2418     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2419     start matching at. This allows lookbehinds to work when searching for multiple
2420     occurrences in a string.
2421    
2422     3. Added additional options to pcretest for testing multiple occurrences:
2423    
2424     /+ outputs the rest of the string that follows a match
2425     /g loops for multiple occurrences, using the new startoffset argument
2426     /G loops for multiple occurrences by passing an incremented pointer
2427    
2428     4. PCRE wasn't doing the "first character" optimization for patterns starting
2429     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2430     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2431     the letter 'x'. On long subject strings, this gives a significant speed-up.
2432    
2433    
2434 nigel 33 Version 2.05 21-Apr-99
2435     ----------------------
2436    
2437     1. Changed the type of magic_number from int to long int so that it works
2438     properly on 16-bit systems.
2439    
2440     2. Fixed a bug which caused patterns starting with .* not to work correctly
2441     when the subject string contained newline characters. PCRE was assuming
2442     anchoring for such patterns in all cases, which is not correct because .* will
2443     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2444     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2445     must be retried after every newline in the subject.
2446    
2447    
2448 nigel 31 Version 2.04 18-Feb-99
2449     ----------------------
2450    
2451     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2452     computation of the store needed to hold the pattern was incorrect (too large).
2453     If such patterns were nested a few deep, this could multiply and become a real
2454     problem.
2455    
2456     2. Added /M option to pcretest to show the memory requirement of a specific
2457     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2458    
2459     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2460     compiled in such a way that the backtracking after subsequent failure was
2461     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2462     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2463    
2464    
2465 nigel 29 Version 2.03 02-Feb-99
2466     ----------------------
2467    
2468     1. Fixed typo and small mistake in man page.
2469    
2470 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2471     LICENCE file containing the conditions.
2472 nigel 29
2473     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2474     Perl, that is the internal \ allows the delimiter to be included in the
2475     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2476     the final delimiter, add \ to the end of the pattern (to test the error).
2477    
2478     4. Added the convenience functions for extracting substrings after a successful
2479     match. Updated pcretest to make it able to test these functions.
2480    
2481    
2482 nigel 27 Version 2.02 14-Jan-99
2483     ----------------------
2484    
2485     1. Initialized the working variables associated with each extraction so that
2486     their saving and restoring doesn't refer to uninitialized store.
2487    
2488     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2489     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2490     fix the problem.
2491    
2492     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2493     calls, and wasn't printing the correct value for compiling calls. Increased the
2494     default value of LOOPREPEAT, and the number of significant figures in the
2495     times.
2496    
2497     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2498    
2499     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2500     a building problem on Windows NT with a FAT file system.
2501    
2502    
2503 nigel 25 Version 2.01 21-Oct-98
2504     ----------------------
2505    
2506     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2507     to character tables built by pcre_maketables() in the current locale. If NULL
2508     is passed, the default tables are used.
2509    
2510    
2511 nigel 23 Version 2.00 24-Sep-98
2512 nigel 21 ----------------------
2513    
2514 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2515     it any more.
2516    
2517     2. Allow quantification of (?>) groups, and make it work correctly.
2518    
2519     3. The first character computation wasn't working for (?>) groups.
2520    
2521     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2522     end of the subject) and add 5.005's \z, which really does match only at the
2523     very end of the subject.
2524    
2525     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2526    
2527     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2528     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2529     localized options. All options to pcre_study() were also removed.
2530    
2531     7. Add other new features from 5.005:
2532    
2533     $(?<= positive lookbehind
2534     $(?<! negative lookbehind
2535     (?imsx-imsx) added the unsetting capability
2536     such a setting is global if at outer level; local otherwise
2537     (?imsx-imsx:) non-capturing groups with option setting
2538     (?(cond)re|re) conditional pattern matching
2539    
2540     A backreference to itself in a repeated group matches the previous
2541     captured string.
2542    
2543     8. General tidying up of studying (both automatic and via "study")
2544     consequential on the addition of new assertions.
2545    
2546     9. As in 5.005, unlimited repeated groups that could match an empty substring
2547     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2548     runtime if any iteration does actually match an empty substring.
2549    
2550     10. Include the RunTest script in the distribution.
2551    
2552     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2553     discrepancies, some of which were old and were also with respect to 5.004. They
2554     have now been fixed.
2555    
2556    
2557     Version 1.09 28-Apr-98
2558     ----------------------
2559    
2560 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2561     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2562     program crashes, or just wrong answers. This did not apply to negated classes
2563     containing more than one character, or to minima other than one.
2564    
2565    
2566 nigel 19 Version 1.08 27-Mar-98
2567     ----------------------
2568    
2569     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2570    
2571     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2572     latter must appear before anything that relies on it in the pattern.
2573    
2574    
2575 nigel 17 Version 1.07 16-Feb-98
2576     ----------------------
2577    
2578     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2579     repeat of a potentially empty string).
2580    
2581    
2582 nigel 15 Version 1.06 23-Jan-98
2583     ----------------------
2584    
2585     1. Added Markus Oberhumer's little patches for C++.
2586    
2587     2. Literal strings longer than 255 characters were broken.
2588    
2589    
2590 nigel 13 Version 1.05 23-Dec-97
2591     ----------------------
2592    
2593     1. Negated character classes containing more than one character were failing if
2594     PCRE_CASELESS was set at run time.
2595    
2596    
2597 nigel 11 Version 1.04 19-Dec-97
2598     ----------------------
2599    
2600     1. Corrected the man page, where some "const" qualifiers had been omitted.
2601    
2602     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2603     input syntax.
2604    
2605     3. Fixed memory leak which occurred when a regex with back references was
2606     matched with an offsets vector that wasn't big enough. The temporary memory
2607     that is used in this case wasn't being freed if the match failed.
2608    
2609     4. Tidied pcretest to ensure it frees memory that it gets.
2610    
2611     5. Temporary memory was being obtained in the case where the passed offsets
2612     vector was exactly big enough.
2613    
2614     6. Corrected definition of offsetof() from change 5 below.
2615    
2616     7. I had screwed up change 6 below and broken the rules for the use of
2617     setjmp(). Now fixed.
2618    
2619    
2620 nigel 9 Version 1.03 18-Dec-97
2621     ----------------------
2622    
2623     1. A erroneous regex with a missing opening parenthesis was correctly
2624     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2625     on some systems.
2626    
2627     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2628     it was reported that one broken compiler failed on the former because "code" is
2629     also an independent variable.
2630    
2631     3. The erroneous regex a[]b caused an array overrun reference.
2632    
2633     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2634     fail on data ending with that character. (It was going on too far, and checking
2635     the next character, typically a binary zero.) This was specific to the
2636     optimized code for single-character negative classes.
2637    
2638     5. Added a contributed patch from the TIN world which does the following:
2639    
2640     + Add an undef for memmove, in case the the system defines a macro for it.
2641    
2642     + Add a definition of offsetof(), in case there isn't one. (I don't know
2643     the reason behind this - offsetof() is part of the ANSI standard - but
2644     it does no harm).
2645    
2646     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2647     most of the places where whitespace preceded '#'. I have given up and
2648     allowed the remaining 2 cases to be at the margin.
2649    
2650     + Rename some variables in pcre to eliminate shadowing. This seems very
2651     pedantic, but does no harm, of course.
2652    
2653     6. Moved the call to setjmp() into its own function, to get rid of warnings
2654     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2655    
2656     7. Constructs such as \d{8,} were compiling into the equivalent of
2657 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2658 nigel 9 outcome, but in this particular case used more store than had been allocated,
2659     which caused the bug to be discovered because it threw up an internal error.
2660    
2661     8. The debugging code in both pcre and pcretest for outputting the compiled
2662     form of a regex was going wrong in the case of back references followed by
2663     curly-bracketed repeats.
2664    
2665    
2666 nigel 7 Version 1.02 12-Dec-97
2667     ----------------------
2668    
2669     1. Typos in pcre.3 and comments in the source fixed.
2670    
2671     2. Applied a contributed patch to get rid of places where it used to remove
2672     'const' from variables, and fixed some signed/unsigned and uninitialized
2673     variable warnings.
2674    
2675     3. Added the "runtest" target to Makefile.
2676    
2677     4. Set default compiler flag to -O2 rather than just -O.
2678    
2679    
2680 nigel 5 Version 1.01 19-Nov-97
2681     ----------------------
2682    
2683     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2684     like /([ab]*)*/, that is, for classes with more than one character in them.
2685    
2686     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2687     as /((?>a*))*/ (a PCRE_EXTRA facility).
2688    
2689    
2690     Version 1.00 18-Nov-97
2691     ----------------------
2692    
2693     1. Added compile-time macros to support systems such as SunOS4 which don't have
2694     memmove() or strerror() but have other things that can be used instead.
2695    
2696     2. Arranged that "make clean" removes the executables.
2697    
2698    
2699 nigel 3 Version 0.99 27-Oct-97
2700     ----------------------
2701    
2702     1. Fixed bug in code for optimizing classes with only one character. It was
2703     initializing a 32-byte map regardless, which could cause it to run off the end
2704     of the memory it had got.
2705    
2706     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2707    
2708    
2709     Version 0.98 22-Oct-97
2710     ----------------------
2711    
2712     1. Fixed bug in code for handling temporary memory usage when there are more
2713     back references than supplied space in the ovector. This could cause segfaults.
2714    
2715    
2716     Version 0.97 21-Oct-97
2717     ----------------------
2718    
2719     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2720    
2721     2. Optimized negated single characters not to use a bit map.
2722    
2723     3. Brought error texts together as macro definitions; clarified some of them;
2724     fixed one that was wrong - it said "range out of order" when it meant "invalid
2725     escape sequence".
2726    
2727     4. Changed some char * arguments to const char *.
2728    
2729     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2730    
2731     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2732     pcretest.
2733    
2734    
2735     Version 0.96 16-Oct-97
2736     ----------------------
2737    
2738     1. Added a simple "pgrep" utility to the distribution.
2739    
2740     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2741     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2742     where "ddd" means "one or more decimal digits".
2743    
2744     3. Fixed serious bug. If a pattern had a back reference, but the call to
2745     pcre_exec() didn't supply a large enough ovector to record the related
2746     identifying subpattern, the match always failed. PCRE now remembers the number
2747     of the largest back reference, and gets some temporary memory in which to save
2748     the offsets during matching if necessary, in order to ensure that
2749     backreferences always work.
2750    
2751     4. Increased the compatibility with Perl in a number of ways:
2752    
2753     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2754     to request this handling. The option can be set at compile or exec time.
2755    
2756     (b) $ matches before a terminating newline by default; an option
2757     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2758     mode). The option can be set at compile or exec time.
2759    
2760     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2761     the same as Perl's. If the decimal number it represents is less than 10
2762     or there aren't that many previous left capturing parentheses, an octal
2763     escape is read. Inside a character class, it's always an octal escape,
2764     even if it is a single digit.
2765    
2766     (d) An escaped but undefined alphabetic character is taken as a literal,
2767     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2768     escapes.
2769    
2770     (e) {0} is now permitted. (The previous item is removed from the compiled
2771     pattern).
2772    
2773     5. Changed all the names of code files so that the basic parts are no longer
2774     than 10 characters, and abolished the teeny "globals.c" file.
2775    
2776     6. Changed the handling of character classes; they are now done with a 32-byte
2777     bit map always.
2778    
2779     7. Added the -d and /D options to pcretest to make it possible to look at the
2780     internals of compilation without having to recompile pcre.
2781    
2782    
2783     Version 0.95 23-Sep-97
2784     ----------------------
2785    
2786     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2787     \x20 at the start of a run of normal characters. These were being treated as
2788     real characters, instead of the source characters being re-checked.
2789    
2790    
2791     Version 0.94 18-Sep-97
2792     ----------------------
2793    
2794     1. The functions are now thread-safe, with the caveat that the global variables
2795     containing pointers to malloc() and free() or alternative functions are the
2796     same for all threads.
2797    
2798     2. Get pcre_study() to generate a bitmap of initial characters for non-
2799     anchored patterns when this is possible, and use it if passed to pcre_exec().
2800    
2801    
2802     Version 0.93 15-Sep-97
2803     ----------------------
2804    
2805     1. /(b)|(:+)/ was computing an incorrect first character.
2806    
2807     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2808     but not actually doing anything yet.
2809    
2810     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2811     as Perl does (e.g. [-az] or [az-]).
2812    
2813     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2814     all possible positions.
2815    
2816     5. Split up into different modules to avoid including unneeded functions in a
2817     compiled binary. However, compile and exec are still in one module. The "study"
2818     function is split off.
2819    
2820     6. The character tables are now in a separate module whose source is generated
2821     by an auxiliary program - but can then be edited by hand if required. There are
2822     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2823     toupper() in the code.
2824    
2825     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2826     make them global. Abolish the function for setting them, as the caller can now
2827     set them directly.
2828    
2829    
2830     Version 0.92 11-Sep-97
2831     ----------------------
2832    
2833     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2834     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2835    
2836     2. Caseless matching was not working in character classes if the characters in
2837     the pattern were in upper case.
2838    
2839     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2840    
2841     4. Make PCRE_ANCHORED public and accept as a compile option.
2842    
2843     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2844     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2845     pass them.
2846    
2847     6. Give an error if bad option bits passed at compile or run time.
2848    
2849     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2850     pcretest to cause it to pass that flag.
2851    
2852     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2853     options, and the first character, if set.
2854    
2855     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2856    
2857    
2858     Version 0.91 10-Sep-97
2859     ----------------------
2860    
2861     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
2862     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
2863    
2864     2. PCRE was looping on encountering an indefinitely repeated back reference to
2865     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
2866     Perl does - treats the match as successful.
2867    
2868     ****

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12