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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 256 - (hide annotations) (download)
Wed Sep 19 08:57:53 2007 UTC (7 years, 1 month ago) by ph10
File size: 141531 byte(s)
Patches to C++ code to avoid compiler warnings.

1 nigel 3 ChangeLog for PCRE
2     ------------------
3    
4 ph10 230 Version 7.4 10-Sep-07
5     ---------------------
6    
7     1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
8 ph10 231 means that a class such as [\s] counted as "explicit reference to CR or
9     LF". That isn't really right - the whole point of the change was to try to
10     help when there was an actual mention of one of the two characters. So now
11     the change happens only if \r or \n (or a literal CR or LF) character is
12 ph10 230 encountered.
13 ph10 231
14     2. The 32-bit options word was also used for 6 internal flags, but the numbers
15     of both had grown to the point where there were only 3 bits left.
16     Fortunately, there was spare space in the data structure, and so I have
17     moved the internal flags into a new 16-bit field to free up more option
18 ph10 230 bits.
19 ph10 231
20     3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
21     but did not set the internal JCHANGED flag - either of these is enough to
22     control the way the "get" function works - but the PCRE_INFO_JCHANGED
23     facility is supposed to tell if (?J) was ever used, so now (?J) at the
24 ph10 230 start sets both bits.
25    
26 ph10 231 4. Added options (at build time, compile time, exec time) to change \R from
27     matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
28 ph10 230
29 ph10 243 5. doc/pcresyntax.html was missing from the distribution.
30    
31     6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
32 ph10 233 compatibility, even though it is no longer used.
33 ph10 243
34     7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
35 ph10 254 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
36 ph10 248 windows.h file is present (where different names are used).
37 ph10 243
38     8. Changed all #include <config.h> to #include "config.h". There were also
39 ph10 236 some further <pcre.h> cases that I changed to "pcre.h".
40 ph10 243
41 ph10 239 9. When pcregrep was used with the --colour option, it missed the line ending
42 ph10 243 sequence off the lines that it output.
43    
44     10. It was pointed out to me that arrays of string pointers cause lots of
45     relocations when a shared library is dynamically loaded. A technique of
46     using a single long string with a table of offsets can drastically reduce
47     these. I have refactored PCRE in four places to do this. The result is
48 ph10 240 dramatic:
49 ph10 243
50 ph10 240 Originally: 290
51     After changing UCP table: 187
52 ph10 243 After changing error message table: 43
53 ph10 240 After changing table of "verbs" 36
54     After changing table of Posix names 22
55 ph10 243
56 ph10 240 Thanks to the folks working on Gregex for glib for this insight.
57 ph10 247
58 ph10 244 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
59 ph10 247 unicode-properties was also set.
60    
61 ph10 248 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
62 ph10 231
63 ph10 253 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
64     checked only for CRLF.
65 ph10 255
66     14. Added casts to pcretest.c to avoid compiler warnings.
67 ph10 233
68 ph10 256 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
69 ph10 253
70 ph10 256
71 ph10 228 Version 7.3 28-Aug-07
72 ph10 157 ---------------------
73    
74 ph10 189 1. In the rejigging of the build system that eventually resulted in 7.1, the
75     line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
76     brackets there is not right, since it causes compilers to look for an
77     installed pcre.h, not the version that is in the source that is being
78     compiled (which of course may be different). I have changed it back to:
79 ph10 197
80 ph10 189 #include "pcre.h"
81 ph10 197
82     I have a vague recollection that the change was concerned with compiling in
83     different directories, but in the new build system, that is taken care of
84     by the VPATH setting the Makefile.
85    
86 ph10 190 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
87     when the subject happened to end in the byte 0x85 (e.g. if the last
88     character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
89     characters but of course it shouldn't be taken as a newline when it is part
90     of another character. The bug was that, for an unlimited repeat of . in
91     not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
92     characters when looking for a newline.
93 ph10 197
94     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
95    
96     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
97     in debug output.
98    
99 ph10 201 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
100     long printing in the pcrecpp unittest when running under MinGW.
101 ph10 189
102 ph10 197 6. ESC_K was left out of the EBCDIC table.
103 ph10 189
104 ph10 197 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
105     parentheses; I made it 1000, which seemed large enough. Unfortunately, the
106     limit also applies to "virtual nesting" when a pattern is recursive, and in
107     this case 1000 isn't so big. I have been able to remove this limit at the
108     expense of backing off one optimization in certain circumstances. Normally,
109     when pcre_exec() would call its internal match() function recursively and
110     immediately return the result unconditionally, it uses a "tail recursion"
111     feature to save stack. However, when a subpattern that can match an empty
112     string has an unlimited repetition quantifier, it no longer makes this
113     optimization. That gives it a stack frame in which to save the data for
114     checking that an empty string has been matched. Previously this was taken
115     from the 1000-entry workspace that had been reserved. So now there is no
116     explicit limit, but more stack is used.
117 ph10 189
118 ph10 200 8. Applied Daniel's patches to solve problems with the import/export magic
119     syntax that is required for Windows, and which was going wrong for the
120     pcreposix and pcrecpp parts of the library. These were overlooked when this
121     problem was solved for the main library.
122 ph10 197
123 ph10 202 9. There were some crude static tests to avoid integer overflow when computing
124     the size of patterns that contain repeated groups with explicit upper
125     limits. As the maximum quantifier is 65535, the maximum group length was
126     set at 30,000 so that the product of these two numbers did not overflow a
127     32-bit integer. However, it turns out that people want to use groups that
128     are longer than 30,000 bytes (though not repeat them that many times).
129     Change 7.0/17 (the refactoring of the way the pattern size is computed) has
130     made it possible to implement the integer overflow checks in a much more
131     dynamic way, which I have now done. The artificial limitation on group
132     length has been removed - we now have only the limit on the total length of
133     the compiled pattern, which depends on the LINK_SIZE setting.
134 ph10 208
135     10. Fixed a bug in the documentation for get/copy named substring when
136     duplicate names are permitted. If none of the named substrings are set, the
137     functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
138     empty string.
139    
140     11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
141     instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
142     because the ] is interpreted as the first data character and the
143     terminating ] is not found. PCRE has been made compatible with Perl in this
144     regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
145     cause memory overwriting.
146    
147 ph10 206 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
148     string has been matched (to stop an infinite loop). It was not recognizing
149 ph10 208 a conditional subpattern that could match an empty string if that
150 ph10 206 subpattern was within another subpattern. For example, it looped when
151 ph10 208 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
152 ph10 206 condition was not nested. This bug has been fixed.
153 ph10 208
154 ph10 207 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
155     past the start of the subject in the presence of bytes with the top bit
156     set, for example "\x8aBCD".
157 ph10 211
158 ph10 210 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
159     (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
160 ph10 200
161 ph10 211 14. Optimized (?!) to (*FAIL).
162 ph10 202
163 ph10 212 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
164     This restricts code points to be within the range 0 to 0x10FFFF, excluding
165     the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
166     full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
167     does: it's just the validity check that is more restrictive.
168 ph10 220
169     16. Inserted checks for integer overflows during escape sequence (backslash)
170     processing, and also fixed erroneous offset values for syntax errors during
171     backslash processing.
172    
173 ph10 214 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
174 ph10 220 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
175    
176 ph10 215 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
177     caused an overrun.
178 ph10 220
179     19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
180     something other than just ASCII characters) inside a group that had an
181     unlimited repeat caused a loop at compile time (while checking to see
182     whether the group could match an empty string).
183    
184 ph10 217 20. Debugging a pattern containing \p or \P could cause a crash. For example,
185     [\P{Any}] did so. (Error in the code for printing property names.)
186 ph10 210
187 ph10 220 21. An orphan \E inside a character class could cause a crash.
188    
189     22. A repeated capturing bracket such as (A)? could cause a wild memory
190 ph10 218 reference during compilation.
191 ph10 220
192     23. There are several functions in pcre_compile() that scan along a compiled
193     expression for various reasons (e.g. to see if it's fixed length for look
194 ph10 218 behind). There were bugs in these functions when a repeated \p or \P was
195 ph10 220 present in the pattern. These operators have additional parameters compared
196     with \d, etc, and these were not being taken into account when moving along
197 ph10 218 the compiled data. Specifically:
198 ph10 220
199     (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
200     length.
201    
202     (b) An item such as \pL+ within a repeated group could cause crashes or
203 ph10 218 loops.
204 ph10 220
205 ph10 218 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
206 ph10 220 "reference to non-existent subpattern" error.
207    
208 ph10 221 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
209    
210 ph10 219 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
211 ph10 220 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
212 ph10 211
213 ph10 222 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
214 ph10 218
215 ph10 227 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
216 ph10 223 character were causing crashes (broken optimization).
217 ph10 227
218 ph10 224 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
219 ph10 227 \p or \P) caused a compile-time loop.
220    
221 ph10 226 28. More problems have arisen in unanchored patterns when CRLF is a valid line
222     break. For example, the unstudied pattern [\r\n]A does not match the string
223     "\r\nA" because change 7.0/46 below moves the current point on by two
224     characters after failing to match at the start. However, the pattern \nA
225 ph10 227 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
226     the same is true. There doesn't seem any very clean way out of this, but
227     what I have chosen to do makes the common cases work: PCRE now takes note
228     of whether there can be an explicit match for \r or \n anywhere in the
229     pattern, and if so, 7.0/46 no longer applies. As part of this change,
230     there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
231     pattern has explicit CR or LF references.
232 ph10 222
233 ph10 227 29. Added (*CR) etc for changing newline setting at start of pattern.
234 ph10 223
235 ph10 227
236 ph10 189 Version 7.2 19-Jun-07
237     ---------------------
238    
239 ph10 157 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
240     which is apparently normally available under Windows.
241 ph10 159
242     2. Re-jig the pcregrep tests with different newline settings in an attempt
243     to make them independent of the local environment's newline setting.
244 ph10 157
245 ph10 160 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
246 ph10 165
247 ph10 161 4. Some of the "internals" tests were previously cut out when the link size
248     was not 2, because the output contained actual offsets. The recent new
249 ph10 165 "Z" feature of pcretest means that these can be cut out, making the tests
250     usable with all link sizes.
251    
252 ph10 164 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
253     stack recursion. This gives a massive performance boost under BSD, but just
254 ph10 165 a small improvement under Linux. However, it saves one field in the frame
255 ph10 164 in all cases.
256 ph10 172
257 ph10 166 6. Added more features from the forthcoming Perl 5.10:
258 ph10 172
259 ph10 166 (a) (?-n) (where n is a string of digits) is a relative subroutine or
260     recursion call. It refers to the nth most recently opened parentheses.
261 ph10 172
262 ph10 166 (b) (?+n) is also a relative subroutine call; it refers to the nth next
263 ph10 172 to be opened parentheses.
264    
265     (c) Conditions that refer to capturing parentheses can be specified
266 ph10 167 relatively, for example, (?(-2)... or (?(+3)...
267 ph10 172
268 ph10 168 (d) \K resets the start of the current match so that everything before
269 ph10 172 is not part of it.
270    
271 ph10 171 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
272 ph10 172
273 ph10 171 (f) \g{name} is another synonym - part of Perl 5.10's unification of
274 ph10 172 reference syntax.
275 ph10 182
276 ph10 175 (g) (?| introduces a group in which the numbering of parentheses in each
277 ph10 182 alternative starts with the same number.
278 ph10 172
279 ph10 182 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
280    
281 ph10 172 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
282     PCRE_INFO_JCHANGED.
283    
284     8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
285     terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
286 ph10 170 for detecting groups that can match an empty string.
287 ph10 159
288 ph10 172 9. A pattern with a very large number of alternatives (more than several
289     hundred) was running out of internal workspace during the pre-compile
290     phase, where pcre_compile() figures out how much memory will be needed. A
291     bit of new cunning has reduced the workspace needed for groups with
292     alternatives. The 1000-alternative test pattern now uses 12 bytes of
293     workspace instead of running out of the 4096 that are available.
294 ph10 182
295 ph10 176 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
296 ph10 172
297 ph10 179 11. Applied patch from Google to remove an optimization that didn't quite work.
298     The report of the bug said:
299 ph10 182
300 ph10 179 pcrecpp::RE("a*").FullMatch("aaa") matches, while
301     pcrecpp::RE("a*?").FullMatch("aaa") does not, and
302 ph10 182 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
303 ph10 185
304 ph10 184 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
305 ph10 185 it matched the wrong number of bytes.
306 ph10 172
307 ph10 179
308 ph10 155 Version 7.1 24-Apr-07
309 ph10 98 ---------------------
310    
311 ph10 111 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
312 ph10 122 that is more "standard", making use of automake and other Autotools. There
313 ph10 99 is some re-arrangement of the files and adjustment of comments consequent
314     on this.
315 ph10 111
316     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
317     for recursive directory scanning broke on some systems because the files
318     are not scanned in any specific order and on different systems the order
319     was different. A call to "sort" has been inserted into RunGrepTest for the
320     approprate test as a short-term fix. In the longer term there may be an
321 ph10 100 alternative.
322 ph10 111
323 ph10 100 3. I had an email from Eric Raymond about problems translating some of PCRE's
324 ph10 111 man pages to HTML (despite the fact that I distribute HTML pages, some
325     people do their own conversions for various reasons). The problems
326     concerned the use of low-level troff macros .br and .in. I have therefore
327     removed all such uses from the man pages (some were redundant, some could
328 ph10 113 be replaced by .nf/.fi pairs). The 132html script that I use to generate
329     HTML has been updated to handle .nf/.fi and to complain if it encounters
330     .br or .in.
331 ph10 111
332 ph10 100 4. Updated comments in configure.ac that get placed in config.h.in and also
333 ph10 123 arranged for config.h to be included in the distribution, with the name
334 ph10 111 config.h.generic, for the benefit of those who have to compile without
335     Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
336    
337     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
338     Weber: (1) pcre_internal.h was missing some function renames; (2) updated
339 ph10 127 makevp.bat for the current PCRE, using the additional files
340 ph10 135 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
341 ph10 111
342     6. A Windows user reported a minor discrepancy with test 2, which turned out
343     to be caused by a trailing space on an input line that had got lost in his
344 ph10 102 copy. The trailing space was an accident, so I've just removed it.
345 ph10 111
346 ph10 104 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
347 ph10 111 that is needed.
348    
349 ph10 105 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
350     as "const" (a) because they are and (b) because it helps the PHP
351     maintainers who have recently made a script to detect big data structures
352 ph10 111 in the php code that should be moved to the .rodata section. I remembered
353     to update Builducptable as well, so it won't revert if ucptable.h is ever
354 ph10 105 re-created.
355 ph10 111
356     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
357 ph10 107 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
358 ph10 111 order to be able to cut out the UTF-8 tables in the latter when UTF-8
359     support is not required. This saves 1.5-2K of code, which is important in
360 ph10 107 some applications.
361 ph10 111
362 ph10 110 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
363 ph10 111 so as not to refer to the tables, even though these functions will never be
364     called when UTF-8 support is disabled. Otherwise there are problems with a
365     shared library.
366    
367 ph10 118 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
368    
369     (a) It was defining its arguments as char * instead of void *.
370    
371 ph10 123 (b) It was assuming that all moves were upwards in memory; this was true
372     a long time ago when I wrote it, but is no longer the case.
373    
374 ph10 118 The emulated memove() is provided for those environments that have neither
375 ph10 123 memmove() nor bcopy(). I didn't think anyone used it these days, but that
376 ph10 118 is clearly not the case, as these two bugs were recently reported.
377 ph10 123
378 ph10 111 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
379 ph10 123 and Detrail to create the HTML documentation, the .txt form of the man
380     pages, and it removes trailing spaces from listed files. It also creates
381     pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
382     case, it wraps all the #defines with #ifndefs. This script should be run
383 ph10 111 before "make dist".
384 ph10 123
385 ph10 115 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
386     with Unicode property support.
387 ph10 123
388     (a) For a maximizing quantifier, if the two different cases of the
389     character were of different lengths in their UTF-8 codings (there are
390     some cases like this - I found 11), and the matching function had to
391 ph10 115 back up over a mixture of the two cases, it incorrectly assumed they
392     were both the same length.
393 ph10 123
394     (b) When PCRE was configured to use the heap rather than the stack for
395     recursion during matching, it was not correctly preserving the data for
396     the other case of a UTF-8 character when checking ahead for a match
397     while processing a minimizing repeat. If the check also involved
398 ph10 115 matching a wide character, but failed, corruption could cause an
399     erroneous result when trying to check for a repeat of the original
400     character.
401 ph10 123
402 ph10 116 13. Some tidying changes to the testing mechanism:
403 ph10 98
404 ph10 116 (a) The RunTest script now detects the internal link size and whether there
405     is UTF-8 and UCP support by running ./pcretest -C instead of relying on
406 ph10 123 values substituted by "configure". (The RunGrepTest script already did
407     this for UTF-8.) The configure.ac script no longer substitutes the
408     relevant variables.
409    
410 ph10 116 (b) The debugging options /B and /D in pcretest show the compiled bytecode
411     with length and offset values. This means that the output is different
412     for different internal link sizes. Test 2 is skipped for link sizes
413     other than 2 because of this, bypassing the problem. Unfortunately,
414     there was also a test in test 3 (the locale tests) that used /B and
415 ph10 123 failed for link sizes other than 2. Rather than cut the whole test out,
416     I have added a new /Z option to pcretest that replaces the length and
417     offset values with spaces. This is now used to make test 3 independent
418 ph10 122 of link size. (Test 2 will be tidied up later.)
419 ph10 123
420     14. If erroroffset was passed as NULL to pcre_compile, it provoked a
421 ph10 122 segmentation fault instead of returning the appropriate error message.
422 ph10 134
423 ph10 131 15. In multiline mode when the newline sequence was set to "any", the pattern
424 ph10 134 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
425     This doesn't seem right; it now treats the CRLF combination as the line
426     ending, and so does not match in that case. It's only a pattern such as ^$
427     that would hit this one: something like ^ABC$ would have failed after \r
428     and then tried again after \r\n.
429    
430 ph10 131 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
431 ph10 134 in an attempt to make files that differ only in their line terminators
432     compare equal. This works on Linux.
433 ph10 142
434 ph10 141 17. Under certain error circumstances pcregrep might try to free random memory
435     as it exited. This is now fixed, thanks to valgrind.
436 ph10 142
437 ph10 141 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
438 ph10 142 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
439     was because its rules for how to advance for /g after matching an empty
440 ph10 143 string at the end of a line did not allow for this case. They now check for
441     it specially.
442 ph10 150
443     20. pcretest is supposed to handle patterns and data of any length, by
444     extending its buffers when necessary. It was getting this wrong when the
445 ph10 147 buffer for a data line had to be extended.
446 ph10 150
447 ph10 149 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
448 ph10 150 CRLF as a newline sequence.
449 ph10 152
450 ph10 151 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
451 ph10 152 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
452     I have nevertheless tidied it up.
453 ph10 111
454 ph10 152 23. Added some casts to kill warnings from HP-UX ia64 compiler.
455 ph10 123
456 ph10 153 24. Added a man page for pcre-config.
457 ph10 152
458 ph10 153
459 nigel 93 Version 7.0 19-Dec-06
460     ---------------------
461    
462     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
463     moving to gcc 4.1.1.
464    
465     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
466     sys/time.h, which is documented as needed for this function. It doesn't
467     seem to matter on Linux, but it showed up on some releases of OS X.
468    
469     3. It seems that there are systems where bytes whose values are greater than
470     127 match isprint() in the "C" locale. The "C" locale should be the
471     default when a C program starts up. In most systems, only ASCII printing
472     characters match isprint(). This difference caused the output from pcretest
473     to vary, making some of the tests fail. I have changed pcretest so that:
474    
475     (a) When it is outputting text in the compiled version of a pattern, bytes
476     other than 32-126 are always shown as hex escapes.
477    
478     (b) When it is outputting text that is a matched part of a subject string,
479     it does the same, unless a different locale has been set for the match
480     (using the /L modifier). In this case, it uses isprint() to decide.
481    
482     4. Fixed a major bug that caused incorrect computation of the amount of memory
483     required for a compiled pattern when options that changed within the
484     pattern affected the logic of the preliminary scan that determines the
485     length. The relevant options are -x, and -i in UTF-8 mode. The result was
486     that the computed length was too small. The symptoms of this bug were
487     either the PCRE error "internal error: code overflow" from pcre_compile(),
488     or a glibc crash with a message such as "pcretest: free(): invalid next
489     size (fast)". Examples of patterns that provoked this bug (shown in
490     pcretest format) are:
491    
492     /(?-x: )/x
493     /(?x)(?-x: \s*#\s*)/
494     /((?i)[\x{c0}])/8
495     /(?i:[\x{c0}])/8
496    
497     HOWEVER: Change 17 below makes this fix obsolete as the memory computation
498     is now done differently.
499    
500     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
501     wrapper classes; (b) implement a new function in the C++ scanner that is
502     more efficient than the old way of doing things because it avoids levels of
503     recursion in the regex matching; (c) add a paragraph to the documentation
504     for the FullMatch() function.
505    
506     6. The escape sequence \n was being treated as whatever was defined as
507     "newline". Not only was this contrary to the documentation, which states
508     that \n is character 10 (hex 0A), but it also went horribly wrong when
509     "newline" was defined as CRLF. This has been fixed.
510    
511     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
512     was being set to -1 for the "end of line" case (supposedly a value that no
513     character can have). Though this value is never used (the check for end of
514     line is "zero bytes in current character"), it caused compiler complaints.
515     I've changed it to 0xffffffff.
516    
517     8. In pcre_version.c, the version string was being built by a sequence of
518     C macros that, in the event of PCRE_PRERELEASE being defined as an empty
519     string (as it is for production releases) called a macro with an empty
520     argument. The C standard says the result of this is undefined. The gcc
521     compiler treats it as an empty string (which was what was wanted) but it is
522     reported that Visual C gives an error. The source has been hacked around to
523     avoid this problem.
524    
525     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
526     builds of pcretest, and changed the call to _setmode() to use _O_BINARY
527     instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
528     of them did).
529    
530     10. Originally, pcretest opened its input and output without "b"; then I was
531     told that "b" was needed in some environments, so it was added for release
532     5.0 to both the input and output. (It makes no difference on Unix-like
533     systems.) Later I was told that it is wrong for the input on Windows. I've
534     now abstracted the modes into two macros, to make it easier to fiddle with
535     them, and removed "b" from the input mode under Windows.
536    
537     11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
538    
539     12. Added -help and --help to pcretest as an official way of being reminded
540     of the options.
541    
542     13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
543     and pcrecpp.cc because they annoy compilers at high warning levels.
544    
545     14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
546    
547     15. Fixed an occurrence of == in configure.ac that should have been = (shell
548     scripts are not C programs :-) and which was not noticed because it works
549     on Linux.
550    
551     16. pcretest is supposed to handle any length of pattern and data line (as one
552     line or as a continued sequence of lines) by extending its input buffer if
553     necessary. This feature was broken for very long pattern lines, leading to
554     a string of junk being passed to pcre_compile() if the pattern was longer
555     than about 50K.
556    
557     17. I have done a major re-factoring of the way pcre_compile() computes the
558     amount of memory needed for a compiled pattern. Previously, there was code
559     that made a preliminary scan of the pattern in order to do this. That was
560     OK when PCRE was new, but as the facilities have expanded, it has become
561     harder and harder to keep it in step with the real compile phase, and there
562     have been a number of bugs (see for example, 4 above). I have now found a
563     cunning way of running the real compile function in a "fake" mode that
564     enables it to compute how much memory it would need, while actually only
565     ever using a few hundred bytes of working memory and without too many
566     tests of the mode. This should make future maintenance and development
567     easier. A side effect of this work is that the limit of 200 on the nesting
568     depth of parentheses has been removed (though this was never a serious
569     limitation, I suspect). However, there is a downside: pcre_compile() now
570     runs more slowly than before (30% or more, depending on the pattern). I
571     hope this isn't a big issue. There is no effect on runtime performance.
572    
573     18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
574     newline (only possible for the last line of a file) and it was a
575     pattern that set a locale (followed by /Lsomething), pcretest crashed.
576    
577     19. Added additional timing features to pcretest. (1) The -tm option now times
578     matching only, not compiling. (2) Both -t and -tm can be followed, as a
579     separate command line item, by a number that specifies the number of
580     repeats to use when timing. The default is 50000; this gives better
581     precision, but takes uncomfortably long for very large patterns.
582    
583     20. Extended pcre_study() to be more clever in cases where a branch of a
584     subpattern has no definite first character. For example, (a*|b*)[cd] would
585     previously give no result from pcre_study(). Now it recognizes that the
586     first character must be a, b, c, or d.
587    
588     21. There was an incorrect error "recursive call could loop indefinitely" if
589     a subpattern (or the entire pattern) that was being tested for matching an
590     empty string contained only one non-empty item after a nested subpattern.
591     For example, the pattern (?>\x{100}*)\d(?R) provoked this error
592     incorrectly, because the \d was being skipped in the check.
593    
594     22. The pcretest program now has a new pattern option /B and a command line
595     option -b, which is equivalent to adding /B to every pattern. This causes
596     it to show the compiled bytecode, without the additional information that
597     -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
598     is the same as /B/I).
599    
600     23. A new optimization is now able automatically to treat some sequences such
601     as a*b as a*+b. More specifically, if something simple (such as a character
602     or a simple class like \d) has an unlimited quantifier, and is followed by
603     something that cannot possibly match the quantified thing, the quantifier
604     is automatically "possessified".
605    
606     24. A recursive reference to a subpattern whose number was greater than 39
607     went wrong under certain circumstances in UTF-8 mode. This bug could also
608     have affected the operation of pcre_study().
609    
610     25. Realized that a little bit of performance could be had by replacing
611     (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
612    
613     26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
614    
615     27. Possessive quantifiers such as a++ were previously implemented by turning
616     them into atomic groups such as ($>a+). Now they have their own opcodes,
617     which improves performance. This includes the automatically created ones
618     from 23 above.
619    
620     28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
621     lookahead was broken if it was not anchored. PCRE was mistakenly expecting
622     the first matched character to be a colon. This applied both to named and
623     numbered groups.
624    
625     29. The ucpinternal.h header file was missing its idempotency #ifdef.
626    
627     30. I was sent a "project" file called libpcre.a.dev which I understand makes
628     building PCRE on Windows easier, so I have included it in the distribution.
629    
630     31. There is now a check in pcretest against a ridiculously large number being
631     returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
632     loop, the loop is abandoned.
633    
634     32. Forward references to subpatterns in conditions such as (?(2)...) where
635     subpattern 2 is defined later cause pcre_compile() to search forwards in
636     the pattern for the relevant set of parentheses. This search went wrong
637     when there were unescaped parentheses in a character class, parentheses
638     escaped with \Q...\E, or parentheses in a #-comment in /x mode.
639    
640     33. "Subroutine" calls and backreferences were previously restricted to
641     referencing subpatterns earlier in the regex. This restriction has now
642     been removed.
643    
644     34. Added a number of extra features that are going to be in Perl 5.10. On the
645     whole, these are just syntactic alternatives for features that PCRE had
646     previously implemented using the Python syntax or my own invention. The
647     other formats are all retained for compatibility.
648    
649     (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
650     as (?P<name>...). The new forms, as well as being in Perl 5.10, are
651     also .NET compatible.
652    
653     (b) A recursion or subroutine call to a named group can now be defined as
654     (?&name) as well as (?P>name).
655    
656     (c) A backreference to a named group can now be defined as \k<name> or
657     \k'name' as well as (?P=name). The new forms, as well as being in Perl
658     5.10, are also .NET compatible.
659    
660     (d) A conditional reference to a named group can now use the syntax
661     (?(<name>) or (?('name') as well as (?(name).
662    
663     (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
664     groups (named and numbered) that are never evaluated inline, but can be
665     called as "subroutines" from elsewhere. In effect, the DEFINE condition
666     is always false. There may be only one alternative in such a group.
667    
668     (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
669     as the simple (?(R). The condition is true only if the most recent
670     recursion is that of the given number or name. It does not search out
671     through the entire recursion stack.
672    
673     (g) The escape \gN or \g{N} has been added, where N is a positive or
674     negative number, specifying an absolute or relative reference.
675    
676     35. Tidied to get rid of some further signed/unsigned compiler warnings and
677     some "unreachable code" warnings.
678    
679     36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
680     things, this adds five new scripts.
681    
682     37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
683     There were also incompatibilities regarding the handling of \Q..\E inside
684     character classes, for example with patterns like [\Qa\E-\Qz\E] where the
685     hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
686    
687     38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
688     matches an empty string, and forcibly breaks the loop. There were bugs in
689     this code in non-simple cases. For a pattern such as ^(a()*)* matched
690     against aaaa the result was just "a" rather than "aaaa", for example. Two
691     separate and independent bugs (that affected different cases) have been
692     fixed.
693    
694     39. Refactored the code to abolish the use of different opcodes for small
695     capturing bracket numbers. This is a tidy that I avoided doing when I
696     removed the limit on the number of capturing brackets for 3.5 back in 2001.
697     The new approach is not only tidier, it makes it possible to reduce the
698     memory needed to fix the previous bug (38).
699    
700     40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
701     sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
702     processing dot, circumflex, or dollar metacharacters, or #-comments in /x
703     mode.
704    
705     41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
706     report.
707    
708     42. Applied patch, originally from Ari Pollak, modified by Google, to allow
709     copy construction and assignment in the C++ wrapper.
710    
711     43. Updated pcregrep to support "--newline=any". In the process, I fixed a
712     couple of bugs that could have given wrong results in the "--newline=crlf"
713     case.
714    
715     44. Added a number of casts and did some reorganization of signed/unsigned int
716     variables following suggestions from Dair Grant. Also renamed the variable
717     "this" as "item" because it is a C++ keyword.
718    
719     45. Arranged for dftables to add
720    
721     #include "pcre_internal.h"
722    
723     to pcre_chartables.c because without it, gcc 4.x may remove the array
724     definition from the final binary if PCRE is built into a static library and
725     dead code stripping is activated.
726    
727     46. For an unanchored pattern, if a match attempt fails at the start of a
728     newline sequence, and the newline setting is CRLF or ANY, and the next two
729     characters are CRLF, advance by two characters instead of one.
730    
731    
732 nigel 91 Version 6.7 04-Jul-06
733     ---------------------
734    
735     1. In order to handle tests when input lines are enormously long, pcretest has
736     been re-factored so that it automatically extends its buffers when
737     necessary. The code is crude, but this _is_ just a test program. The
738     default size has been increased from 32K to 50K.
739    
740     2. The code in pcre_study() was using the value of the re argument before
741     testing it for NULL. (Of course, in any sensible call of the function, it
742     won't be NULL.)
743    
744     3. The memmove() emulation function in pcre_internal.h, which is used on
745     systems that lack both memmove() and bcopy() - that is, hardly ever -
746     was missing a "static" storage class specifier.
747    
748     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
749     containing an extended class (one that cannot be represented by a bitmap
750     because it contains high-valued characters or Unicode property items, e.g.
751     [\pZ]). Almost always one would set UTF-8 mode when processing such a
752     pattern, but PCRE should not loop if you do not (it no longer does).
753     [Detail: two cases were found: (a) a repeated subpattern containing an
754     extended class; (b) a recursive reference to a subpattern that followed a
755     previous extended class. It wasn't skipping over the extended class
756     correctly when UTF-8 mode was not set.]
757    
758     5. A negated single-character class was not being recognized as fixed-length
759     in lookbehind assertions such as (?<=[^f]), leading to an incorrect
760     compile error "lookbehind assertion is not fixed length".
761    
762     6. The RunPerlTest auxiliary script was showing an unexpected difference
763     between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
764     write a Perl script that can interpret lines of an input file either as
765     byte characters or as UTF-8, which is what "perltest" was being required to
766     do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
767     can't do is switch easily at run time between having the "use utf8;" pragma
768     or not. In the end, I fudged it by using the RunPerlTest script to insert
769     "use utf8;" explicitly for the UTF-8 tests.
770    
771     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
772     the end of the subject string, contrary to the documentation and to what
773     Perl does. This was true of both matching functions. Now it matches only at
774     the start of the subject and immediately after *internal* newlines.
775    
776     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
777     a pointer to an int instead of a pointer to an unsigned long int. This
778     caused problems on 64-bit systems.
779    
780     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
781     instance of the 'standard' template library not being so standard".
782    
783     10. There was no check on the number of named subpatterns nor the maximum
784     length of a subpattern name. The product of these values is used to compute
785     the size of the memory block for a compiled pattern. By supplying a very
786     long subpattern name and a large number of named subpatterns, the size
787     computation could be caused to overflow. This is now prevented by limiting
788     the length of names to 32 characters, and the number of named subpatterns
789     to 10,000.
790    
791     11. Subpatterns that are repeated with specific counts have to be replicated in
792     the compiled pattern. The size of memory for this was computed from the
793     length of the subpattern and the repeat count. The latter is limited to
794     65535, but there was no limit on the former, meaning that integer overflow
795     could in principle occur. The compiled length of a repeated subpattern is
796     now limited to 30,000 bytes in order to prevent this.
797    
798     12. Added the optional facility to have named substrings with the same name.
799    
800     13. Added the ability to use a named substring as a condition, using the
801     Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
802     are numbers (not recommended). Forward references are permitted.
803    
804     14. Added forward references in named backreferences (if you see what I mean).
805    
806     15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
807     pattern could run off the end of the subject. For example, the pattern
808     "(?s)(.{1,5})"8 did this with the subject "ab".
809    
810     16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
811     PCRE_CASELESS was set when matching characters that were quantified with ?
812     or *.
813    
814     17. A character class other than a single negated character that had a minimum
815     but no maximum quantifier - for example [ab]{6,} - was not handled
816     correctly by pce_dfa_exec(). It would match only one character.
817    
818     18. A valid (though odd) pattern that looked like a POSIX character
819     class but used an invalid character after [ (for example [[,abc,]]) caused
820     pcre_compile() to give the error "Failed: internal error: code overflow" or
821     in some cases to crash with a glibc free() error. This could even happen if
822     the pattern terminated after [[ but there just happened to be a sequence of
823     letters, a binary zero, and a closing ] in the memory that followed.
824    
825     19. Perl's treatment of octal escapes in the range \400 to \777 has changed
826     over the years. Originally (before any Unicode support), just the bottom 8
827     bits were taken. Thus, for example, \500 really meant \100. Nowadays the
828     output from "man perlunicode" includes this:
829    
830     The regular expression compiler produces polymorphic opcodes. That
831     is, the pattern adapts to the data and automatically switches to
832     the Unicode character scheme when presented with Unicode data--or
833     instead uses a traditional byte scheme when presented with byte
834     data.
835    
836     Sadly, a wide octal escape does not cause a switch, and in a string with
837     no other multibyte characters, these octal escapes are treated as before.
838     Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
839     /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
840     Unicode string.
841    
842     I have not perpetrated such confusion in PCRE. Up till now, it took just
843     the bottom 8 bits, as in old Perl. I have now made octal escapes with
844     values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
845     translate to the appropriate multibyte character.
846    
847     29. Applied some refactoring to reduce the number of warnings from Microsoft
848     and Borland compilers. This has included removing the fudge introduced
849     seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
850     a warning about an unused variable.
851    
852     21. PCRE has not included VT (character 0x0b) in the set of whitespace
853     characters since release 4.0, because Perl (from release 5.004) does not.
854     [Or at least, is documented not to: some releases seem to be in conflict
855     with the documentation.] However, when a pattern was studied with
856     pcre_study() and all its branches started with \s, PCRE still included VT
857     as a possible starting character. Of course, this did no harm; it just
858     caused an unnecessary match attempt.
859    
860     22. Removed a now-redundant internal flag bit that recorded the fact that case
861     dependency changed within the pattern. This was once needed for "required
862     byte" processing, but is no longer used. This recovers a now-scarce options
863     bit. Also moved the least significant internal flag bit to the most-
864     significant bit of the word, which was not previously used (hangover from
865     the days when it was an int rather than a uint) to free up another bit for
866     the future.
867    
868     23. Added support for CRLF line endings as well as CR and LF. As well as the
869     default being selectable at build time, it can now be changed at runtime
870     via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
871     specify that it is scanning data with non-default line endings.
872    
873     24. Changed the definition of CXXLINK to make it agree with the definition of
874     LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
875    
876     25. Applied Ian Taylor's patches to avoid using another stack frame for tail
877     recursions. This makes a big different to stack usage for some patterns.
878    
879     26. If a subpattern containing a named recursion or subroutine reference such
880     as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
881     the space required for the compiled pattern went wrong and gave too small a
882     value. Depending on the environment, this could lead to "Failed: internal
883     error: code overflow at offset 49" or "glibc detected double free or
884     corruption" errors.
885    
886     27. Applied patches from Google (a) to support the new newline modes and (b) to
887     advance over multibyte UTF-8 characters in GlobalReplace.
888    
889     28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
890     difference for some implementation of PCRE in some Windows version.
891    
892     29. Added some extra testing facilities to pcretest:
893    
894     \q<number> in a data line sets the "match limit" value
895     \Q<number> in a data line sets the "match recursion limt" value
896     -S <number> sets the stack size, where <number> is in megabytes
897    
898     The -S option isn't available for Windows.
899    
900    
901 nigel 89 Version 6.6 06-Feb-06
902     ---------------------
903    
904     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
905     in pcreposix.h. I have copied the definition from pcre.h.
906    
907     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
908     because pcre.h is no longer a built file.
909    
910     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
911     not normally included in the compiled code.
912    
913    
914 nigel 87 Version 6.5 01-Feb-06
915     ---------------------
916    
917     1. When using the partial match feature with pcre_dfa_exec(), it was not
918     anchoring the second and subsequent partial matches at the new starting
919     point. This could lead to incorrect results. For example, with the pattern
920     /1234/, partially matching against "123" and then "a4" gave a match.
921    
922     2. Changes to pcregrep:
923    
924     (a) All non-match returns from pcre_exec() were being treated as failures
925     to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
926     error message is output. Some extra information is given for the
927     PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
928     probably the only errors that are likely to be caused by users (by
929     specifying a regex that has nested indefinite repeats, for instance).
930     If there are more than 20 of these errors, pcregrep is abandoned.
931    
932     (b) A binary zero was treated as data while matching, but terminated the
933     output line if it was written out. This has been fixed: binary zeroes
934     are now no different to any other data bytes.
935    
936     (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
937     used to set a locale for matching. The --locale=xxxx long option has
938     been added (no short equivalent) to specify a locale explicitly on the
939     pcregrep command, overriding the environment variables.
940    
941     (d) When -B was used with -n, some line numbers in the output were one less
942     than they should have been.
943    
944     (e) Added the -o (--only-matching) option.
945    
946     (f) If -A or -C was used with -c (count only), some lines of context were
947     accidentally printed for the final match.
948    
949     (g) Added the -H (--with-filename) option.
950    
951     (h) The combination of options -rh failed to suppress file names for files
952     that were found from directory arguments.
953    
954     (i) Added the -D (--devices) and -d (--directories) options.
955    
956     (j) Added the -F (--fixed-strings) option.
957    
958     (k) Allow "-" to be used as a file name for -f as well as for a data file.
959    
960     (l) Added the --colo(u)r option.
961    
962     (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
963     is not present by default.
964    
965     3. A nasty bug was discovered in the handling of recursive patterns, that is,
966     items such as (?R) or (?1), when the recursion could match a number of
967     alternatives. If it matched one of the alternatives, but subsequently,
968     outside the recursion, there was a failure, the code tried to back up into
969     the recursion. However, because of the way PCRE is implemented, this is not
970     possible, and the result was an incorrect result from the match.
971    
972     In order to prevent this happening, the specification of recursion has
973     been changed so that all such subpatterns are automatically treated as
974     atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
975    
976     4. I had overlooked the fact that, in some locales, there are characters for
977     which isalpha() is true but neither isupper() nor islower() are true. In
978     the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
979     and ordfeminine) are like this. This affected the treatment of \w and \W
980     when they appeared in character classes, but not when they appeared outside
981     a character class. The bit map for "word" characters is now created
982     separately from the results of isalnum() instead of just taking it from the
983     upper, lower, and digit maps. (Plus the underscore character, of course.)
984    
985     5. The above bug also affected the handling of POSIX character classes such as
986     [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
987     permanent tables. Instead, the bit maps for such a class were previously
988     created as the appropriate unions of the upper, lower, and digit bitmaps.
989     Now they are created by subtraction from the [[:word:]] class, which has
990     its own bitmap.
991    
992     6. The [[:blank:]] character class matches horizontal, but not vertical space.
993     It is created by subtracting the vertical space characters (\x09, \x0a,
994     \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
995     subtraction was done in the overall bitmap for a character class, meaning
996     that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
997     be recognized. This bug has been fixed.
998    
999     7. Patches from the folks at Google:
1000    
1001     (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
1002     real life, but is still worth protecting against".
1003    
1004     (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
1005     regular expressions".
1006    
1007     (c) pcre_scanner.cc: avoid use of std::count() because not all systems
1008     have it.
1009    
1010     (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
1011     "configure" and the latter not, in order to fix a problem somebody had
1012     with compiling the Arg class on HP-UX.
1013    
1014     (e) Improve the error-handling of the C++ wrapper a little bit.
1015    
1016     (f) New tests for checking recursion limiting.
1017    
1018     8. The pcre_memmove() function, which is used only if the environment does not
1019     have a standard memmove() function (and is therefore rarely compiled),
1020     contained two bugs: (a) use of int instead of size_t, and (b) it was not
1021     returning a result (though PCRE never actually uses the result).
1022    
1023     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
1024     large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
1025     returned instead of calling malloc() with an overflowing number that would
1026     most likely cause subsequent chaos.
1027    
1028     10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
1029    
1030     11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
1031     with this option is matched, the nmatch and pmatch options of regexec() are
1032     ignored.
1033    
1034     12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
1035     provided in case anyone wants to the the POSIX interface with UTF-8
1036     strings.
1037    
1038     13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
1039     C++ linking (needed for some HP-UX environments).
1040    
1041     14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
1042     (unused parameter) and in the pcre_printint() function (omitted "default"
1043     switch label when the default is to do nothing).
1044    
1045     15. Added some code to make it possible, when PCRE is compiled as a C++
1046     library, to replace subject pointers for pcre_exec() with a smart pointer
1047     class, thus making it possible to process discontinuous strings.
1048    
1049     16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
1050     much the same function. They were added by different people who were trying
1051     to make PCRE easy to compile on non-Unix systems. It has been suggested
1052     that PCRE_EXPORT be abolished now that there is more automatic apparatus
1053     for compiling on Windows systems. I have therefore replaced it with
1054     PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
1055     defaults to "extern" for C or "extern C" for C++, which works fine on
1056     Unix-like systems. It is now possible to override the value of PCRE_DATA_
1057     SCOPE with something explicit in config.h. In addition:
1058    
1059     (a) pcreposix.h still had just "extern" instead of either of these macros;
1060     I have replaced it with PCRE_DATA_SCOPE.
1061    
1062     (b) Functions such as _pcre_xclass(), which are internal to the library,
1063     but external in the C sense, all had PCRE_EXPORT in their definitions.
1064     This is apparently wrong for the Windows case, so I have removed it.
1065     (It makes no difference on Unix-like systems.)
1066    
1067     17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
1068     of recursive calls to match(). This is different to MATCH_LIMIT because
1069     that limits the total number of calls to match(), not all of which increase
1070     the depth of recursion. Limiting the recursion depth limits the amount of
1071     stack (or heap if NO_RECURSE is set) that is used. The default can be set
1072     when PCRE is compiled, and changed at run time. A patch from Google adds
1073     this functionality to the C++ interface.
1074    
1075     18. Changes to the handling of Unicode character properties:
1076    
1077     (a) Updated the table to Unicode 4.1.0.
1078    
1079     (b) Recognize characters that are not in the table as "Cn" (undefined).
1080    
1081     (c) I revised the way the table is implemented to a much improved format
1082     which includes recognition of ranges. It now supports the ranges that
1083     are defined in UnicodeData.txt, and it also amalgamates other
1084     characters into ranges. This has reduced the number of entries in the
1085     table from around 16,000 to around 3,000, thus reducing its size
1086     considerably. I realized I did not need to use a tree structure after
1087     all - a binary chop search is just as efficient. Having reduced the
1088     number of entries, I extended their size from 6 bytes to 8 bytes to
1089     allow for more data.
1090    
1091     (d) Added support for Unicode script names via properties such as \p{Han}.
1092    
1093     19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
1094     matching that character.
1095    
1096     20. When matching a repeated Unicode property with a minimum greater than zero,
1097     (for example \pL{2,}), PCRE could look past the end of the subject if it
1098     reached it while seeking the minimum number of characters. This could
1099     happen only if some of the characters were more than one byte long, because
1100     there is a check for at least the minimum number of bytes.
1101    
1102     21. Refactored the implementation of \p and \P so as to be more general, to
1103     allow for more different types of property in future. This has changed the
1104     compiled form incompatibly. Anybody with saved compiled patterns that use
1105     \p or \P will have to recompile them.
1106    
1107     22. Added "Any" and "L&" to the supported property types.
1108    
1109     23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
1110     but give a compile time error if the value is greater than 0xff.
1111    
1112     24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
1113     accidentally not being installed or uninstalled.
1114    
1115     25. The pcre.h file was built from pcre.h.in, but the only changes that were
1116     made were to insert the current release number. This seemed silly, because
1117     it made things harder for people building PCRE on systems that don't run
1118     "configure". I have turned pcre.h into a distributed file, no longer built
1119     by "configure", with the version identification directly included. There is
1120     no longer a pcre.h.in file.
1121    
1122     However, this change necessitated a change to the pcre-config script as
1123     well. It is built from pcre-config.in, and one of the substitutions was the
1124     release number. I have updated configure.ac so that ./configure now finds
1125     the release number by grepping pcre.h.
1126    
1127     26. Added the ability to run the tests under valgrind.
1128    
1129    
1130 nigel 85 Version 6.4 05-Sep-05
1131     ---------------------
1132    
1133     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
1134     "--" to be printed when multiple files were scanned, even when none of the
1135     -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
1136     consider it to be a bug, and have restored the previous behaviour.
1137    
1138     2. A couple of code tidies to get rid of compiler warnings.
1139    
1140     3. The pcretest program used to cheat by referring to symbols in the library
1141     whose names begin with _pcre_. These are internal symbols that are not
1142     really supposed to be visible externally, and in some environments it is
1143     possible to suppress them. The cheating is now confined to including
1144     certain files from the library's source, which is a bit cleaner.
1145    
1146     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
1147     file's purpose clearer.
1148    
1149     5. Reorganized pcre_ucp_findchar().
1150    
1151    
1152 nigel 83 Version 6.3 15-Aug-05
1153     ---------------------
1154    
1155     1. The file libpcre.pc.in did not have general read permission in the tarball.
1156    
1157     2. There were some problems when building without C++ support:
1158    
1159     (a) If C++ support was not built, "make install" and "make test" still
1160     tried to test it.
1161    
1162     (b) There were problems when the value of CXX was explicitly set. Some
1163     changes have been made to try to fix these, and ...
1164    
1165     (c) --disable-cpp can now be used to explicitly disable C++ support.
1166    
1167     (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
1168     backslash in a target when C++ was disabled. This confuses some
1169     versions of "make", apparently. Using an intermediate variable solves
1170     this. (Same for CPP_LOBJ.)
1171    
1172     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
1173     (non-Windows) now includes $(CFLAGS) because these flags are sometimes
1174     necessary on certain architectures.
1175    
1176     4. Added a setting of -export-symbols-regex to the link command to remove
1177     those symbols that are exported in the C sense, but actually are local
1178     within the library, and not documented. Their names all begin with
1179     "_pcre_". This is not a perfect job, because (a) we have to except some
1180     symbols that pcretest ("illegally") uses, and (b) the facility isn't always
1181     available (and never for static libraries). I have made a note to try to
1182     find a way round (a) in the future.
1183    
1184    
1185 nigel 81 Version 6.2 01-Aug-05
1186     ---------------------
1187    
1188     1. There was no test for integer overflow of quantifier values. A construction
1189     such as {1111111111111111} would give undefined results. What is worse, if
1190     a minimum quantifier for a parenthesized subpattern overflowed and became
1191     negative, the calculation of the memory size went wrong. This could have
1192     led to memory overwriting.
1193    
1194     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
1195    
1196     3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
1197     operating environments where this matters.
1198    
1199     4. Applied Giuseppe Maxia's patch to add additional features for controlling
1200     PCRE options from within the C++ wrapper.
1201    
1202     5. Named capturing subpatterns were not being correctly counted when a pattern
1203     was compiled. This caused two problems: (a) If there were more than 100
1204     such subpatterns, the calculation of the memory needed for the whole
1205     compiled pattern went wrong, leading to an overflow error. (b) Numerical
1206     back references of the form \12, where the number was greater than 9, were
1207     not recognized as back references, even though there were sufficient
1208     previous subpatterns.
1209    
1210     6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
1211     versions of gcc, e.g. 2.95.4.
1212    
1213    
1214 nigel 79 Version 6.1 21-Jun-05
1215     ---------------------
1216    
1217     1. There was one reference to the variable "posix" in pcretest.c that was not
1218     surrounded by "#if !defined NOPOSIX".
1219    
1220     2. Make it possible to compile pcretest without DFA support, UTF8 support, or
1221     the cross-check on the old pcre_info() function, for the benefit of the
1222     cut-down version of PCRE that is currently imported into Exim.
1223    
1224     3. A (silly) pattern starting with (?i)(?-i) caused an internal space
1225     allocation error. I've done the easy fix, which wastes 2 bytes for sensible
1226     patterns that start (?i) but I don't think that matters. The use of (?i) is
1227     just an example; this all applies to the other options as well.
1228    
1229     4. Since libtool seems to echo the compile commands it is issuing, the output
1230     from "make" can be reduced a bit by putting "@" in front of each libtool
1231     compile command.
1232    
1233     5. Patch from the folks at Google for configure.in to be a bit more thorough
1234     in checking for a suitable C++ installation before trying to compile the
1235     C++ stuff. This should fix a reported problem when a compiler was present,
1236     but no suitable headers.
1237    
1238     6. The man pages all had just "PCRE" as their title. I have changed them to
1239     be the relevant file name. I have also arranged that these names are
1240     retained in the file doc/pcre.txt, which is a concatenation in text format
1241     of all the man pages except the little individual ones for each function.
1242    
1243     7. The NON-UNIX-USE file had not been updated for the different set of source
1244     files that come with release 6. I also added a few comments about the C++
1245     wrapper.
1246    
1247    
1248 nigel 77 Version 6.0 07-Jun-05
1249     ---------------------
1250    
1251     1. Some minor internal re-organization to help with my DFA experiments.
1252    
1253     2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
1254     didn't matter for the library itself when fully configured, but did matter
1255     when compiling without UCP support, or within Exim, where the ucp files are
1256     not imported.
1257    
1258     3. Refactoring of the library code to split up the various functions into
1259     different source modules. The addition of the new DFA matching code (see
1260     below) to a single monolithic source would have made it really too
1261     unwieldy, quite apart from causing all the code to be include in a
1262     statically linked application, when only some functions are used. This is
1263     relevant even without the DFA addition now that patterns can be compiled in
1264     one application and matched in another.
1265    
1266     The downside of splitting up is that there have to be some external
1267     functions and data tables that are used internally in different modules of
1268     the library but which are not part of the API. These have all had their
1269     names changed to start with "_pcre_" so that they are unlikely to clash
1270     with other external names.
1271    
1272     4. Added an alternate matching function, pcre_dfa_exec(), which matches using
1273     a different (DFA) algorithm. Although it is slower than the original
1274     function, it does have some advantages for certain types of matching
1275     problem.
1276    
1277     5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
1278     including restarting after a partial match.
1279    
1280     6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
1281     defined when compiling for Windows was sent to me. I have put it into the
1282     code, though I have no means of testing or verifying it.
1283    
1284     7. Added the pcre_refcount() auxiliary function.
1285    
1286     8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
1287     match before or at the first newline in the subject string. In pcretest,
1288     the /f option on a pattern can be used to set this.
1289    
1290     9. A repeated \w when used in UTF-8 mode with characters greater than 256
1291     would behave wrongly. This has been present in PCRE since release 4.0.
1292    
1293     10. A number of changes to the pcregrep command:
1294    
1295     (a) Refactored how -x works; insert ^(...)$ instead of setting
1296     PCRE_ANCHORED and checking the length, in preparation for adding
1297     something similar for -w.
1298    
1299     (b) Added the -w (match as a word) option.
1300    
1301     (c) Refactored the way lines are read and buffered so as to have more
1302     than one at a time available.
1303    
1304     (d) Implemented a pcregrep test script.
1305    
1306     (e) Added the -M (multiline match) option. This allows patterns to match
1307     over several lines of the subject. The buffering ensures that at least
1308     8K, or the rest of the document (whichever is the shorter) is available
1309     for matching (and similarly the previous 8K for lookbehind assertions).
1310    
1311     (f) Changed the --help output so that it now says
1312    
1313     -w, --word-regex(p)
1314    
1315     instead of two lines, one with "regex" and the other with "regexp"
1316     because that confused at least one person since the short forms are the
1317     same. (This required a bit of code, as the output is generated
1318     automatically from a table. It wasn't just a text change.)
1319    
1320     (g) -- can be used to terminate pcregrep options if the next thing isn't an
1321     option but starts with a hyphen. Could be a pattern or a path name
1322     starting with a hyphen, for instance.
1323    
1324     (h) "-" can be given as a file name to represent stdin.
1325    
1326     (i) When file names are being printed, "(standard input)" is used for
1327     the standard input, for compatibility with GNU grep. Previously
1328     "<stdin>" was used.
1329    
1330     (j) The option --label=xxx can be used to supply a name to be used for
1331     stdin when file names are being printed. There is no short form.
1332    
1333     (k) Re-factored the options decoding logic because we are going to add
1334     two more options that take data. Such options can now be given in four
1335     different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
1336    
1337     (l) Added the -A, -B, and -C options for requesting that lines of context
1338     around matches be printed.
1339    
1340     (m) Added the -L option to print the names of files that do not contain
1341     any matching lines, that is, the complement of -l.
1342    
1343     (n) The return code is 2 if any file cannot be opened, but pcregrep does
1344     continue to scan other files.
1345    
1346     (o) The -s option was incorrectly implemented. For compatibility with other
1347     greps, it now suppresses the error message for a non-existent or non-
1348     accessible file (but not the return code). There is a new option called
1349     -q that suppresses the output of matching lines, which was what -s was
1350     previously doing.
1351    
1352     (p) Added --include and --exclude options to specify files for inclusion
1353     and exclusion when recursing.
1354    
1355     11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
1356     Hopefully, it now does.
1357    
1358     12. Missing cast in pcre_study().
1359    
1360     13. Added an "uninstall" target to the makefile.
1361    
1362     14. Replaced "extern" in the function prototypes in Makefile.in with
1363     "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
1364     world, but is set differently for Windows.
1365    
1366     15. Added a second compiling function called pcre_compile2(). The only
1367     difference is that it has an extra argument, which is a pointer to an
1368     integer error code. When there is a compile-time failure, this is set
1369     non-zero, in addition to the error test pointer being set to point to an
1370     error message. The new argument may be NULL if no error number is required
1371     (but then you may as well call pcre_compile(), which is now just a
1372     wrapper). This facility is provided because some applications need a
1373     numeric error indication, but it has also enabled me to tidy up the way
1374     compile-time errors are handled in the POSIX wrapper.
1375    
1376     16. Added VPATH=.libs to the makefile; this should help when building with one
1377     prefix path and installing with another. (Or so I'm told by someone who
1378     knows more about this stuff than I do.)
1379    
1380     17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
1381     passes PCRE_DOTALL to the pcre_compile() function, making the "." character
1382     match everything, including newlines. This is not POSIX-compatible, but
1383     somebody wanted the feature. From pcretest it can be activated by using
1384     both the P and the s flags.
1385    
1386     18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
1387    
1388     19. libpcre.pc was being incorrectly installed as executable.
1389    
1390     20. A couple of places in pcretest check for end-of-line by looking for '\n';
1391     it now also looks for '\r' so that it will work unmodified on Windows.
1392    
1393     21. Added Google's contributed C++ wrapper to the distribution.
1394    
1395     22. Added some untidy missing memory free() calls in pcretest, to keep
1396     Electric Fence happy when testing.
1397    
1398    
1399    
1400 nigel 75 Version 5.0 13-Sep-04
1401     ---------------------
1402    
1403     1. Internal change: literal characters are no longer packed up into items
1404     containing multiple characters in a single byte-string. Each character
1405     is now matched using a separate opcode. However, there may be more than one
1406     byte in the character in UTF-8 mode.
1407    
1408     2. The pcre_callout_block structure has two new fields: pattern_position and
1409     next_item_length. These contain the offset in the pattern to the next match
1410     item, and its length, respectively.
1411    
1412     3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
1413     insertion of callouts before each pattern item. Added the /C option to
1414     pcretest to make use of this.
1415    
1416     4. On the advice of a Windows user, the lines
1417    
1418     #if defined(_WIN32) || defined(WIN32)
1419     _setmode( _fileno( stdout ), 0x8000 );
1420     #endif /* defined(_WIN32) || defined(WIN32) */
1421    
1422     have been added to the source of pcretest. This apparently does useful
1423     magic in relation to line terminators.
1424    
1425     5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
1426     for the benefit of those environments where the "b" makes a difference.
1427    
1428     6. The icc compiler has the same options as gcc, but "configure" doesn't seem
1429     to know about it. I have put a hack into configure.in that adds in code
1430     to set GCC=yes if CC=icc. This seems to end up at a point in the
1431     generated configure script that is early enough to affect the setting of
1432     compiler options, which is what is needed, but I have no means of testing
1433     whether it really works. (The user who reported this had patched the
1434     generated configure script, which of course I cannot do.)
1435    
1436     LATER: After change 22 below (new libtool files), the configure script
1437     seems to know about icc (and also ecc). Therefore, I have commented out
1438     this hack in configure.in.
1439    
1440     7. Added support for pkg-config (2 patches were sent in).
1441    
1442     8. Negated POSIX character classes that used a combination of internal tables
1443     were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
1444     [[:^ascii]]. Typically, they would match almost any characters. The other
1445     POSIX classes were not broken in this way.
1446    
1447     9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
1448     to find the match, as PCRE was deluded into thinking that the match had to
1449     start at the start point or following a newline. The same bug applied to
1450     patterns with negative forward assertions or any backward assertions
1451     preceding ".*" at the start, unless the pattern required a fixed first
1452     character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
1453    
1454     10. In UTF-8 mode, when moving forwards in the subject after a failed match
1455     starting at the last subject character, bytes beyond the end of the subject
1456     string were read.
1457    
1458     11. Renamed the variable "class" as "classbits" to make life easier for C++
1459     users. (Previously there was a macro definition, but it apparently wasn't
1460     enough.)
1461    
1462     12. Added the new field "tables" to the extra data so that tables can be passed
1463     in at exec time, or the internal tables can be re-selected. This allows
1464     a compiled regex to be saved and re-used at a later time by a different
1465     program that might have everything at different addresses.
1466    
1467     13. Modified the pcre-config script so that, when run on Solaris, it shows a
1468     -R library as well as a -L library.
1469    
1470     14. The debugging options of pcretest (-d on the command line or D on a
1471     pattern) showed incorrect output for anything following an extended class
1472     that contained multibyte characters and which was followed by a quantifier.
1473    
1474     15. Added optional support for general category Unicode character properties
1475     via the \p, \P, and \X escapes. Unicode property support implies UTF-8
1476     support. It adds about 90K to the size of the library. The meanings of the
1477     inbuilt class escapes such as \d and \s have NOT been changed.
1478    
1479     16. Updated pcredemo.c to include calls to free() to release the memory for the
1480     compiled pattern.
1481    
1482     17. The generated file chartables.c was being created in the source directory
1483     instead of in the building directory. This caused the build to fail if the
1484     source directory was different from the building directory, and was
1485     read-only.
1486    
1487     18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
1488     file. No doubt somebody will tell me if they don't make sense... Also added
1489     Dan Mooney's comments about building on OpenVMS.
1490    
1491     19. Added support for partial matching via the PCRE_PARTIAL option for
1492     pcre_exec() and the \P data escape in pcretest.
1493    
1494     20. Extended pcretest with 3 new pattern features:
1495    
1496     (i) A pattern option of the form ">rest-of-line" causes pcretest to
1497     write the compiled pattern to the file whose name is "rest-of-line".
1498     This is a straight binary dump of the data, with the saved pointer to
1499     the character tables forced to be NULL. The study data, if any, is
1500     written too. After writing, pcretest reads a new pattern.
1501    
1502     (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
1503     compiled pattern from the given file. There must not be any
1504     occurrences of "<" in the file name (pretty unlikely); if there are,
1505     pcretest will instead treat the initial "<" as a pattern delimiter.
1506     After reading in the pattern, pcretest goes on to read data lines as
1507     usual.
1508    
1509     (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
1510     and 16-bit fields in a compiled pattern, to simulate a pattern that
1511     was compiled on a host of opposite endianness.
1512    
1513     21. The pcre-exec() function can now cope with patterns that were compiled on
1514     hosts of opposite endianness, with this restriction:
1515    
1516     As for any compiled expression that is saved and used later, the tables
1517     pointer field cannot be preserved; the extra_data field in the arguments
1518     to pcre_exec() should be used to pass in a tables address if a value
1519     other than the default internal tables were used at compile time.
1520    
1521     22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
1522     now diagnosed as an error. Previously, most of the time, a negative number
1523     would have been treated as zero, but if in addition "ovector" was passed as
1524     NULL, a crash could occur.
1525    
1526     23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
1527     new versions from the libtool 1.5 distribution (the last one is a copy of
1528     a file called libtool.m4). This seems to have fixed the need to patch
1529     "configure" to support Darwin 1.3 (which I used to do). However, I still
1530     had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
1531     workstation).
1532    
1533     24. Changed the PCRE licence to be the more standard "BSD" licence.
1534    
1535    
1536 nigel 73 Version 4.5 01-Dec-03
1537     ---------------------
1538    
1539     1. There has been some re-arrangement of the code for the match() function so
1540     that it can be compiled in a version that does not call itself recursively.
1541     Instead, it keeps those local variables that need separate instances for
1542     each "recursion" in a frame on the heap, and gets/frees frames whenever it
1543     needs to "recurse". Keeping track of where control must go is done by means
1544     of setjmp/longjmp. The whole thing is implemented by a set of macros that
1545     hide most of the details from the main code, and operates only if
1546     NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
1547     "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
1548     operating.
1549    
1550     To make it easier for callers to provide specially tailored get/free
1551     functions for this usage, two new functions, pcre_stack_malloc, and
1552     pcre_stack_free, are used. They are always called in strict stacking order,
1553     and the size of block requested is always the same.
1554    
1555     The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
1556     PCRE has been compiled to use the stack or the heap for recursion. The
1557     -C option of pcretest uses this to show which version is compiled.
1558    
1559     A new data escape \S, is added to pcretest; it causes the amounts of store
1560     obtained and freed by both kinds of malloc/free at match time to be added
1561     to the output.
1562    
1563     2. Changed the locale test to use "fr_FR" instead of "fr" because that's
1564     what's available on my current Linux desktop machine.
1565    
1566     3. When matching a UTF-8 string, the test for a valid string at the start has
1567     been extended. If start_offset is not zero, PCRE now checks that it points
1568     to a byte that is the start of a UTF-8 character. If not, it returns
1569     PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
1570     this is necessary because there may be backward assertions in the pattern.
1571     When matching the same subject several times, it may save resources to use
1572     PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
1573    
1574     4. The code for checking the validity of UTF-8 strings has been tightened so
1575     that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
1576     containing "overlong sequences".
1577    
1578     5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
1579     I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
1580     should have been "&", but it just so happened that all the cases this let
1581     through by mistake were picked up later in the function.
1582    
1583     6. I had used a variable called "isblank" - this is a C99 function, causing
1584     some compilers to warn. To avoid this, I renamed it (as "blankclass").
1585    
1586     7. Cosmetic: (a) only output another newline at the end of pcretest if it is
1587     prompting; (b) run "./pcretest /dev/null" at the start of the test script
1588     so the version is shown; (c) stop "make test" echoing "./RunTest".
1589    
1590     8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
1591    
1592     9. The prototype for memmove() for systems that don't have it was using
1593     size_t, but the inclusion of the header that defines size_t was later. I've
1594     moved the #includes for the C headers earlier to avoid this.
1595    
1596     10. Added some adjustments to the code to make it easier to compiler on certain
1597     special systems:
1598    
1599     (a) Some "const" qualifiers were missing.
1600     (b) Added the macro EXPORT before all exported functions; by default this
1601     is defined to be empty.
1602     (c) Changed the dftables auxiliary program (that builds chartables.c) so
1603     that it reads its output file name as an argument instead of writing
1604     to the standard output and assuming this can be redirected.
1605    
1606     11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
1607     class containing characters with values greater than 255, PCRE compilation
1608     went into a loop.
1609    
1610     12. A recursive reference to a subpattern that was within another subpattern
1611     that had a minimum quantifier of zero caused PCRE to crash. For example,
1612     (x(y(?2))z)? provoked this bug with a subject that got as far as the
1613     recursion. If the recursively-called subpattern itself had a zero repeat,
1614     that was OK.
1615    
1616     13. In pcretest, the buffer for reading a data line was set at 30K, but the
1617     buffer into which it was copied (for escape processing) was still set at
1618     1024, so long lines caused crashes.
1619    
1620     14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
1621     "internal error: code overflow...". This applied to any character class
1622     that was followed by a possessive quantifier.
1623    
1624     15. Modified the Makefile to add libpcre.la as a prerequisite for
1625     libpcreposix.la because I was told this is needed for a parallel build to
1626     work.
1627    
1628     16. If a pattern that contained .* following optional items at the start was
1629     studied, the wrong optimizing data was generated, leading to matching
1630     errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
1631     matching string must start with a or b or c. The correct conclusion for
1632     this pattern is that a match can start with any character.
1633    
1634    
1635 nigel 71 Version 4.4 13-Aug-03
1636     ---------------------
1637    
1638     1. In UTF-8 mode, a character class containing characters with values between
1639     127 and 255 was not handled correctly if the compiled pattern was studied.
1640     In fixing this, I have also improved the studying algorithm for such
1641     classes (slightly).
1642    
1643     2. Three internal functions had redundant arguments passed to them. Removal
1644     might give a very teeny performance improvement.
1645    
1646     3. Documentation bug: the value of the capture_top field in a callout is *one
1647     more than* the number of the hightest numbered captured substring.
1648    
1649     4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
1650     in incorrectly linking with a previously installed version. They now link
1651     explicitly with libpcre.la.
1652    
1653     5. configure.in no longer needs to recognize Cygwin specially.
1654    
1655     6. A problem in pcre.in for Windows platforms is fixed.
1656    
1657     7. If a pattern was successfully studied, and the -d (or /D) flag was given to
1658     pcretest, it used to include the size of the study block as part of its
1659     output. Unfortunately, the structure contains a field that has a different
1660     size on different hardware architectures. This meant that the tests that
1661     showed this size failed. As the block is currently always of a fixed size,
1662     this information isn't actually particularly useful in pcretest output, so
1663     I have just removed it.
1664    
1665     8. Three pre-processor statements accidentally did not start in column 1.
1666     Sadly, there are *still* compilers around that complain, even though
1667     standard C has not required this for well over a decade. Sigh.
1668    
1669     9. In pcretest, the code for checking callouts passed small integers in the
1670     callout_data field, which is a void * field. However, some picky compilers
1671     complained about the casts involved for this on 64-bit systems. Now
1672     pcretest passes the address of the small integer instead, which should get
1673     rid of the warnings.
1674    
1675     10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
1676     both compile and run time, and gives an error if an invalid UTF-8 sequence
1677     is found. There is a option for disabling this check in cases where the
1678     string is known to be correct and/or the maximum performance is wanted.
1679    
1680     11. In response to a bug report, I changed one line in Makefile.in from
1681    
1682     -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
1683     to
1684     -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
1685    
1686     to look similar to other lines, but I have no way of telling whether this
1687     is the right thing to do, as I do not use Windows. No doubt I'll get told
1688     if it's wrong...
1689    
1690    
1691 nigel 69 Version 4.3 21-May-03
1692     ---------------------
1693    
1694     1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
1695     Makefile.
1696    
1697     2. Some refactoring to improve the quality of the code:
1698    
1699     (i) The utf8_table... variables are now declared "const".
1700    
1701     (ii) The code for \cx, which used the "case flipping" table to upper case
1702     lower case letters, now just substracts 32. This is ASCII-specific,
1703     but the whole concept of \cx is ASCII-specific, so it seems
1704     reasonable.
1705    
1706     (iii) PCRE was using its character types table to recognize decimal and
1707     hexadecimal digits in the pattern. This is silly, because it handles
1708     only 0-9, a-f, and A-F, but the character types table is locale-
1709     specific, which means strange things might happen. A private
1710     table is now used for this - though it costs 256 bytes, a table is
1711     much faster than multiple explicit tests. Of course, the standard
1712     character types table is still used for matching digits in subject
1713     strings against \d.
1714    
1715     (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
1716     ending in _t are). So I've renamed it as ESC_tee.
1717    
1718     3. The first argument for regexec() in the POSIX wrapper should have been
1719     defined as "const".
1720    
1721     4. Changed pcretest to use malloc() for its buffers so that they can be
1722     Electric Fenced for debugging.
1723    
1724     5. There were several places in the code where, in UTF-8 mode, PCRE would try
1725     to read one or more bytes before the start of the subject string. Often this
1726     had no effect on PCRE's behaviour, but in some circumstances it could
1727     provoke a segmentation fault.
1728    
1729     6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
1730     to try to read one or more bytes before the start of the subject string.
1731    
1732     7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
1733     UTF-8 support could misbehave in various ways if the subject string
1734     contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
1735     area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
1736     back over UTF-8 characters.)
1737    
1738    
1739 nigel 67 Version 4.2 14-Apr-03
1740     ---------------------
1741    
1742     1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
1743    
1744     2. Changes to the building process, supplied by Ronald Landheer-Cieslak
1745     [ON_WINDOWS]: new variable, "#" on non-Windows platforms
1746     [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
1747     [WIN_PREFIX]: new variable, "cyg" for Cygwin
1748     * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
1749     and BUILD_EXEEXT
1750     Note: automatic setting of the BUILD variables is not yet working
1751     set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
1752     compile-time but not at link-time
1753     [LINK]: use for linking executables only
1754     make different versions for Windows and non-Windows
1755     [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
1756     libraries
1757     [LINK_FOR_BUILD]: new variable
1758     [OBJEXT]: use throughout
1759     [EXEEXT]: use throughout
1760     <winshared>: new target
1761     <wininstall>: new target
1762     <dftables.o>: use native compiler
1763     <dftables>: use native linker
1764     <install>: handle Windows platform correctly
1765     <clean>: ditto
1766     <check>: ditto
1767     copy DLL to top builddir before testing
1768    
1769     As part of these changes, -no-undefined was removed again. This was reported
1770     to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
1771     in any case.
1772    
1773     3. Some tidies to get rid of compiler warnings:
1774    
1775     . In the match_data structure, match_limit was an unsigned long int, whereas
1776     match_call_count was an int. I've made them both unsigned long ints.
1777    
1778     . In pcretest the fact that a const uschar * doesn't automatically cast to
1779     a void * provoked a warning.
1780    
1781     . Turning on some more compiler warnings threw up some "shadow" variables
1782     and a few more missing casts.
1783    
1784     4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
1785     option, a class that contained a single character with a value between 128
1786     and 255 (e.g. /[\xFF]/) caused PCRE to crash.
1787    
1788     5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
1789     option, a class that contained several characters, but with at least one
1790     whose value was between 128 and 255 caused PCRE to crash.
1791    
1792    
1793 nigel 65 Version 4.1 12-Mar-03
1794     ---------------------
1795 nigel 63
1796 nigel 65 1. Compiling with gcc -pedantic found a couple of places where casts were
1797     needed, and a string in dftables.c that was longer than standard compilers are
1798     required to support.
1799    
1800     2. Compiling with Sun's compiler found a few more places where the code could
1801     be tidied up in order to avoid warnings.
1802    
1803     3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
1804     first of these names is deprecated in the latest Autoconf in favour of the name
1805     CC_FOR_BUILD, because "host" is typically used to mean the system on which the
1806     compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
1807     analogy I have changed it to CFLAGS_FOR_BUILD.
1808    
1809     4. Added -no-undefined to the linking command in the Makefile, because this is
1810     apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
1811     linking step for the pcreposix library.
1812    
1813     5. PCRE was failing to diagnose the case of two named groups with the same
1814     name.
1815    
1816     6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
1817     literal character that is needed in the subject for a match, and scans along to
1818     ensure that it is present before embarking on the full matching process. This
1819     saves time in cases of nested unlimited repeats that are never going to match.
1820     Problem: the scan can take a lot of time if the subject is very long (e.g.
1821     megabytes), thus penalizing straightforward matches. It is now done only if the
1822     amount of subject to be scanned is less than 1000 bytes.
1823    
1824     7. A lesser problem with the same optimization is that it was recording the
1825     first character of an anchored pattern as "needed", thus provoking a search
1826     right along the subject, even when the first match of the pattern was going to
1827     fail. The "needed" character is now not set for anchored patterns, unless it
1828     follows something in the pattern that is of non-fixed length. Thus, it still
1829     fulfils its original purpose of finding quick non-matches in cases of nested
1830     unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
1831    
1832    
1833     Version 4.0 17-Feb-03
1834     ---------------------
1835    
1836 nigel 63 1. If a comment in an extended regex that started immediately after a meta-item
1837     extended to the end of string, PCRE compiled incorrect data. This could lead to
1838     all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
1839    
1840     2. Moved to autoconf 2.53 and libtool 1.4.2.
1841    
1842     3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
1843     the special perltest8 script is no longer needed - all the tests can be run
1844     from a single perltest script.
1845    
1846     4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
1847     by \s. It has now been removed in PCRE. This means it isn't recognized as
1848     whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
1849     class [:space:] *does* include VT, thereby creating a mess.
1850    
1851     5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
1852     space and tab.
1853    
1854     6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
1855     its new features into the main test script, reducing the number of scripts.
1856    
1857     7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
1858     were backward compatible, and made the (?i) apply to the whole pattern, as if
1859     /i were given. Now it behaves more logically, and applies the option setting
1860     only to what follows. PCRE has been changed to follow suit. However, if it
1861     finds options settings right at the start of the pattern, it extracts them into
1862     the global options, as before. Thus, they show up in the info data.
1863    
1864     8. Added support for the \Q...\E escape sequence. Characters in between are
1865     treated as literals. This is slightly different from Perl in that $ and @ are
1866     also handled as literals inside the quotes. In Perl, they will cause variable
1867     interpolation. Note the following examples:
1868    
1869     Pattern PCRE matches Perl matches
1870    
1871     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
1872     \Qabc\$xyz\E abc\$xyz abc\$xyz
1873     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
1874    
1875     For compatibility with Perl, \Q...\E sequences are recognized inside character
1876     classes as well as outside them.
1877    
1878     9. Re-organized 3 code statements in pcretest to avoid "overflow in
1879     floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
1880     (size_t) cast to one statement in pcretest and one in pcreposix to avoid
1881     signed/unsigned warnings.
1882    
1883     10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
1884     option for pcretest, so I've replaced it by a simple function that does just
1885     that job.
1886    
1887     11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
1888     "pcregrep -".
1889    
1890     12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
1891     Java package. This provides some syntactic sugar for simple cases of what my
1892     documentation calls "once-only subpatterns". A pattern such as x*+ is the same
1893     as (?>x*). In other words, if what is inside (?>...) is just a single repeated
1894     item, you can use this simplified notation. Note that only makes sense with
1895     greedy quantifiers. Consequently, the use of the possessive quantifier forces
1896     greediness, whatever the setting of the PCRE_UNGREEDY option.
1897    
1898     13. A change of greediness default within a pattern was not taking effect at
1899     the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
1900     subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
1901     was abstracted outside.
1902    
1903     14. PCRE now supports the \G assertion. It is true when the current matching
1904     position is at the start point of the match. This differs from \A when the
1905     starting offset is non-zero. Used with the /g option of pcretest (or similar
1906     code), it works in the same way as it does for Perl's /g option. If all
1907     alternatives of a regex begin with \G, the expression is anchored to the start
1908     match position, and the "anchored" flag is set in the compiled expression.
1909    
1910     15. Some bugs concerning the handling of certain option changes within patterns
1911     have been fixed. These applied to options other than (?ims). For example,
1912     "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
1913     been the other way round. Some of this was related to change 7 above.
1914    
1915     16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
1916     features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
1917     and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
1918     POSIX classes only within a class (e.g. /[[:alpha:]]/).
1919    
1920     17. Added support for Perl's \C escape. This matches one byte, even in UTF8
1921     mode. Unlike ".", it always matches newline, whatever the setting of
1922     PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
1923     assertions. Perl allows it, but it doesn't (in general) work because it can't
1924     calculate the length of the lookbehind. At least, that's the case for Perl
1925     5.8.0 - I've been told they are going to document that it doesn't work in
1926     future.
1927    
1928     18. Added an error diagnosis for escapes that PCRE does not support: these are
1929     \L, \l, \N, \P, \p, \U, \u, and \X.
1930    
1931     19. Although correctly diagnosing a missing ']' in a character class, PCRE was
1932     reading past the end of the pattern in cases such as /[abcd/.
1933    
1934     20. PCRE was getting more memory than necessary for patterns with classes that
1935     contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
1936    
1937     21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
1938     compiling PCRE for use with Virtual Pascal.
1939    
1940     22. Small fix to the Makefile to make it work properly if the build is done
1941     outside the source tree.
1942    
1943     23. Added a new extension: a condition to go with recursion. If a conditional
1944     subpattern starts with (?(R) the "true" branch is used if recursion has
1945     happened, whereas the "false" branch is used only at the top level.
1946    
1947     24. When there was a very long string of literal characters (over 255 bytes
1948     without UTF support, over 250 bytes with UTF support), the computation of how
1949     much memory was required could be incorrect, leading to segfaults or other
1950     strange effects.
1951    
1952     25. PCRE was incorrectly assuming anchoring (either to start of subject or to
1953     start of line for a non-DOTALL pattern) when a pattern started with (.*) and
1954     there was a subsequent back reference to those brackets. This meant that, for
1955     example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
1956     possible to check for precisely this case. All we can do is abandon the
1957     optimization if .* occurs inside capturing brackets when there are any back
1958     references whatsoever. (See below for a better fix that came later.)
1959    
1960     26. The handling of the optimization for finding the first character of a
1961     non-anchored pattern, and for finding a character that is required later in the
1962     match were failing in some cases. This didn't break the matching; it just
1963     failed to optimize when it could. The way this is done has been re-implemented.
1964    
1965     27. Fixed typo in error message for invalid (?R item (it said "(?p").
1966    
1967     28. Added a new feature that provides some of the functionality that Perl
1968     provides with (?{...}). The facility is termed a "callout". The way it is done
1969     in PCRE is for the caller to provide an optional function, by setting
1970     pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
1971     global variable. By default it is unset, which disables all calling out. To get
1972     the function called, the regex must include (?C) at appropriate points. This
1973     is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
1974     This provides a means of identifying different callout points. When PCRE
1975     reaches such a point in the regex, if pcre_callout has been set, the external
1976     function is called. It is provided with data in a structure called
1977     pcre_callout_block, which is defined in pcre.h. If the function returns 0,
1978     matching continues; if it returns a non-zero value, the match at the current
1979     point fails. However, backtracking will occur if possible. [This was changed
1980     later and other features added - see item 49 below.]
1981    
1982     29. pcretest is upgraded to test the callout functionality. It provides a
1983     callout function that displays information. By default, it shows the start of
1984     the match and the current position in the text. There are some new data escapes
1985     to vary what happens:
1986    
1987     \C+ in addition, show current contents of captured substrings
1988     \C- do not supply a callout function
1989     \C!n return 1 when callout number n is reached
1990     \C!n!m return 1 when callout number n is reached for the mth time
1991    
1992     30. If pcregrep was called with the -l option and just a single file name, it
1993     output "<stdin>" if a match was found, instead of the file name.
1994    
1995     31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
1996     slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
1997     pcre_exec(). This saves a malloc/free per call. The default value of
1998     POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
1999     when configuring.
2000    
2001     32. The default maximum size of a compiled pattern is 64K. There have been a
2002     few cases of people hitting this limit. The code now uses macros to handle the
2003     storing of links as offsets within the compiled pattern. It defaults to 2-byte
2004     links, but this can be changed to 3 or 4 bytes by --with-link-size when
2005     configuring. Tests 2 and 5 work only with 2-byte links because they output
2006     debugging information about compiled patterns.
2007    
2008     33. Internal code re-arrangements:
2009    
2010     (a) Moved the debugging function for printing out a compiled regex into
2011     its own source file (printint.c) and used #include to pull it into
2012     pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
2013     separate copies.
2014    
2015     (b) Defined the list of op-code names for debugging as a macro in
2016     internal.h so that it is next to the definition of the opcodes.
2017    
2018     (c) Defined a table of op-code lengths for simpler skipping along compiled
2019     code. This is again a macro in internal.h so that it is next to the
2020     definition of the opcodes.
2021    
2022     34. Added support for recursive calls to individual subpatterns, along the
2023     lines of Robin Houston's patch (but implemented somewhat differently).
2024    
2025     35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
2026     allow it to read and process whole directories in Win32. This code was
2027     contributed by Lionel Fourquaux; it has not been tested by me.
2028    
2029     36. Added support for named subpatterns. The Python syntax (?P<name>...) is
2030     used to name a group. Names consist of alphanumerics and underscores, and must
2031     be unique. Back references use the syntax (?P=name) and recursive calls use
2032     (?P>name) which is a PCRE extension to the Python extension. Groups still have
2033     numbers. The function pcre_fullinfo() can be used after compilation to extract
2034     a name/number map. There are three relevant calls:
2035    
2036     PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
2037     PCRE_INFO_NAMECOUNT yields the number of entries
2038     PCRE_INFO_NAMETABLE yields a pointer to the map.
2039    
2040     The map is a vector of fixed-size entries. The size of each entry depends on
2041     the length of the longest name used. The first two bytes of each entry are the
2042     group number, most significant byte first. There follows the corresponding
2043     name, zero terminated. The names are in alphabetical order.
2044    
2045     37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
2046     case instead of 255. Making it the same both with and without UTF-8 support
2047     means that the same test output works with both.
2048    
2049     38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
2050     calling malloc() with a zero argument.
2051    
2052     39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
2053     optimization. I've improved things by keeping a bitmap of backreferences with
2054     numbers 1-31 so that if .* occurs inside capturing brackets that are not in
2055     fact referenced, the optimization can be applied. It is unlikely that a
2056     relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
2057     the match to follow \n) will appear inside brackets with a number greater than
2058     31, but if it does, any back reference > 31 suppresses the optimization.
2059    
2060     40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
2061     of disabling numbered capturing parentheses. Any opening parenthesis that is
2062     not followed by ? behaves as if it were followed by ?: but named parentheses
2063     can still be used for capturing (and they will acquire numbers in the usual
2064     way).
2065    
2066     41. Redesigned the return codes from the match() function into yes/no/error so
2067     that errors can be passed back from deep inside the nested calls. A malloc
2068     failure while inside a recursive subpattern call now causes the
2069     PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
2070    
2071     42. It is now possible to set a limit on the number of times the match()
2072     function is called in a call to pcre_exec(). This facility makes it possible to
2073     limit the amount of recursion and backtracking, though not in a directly
2074     obvious way, because the match() function is used in a number of different
2075     circumstances. The count starts from zero for each position in the subject
2076     string (for non-anchored patterns). The default limit is, for compatibility, a
2077     large number, namely 10 000 000. You can change this in two ways:
2078    
2079     (a) When configuring PCRE before making, you can use --with-match-limit=n
2080     to set a default value for the compiled library.
2081    
2082     (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
2083     a different value is set. See 45 below.
2084    
2085     If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
2086    
2087     43. Added a new function pcre_config(int, void *) to enable run-time extraction
2088     of things that can be changed at compile time. The first argument specifies
2089     what is wanted and the second points to where the information is to be placed.
2090     The current list of available information is:
2091    
2092     PCRE_CONFIG_UTF8
2093    
2094     The output is an integer that is set to one if UTF-8 support is available;
2095     otherwise it is set to zero.
2096    
2097     PCRE_CONFIG_NEWLINE
2098    
2099     The output is an integer that it set to the value of the code that is used for
2100     newline. It is either LF (10) or CR (13).
2101    
2102     PCRE_CONFIG_LINK_SIZE
2103    
2104     The output is an integer that contains the number of bytes used for internal
2105     linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
2106    
2107     PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
2108    
2109     The output is an integer that contains the threshold above which the POSIX
2110     interface uses malloc() for output vectors. See item 31 above.
2111    
2112     PCRE_CONFIG_MATCH_LIMIT
2113    
2114     The output is an unsigned integer that contains the default limit of the number
2115     of match() calls in a pcre_exec() execution. See 42 above.
2116    
2117     44. pcretest has been upgraded by the addition of the -C option. This causes it
2118     to extract all the available output from the new pcre_config() function, and to
2119     output it. The program then exits immediately.
2120    
2121     45. A need has arisen to pass over additional data with calls to pcre_exec() in
2122     order to support additional features. One way would have been to define
2123     pcre_exec2() (for example) with extra arguments, but this would not have been
2124     extensible, and would also have required all calls to the original function to
2125     be mapped to the new one. Instead, I have chosen to extend the mechanism that
2126     is used for passing in "extra" data from pcre_study().
2127    
2128     The pcre_extra structure is now exposed and defined in pcre.h. It currently
2129     contains the following fields:
2130    
2131     flags a bitmap indicating which of the following fields are set
2132     study_data opaque data from pcre_study()
2133     match_limit a way of specifying a limit on match() calls for a specific
2134     call to pcre_exec()
2135     callout_data data for callouts (see 49 below)
2136    
2137     The flag bits are also defined in pcre.h, and are
2138    
2139     PCRE_EXTRA_STUDY_DATA
2140     PCRE_EXTRA_MATCH_LIMIT
2141     PCRE_EXTRA_CALLOUT_DATA
2142    
2143     The pcre_study() function now returns one of these new pcre_extra blocks, with
2144     the actual study data pointed to by the study_data field, and the
2145     PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
2146     before. That is, this change is entirely upwards-compatible and requires no
2147     change to existing code.
2148    
2149     If you want to pass in additional data to pcre_exec(), you can either place it
2150     in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
2151     block.
2152    
2153     46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
2154     data string contains the escape sequence \M, pcretest calls pcre_exec() several
2155     times with different match limits, until it finds the minimum value needed for
2156     pcre_exec() to complete. The value is then output. This can be instructive; for
2157     most simple matches the number is quite small, but for pathological cases it
2158     gets very large very quickly.
2159    
2160     47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
2161     returns the size of the data block pointed to by the study_data field in a
2162     pcre_extra block, that is, the value that was passed as the argument to
2163     pcre_malloc() when PCRE was getting memory in which to place the information
2164     created by pcre_study(). The fourth argument should point to a size_t variable.
2165     pcretest has been extended so that this information is shown after a successful
2166     pcre_study() call when information about the compiled regex is being displayed.
2167    
2168     48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
2169     because what follows is always an absolute path. (Later: it turns out that this
2170     is more than cosmetic for MinGW, because it doesn't like empty path
2171     components.)
2172    
2173     49. Some changes have been made to the callout feature (see 28 above):
2174    
2175     (i) A callout function now has three choices for what it returns:
2176    
2177     0 => success, carry on matching
2178     > 0 => failure at this point, but backtrack if possible
2179     < 0 => serious error, return this value from pcre_exec()
2180    
2181     Negative values should normally be chosen from the set of PCRE_ERROR_xxx
2182     values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
2183     "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
2184     use by callout functions. It will never be used by PCRE itself.
2185    
2186     (ii) The pcre_extra structure (see 45 above) has a void * field called
2187     callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
2188     pcre_callout_block structure has a field of the same name. The contents of
2189     the field passed in the pcre_extra structure are passed to the callout
2190     function in the corresponding field in the callout block. This makes it
2191     easier to use the same callout-containing regex from multiple threads. For
2192     testing, the pcretest program has a new data escape
2193    
2194     \C*n pass the number n (may be negative) as callout_data
2195    
2196     If the callout function in pcretest receives a non-zero value as
2197     callout_data, it returns that value.
2198    
2199     50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
2200     there were some redundant $(CFLAGS) in commands that are now specified as
2201     $(LINK), which already includes $(CFLAGS).
2202    
2203     51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
2204     has been compiled with UTF-8 support *and* pcre_compile() has been compiled
2205     with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
2206     one-byte characters throughout. Note that case-insensitive matching applies
2207     only to characters whose values are less than 256. PCRE doesn't support the
2208     notion of cases for higher-valued characters.
2209    
2210     (i) A character class whose characters are all within 0-255 is handled as
2211     a bit map, and the map is inverted for negative classes. Previously, a
2212     character > 255 always failed to match such a class; however it should
2213     match if the class was a negative one (e.g. [^ab]). This has been fixed.
2214    
2215     (ii) A negated character class with a single character < 255 is coded as
2216     "not this character" (OP_NOT). This wasn't working properly when the test
2217     character was multibyte, either singly or repeated.
2218    
2219     (iii) Repeats of multibyte characters are now handled correctly in UTF-8
2220     mode, for example: \x{100}{2,3}.
2221    
2222     (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
2223     singly or repeated) now correctly test multibyte characters. However,
2224     PCRE doesn't recognize any characters with values greater than 255 as
2225     digits, spaces, or word characters. Such characters always match \D, \S,
2226     and \W, and never match \d, \s, or \w.
2227    
2228     (v) Classes may now contain characters and character ranges with values
2229     greater than 255. For example: [ab\x{100}-\x{400}].
2230    
2231     (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
2232     PCRE in UTF-8 mode.
2233    
2234     52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
2235     PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
2236     retained for backwards compatibility. (Note that LASTLITERAL is also a byte
2237     value.)
2238    
2239     53. The single man page has become too large. I have therefore split it up into
2240     a number of separate man pages. These also give rise to individual HTML pages;
2241     these are now put in a separate directory, and there is an index.html page that
2242     lists them all. Some hyperlinking between the pages has been installed.
2243    
2244     54. Added convenience functions for handling named capturing parentheses.
2245    
2246     55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
2247     aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
2248     true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
2249     are faulted.
2250    
2251     56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
2252     calling configure. These values are used when compiling the dftables.c program
2253     which is run to generate the source of the default character tables. They
2254     default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
2255     you will need to set these values.
2256    
2257     57. Updated the building process for Windows DLL, as provided by Fred Cox.
2258    
2259    
2260     Version 3.9 02-Jan-02
2261 nigel 61 ---------------------
2262    
2263     1. A bit of extraneous text had somehow crept into the pcregrep documentation.
2264    
2265     2. If --disable-static was given, the building process failed when trying to
2266     build pcretest and pcregrep. (For some reason it was using libtool to compile
2267     them, which is not right, as they aren't part of the library.)
2268    
2269    
2270 nigel 59 Version 3.8 18-Dec-01
2271     ---------------------
2272    
2273     1. The experimental UTF-8 code was completely screwed up. It was packing the
2274     bytes in the wrong order. How dumb can you get?
2275    
2276    
2277 nigel 57 Version 3.7 29-Oct-01
2278     ---------------------
2279    
2280     1. In updating pcretest to check change 1 of version 3.6, I screwed up.
2281     This caused pcretest, when used on the test data, to segfault. Unfortunately,
2282     this didn't happen under Solaris 8, where I normally test things.
2283    
2284 nigel 59 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
2285     doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
2286     isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
2287     this fix an hour or so after the initial 3.7 release.)
2288 nigel 57
2289 nigel 59
2290 nigel 55 Version 3.6 23-Oct-01
2291     ---------------------
2292    
2293     1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
2294     offsets passed as NULL with zero offset count.
2295    
2296     2. The config.guess and config.sub files had not been updated when I moved to
2297     the latest autoconf.
2298    
2299    
2300 nigel 53 Version 3.5 15-Aug-01
2301     ---------------------
2302 nigel 5
2303 nigel 53 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
2304     had been forgotten.
2305    
2306     2. By using declared but undefined structures, we can avoid using "void"
2307     definitions in pcre.h while keeping the internal definitions of the structures
2308     private.
2309    
2310     3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
2311     user point of view, this means that both static and shared libraries are built
2312     by default, but this can be individually controlled. More of the work of
2313     handling this static/shared cases is now inside libtool instead of PCRE's make
2314     file.
2315    
2316     4. The pcretest utility is now installed along with pcregrep because it is
2317     useful for users (to test regexs) and by doing this, it automatically gets
2318     relinked by libtool. The documentation has been turned into a man page, so
2319     there are now .1, .txt, and .html versions in /doc.
2320    
2321     5. Upgrades to pcregrep:
2322     (i) Added long-form option names like gnu grep.
2323     (ii) Added --help to list all options with an explanatory phrase.
2324     (iii) Added -r, --recursive to recurse into sub-directories.
2325     (iv) Added -f, --file to read patterns from a file.
2326    
2327     6. pcre_exec() was referring to its "code" argument before testing that
2328     argument for NULL (and giving an error if it was NULL).
2329    
2330     7. Upgraded Makefile.in to allow for compiling in a different directory from
2331     the source directory.
2332    
2333     8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
2334     options bits, the pointer it was passed was to an int instead of to an unsigned
2335     long int. This mattered only on 64-bit systems.
2336    
2337     9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
2338     generated) instead of pcre.in, which it its source. Also made the same change
2339     in several of the .c files.
2340    
2341     10. A new release of gcc defines printf() as a macro, which broke pcretest
2342     because it had an ifdef in the middle of a string argument for printf(). Fixed
2343     by using separate calls to printf().
2344    
2345     11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
2346     script, to force use of CR or LF instead of \n in the source. On non-Unix
2347     systems, the value can be set in config.h.
2348    
2349     12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
2350     absolute limit. Changed the text of the error message to make this clear, and
2351     likewise updated the man page.
2352    
2353     13. The limit of 99 on the number of capturing subpatterns has been removed.
2354     The new limit is 65535, which I hope will not be a "real" limit.
2355    
2356    
2357 nigel 51 Version 3.4 22-Aug-00
2358     ---------------------
2359    
2360     1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
2361    
2362     2. Diagnose condition (?(0) as an error instead of crashing on matching.
2363    
2364    
2365 nigel 49 Version 3.3 01-Aug-00
2366     ---------------------
2367    
2368     1. If an octal character was given, but the value was greater than \377, it
2369     was not getting masked to the least significant bits, as documented. This could
2370     lead to crashes in some systems.
2371    
2372     2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
2373     the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
2374    
2375     3. Added the functions pcre_free_substring() and pcre_free_substring_list().
2376     These just pass their arguments on to (pcre_free)(), but they are provided
2377     because some uses of PCRE bind it to non-C systems that can call its functions,
2378     but cannot call free() or pcre_free() directly.
2379    
2380     4. Add "make test" as a synonym for "make check". Corrected some comments in
2381     the Makefile.
2382    
2383     5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
2384     Makefile.
2385    
2386     6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
2387     command called pgrep for grepping around the active processes.
2388    
2389     7. Added the beginnings of support for UTF-8 character strings.
2390    
2391     8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
2392     RANLIB to ./ltconfig so that they are used by libtool. I think these are all
2393     the relevant ones. (AR is not passed because ./ltconfig does its own figuring
2394     out for the ar command.)
2395    
2396    
2397 nigel 47 Version 3.2 12-May-00
2398     ---------------------
2399    
2400     This is purely a bug fixing release.
2401    
2402     1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
2403     of ZA. This was just one example of several cases that could provoke this bug,
2404     which was introduced by change 9 of version 2.00. The code for breaking
2405     infinite loops after an iteration that matches an empty string was't working
2406     correctly.
2407    
2408     2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
2409     when matched against abbab (for example). After matching an empty string, it
2410     wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
2411     caused it to match further down the string than it should.
2412    
2413     3. The code contained an inclusion of sys/types.h. It isn't clear why this
2414     was there because it doesn't seem to be needed, and it causes trouble on some
2415     systems, as it is not a Standard C header. It has been removed.
2416    
2417     4. Made 4 silly changes to the source to avoid stupid compiler warnings that
2418     were reported on the Macintosh. The changes were from
2419    
2420     while ((c = *(++ptr)) != 0 && c != '\n');
2421     to
2422     while ((c = *(++ptr)) != 0 && c != '\n') ;
2423    
2424     Totally extraordinary, but if that's what it takes...
2425    
2426     5. PCRE is being used in one environment where neither memmove() nor bcopy() is
2427     available. Added HAVE_BCOPY and an autoconf test for it; if neither
2428     HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
2429     assumes the way PCRE uses memmove() (always moving upwards).
2430    
2431     6. PCRE is being used in one environment where strchr() is not available. There
2432     was only one use in pcre.c, and writing it out to avoid strchr() probably gives
2433     faster code anyway.
2434    
2435    
2436 nigel 45 Version 3.1 09-Feb-00
2437     ---------------------
2438    
2439     The only change in this release is the fixing of some bugs in Makefile.in for
2440     the "install" target:
2441    
2442     (1) It was failing to install pcreposix.h.
2443    
2444     (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
2445    
2446    
2447 nigel 43 Version 3.0 01-Feb-00
2448     ---------------------
2449 nigel 41
2450     1. Add support for the /+ modifier to perltest (to output $` like it does in
2451     pcretest).
2452    
2453     2. Add support for the /g modifier to perltest.
2454    
2455     3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
2456     matches null strings.
2457    
2458     4. Fix perltest so that it doesn't do unwanted things when fed an empty
2459     pattern. Perl treats empty patterns specially - it reuses the most recent
2460     pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
2461     effect.
2462    
2463     5. The POSIX interface was broken in that it was just handing over the POSIX
2464     captured string vector to pcre_exec(), but (since release 2.00) PCRE has
2465     required a bigger vector, with some working space on the end. This means that
2466     the POSIX wrapper now has to get and free some memory, and copy the results.
2467    
2468 nigel 43 6. Added some simple autoconf support, placing the test data and the
2469     documentation in separate directories, re-organizing some of the
2470     information files, and making it build pcre-config (a GNU standard). Also added
2471     libtool support for building PCRE as a shared library, which is now the
2472     default.
2473 nigel 41
2474 nigel 43 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
2475     09 are not valid octal constants. Single digits will be used for minor values
2476     less than 10.
2477    
2478     8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
2479     existing programs that set these in the POSIX interface can use PCRE without
2480     modification.
2481    
2482     9. Added a new function, pcre_fullinfo() with an extensible interface. It can
2483     return all that pcre_info() returns, plus additional data. The pcre_info()
2484     function is retained for compatibility, but is considered to be obsolete.
2485    
2486     10. Added experimental recursion feature (?R) to handle one common case that
2487     Perl 5.6 will be able to do with (?p{...}).
2488    
2489     11. Added support for POSIX character classes like [:alpha:], which Perl is
2490     adopting.
2491    
2492    
2493 nigel 39 Version 2.08 31-Aug-99
2494     ----------------------
2495    
2496     1. When startoffset was not zero and the pattern began with ".*", PCRE was not
2497     trying to match at the startoffset position, but instead was moving forward to
2498     the next newline as if a previous match had failed.
2499    
2500     2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
2501     and could get into a loop if a null string was matched other than at the start
2502     of the subject.
2503    
2504     3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
2505     be distinguished at compile time, and for completeness also added PCRE_DATE.
2506    
2507     5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
2508     in GnuWin32 environments.
2509    
2510    
2511 nigel 37 Version 2.07 29-Jul-99
2512     ----------------------
2513    
2514     1. The documentation is now supplied in plain text form and HTML as well as in
2515     the form of man page sources.
2516    
2517     2. C++ compilers don't like assigning (void *) values to other pointer types.
2518     In particular this affects malloc(). Although there is no problem in Standard
2519     C, I've put in casts to keep C++ compilers happy.
2520    
2521     3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
2522     should be (const char *).
2523    
2524     4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
2525     be useful for non-Unix systems who don't want to bother with the POSIX stuff.
2526     However, I haven't made this a standard facility. The documentation doesn't
2527     mention it, and the Makefile doesn't support it.
2528    
2529     5. The Makefile now contains an "install" target, with editable destinations at
2530     the top of the file. The pcretest program is not installed.
2531    
2532     6. pgrep -V now gives the PCRE version number and date.
2533    
2534     7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
2535     causing the entire string to be ignored, instead of just the last character.
2536    
2537     8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
2538     non-matching string, it can take a very, very long time, even for strings of
2539     quite modest length, because of the nested recursion. PCRE now does better in
2540     some of these cases. It does this by remembering the last required literal
2541     character in the pattern, and pre-searching the subject to ensure it is present
2542     before running the real match. In other words, it applies a heuristic to detect
2543     some types of certain failure quickly, and in the above example, if presented
2544     with a string that has no trailing " it gives "no match" very quickly.
2545    
2546     9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
2547     other alternatives are tried instead.
2548    
2549    
2550 nigel 35 Version 2.06 09-Jun-99
2551     ----------------------
2552    
2553     1. Change pcretest's output for amount of store used to show just the code
2554     space, because the remainder (the data block) varies in size between 32-bit and
2555     64-bit systems.
2556    
2557     2. Added an extra argument to pcre_exec() to supply an offset in the subject to
2558     start matching at. This allows lookbehinds to work when searching for multiple
2559     occurrences in a string.
2560    
2561     3. Added additional options to pcretest for testing multiple occurrences:
2562    
2563     /+ outputs the rest of the string that follows a match
2564     /g loops for multiple occurrences, using the new startoffset argument
2565     /G loops for multiple occurrences by passing an incremented pointer
2566    
2567     4. PCRE wasn't doing the "first character" optimization for patterns starting
2568     with \b or \B, though it was doing it for other lookbehind assertions. That is,
2569     it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
2570     the letter 'x'. On long subject strings, this gives a significant speed-up.
2571    
2572    
2573 nigel 33 Version 2.05 21-Apr-99
2574     ----------------------
2575    
2576     1. Changed the type of magic_number from int to long int so that it works
2577     properly on 16-bit systems.
2578    
2579     2. Fixed a bug which caused patterns starting with .* not to work correctly
2580     when the subject string contained newline characters. PCRE was assuming
2581     anchoring for such patterns in all cases, which is not correct because .* will
2582     not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
2583     DOTALL is set at top level; otherwise it knows that patterns starting with .*
2584     must be retried after every newline in the subject.
2585    
2586    
2587 nigel 31 Version 2.04 18-Feb-99
2588     ----------------------
2589    
2590     1. For parenthesized subpatterns with repeats whose minimum was zero, the
2591     computation of the store needed to hold the pattern was incorrect (too large).
2592     If such patterns were nested a few deep, this could multiply and become a real
2593     problem.
2594    
2595     2. Added /M option to pcretest to show the memory requirement of a specific
2596     pattern. Made -m a synonym of -s (which does this globally) for compatibility.
2597    
2598     3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
2599     compiled in such a way that the backtracking after subsequent failure was
2600     pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
2601     ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
2602    
2603    
2604 nigel 29 Version 2.03 02-Feb-99
2605     ----------------------
2606    
2607     1. Fixed typo and small mistake in man page.
2608    
2609 nigel 31 2. Added 4th condition (GPL supersedes if conflict) and created separate
2610     LICENCE file containing the conditions.
2611 nigel 29
2612     3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
2613     Perl, that is the internal \ allows the delimiter to be included in the
2614     pattern. Locked out the use of \ as a delimiter. If \ immediately follows
2615     the final delimiter, add \ to the end of the pattern (to test the error).
2616    
2617     4. Added the convenience functions for extracting substrings after a successful
2618     match. Updated pcretest to make it able to test these functions.
2619    
2620    
2621 nigel 27 Version 2.02 14-Jan-99
2622     ----------------------
2623    
2624     1. Initialized the working variables associated with each extraction so that
2625     their saving and restoring doesn't refer to uninitialized store.
2626    
2627     2. Put dummy code into study.c in order to trick the optimizer of the IBM C
2628     compiler for OS/2 into generating correct code. Apparently IBM isn't going to
2629     fix the problem.
2630    
2631     3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
2632     calls, and wasn't printing the correct value for compiling calls. Increased the
2633     default value of LOOPREPEAT, and the number of significant figures in the
2634     times.
2635    
2636     4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
2637    
2638     5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
2639     a building problem on Windows NT with a FAT file system.
2640    
2641    
2642 nigel 25 Version 2.01 21-Oct-98
2643     ----------------------
2644    
2645     1. Changed the API for pcre_compile() to allow for the provision of a pointer
2646     to character tables built by pcre_maketables() in the current locale. If NULL
2647     is passed, the default tables are used.
2648    
2649    
2650 nigel 23 Version 2.00 24-Sep-98
2651 nigel 21 ----------------------
2652    
2653 nigel 23 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
2654     it any more.
2655    
2656     2. Allow quantification of (?>) groups, and make it work correctly.
2657    
2658     3. The first character computation wasn't working for (?>) groups.
2659    
2660     4. Correct the implementation of \Z (it is permitted to match on the \n at the
2661     end of the subject) and add 5.005's \z, which really does match only at the
2662     very end of the subject.
2663    
2664     5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
2665    
2666     6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
2667     DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
2668     localized options. All options to pcre_study() were also removed.
2669    
2670     7. Add other new features from 5.005:
2671    
2672     $(?<= positive lookbehind
2673     $(?<! negative lookbehind
2674     (?imsx-imsx) added the unsetting capability
2675     such a setting is global if at outer level; local otherwise
2676     (?imsx-imsx:) non-capturing groups with option setting
2677     (?(cond)re|re) conditional pattern matching
2678    
2679     A backreference to itself in a repeated group matches the previous
2680     captured string.
2681    
2682     8. General tidying up of studying (both automatic and via "study")
2683     consequential on the addition of new assertions.
2684    
2685     9. As in 5.005, unlimited repeated groups that could match an empty substring
2686     are no longer faulted at compile time. Instead, the loop is forcibly broken at
2687     runtime if any iteration does actually match an empty substring.
2688    
2689     10. Include the RunTest script in the distribution.
2690    
2691     11. Added tests from the Perl 5.005_02 distribution. This showed up a few
2692     discrepancies, some of which were old and were also with respect to 5.004. They
2693     have now been fixed.
2694    
2695    
2696     Version 1.09 28-Apr-98
2697     ----------------------
2698    
2699 nigel 21 1. A negated single character class followed by a quantifier with a minimum
2700     value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
2701     program crashes, or just wrong answers. This did not apply to negated classes
2702     containing more than one character, or to minima other than one.
2703    
2704    
2705 nigel 19 Version 1.08 27-Mar-98
2706     ----------------------
2707    
2708     1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
2709    
2710     2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
2711     latter must appear before anything that relies on it in the pattern.
2712    
2713    
2714 nigel 17 Version 1.07 16-Feb-98
2715     ----------------------
2716    
2717     1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
2718     repeat of a potentially empty string).
2719    
2720    
2721 nigel 15 Version 1.06 23-Jan-98
2722     ----------------------
2723    
2724     1. Added Markus Oberhumer's little patches for C++.
2725    
2726     2. Literal strings longer than 255 characters were broken.
2727    
2728    
2729 nigel 13 Version 1.05 23-Dec-97
2730     ----------------------
2731    
2732     1. Negated character classes containing more than one character were failing if
2733     PCRE_CASELESS was set at run time.
2734    
2735    
2736 nigel 11 Version 1.04 19-Dec-97
2737     ----------------------
2738    
2739     1. Corrected the man page, where some "const" qualifiers had been omitted.
2740    
2741     2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
2742     input syntax.
2743    
2744     3. Fixed memory leak which occurred when a regex with back references was
2745     matched with an offsets vector that wasn't big enough. The temporary memory
2746     that is used in this case wasn't being freed if the match failed.
2747    
2748     4. Tidied pcretest to ensure it frees memory that it gets.
2749    
2750     5. Temporary memory was being obtained in the case where the passed offsets
2751     vector was exactly big enough.
2752    
2753     6. Corrected definition of offsetof() from change 5 below.
2754    
2755     7. I had screwed up change 6 below and broken the rules for the use of
2756     setjmp(). Now fixed.
2757    
2758    
2759 nigel 9 Version 1.03 18-Dec-97
2760     ----------------------
2761    
2762     1. A erroneous regex with a missing opening parenthesis was correctly
2763     diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
2764     on some systems.
2765    
2766     2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
2767     it was reported that one broken compiler failed on the former because "code" is
2768     also an independent variable.
2769    
2770     3. The erroneous regex a[]b caused an array overrun reference.
2771    
2772     4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
2773     fail on data ending with that character. (It was going on too far, and checking
2774     the next character, typically a binary zero.) This was specific to the
2775     optimized code for single-character negative classes.
2776    
2777     5. Added a contributed patch from the TIN world which does the following:
2778    
2779     + Add an undef for memmove, in case the the system defines a macro for it.
2780    
2781     + Add a definition of offsetof(), in case there isn't one. (I don't know
2782     the reason behind this - offsetof() is part of the ANSI standard - but
2783     it does no harm).
2784    
2785     + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
2786     most of the places where whitespace preceded '#'. I have given up and
2787     allowed the remaining 2 cases to be at the margin.
2788    
2789     + Rename some variables in pcre to eliminate shadowing. This seems very
2790     pedantic, but does no harm, of course.
2791    
2792     6. Moved the call to setjmp() into its own function, to get rid of warnings
2793     from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
2794    
2795     7. Constructs such as \d{8,} were compiling into the equivalent of
2796 nigel 11 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
2797 nigel 9 outcome, but in this particular case used more store than had been allocated,
2798     which caused the bug to be discovered because it threw up an internal error.
2799    
2800     8. The debugging code in both pcre and pcretest for outputting the compiled
2801     form of a regex was going wrong in the case of back references followed by
2802     curly-bracketed repeats.
2803    
2804    
2805 nigel 7 Version 1.02 12-Dec-97
2806     ----------------------
2807    
2808     1. Typos in pcre.3 and comments in the source fixed.
2809    
2810     2. Applied a contributed patch to get rid of places where it used to remove
2811     'const' from variables, and fixed some signed/unsigned and uninitialized
2812     variable warnings.
2813    
2814     3. Added the "runtest" target to Makefile.
2815    
2816     4. Set default compiler flag to -O2 rather than just -O.
2817    
2818    
2819 nigel 5 Version 1.01 19-Nov-97
2820     ----------------------
2821    
2822     1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
2823     like /([ab]*)*/, that is, for classes with more than one character in them.
2824    
2825     2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
2826     as /((?>a*))*/ (a PCRE_EXTRA facility).
2827    
2828    
2829     Version 1.00 18-Nov-97
2830     ----------------------
2831    
2832     1. Added compile-time macros to support systems such as SunOS4 which don't have
2833     memmove() or strerror() but have other things that can be used instead.
2834    
2835     2. Arranged that "make clean" removes the executables.
2836    
2837    
2838 nigel 3 Version 0.99 27-Oct-97
2839     ----------------------
2840    
2841     1. Fixed bug in code for optimizing classes with only one character. It was
2842     initializing a 32-byte map regardless, which could cause it to run off the end
2843     of the memory it had got.
2844    
2845     2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
2846    
2847    
2848     Version 0.98 22-Oct-97
2849     ----------------------
2850    
2851     1. Fixed bug in code for handling temporary memory usage when there are more
2852     back references than supplied space in the ovector. This could cause segfaults.
2853    
2854    
2855     Version 0.97 21-Oct-97
2856     ----------------------
2857    
2858     1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
2859    
2860     2. Optimized negated single characters not to use a bit map.
2861    
2862     3. Brought error texts together as macro definitions; clarified some of them;
2863     fixed one that was wrong - it said "range out of order" when it meant "invalid
2864     escape sequence".
2865    
2866     4. Changed some char * arguments to const char *.
2867    
2868     5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
2869    
2870     6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
2871     pcretest.
2872    
2873    
2874     Version 0.96 16-Oct-97
2875     ----------------------
2876    
2877     1. Added a simple "pgrep" utility to the distribution.
2878    
2879     2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
2880     unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
2881     where "ddd" means "one or more decimal digits".
2882    
2883     3. Fixed serious bug. If a pattern had a back reference, but the call to
2884     pcre_exec() didn't supply a large enough ovector to record the related
2885     identifying subpattern, the match always failed. PCRE now remembers the number
2886     of the largest back reference, and gets some temporary memory in which to save
2887     the offsets during matching if necessary, in order to ensure that
2888     backreferences always work.
2889    
2890     4. Increased the compatibility with Perl in a number of ways:
2891    
2892     (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
2893     to request this handling. The option can be set at compile or exec time.
2894    
2895     (b) $ matches before a terminating newline by default; an option
2896     PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
2897     mode). The option can be set at compile or exec time.
2898    
2899     (c) The handling of \ followed by a digit other than 0 is now supposed to be
2900     the same as Perl's. If the decimal number it represents is less than 10
2901     or there aren't that many previous left capturing parentheses, an octal
2902     escape is read. Inside a character class, it's always an octal escape,
2903     even if it is a single digit.
2904    
2905     (d) An escaped but undefined alphabetic character is taken as a literal,
2906     unless PCRE_EXTRA is set. Currently this just reserves the remaining
2907     escapes.
2908    
2909     (e) {0} is now permitted. (The previous item is removed from the compiled
2910     pattern).
2911    
2912     5. Changed all the names of code files so that the basic parts are no longer
2913     than 10 characters, and abolished the teeny "globals.c" file.
2914    
2915     6. Changed the handling of character classes; they are now done with a 32-byte
2916     bit map always.
2917    
2918     7. Added the -d and /D options to pcretest to make it possible to look at the
2919     internals of compilation without having to recompile pcre.
2920    
2921    
2922     Version 0.95 23-Sep-97
2923     ----------------------
2924    
2925     1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
2926     \x20 at the start of a run of normal characters. These were being treated as
2927     real characters, instead of the source characters being re-checked.
2928    
2929    
2930     Version 0.94 18-Sep-97
2931     ----------------------
2932    
2933     1. The functions are now thread-safe, with the caveat that the global variables
2934     containing pointers to malloc() and free() or alternative functions are the
2935     same for all threads.
2936    
2937     2. Get pcre_study() to generate a bitmap of initial characters for non-
2938     anchored patterns when this is possible, and use it if passed to pcre_exec().
2939    
2940    
2941     Version 0.93 15-Sep-97
2942     ----------------------
2943    
2944     1. /(b)|(:+)/ was computing an incorrect first character.
2945    
2946     2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
2947     but not actually doing anything yet.
2948    
2949     3. Treat "-" characters in classes that cannot be part of ranges as literals,
2950     as Perl does (e.g. [-az] or [az-]).
2951    
2952     4. Set the anchored flag if a branch starts with .* or .*? because that tests
2953     all possible positions.
2954    
2955     5. Split up into different modules to avoid including unneeded functions in a
2956     compiled binary. However, compile and exec are still in one module. The "study"
2957     function is split off.
2958    
2959     6. The character tables are now in a separate module whose source is generated
2960     by an auxiliary program - but can then be edited by hand if required. There are
2961     now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
2962     toupper() in the code.
2963    
2964     7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
2965     make them global. Abolish the function for setting them, as the caller can now
2966     set them directly.
2967    
2968    
2969     Version 0.92 11-Sep-97
2970     ----------------------
2971    
2972     1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
2973     (e.g. /a{1,3}/) was broken (I mis-optimized it).
2974    
2975     2. Caseless matching was not working in character classes if the characters in
2976     the pattern were in upper case.
2977    
2978     3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
2979    
2980     4. Make PCRE_ANCHORED public and accept as a compile option.
2981    
2982     5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
2983     PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
2984     pass them.
2985    
2986     6. Give an error if bad option bits passed at compile or run time.
2987    
2988     7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
2989     pcretest to cause it to pass that flag.
2990    
2991     8. Add pcre_info(), to get the number of identifying subpatterns, the stored
2992     options, and the first character, if set.
2993    
2994     9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
2995    
2996    
2997     Version 0.91 10-Sep-97
2998     ----------------------
2999    
3000     1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
3001     match the empty string as in /(a*)*/. It was looping and ultimately crashing.
3002    
3003     2. PCRE was looping on encountering an indefinitely repeated back reference to
3004     a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
3005     Perl does - treats the match as successful.
3006    
3007     ****

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