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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 65 - (show annotations) (download)
Sat Feb 24 21:40:08 2007 UTC (7 years, 7 months ago) by nigel
File size: 52643 byte(s)
Load pcre-4.1 into code/trunk.

1 ChangeLog for PCRE
2 ------------------
3
4 Version 4.1 12-Mar-03
5 ---------------------
6
7 1. Compiling with gcc -pedantic found a couple of places where casts were
8 needed, and a string in dftables.c that was longer than standard compilers are
9 required to support.
10
11 2. Compiling with Sun's compiler found a few more places where the code could
12 be tidied up in order to avoid warnings.
13
14 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
15 first of these names is deprecated in the latest Autoconf in favour of the name
16 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
17 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
18 analogy I have changed it to CFLAGS_FOR_BUILD.
19
20 4. Added -no-undefined to the linking command in the Makefile, because this is
21 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
22 linking step for the pcreposix library.
23
24 5. PCRE was failing to diagnose the case of two named groups with the same
25 name.
26
27 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
28 literal character that is needed in the subject for a match, and scans along to
29 ensure that it is present before embarking on the full matching process. This
30 saves time in cases of nested unlimited repeats that are never going to match.
31 Problem: the scan can take a lot of time if the subject is very long (e.g.
32 megabytes), thus penalizing straightforward matches. It is now done only if the
33 amount of subject to be scanned is less than 1000 bytes.
34
35 7. A lesser problem with the same optimization is that it was recording the
36 first character of an anchored pattern as "needed", thus provoking a search
37 right along the subject, even when the first match of the pattern was going to
38 fail. The "needed" character is now not set for anchored patterns, unless it
39 follows something in the pattern that is of non-fixed length. Thus, it still
40 fulfils its original purpose of finding quick non-matches in cases of nested
41 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
42
43
44 Version 4.0 17-Feb-03
45 ---------------------
46
47 1. If a comment in an extended regex that started immediately after a meta-item
48 extended to the end of string, PCRE compiled incorrect data. This could lead to
49 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
50
51 2. Moved to autoconf 2.53 and libtool 1.4.2.
52
53 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
54 the special perltest8 script is no longer needed - all the tests can be run
55 from a single perltest script.
56
57 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
58 by \s. It has now been removed in PCRE. This means it isn't recognized as
59 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
60 class [:space:] *does* include VT, thereby creating a mess.
61
62 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
63 space and tab.
64
65 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
66 its new features into the main test script, reducing the number of scripts.
67
68 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
69 were backward compatible, and made the (?i) apply to the whole pattern, as if
70 /i were given. Now it behaves more logically, and applies the option setting
71 only to what follows. PCRE has been changed to follow suit. However, if it
72 finds options settings right at the start of the pattern, it extracts them into
73 the global options, as before. Thus, they show up in the info data.
74
75 8. Added support for the \Q...\E escape sequence. Characters in between are
76 treated as literals. This is slightly different from Perl in that $ and @ are
77 also handled as literals inside the quotes. In Perl, they will cause variable
78 interpolation. Note the following examples:
79
80 Pattern PCRE matches Perl matches
81
82 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
83 \Qabc\$xyz\E abc\$xyz abc\$xyz
84 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
85
86 For compatibility with Perl, \Q...\E sequences are recognized inside character
87 classes as well as outside them.
88
89 9. Re-organized 3 code statements in pcretest to avoid "overflow in
90 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
91 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
92 signed/unsigned warnings.
93
94 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
95 option for pcretest, so I've replaced it by a simple function that does just
96 that job.
97
98 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
99 "pcregrep -".
100
101 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
102 Java package. This provides some syntactic sugar for simple cases of what my
103 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
104 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
105 item, you can use this simplified notation. Note that only makes sense with
106 greedy quantifiers. Consequently, the use of the possessive quantifier forces
107 greediness, whatever the setting of the PCRE_UNGREEDY option.
108
109 13. A change of greediness default within a pattern was not taking effect at
110 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
111 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
112 was abstracted outside.
113
114 14. PCRE now supports the \G assertion. It is true when the current matching
115 position is at the start point of the match. This differs from \A when the
116 starting offset is non-zero. Used with the /g option of pcretest (or similar
117 code), it works in the same way as it does for Perl's /g option. If all
118 alternatives of a regex begin with \G, the expression is anchored to the start
119 match position, and the "anchored" flag is set in the compiled expression.
120
121 15. Some bugs concerning the handling of certain option changes within patterns
122 have been fixed. These applied to options other than (?ims). For example,
123 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
124 been the other way round. Some of this was related to change 7 above.
125
126 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
127 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
128 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
129 POSIX classes only within a class (e.g. /[[:alpha:]]/).
130
131 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
132 mode. Unlike ".", it always matches newline, whatever the setting of
133 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
134 assertions. Perl allows it, but it doesn't (in general) work because it can't
135 calculate the length of the lookbehind. At least, that's the case for Perl
136 5.8.0 - I've been told they are going to document that it doesn't work in
137 future.
138
139 18. Added an error diagnosis for escapes that PCRE does not support: these are
140 \L, \l, \N, \P, \p, \U, \u, and \X.
141
142 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
143 reading past the end of the pattern in cases such as /[abcd/.
144
145 20. PCRE was getting more memory than necessary for patterns with classes that
146 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
147
148 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
149 compiling PCRE for use with Virtual Pascal.
150
151 22. Small fix to the Makefile to make it work properly if the build is done
152 outside the source tree.
153
154 23. Added a new extension: a condition to go with recursion. If a conditional
155 subpattern starts with (?(R) the "true" branch is used if recursion has
156 happened, whereas the "false" branch is used only at the top level.
157
158 24. When there was a very long string of literal characters (over 255 bytes
159 without UTF support, over 250 bytes with UTF support), the computation of how
160 much memory was required could be incorrect, leading to segfaults or other
161 strange effects.
162
163 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
164 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
165 there was a subsequent back reference to those brackets. This meant that, for
166 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
167 possible to check for precisely this case. All we can do is abandon the
168 optimization if .* occurs inside capturing brackets when there are any back
169 references whatsoever. (See below for a better fix that came later.)
170
171 26. The handling of the optimization for finding the first character of a
172 non-anchored pattern, and for finding a character that is required later in the
173 match were failing in some cases. This didn't break the matching; it just
174 failed to optimize when it could. The way this is done has been re-implemented.
175
176 27. Fixed typo in error message for invalid (?R item (it said "(?p").
177
178 28. Added a new feature that provides some of the functionality that Perl
179 provides with (?{...}). The facility is termed a "callout". The way it is done
180 in PCRE is for the caller to provide an optional function, by setting
181 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
182 global variable. By default it is unset, which disables all calling out. To get
183 the function called, the regex must include (?C) at appropriate points. This
184 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
185 This provides a means of identifying different callout points. When PCRE
186 reaches such a point in the regex, if pcre_callout has been set, the external
187 function is called. It is provided with data in a structure called
188 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
189 matching continues; if it returns a non-zero value, the match at the current
190 point fails. However, backtracking will occur if possible. [This was changed
191 later and other features added - see item 49 below.]
192
193 29. pcretest is upgraded to test the callout functionality. It provides a
194 callout function that displays information. By default, it shows the start of
195 the match and the current position in the text. There are some new data escapes
196 to vary what happens:
197
198 \C+ in addition, show current contents of captured substrings
199 \C- do not supply a callout function
200 \C!n return 1 when callout number n is reached
201 \C!n!m return 1 when callout number n is reached for the mth time
202
203 30. If pcregrep was called with the -l option and just a single file name, it
204 output "<stdin>" if a match was found, instead of the file name.
205
206 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
207 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
208 pcre_exec(). This saves a malloc/free per call. The default value of
209 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
210 when configuring.
211
212 32. The default maximum size of a compiled pattern is 64K. There have been a
213 few cases of people hitting this limit. The code now uses macros to handle the
214 storing of links as offsets within the compiled pattern. It defaults to 2-byte
215 links, but this can be changed to 3 or 4 bytes by --with-link-size when
216 configuring. Tests 2 and 5 work only with 2-byte links because they output
217 debugging information about compiled patterns.
218
219 33. Internal code re-arrangements:
220
221 (a) Moved the debugging function for printing out a compiled regex into
222 its own source file (printint.c) and used #include to pull it into
223 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
224 separate copies.
225
226 (b) Defined the list of op-code names for debugging as a macro in
227 internal.h so that it is next to the definition of the opcodes.
228
229 (c) Defined a table of op-code lengths for simpler skipping along compiled
230 code. This is again a macro in internal.h so that it is next to the
231 definition of the opcodes.
232
233 34. Added support for recursive calls to individual subpatterns, along the
234 lines of Robin Houston's patch (but implemented somewhat differently).
235
236 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
237 allow it to read and process whole directories in Win32. This code was
238 contributed by Lionel Fourquaux; it has not been tested by me.
239
240 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
241 used to name a group. Names consist of alphanumerics and underscores, and must
242 be unique. Back references use the syntax (?P=name) and recursive calls use
243 (?P>name) which is a PCRE extension to the Python extension. Groups still have
244 numbers. The function pcre_fullinfo() can be used after compilation to extract
245 a name/number map. There are three relevant calls:
246
247 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
248 PCRE_INFO_NAMECOUNT yields the number of entries
249 PCRE_INFO_NAMETABLE yields a pointer to the map.
250
251 The map is a vector of fixed-size entries. The size of each entry depends on
252 the length of the longest name used. The first two bytes of each entry are the
253 group number, most significant byte first. There follows the corresponding
254 name, zero terminated. The names are in alphabetical order.
255
256 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
257 case instead of 255. Making it the same both with and without UTF-8 support
258 means that the same test output works with both.
259
260 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
261 calling malloc() with a zero argument.
262
263 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
264 optimization. I've improved things by keeping a bitmap of backreferences with
265 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
266 fact referenced, the optimization can be applied. It is unlikely that a
267 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
268 the match to follow \n) will appear inside brackets with a number greater than
269 31, but if it does, any back reference > 31 suppresses the optimization.
270
271 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
272 of disabling numbered capturing parentheses. Any opening parenthesis that is
273 not followed by ? behaves as if it were followed by ?: but named parentheses
274 can still be used for capturing (and they will acquire numbers in the usual
275 way).
276
277 41. Redesigned the return codes from the match() function into yes/no/error so
278 that errors can be passed back from deep inside the nested calls. A malloc
279 failure while inside a recursive subpattern call now causes the
280 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
281
282 42. It is now possible to set a limit on the number of times the match()
283 function is called in a call to pcre_exec(). This facility makes it possible to
284 limit the amount of recursion and backtracking, though not in a directly
285 obvious way, because the match() function is used in a number of different
286 circumstances. The count starts from zero for each position in the subject
287 string (for non-anchored patterns). The default limit is, for compatibility, a
288 large number, namely 10 000 000. You can change this in two ways:
289
290 (a) When configuring PCRE before making, you can use --with-match-limit=n
291 to set a default value for the compiled library.
292
293 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
294 a different value is set. See 45 below.
295
296 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
297
298 43. Added a new function pcre_config(int, void *) to enable run-time extraction
299 of things that can be changed at compile time. The first argument specifies
300 what is wanted and the second points to where the information is to be placed.
301 The current list of available information is:
302
303 PCRE_CONFIG_UTF8
304
305 The output is an integer that is set to one if UTF-8 support is available;
306 otherwise it is set to zero.
307
308 PCRE_CONFIG_NEWLINE
309
310 The output is an integer that it set to the value of the code that is used for
311 newline. It is either LF (10) or CR (13).
312
313 PCRE_CONFIG_LINK_SIZE
314
315 The output is an integer that contains the number of bytes used for internal
316 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
317
318 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
319
320 The output is an integer that contains the threshold above which the POSIX
321 interface uses malloc() for output vectors. See item 31 above.
322
323 PCRE_CONFIG_MATCH_LIMIT
324
325 The output is an unsigned integer that contains the default limit of the number
326 of match() calls in a pcre_exec() execution. See 42 above.
327
328 44. pcretest has been upgraded by the addition of the -C option. This causes it
329 to extract all the available output from the new pcre_config() function, and to
330 output it. The program then exits immediately.
331
332 45. A need has arisen to pass over additional data with calls to pcre_exec() in
333 order to support additional features. One way would have been to define
334 pcre_exec2() (for example) with extra arguments, but this would not have been
335 extensible, and would also have required all calls to the original function to
336 be mapped to the new one. Instead, I have chosen to extend the mechanism that
337 is used for passing in "extra" data from pcre_study().
338
339 The pcre_extra structure is now exposed and defined in pcre.h. It currently
340 contains the following fields:
341
342 flags a bitmap indicating which of the following fields are set
343 study_data opaque data from pcre_study()
344 match_limit a way of specifying a limit on match() calls for a specific
345 call to pcre_exec()
346 callout_data data for callouts (see 49 below)
347
348 The flag bits are also defined in pcre.h, and are
349
350 PCRE_EXTRA_STUDY_DATA
351 PCRE_EXTRA_MATCH_LIMIT
352 PCRE_EXTRA_CALLOUT_DATA
353
354 The pcre_study() function now returns one of these new pcre_extra blocks, with
355 the actual study data pointed to by the study_data field, and the
356 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
357 before. That is, this change is entirely upwards-compatible and requires no
358 change to existing code.
359
360 If you want to pass in additional data to pcre_exec(), you can either place it
361 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
362 block.
363
364 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
365 data string contains the escape sequence \M, pcretest calls pcre_exec() several
366 times with different match limits, until it finds the minimum value needed for
367 pcre_exec() to complete. The value is then output. This can be instructive; for
368 most simple matches the number is quite small, but for pathological cases it
369 gets very large very quickly.
370
371 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
372 returns the size of the data block pointed to by the study_data field in a
373 pcre_extra block, that is, the value that was passed as the argument to
374 pcre_malloc() when PCRE was getting memory in which to place the information
375 created by pcre_study(). The fourth argument should point to a size_t variable.
376 pcretest has been extended so that this information is shown after a successful
377 pcre_study() call when information about the compiled regex is being displayed.
378
379 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
380 because what follows is always an absolute path. (Later: it turns out that this
381 is more than cosmetic for MinGW, because it doesn't like empty path
382 components.)
383
384 49. Some changes have been made to the callout feature (see 28 above):
385
386 (i) A callout function now has three choices for what it returns:
387
388 0 => success, carry on matching
389 > 0 => failure at this point, but backtrack if possible
390 < 0 => serious error, return this value from pcre_exec()
391
392 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
393 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
394 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
395 use by callout functions. It will never be used by PCRE itself.
396
397 (ii) The pcre_extra structure (see 45 above) has a void * field called
398 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
399 pcre_callout_block structure has a field of the same name. The contents of
400 the field passed in the pcre_extra structure are passed to the callout
401 function in the corresponding field in the callout block. This makes it
402 easier to use the same callout-containing regex from multiple threads. For
403 testing, the pcretest program has a new data escape
404
405 \C*n pass the number n (may be negative) as callout_data
406
407 If the callout function in pcretest receives a non-zero value as
408 callout_data, it returns that value.
409
410 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
411 there were some redundant $(CFLAGS) in commands that are now specified as
412 $(LINK), which already includes $(CFLAGS).
413
414 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
415 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
416 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
417 one-byte characters throughout. Note that case-insensitive matching applies
418 only to characters whose values are less than 256. PCRE doesn't support the
419 notion of cases for higher-valued characters.
420
421 (i) A character class whose characters are all within 0-255 is handled as
422 a bit map, and the map is inverted for negative classes. Previously, a
423 character > 255 always failed to match such a class; however it should
424 match if the class was a negative one (e.g. [^ab]). This has been fixed.
425
426 (ii) A negated character class with a single character < 255 is coded as
427 "not this character" (OP_NOT). This wasn't working properly when the test
428 character was multibyte, either singly or repeated.
429
430 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
431 mode, for example: \x{100}{2,3}.
432
433 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
434 singly or repeated) now correctly test multibyte characters. However,
435 PCRE doesn't recognize any characters with values greater than 255 as
436 digits, spaces, or word characters. Such characters always match \D, \S,
437 and \W, and never match \d, \s, or \w.
438
439 (v) Classes may now contain characters and character ranges with values
440 greater than 255. For example: [ab\x{100}-\x{400}].
441
442 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
443 PCRE in UTF-8 mode.
444
445 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
446 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
447 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
448 value.)
449
450 53. The single man page has become too large. I have therefore split it up into
451 a number of separate man pages. These also give rise to individual HTML pages;
452 these are now put in a separate directory, and there is an index.html page that
453 lists them all. Some hyperlinking between the pages has been installed.
454
455 54. Added convenience functions for handling named capturing parentheses.
456
457 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
458 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
459 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
460 are faulted.
461
462 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
463 calling configure. These values are used when compiling the dftables.c program
464 which is run to generate the source of the default character tables. They
465 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
466 you will need to set these values.
467
468 57. Updated the building process for Windows DLL, as provided by Fred Cox.
469
470
471 Version 3.9 02-Jan-02
472 ---------------------
473
474 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
475
476 2. If --disable-static was given, the building process failed when trying to
477 build pcretest and pcregrep. (For some reason it was using libtool to compile
478 them, which is not right, as they aren't part of the library.)
479
480
481 Version 3.8 18-Dec-01
482 ---------------------
483
484 1. The experimental UTF-8 code was completely screwed up. It was packing the
485 bytes in the wrong order. How dumb can you get?
486
487
488 Version 3.7 29-Oct-01
489 ---------------------
490
491 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
492 This caused pcretest, when used on the test data, to segfault. Unfortunately,
493 this didn't happen under Solaris 8, where I normally test things.
494
495 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
496 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
497 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
498 this fix an hour or so after the initial 3.7 release.)
499
500
501 Version 3.6 23-Oct-01
502 ---------------------
503
504 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
505 offsets passed as NULL with zero offset count.
506
507 2. The config.guess and config.sub files had not been updated when I moved to
508 the latest autoconf.
509
510
511 Version 3.5 15-Aug-01
512 ---------------------
513
514 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
515 had been forgotten.
516
517 2. By using declared but undefined structures, we can avoid using "void"
518 definitions in pcre.h while keeping the internal definitions of the structures
519 private.
520
521 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
522 user point of view, this means that both static and shared libraries are built
523 by default, but this can be individually controlled. More of the work of
524 handling this static/shared cases is now inside libtool instead of PCRE's make
525 file.
526
527 4. The pcretest utility is now installed along with pcregrep because it is
528 useful for users (to test regexs) and by doing this, it automatically gets
529 relinked by libtool. The documentation has been turned into a man page, so
530 there are now .1, .txt, and .html versions in /doc.
531
532 5. Upgrades to pcregrep:
533 (i) Added long-form option names like gnu grep.
534 (ii) Added --help to list all options with an explanatory phrase.
535 (iii) Added -r, --recursive to recurse into sub-directories.
536 (iv) Added -f, --file to read patterns from a file.
537
538 6. pcre_exec() was referring to its "code" argument before testing that
539 argument for NULL (and giving an error if it was NULL).
540
541 7. Upgraded Makefile.in to allow for compiling in a different directory from
542 the source directory.
543
544 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
545 options bits, the pointer it was passed was to an int instead of to an unsigned
546 long int. This mattered only on 64-bit systems.
547
548 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
549 generated) instead of pcre.in, which it its source. Also made the same change
550 in several of the .c files.
551
552 10. A new release of gcc defines printf() as a macro, which broke pcretest
553 because it had an ifdef in the middle of a string argument for printf(). Fixed
554 by using separate calls to printf().
555
556 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
557 script, to force use of CR or LF instead of \n in the source. On non-Unix
558 systems, the value can be set in config.h.
559
560 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
561 absolute limit. Changed the text of the error message to make this clear, and
562 likewise updated the man page.
563
564 13. The limit of 99 on the number of capturing subpatterns has been removed.
565 The new limit is 65535, which I hope will not be a "real" limit.
566
567
568 Version 3.4 22-Aug-00
569 ---------------------
570
571 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
572
573 2. Diagnose condition (?(0) as an error instead of crashing on matching.
574
575
576 Version 3.3 01-Aug-00
577 ---------------------
578
579 1. If an octal character was given, but the value was greater than \377, it
580 was not getting masked to the least significant bits, as documented. This could
581 lead to crashes in some systems.
582
583 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
584 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
585
586 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
587 These just pass their arguments on to (pcre_free)(), but they are provided
588 because some uses of PCRE bind it to non-C systems that can call its functions,
589 but cannot call free() or pcre_free() directly.
590
591 4. Add "make test" as a synonym for "make check". Corrected some comments in
592 the Makefile.
593
594 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
595 Makefile.
596
597 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
598 command called pgrep for grepping around the active processes.
599
600 7. Added the beginnings of support for UTF-8 character strings.
601
602 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
603 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
604 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
605 out for the ar command.)
606
607
608 Version 3.2 12-May-00
609 ---------------------
610
611 This is purely a bug fixing release.
612
613 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
614 of ZA. This was just one example of several cases that could provoke this bug,
615 which was introduced by change 9 of version 2.00. The code for breaking
616 infinite loops after an iteration that matches an empty string was't working
617 correctly.
618
619 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
620 when matched against abbab (for example). After matching an empty string, it
621 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
622 caused it to match further down the string than it should.
623
624 3. The code contained an inclusion of sys/types.h. It isn't clear why this
625 was there because it doesn't seem to be needed, and it causes trouble on some
626 systems, as it is not a Standard C header. It has been removed.
627
628 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
629 were reported on the Macintosh. The changes were from
630
631 while ((c = *(++ptr)) != 0 && c != '\n');
632 to
633 while ((c = *(++ptr)) != 0 && c != '\n') ;
634
635 Totally extraordinary, but if that's what it takes...
636
637 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
638 available. Added HAVE_BCOPY and an autoconf test for it; if neither
639 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
640 assumes the way PCRE uses memmove() (always moving upwards).
641
642 6. PCRE is being used in one environment where strchr() is not available. There
643 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
644 faster code anyway.
645
646
647 Version 3.1 09-Feb-00
648 ---------------------
649
650 The only change in this release is the fixing of some bugs in Makefile.in for
651 the "install" target:
652
653 (1) It was failing to install pcreposix.h.
654
655 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
656
657
658 Version 3.0 01-Feb-00
659 ---------------------
660
661 1. Add support for the /+ modifier to perltest (to output $` like it does in
662 pcretest).
663
664 2. Add support for the /g modifier to perltest.
665
666 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
667 matches null strings.
668
669 4. Fix perltest so that it doesn't do unwanted things when fed an empty
670 pattern. Perl treats empty patterns specially - it reuses the most recent
671 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
672 effect.
673
674 5. The POSIX interface was broken in that it was just handing over the POSIX
675 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
676 required a bigger vector, with some working space on the end. This means that
677 the POSIX wrapper now has to get and free some memory, and copy the results.
678
679 6. Added some simple autoconf support, placing the test data and the
680 documentation in separate directories, re-organizing some of the
681 information files, and making it build pcre-config (a GNU standard). Also added
682 libtool support for building PCRE as a shared library, which is now the
683 default.
684
685 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
686 09 are not valid octal constants. Single digits will be used for minor values
687 less than 10.
688
689 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
690 existing programs that set these in the POSIX interface can use PCRE without
691 modification.
692
693 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
694 return all that pcre_info() returns, plus additional data. The pcre_info()
695 function is retained for compatibility, but is considered to be obsolete.
696
697 10. Added experimental recursion feature (?R) to handle one common case that
698 Perl 5.6 will be able to do with (?p{...}).
699
700 11. Added support for POSIX character classes like [:alpha:], which Perl is
701 adopting.
702
703
704 Version 2.08 31-Aug-99
705 ----------------------
706
707 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
708 trying to match at the startoffset position, but instead was moving forward to
709 the next newline as if a previous match had failed.
710
711 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
712 and could get into a loop if a null string was matched other than at the start
713 of the subject.
714
715 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
716 be distinguished at compile time, and for completeness also added PCRE_DATE.
717
718 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
719 in GnuWin32 environments.
720
721
722 Version 2.07 29-Jul-99
723 ----------------------
724
725 1. The documentation is now supplied in plain text form and HTML as well as in
726 the form of man page sources.
727
728 2. C++ compilers don't like assigning (void *) values to other pointer types.
729 In particular this affects malloc(). Although there is no problem in Standard
730 C, I've put in casts to keep C++ compilers happy.
731
732 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
733 should be (const char *).
734
735 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
736 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
737 However, I haven't made this a standard facility. The documentation doesn't
738 mention it, and the Makefile doesn't support it.
739
740 5. The Makefile now contains an "install" target, with editable destinations at
741 the top of the file. The pcretest program is not installed.
742
743 6. pgrep -V now gives the PCRE version number and date.
744
745 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
746 causing the entire string to be ignored, instead of just the last character.
747
748 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
749 non-matching string, it can take a very, very long time, even for strings of
750 quite modest length, because of the nested recursion. PCRE now does better in
751 some of these cases. It does this by remembering the last required literal
752 character in the pattern, and pre-searching the subject to ensure it is present
753 before running the real match. In other words, it applies a heuristic to detect
754 some types of certain failure quickly, and in the above example, if presented
755 with a string that has no trailing " it gives "no match" very quickly.
756
757 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
758 other alternatives are tried instead.
759
760
761 Version 2.06 09-Jun-99
762 ----------------------
763
764 1. Change pcretest's output for amount of store used to show just the code
765 space, because the remainder (the data block) varies in size between 32-bit and
766 64-bit systems.
767
768 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
769 start matching at. This allows lookbehinds to work when searching for multiple
770 occurrences in a string.
771
772 3. Added additional options to pcretest for testing multiple occurrences:
773
774 /+ outputs the rest of the string that follows a match
775 /g loops for multiple occurrences, using the new startoffset argument
776 /G loops for multiple occurrences by passing an incremented pointer
777
778 4. PCRE wasn't doing the "first character" optimization for patterns starting
779 with \b or \B, though it was doing it for other lookbehind assertions. That is,
780 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
781 the letter 'x'. On long subject strings, this gives a significant speed-up.
782
783
784 Version 2.05 21-Apr-99
785 ----------------------
786
787 1. Changed the type of magic_number from int to long int so that it works
788 properly on 16-bit systems.
789
790 2. Fixed a bug which caused patterns starting with .* not to work correctly
791 when the subject string contained newline characters. PCRE was assuming
792 anchoring for such patterns in all cases, which is not correct because .* will
793 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
794 DOTALL is set at top level; otherwise it knows that patterns starting with .*
795 must be retried after every newline in the subject.
796
797
798 Version 2.04 18-Feb-99
799 ----------------------
800
801 1. For parenthesized subpatterns with repeats whose minimum was zero, the
802 computation of the store needed to hold the pattern was incorrect (too large).
803 If such patterns were nested a few deep, this could multiply and become a real
804 problem.
805
806 2. Added /M option to pcretest to show the memory requirement of a specific
807 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
808
809 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
810 compiled in such a way that the backtracking after subsequent failure was
811 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
812 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
813
814
815 Version 2.03 02-Feb-99
816 ----------------------
817
818 1. Fixed typo and small mistake in man page.
819
820 2. Added 4th condition (GPL supersedes if conflict) and created separate
821 LICENCE file containing the conditions.
822
823 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
824 Perl, that is the internal \ allows the delimiter to be included in the
825 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
826 the final delimiter, add \ to the end of the pattern (to test the error).
827
828 4. Added the convenience functions for extracting substrings after a successful
829 match. Updated pcretest to make it able to test these functions.
830
831
832 Version 2.02 14-Jan-99
833 ----------------------
834
835 1. Initialized the working variables associated with each extraction so that
836 their saving and restoring doesn't refer to uninitialized store.
837
838 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
839 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
840 fix the problem.
841
842 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
843 calls, and wasn't printing the correct value for compiling calls. Increased the
844 default value of LOOPREPEAT, and the number of significant figures in the
845 times.
846
847 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
848
849 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
850 a building problem on Windows NT with a FAT file system.
851
852
853 Version 2.01 21-Oct-98
854 ----------------------
855
856 1. Changed the API for pcre_compile() to allow for the provision of a pointer
857 to character tables built by pcre_maketables() in the current locale. If NULL
858 is passed, the default tables are used.
859
860
861 Version 2.00 24-Sep-98
862 ----------------------
863
864 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
865 it any more.
866
867 2. Allow quantification of (?>) groups, and make it work correctly.
868
869 3. The first character computation wasn't working for (?>) groups.
870
871 4. Correct the implementation of \Z (it is permitted to match on the \n at the
872 end of the subject) and add 5.005's \z, which really does match only at the
873 very end of the subject.
874
875 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
876
877 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
878 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
879 localized options. All options to pcre_study() were also removed.
880
881 7. Add other new features from 5.005:
882
883 $(?<= positive lookbehind
884 $(?<! negative lookbehind
885 (?imsx-imsx) added the unsetting capability
886 such a setting is global if at outer level; local otherwise
887 (?imsx-imsx:) non-capturing groups with option setting
888 (?(cond)re|re) conditional pattern matching
889
890 A backreference to itself in a repeated group matches the previous
891 captured string.
892
893 8. General tidying up of studying (both automatic and via "study")
894 consequential on the addition of new assertions.
895
896 9. As in 5.005, unlimited repeated groups that could match an empty substring
897 are no longer faulted at compile time. Instead, the loop is forcibly broken at
898 runtime if any iteration does actually match an empty substring.
899
900 10. Include the RunTest script in the distribution.
901
902 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
903 discrepancies, some of which were old and were also with respect to 5.004. They
904 have now been fixed.
905
906
907 Version 1.09 28-Apr-98
908 ----------------------
909
910 1. A negated single character class followed by a quantifier with a minimum
911 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
912 program crashes, or just wrong answers. This did not apply to negated classes
913 containing more than one character, or to minima other than one.
914
915
916 Version 1.08 27-Mar-98
917 ----------------------
918
919 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
920
921 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
922 latter must appear before anything that relies on it in the pattern.
923
924
925 Version 1.07 16-Feb-98
926 ----------------------
927
928 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
929 repeat of a potentially empty string).
930
931
932 Version 1.06 23-Jan-98
933 ----------------------
934
935 1. Added Markus Oberhumer's little patches for C++.
936
937 2. Literal strings longer than 255 characters were broken.
938
939
940 Version 1.05 23-Dec-97
941 ----------------------
942
943 1. Negated character classes containing more than one character were failing if
944 PCRE_CASELESS was set at run time.
945
946
947 Version 1.04 19-Dec-97
948 ----------------------
949
950 1. Corrected the man page, where some "const" qualifiers had been omitted.
951
952 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
953 input syntax.
954
955 3. Fixed memory leak which occurred when a regex with back references was
956 matched with an offsets vector that wasn't big enough. The temporary memory
957 that is used in this case wasn't being freed if the match failed.
958
959 4. Tidied pcretest to ensure it frees memory that it gets.
960
961 5. Temporary memory was being obtained in the case where the passed offsets
962 vector was exactly big enough.
963
964 6. Corrected definition of offsetof() from change 5 below.
965
966 7. I had screwed up change 6 below and broken the rules for the use of
967 setjmp(). Now fixed.
968
969
970 Version 1.03 18-Dec-97
971 ----------------------
972
973 1. A erroneous regex with a missing opening parenthesis was correctly
974 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
975 on some systems.
976
977 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
978 it was reported that one broken compiler failed on the former because "code" is
979 also an independent variable.
980
981 3. The erroneous regex a[]b caused an array overrun reference.
982
983 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
984 fail on data ending with that character. (It was going on too far, and checking
985 the next character, typically a binary zero.) This was specific to the
986 optimized code for single-character negative classes.
987
988 5. Added a contributed patch from the TIN world which does the following:
989
990 + Add an undef for memmove, in case the the system defines a macro for it.
991
992 + Add a definition of offsetof(), in case there isn't one. (I don't know
993 the reason behind this - offsetof() is part of the ANSI standard - but
994 it does no harm).
995
996 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
997 most of the places where whitespace preceded '#'. I have given up and
998 allowed the remaining 2 cases to be at the margin.
999
1000 + Rename some variables in pcre to eliminate shadowing. This seems very
1001 pedantic, but does no harm, of course.
1002
1003 6. Moved the call to setjmp() into its own function, to get rid of warnings
1004 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
1005
1006 7. Constructs such as \d{8,} were compiling into the equivalent of
1007 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
1008 outcome, but in this particular case used more store than had been allocated,
1009 which caused the bug to be discovered because it threw up an internal error.
1010
1011 8. The debugging code in both pcre and pcretest for outputting the compiled
1012 form of a regex was going wrong in the case of back references followed by
1013 curly-bracketed repeats.
1014
1015
1016 Version 1.02 12-Dec-97
1017 ----------------------
1018
1019 1. Typos in pcre.3 and comments in the source fixed.
1020
1021 2. Applied a contributed patch to get rid of places where it used to remove
1022 'const' from variables, and fixed some signed/unsigned and uninitialized
1023 variable warnings.
1024
1025 3. Added the "runtest" target to Makefile.
1026
1027 4. Set default compiler flag to -O2 rather than just -O.
1028
1029
1030 Version 1.01 19-Nov-97
1031 ----------------------
1032
1033 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
1034 like /([ab]*)*/, that is, for classes with more than one character in them.
1035
1036 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
1037 as /((?>a*))*/ (a PCRE_EXTRA facility).
1038
1039
1040 Version 1.00 18-Nov-97
1041 ----------------------
1042
1043 1. Added compile-time macros to support systems such as SunOS4 which don't have
1044 memmove() or strerror() but have other things that can be used instead.
1045
1046 2. Arranged that "make clean" removes the executables.
1047
1048
1049 Version 0.99 27-Oct-97
1050 ----------------------
1051
1052 1. Fixed bug in code for optimizing classes with only one character. It was
1053 initializing a 32-byte map regardless, which could cause it to run off the end
1054 of the memory it had got.
1055
1056 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1057
1058
1059 Version 0.98 22-Oct-97
1060 ----------------------
1061
1062 1. Fixed bug in code for handling temporary memory usage when there are more
1063 back references than supplied space in the ovector. This could cause segfaults.
1064
1065
1066 Version 0.97 21-Oct-97
1067 ----------------------
1068
1069 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1070
1071 2. Optimized negated single characters not to use a bit map.
1072
1073 3. Brought error texts together as macro definitions; clarified some of them;
1074 fixed one that was wrong - it said "range out of order" when it meant "invalid
1075 escape sequence".
1076
1077 4. Changed some char * arguments to const char *.
1078
1079 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1080
1081 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1082 pcretest.
1083
1084
1085 Version 0.96 16-Oct-97
1086 ----------------------
1087
1088 1. Added a simple "pgrep" utility to the distribution.
1089
1090 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1091 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1092 where "ddd" means "one or more decimal digits".
1093
1094 3. Fixed serious bug. If a pattern had a back reference, but the call to
1095 pcre_exec() didn't supply a large enough ovector to record the related
1096 identifying subpattern, the match always failed. PCRE now remembers the number
1097 of the largest back reference, and gets some temporary memory in which to save
1098 the offsets during matching if necessary, in order to ensure that
1099 backreferences always work.
1100
1101 4. Increased the compatibility with Perl in a number of ways:
1102
1103 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1104 to request this handling. The option can be set at compile or exec time.
1105
1106 (b) $ matches before a terminating newline by default; an option
1107 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
1108 mode). The option can be set at compile or exec time.
1109
1110 (c) The handling of \ followed by a digit other than 0 is now supposed to be
1111 the same as Perl's. If the decimal number it represents is less than 10
1112 or there aren't that many previous left capturing parentheses, an octal
1113 escape is read. Inside a character class, it's always an octal escape,
1114 even if it is a single digit.
1115
1116 (d) An escaped but undefined alphabetic character is taken as a literal,
1117 unless PCRE_EXTRA is set. Currently this just reserves the remaining
1118 escapes.
1119
1120 (e) {0} is now permitted. (The previous item is removed from the compiled
1121 pattern).
1122
1123 5. Changed all the names of code files so that the basic parts are no longer
1124 than 10 characters, and abolished the teeny "globals.c" file.
1125
1126 6. Changed the handling of character classes; they are now done with a 32-byte
1127 bit map always.
1128
1129 7. Added the -d and /D options to pcretest to make it possible to look at the
1130 internals of compilation without having to recompile pcre.
1131
1132
1133 Version 0.95 23-Sep-97
1134 ----------------------
1135
1136 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
1137 \x20 at the start of a run of normal characters. These were being treated as
1138 real characters, instead of the source characters being re-checked.
1139
1140
1141 Version 0.94 18-Sep-97
1142 ----------------------
1143
1144 1. The functions are now thread-safe, with the caveat that the global variables
1145 containing pointers to malloc() and free() or alternative functions are the
1146 same for all threads.
1147
1148 2. Get pcre_study() to generate a bitmap of initial characters for non-
1149 anchored patterns when this is possible, and use it if passed to pcre_exec().
1150
1151
1152 Version 0.93 15-Sep-97
1153 ----------------------
1154
1155 1. /(b)|(:+)/ was computing an incorrect first character.
1156
1157 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
1158 but not actually doing anything yet.
1159
1160 3. Treat "-" characters in classes that cannot be part of ranges as literals,
1161 as Perl does (e.g. [-az] or [az-]).
1162
1163 4. Set the anchored flag if a branch starts with .* or .*? because that tests
1164 all possible positions.
1165
1166 5. Split up into different modules to avoid including unneeded functions in a
1167 compiled binary. However, compile and exec are still in one module. The "study"
1168 function is split off.
1169
1170 6. The character tables are now in a separate module whose source is generated
1171 by an auxiliary program - but can then be edited by hand if required. There are
1172 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
1173 toupper() in the code.
1174
1175 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
1176 make them global. Abolish the function for setting them, as the caller can now
1177 set them directly.
1178
1179
1180 Version 0.92 11-Sep-97
1181 ----------------------
1182
1183 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
1184 (e.g. /a{1,3}/) was broken (I mis-optimized it).
1185
1186 2. Caseless matching was not working in character classes if the characters in
1187 the pattern were in upper case.
1188
1189 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
1190
1191 4. Make PCRE_ANCHORED public and accept as a compile option.
1192
1193 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
1194 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
1195 pass them.
1196
1197 6. Give an error if bad option bits passed at compile or run time.
1198
1199 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
1200 pcretest to cause it to pass that flag.
1201
1202 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
1203 options, and the first character, if set.
1204
1205 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
1206
1207
1208 Version 0.91 10-Sep-97
1209 ----------------------
1210
1211 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
1212 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
1213
1214 2. PCRE was looping on encountering an indefinitely repeated back reference to
1215 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
1216 Perl does - treats the match as successful.
1217
1218 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12