ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 63 - (show annotations)
Sat Feb 24 21:40:03 2007 UTC (14 years, 2 months ago) by nigel
File size: 50590 byte(s)
Load pcre-4.0 into code/trunk.
1 ChangeLog for PCRE
2 ------------------
4 Version 4.00 17-Feb-03
5 ----------------------
7 1. If a comment in an extended regex that started immediately after a meta-item
8 extended to the end of string, PCRE compiled incorrect data. This could lead to
9 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
11 2. Moved to autoconf 2.53 and libtool 1.4.2.
13 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
14 the special perltest8 script is no longer needed - all the tests can be run
15 from a single perltest script.
17 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
18 by \s. It has now been removed in PCRE. This means it isn't recognized as
19 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
20 class [:space:] *does* include VT, thereby creating a mess.
22 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
23 space and tab.
25 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
26 its new features into the main test script, reducing the number of scripts.
28 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
29 were backward compatible, and made the (?i) apply to the whole pattern, as if
30 /i were given. Now it behaves more logically, and applies the option setting
31 only to what follows. PCRE has been changed to follow suit. However, if it
32 finds options settings right at the start of the pattern, it extracts them into
33 the global options, as before. Thus, they show up in the info data.
35 8. Added support for the \Q...\E escape sequence. Characters in between are
36 treated as literals. This is slightly different from Perl in that $ and @ are
37 also handled as literals inside the quotes. In Perl, they will cause variable
38 interpolation. Note the following examples:
40 Pattern PCRE matches Perl matches
42 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
43 \Qabc\$xyz\E abc\$xyz abc\$xyz
44 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
46 For compatibility with Perl, \Q...\E sequences are recognized inside character
47 classes as well as outside them.
49 9. Re-organized 3 code statements in pcretest to avoid "overflow in
50 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
51 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
52 signed/unsigned warnings.
54 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
55 option for pcretest, so I've replaced it by a simple function that does just
56 that job.
58 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
59 "pcregrep -".
61 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
62 Java package. This provides some syntactic sugar for simple cases of what my
63 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
64 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
65 item, you can use this simplified notation. Note that only makes sense with
66 greedy quantifiers. Consequently, the use of the possessive quantifier forces
67 greediness, whatever the setting of the PCRE_UNGREEDY option.
69 13. A change of greediness default within a pattern was not taking effect at
70 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
71 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
72 was abstracted outside.
74 14. PCRE now supports the \G assertion. It is true when the current matching
75 position is at the start point of the match. This differs from \A when the
76 starting offset is non-zero. Used with the /g option of pcretest (or similar
77 code), it works in the same way as it does for Perl's /g option. If all
78 alternatives of a regex begin with \G, the expression is anchored to the start
79 match position, and the "anchored" flag is set in the compiled expression.
81 15. Some bugs concerning the handling of certain option changes within patterns
82 have been fixed. These applied to options other than (?ims). For example,
83 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
84 been the other way round. Some of this was related to change 7 above.
86 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
87 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
88 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
89 POSIX classes only within a class (e.g. /[[:alpha:]]/).
91 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
92 mode. Unlike ".", it always matches newline, whatever the setting of
93 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
94 assertions. Perl allows it, but it doesn't (in general) work because it can't
95 calculate the length of the lookbehind. At least, that's the case for Perl
96 5.8.0 - I've been told they are going to document that it doesn't work in
97 future.
99 18. Added an error diagnosis for escapes that PCRE does not support: these are
100 \L, \l, \N, \P, \p, \U, \u, and \X.
102 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
103 reading past the end of the pattern in cases such as /[abcd/.
105 20. PCRE was getting more memory than necessary for patterns with classes that
106 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
108 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
109 compiling PCRE for use with Virtual Pascal.
111 22. Small fix to the Makefile to make it work properly if the build is done
112 outside the source tree.
114 23. Added a new extension: a condition to go with recursion. If a conditional
115 subpattern starts with (?(R) the "true" branch is used if recursion has
116 happened, whereas the "false" branch is used only at the top level.
118 24. When there was a very long string of literal characters (over 255 bytes
119 without UTF support, over 250 bytes with UTF support), the computation of how
120 much memory was required could be incorrect, leading to segfaults or other
121 strange effects.
123 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
124 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
125 there was a subsequent back reference to those brackets. This meant that, for
126 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
127 possible to check for precisely this case. All we can do is abandon the
128 optimization if .* occurs inside capturing brackets when there are any back
129 references whatsoever. (See below for a better fix that came later.)
131 26. The handling of the optimization for finding the first character of a
132 non-anchored pattern, and for finding a character that is required later in the
133 match were failing in some cases. This didn't break the matching; it just
134 failed to optimize when it could. The way this is done has been re-implemented.
136 27. Fixed typo in error message for invalid (?R item (it said "(?p").
138 28. Added a new feature that provides some of the functionality that Perl
139 provides with (?{...}). The facility is termed a "callout". The way it is done
140 in PCRE is for the caller to provide an optional function, by setting
141 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
142 global variable. By default it is unset, which disables all calling out. To get
143 the function called, the regex must include (?C) at appropriate points. This
144 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
145 This provides a means of identifying different callout points. When PCRE
146 reaches such a point in the regex, if pcre_callout has been set, the external
147 function is called. It is provided with data in a structure called
148 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
149 matching continues; if it returns a non-zero value, the match at the current
150 point fails. However, backtracking will occur if possible. [This was changed
151 later and other features added - see item 49 below.]
153 29. pcretest is upgraded to test the callout functionality. It provides a
154 callout function that displays information. By default, it shows the start of
155 the match and the current position in the text. There are some new data escapes
156 to vary what happens:
158 \C+ in addition, show current contents of captured substrings
159 \C- do not supply a callout function
160 \C!n return 1 when callout number n is reached
161 \C!n!m return 1 when callout number n is reached for the mth time
163 30. If pcregrep was called with the -l option and just a single file name, it
164 output "<stdin>" if a match was found, instead of the file name.
166 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
167 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
168 pcre_exec(). This saves a malloc/free per call. The default value of
169 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
170 when configuring.
172 32. The default maximum size of a compiled pattern is 64K. There have been a
173 few cases of people hitting this limit. The code now uses macros to handle the
174 storing of links as offsets within the compiled pattern. It defaults to 2-byte
175 links, but this can be changed to 3 or 4 bytes by --with-link-size when
176 configuring. Tests 2 and 5 work only with 2-byte links because they output
177 debugging information about compiled patterns.
179 33. Internal code re-arrangements:
181 (a) Moved the debugging function for printing out a compiled regex into
182 its own source file (printint.c) and used #include to pull it into
183 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
184 separate copies.
186 (b) Defined the list of op-code names for debugging as a macro in
187 internal.h so that it is next to the definition of the opcodes.
189 (c) Defined a table of op-code lengths for simpler skipping along compiled
190 code. This is again a macro in internal.h so that it is next to the
191 definition of the opcodes.
193 34. Added support for recursive calls to individual subpatterns, along the
194 lines of Robin Houston's patch (but implemented somewhat differently).
196 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
197 allow it to read and process whole directories in Win32. This code was
198 contributed by Lionel Fourquaux; it has not been tested by me.
200 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
201 used to name a group. Names consist of alphanumerics and underscores, and must
202 be unique. Back references use the syntax (?P=name) and recursive calls use
203 (?P>name) which is a PCRE extension to the Python extension. Groups still have
204 numbers. The function pcre_fullinfo() can be used after compilation to extract
205 a name/number map. There are three relevant calls:
207 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
208 PCRE_INFO_NAMECOUNT yields the number of entries
209 PCRE_INFO_NAMETABLE yields a pointer to the map.
211 The map is a vector of fixed-size entries. The size of each entry depends on
212 the length of the longest name used. The first two bytes of each entry are the
213 group number, most significant byte first. There follows the corresponding
214 name, zero terminated. The names are in alphabetical order.
216 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
217 case instead of 255. Making it the same both with and without UTF-8 support
218 means that the same test output works with both.
220 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
221 calling malloc() with a zero argument.
223 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
224 optimization. I've improved things by keeping a bitmap of backreferences with
225 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
226 fact referenced, the optimization can be applied. It is unlikely that a
227 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
228 the match to follow \n) will appear inside brackets with a number greater than
229 31, but if it does, any back reference > 31 suppresses the optimization.
231 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
232 of disabling numbered capturing parentheses. Any opening parenthesis that is
233 not followed by ? behaves as if it were followed by ?: but named parentheses
234 can still be used for capturing (and they will acquire numbers in the usual
235 way).
237 41. Redesigned the return codes from the match() function into yes/no/error so
238 that errors can be passed back from deep inside the nested calls. A malloc
239 failure while inside a recursive subpattern call now causes the
240 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
242 42. It is now possible to set a limit on the number of times the match()
243 function is called in a call to pcre_exec(). This facility makes it possible to
244 limit the amount of recursion and backtracking, though not in a directly
245 obvious way, because the match() function is used in a number of different
246 circumstances. The count starts from zero for each position in the subject
247 string (for non-anchored patterns). The default limit is, for compatibility, a
248 large number, namely 10 000 000. You can change this in two ways:
250 (a) When configuring PCRE before making, you can use --with-match-limit=n
251 to set a default value for the compiled library.
253 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
254 a different value is set. See 45 below.
256 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
258 43. Added a new function pcre_config(int, void *) to enable run-time extraction
259 of things that can be changed at compile time. The first argument specifies
260 what is wanted and the second points to where the information is to be placed.
261 The current list of available information is:
265 The output is an integer that is set to one if UTF-8 support is available;
266 otherwise it is set to zero.
270 The output is an integer that it set to the value of the code that is used for
271 newline. It is either LF (10) or CR (13).
275 The output is an integer that contains the number of bytes used for internal
276 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
280 The output is an integer that contains the threshold above which the POSIX
281 interface uses malloc() for output vectors. See item 31 above.
285 The output is an unsigned integer that contains the default limit of the number
286 of match() calls in a pcre_exec() execution. See 42 above.
288 44. pcretest has been upgraded by the addition of the -C option. This causes it
289 to extract all the available output from the new pcre_config() function, and to
290 output it. The program then exits immediately.
292 45. A need has arisen to pass over additional data with calls to pcre_exec() in
293 order to support additional features. One way would have been to define
294 pcre_exec2() (for example) with extra arguments, but this would not have been
295 extensible, and would also have required all calls to the original function to
296 be mapped to the new one. Instead, I have chosen to extend the mechanism that
297 is used for passing in "extra" data from pcre_study().
299 The pcre_extra structure is now exposed and defined in pcre.h. It currently
300 contains the following fields:
302 flags a bitmap indicating which of the following fields are set
303 study_data opaque data from pcre_study()
304 match_limit a way of specifying a limit on match() calls for a specific
305 call to pcre_exec()
306 callout_data data for callouts (see 49 below)
308 The flag bits are also defined in pcre.h, and are
314 The pcre_study() function now returns one of these new pcre_extra blocks, with
315 the actual study data pointed to by the study_data field, and the
316 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
317 before. That is, this change is entirely upwards-compatible and requires no
318 change to existing code.
320 If you want to pass in additional data to pcre_exec(), you can either place it
321 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
322 block.
324 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
325 data string contains the escape sequence \M, pcretest calls pcre_exec() several
326 times with different match limits, until it finds the minimum value needed for
327 pcre_exec() to complete. The value is then output. This can be instructive; for
328 most simple matches the number is quite small, but for pathological cases it
329 gets very large very quickly.
331 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
332 returns the size of the data block pointed to by the study_data field in a
333 pcre_extra block, that is, the value that was passed as the argument to
334 pcre_malloc() when PCRE was getting memory in which to place the information
335 created by pcre_study(). The fourth argument should point to a size_t variable.
336 pcretest has been extended so that this information is shown after a successful
337 pcre_study() call when information about the compiled regex is being displayed.
339 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
340 because what follows is always an absolute path. (Later: it turns out that this
341 is more than cosmetic for MinGW, because it doesn't like empty path
342 components.)
344 49. Some changes have been made to the callout feature (see 28 above):
346 (i) A callout function now has three choices for what it returns:
348 0 => success, carry on matching
349 > 0 => failure at this point, but backtrack if possible
350 < 0 => serious error, return this value from pcre_exec()
352 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
353 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
354 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
355 use by callout functions. It will never be used by PCRE itself.
357 (ii) The pcre_extra structure (see 45 above) has a void * field called
358 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
359 pcre_callout_block structure has a field of the same name. The contents of
360 the field passed in the pcre_extra structure are passed to the callout
361 function in the corresponding field in the callout block. This makes it
362 easier to use the same callout-containing regex from multiple threads. For
363 testing, the pcretest program has a new data escape
365 \C*n pass the number n (may be negative) as callout_data
367 If the callout function in pcretest receives a non-zero value as
368 callout_data, it returns that value.
370 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
371 there were some redundant $(CFLAGS) in commands that are now specified as
372 $(LINK), which already includes $(CFLAGS).
374 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
375 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
376 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
377 one-byte characters throughout. Note that case-insensitive matching applies
378 only to characters whose values are less than 256. PCRE doesn't support the
379 notion of cases for higher-valued characters.
381 (i) A character class whose characters are all within 0-255 is handled as
382 a bit map, and the map is inverted for negative classes. Previously, a
383 character > 255 always failed to match such a class; however it should
384 match if the class was a negative one (e.g. [^ab]). This has been fixed.
386 (ii) A negated character class with a single character < 255 is coded as
387 "not this character" (OP_NOT). This wasn't working properly when the test
388 character was multibyte, either singly or repeated.
390 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
391 mode, for example: \x{100}{2,3}.
393 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
394 singly or repeated) now correctly test multibyte characters. However,
395 PCRE doesn't recognize any characters with values greater than 255 as
396 digits, spaces, or word characters. Such characters always match \D, \S,
397 and \W, and never match \d, \s, or \w.
399 (v) Classes may now contain characters and character ranges with values
400 greater than 255. For example: [ab\x{100}-\x{400}].
402 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
403 PCRE in UTF-8 mode.
405 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
406 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
407 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
408 value.)
410 53. The single man page has become too large. I have therefore split it up into
411 a number of separate man pages. These also give rise to individual HTML pages;
412 these are now put in a separate directory, and there is an index.html page that
413 lists them all. Some hyperlinking between the pages has been installed.
415 54. Added convenience functions for handling named capturing parentheses.
417 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
418 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
419 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
420 are faulted.
422 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
423 calling configure. These values are used when compiling the dftables.c program
424 which is run to generate the source of the default character tables. They
425 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
426 you will need to set these values.
428 57. Updated the building process for Windows DLL, as provided by Fred Cox.
431 Version 3.9 02-Jan-02
432 ---------------------
434 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
436 2. If --disable-static was given, the building process failed when trying to
437 build pcretest and pcregrep. (For some reason it was using libtool to compile
438 them, which is not right, as they aren't part of the library.)
441 Version 3.8 18-Dec-01
442 ---------------------
444 1. The experimental UTF-8 code was completely screwed up. It was packing the
445 bytes in the wrong order. How dumb can you get?
448 Version 3.7 29-Oct-01
449 ---------------------
451 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
452 This caused pcretest, when used on the test data, to segfault. Unfortunately,
453 this didn't happen under Solaris 8, where I normally test things.
455 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
456 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
457 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
458 this fix an hour or so after the initial 3.7 release.)
461 Version 3.6 23-Oct-01
462 ---------------------
464 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
465 offsets passed as NULL with zero offset count.
467 2. The config.guess and config.sub files had not been updated when I moved to
468 the latest autoconf.
471 Version 3.5 15-Aug-01
472 ---------------------
474 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
475 had been forgotten.
477 2. By using declared but undefined structures, we can avoid using "void"
478 definitions in pcre.h while keeping the internal definitions of the structures
479 private.
481 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
482 user point of view, this means that both static and shared libraries are built
483 by default, but this can be individually controlled. More of the work of
484 handling this static/shared cases is now inside libtool instead of PCRE's make
485 file.
487 4. The pcretest utility is now installed along with pcregrep because it is
488 useful for users (to test regexs) and by doing this, it automatically gets
489 relinked by libtool. The documentation has been turned into a man page, so
490 there are now .1, .txt, and .html versions in /doc.
492 5. Upgrades to pcregrep:
493 (i) Added long-form option names like gnu grep.
494 (ii) Added --help to list all options with an explanatory phrase.
495 (iii) Added -r, --recursive to recurse into sub-directories.
496 (iv) Added -f, --file to read patterns from a file.
498 6. pcre_exec() was referring to its "code" argument before testing that
499 argument for NULL (and giving an error if it was NULL).
501 7. Upgraded Makefile.in to allow for compiling in a different directory from
502 the source directory.
504 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
505 options bits, the pointer it was passed was to an int instead of to an unsigned
506 long int. This mattered only on 64-bit systems.
508 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
509 generated) instead of pcre.in, which it its source. Also made the same change
510 in several of the .c files.
512 10. A new release of gcc defines printf() as a macro, which broke pcretest
513 because it had an ifdef in the middle of a string argument for printf(). Fixed
514 by using separate calls to printf().
516 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
517 script, to force use of CR or LF instead of \n in the source. On non-Unix
518 systems, the value can be set in config.h.
520 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
521 absolute limit. Changed the text of the error message to make this clear, and
522 likewise updated the man page.
524 13. The limit of 99 on the number of capturing subpatterns has been removed.
525 The new limit is 65535, which I hope will not be a "real" limit.
528 Version 3.4 22-Aug-00
529 ---------------------
531 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
533 2. Diagnose condition (?(0) as an error instead of crashing on matching.
536 Version 3.3 01-Aug-00
537 ---------------------
539 1. If an octal character was given, but the value was greater than \377, it
540 was not getting masked to the least significant bits, as documented. This could
541 lead to crashes in some systems.
543 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
544 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
546 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
547 These just pass their arguments on to (pcre_free)(), but they are provided
548 because some uses of PCRE bind it to non-C systems that can call its functions,
549 but cannot call free() or pcre_free() directly.
551 4. Add "make test" as a synonym for "make check". Corrected some comments in
552 the Makefile.
554 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
555 Makefile.
557 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
558 command called pgrep for grepping around the active processes.
560 7. Added the beginnings of support for UTF-8 character strings.
562 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
563 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
564 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
565 out for the ar command.)
568 Version 3.2 12-May-00
569 ---------------------
571 This is purely a bug fixing release.
573 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
574 of ZA. This was just one example of several cases that could provoke this bug,
575 which was introduced by change 9 of version 2.00. The code for breaking
576 infinite loops after an iteration that matches an empty string was't working
577 correctly.
579 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
580 when matched against abbab (for example). After matching an empty string, it
581 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
582 caused it to match further down the string than it should.
584 3. The code contained an inclusion of sys/types.h. It isn't clear why this
585 was there because it doesn't seem to be needed, and it causes trouble on some
586 systems, as it is not a Standard C header. It has been removed.
588 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
589 were reported on the Macintosh. The changes were from
591 while ((c = *(++ptr)) != 0 && c != '\n');
592 to
593 while ((c = *(++ptr)) != 0 && c != '\n') ;
595 Totally extraordinary, but if that's what it takes...
597 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
598 available. Added HAVE_BCOPY and an autoconf test for it; if neither
599 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
600 assumes the way PCRE uses memmove() (always moving upwards).
602 6. PCRE is being used in one environment where strchr() is not available. There
603 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
604 faster code anyway.
607 Version 3.1 09-Feb-00
608 ---------------------
610 The only change in this release is the fixing of some bugs in Makefile.in for
611 the "install" target:
613 (1) It was failing to install pcreposix.h.
615 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
618 Version 3.0 01-Feb-00
619 ---------------------
621 1. Add support for the /+ modifier to perltest (to output $` like it does in
622 pcretest).
624 2. Add support for the /g modifier to perltest.
626 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
627 matches null strings.
629 4. Fix perltest so that it doesn't do unwanted things when fed an empty
630 pattern. Perl treats empty patterns specially - it reuses the most recent
631 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
632 effect.
634 5. The POSIX interface was broken in that it was just handing over the POSIX
635 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
636 required a bigger vector, with some working space on the end. This means that
637 the POSIX wrapper now has to get and free some memory, and copy the results.
639 6. Added some simple autoconf support, placing the test data and the
640 documentation in separate directories, re-organizing some of the
641 information files, and making it build pcre-config (a GNU standard). Also added
642 libtool support for building PCRE as a shared library, which is now the
643 default.
645 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
646 09 are not valid octal constants. Single digits will be used for minor values
647 less than 10.
649 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
650 existing programs that set these in the POSIX interface can use PCRE without
651 modification.
653 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
654 return all that pcre_info() returns, plus additional data. The pcre_info()
655 function is retained for compatibility, but is considered to be obsolete.
657 10. Added experimental recursion feature (?R) to handle one common case that
658 Perl 5.6 will be able to do with (?p{...}).
660 11. Added support for POSIX character classes like [:alpha:], which Perl is
661 adopting.
664 Version 2.08 31-Aug-99
665 ----------------------
667 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
668 trying to match at the startoffset position, but instead was moving forward to
669 the next newline as if a previous match had failed.
671 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
672 and could get into a loop if a null string was matched other than at the start
673 of the subject.
675 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
676 be distinguished at compile time, and for completeness also added PCRE_DATE.
678 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
679 in GnuWin32 environments.
682 Version 2.07 29-Jul-99
683 ----------------------
685 1. The documentation is now supplied in plain text form and HTML as well as in
686 the form of man page sources.
688 2. C++ compilers don't like assigning (void *) values to other pointer types.
689 In particular this affects malloc(). Although there is no problem in Standard
690 C, I've put in casts to keep C++ compilers happy.
692 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
693 should be (const char *).
695 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
696 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
697 However, I haven't made this a standard facility. The documentation doesn't
698 mention it, and the Makefile doesn't support it.
700 5. The Makefile now contains an "install" target, with editable destinations at
701 the top of the file. The pcretest program is not installed.
703 6. pgrep -V now gives the PCRE version number and date.
705 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
706 causing the entire string to be ignored, instead of just the last character.
708 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
709 non-matching string, it can take a very, very long time, even for strings of
710 quite modest length, because of the nested recursion. PCRE now does better in
711 some of these cases. It does this by remembering the last required literal
712 character in the pattern, and pre-searching the subject to ensure it is present
713 before running the real match. In other words, it applies a heuristic to detect
714 some types of certain failure quickly, and in the above example, if presented
715 with a string that has no trailing " it gives "no match" very quickly.
717 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
718 other alternatives are tried instead.
721 Version 2.06 09-Jun-99
722 ----------------------
724 1. Change pcretest's output for amount of store used to show just the code
725 space, because the remainder (the data block) varies in size between 32-bit and
726 64-bit systems.
728 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
729 start matching at. This allows lookbehinds to work when searching for multiple
730 occurrences in a string.
732 3. Added additional options to pcretest for testing multiple occurrences:
734 /+ outputs the rest of the string that follows a match
735 /g loops for multiple occurrences, using the new startoffset argument
736 /G loops for multiple occurrences by passing an incremented pointer
738 4. PCRE wasn't doing the "first character" optimization for patterns starting
739 with \b or \B, though it was doing it for other lookbehind assertions. That is,
740 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
741 the letter 'x'. On long subject strings, this gives a significant speed-up.
744 Version 2.05 21-Apr-99
745 ----------------------
747 1. Changed the type of magic_number from int to long int so that it works
748 properly on 16-bit systems.
750 2. Fixed a bug which caused patterns starting with .* not to work correctly
751 when the subject string contained newline characters. PCRE was assuming
752 anchoring for such patterns in all cases, which is not correct because .* will
753 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
754 DOTALL is set at top level; otherwise it knows that patterns starting with .*
755 must be retried after every newline in the subject.
758 Version 2.04 18-Feb-99
759 ----------------------
761 1. For parenthesized subpatterns with repeats whose minimum was zero, the
762 computation of the store needed to hold the pattern was incorrect (too large).
763 If such patterns were nested a few deep, this could multiply and become a real
764 problem.
766 2. Added /M option to pcretest to show the memory requirement of a specific
767 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
769 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
770 compiled in such a way that the backtracking after subsequent failure was
771 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
772 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
775 Version 2.03 02-Feb-99
776 ----------------------
778 1. Fixed typo and small mistake in man page.
780 2. Added 4th condition (GPL supersedes if conflict) and created separate
781 LICENCE file containing the conditions.
783 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
784 Perl, that is the internal \ allows the delimiter to be included in the
785 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
786 the final delimiter, add \ to the end of the pattern (to test the error).
788 4. Added the convenience functions for extracting substrings after a successful
789 match. Updated pcretest to make it able to test these functions.
792 Version 2.02 14-Jan-99
793 ----------------------
795 1. Initialized the working variables associated with each extraction so that
796 their saving and restoring doesn't refer to uninitialized store.
798 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
799 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
800 fix the problem.
802 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
803 calls, and wasn't printing the correct value for compiling calls. Increased the
804 default value of LOOPREPEAT, and the number of significant figures in the
805 times.
807 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
809 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
810 a building problem on Windows NT with a FAT file system.
813 Version 2.01 21-Oct-98
814 ----------------------
816 1. Changed the API for pcre_compile() to allow for the provision of a pointer
817 to character tables built by pcre_maketables() in the current locale. If NULL
818 is passed, the default tables are used.
821 Version 2.00 24-Sep-98
822 ----------------------
824 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
825 it any more.
827 2. Allow quantification of (?>) groups, and make it work correctly.
829 3. The first character computation wasn't working for (?>) groups.
831 4. Correct the implementation of \Z (it is permitted to match on the \n at the
832 end of the subject) and add 5.005's \z, which really does match only at the
833 very end of the subject.
835 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
837 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
838 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
839 localized options. All options to pcre_study() were also removed.
841 7. Add other new features from 5.005:
843 $(?<= positive lookbehind
844 $(?<! negative lookbehind
845 (?imsx-imsx) added the unsetting capability
846 such a setting is global if at outer level; local otherwise
847 (?imsx-imsx:) non-capturing groups with option setting
848 (?(cond)re|re) conditional pattern matching
850 A backreference to itself in a repeated group matches the previous
851 captured string.
853 8. General tidying up of studying (both automatic and via "study")
854 consequential on the addition of new assertions.
856 9. As in 5.005, unlimited repeated groups that could match an empty substring
857 are no longer faulted at compile time. Instead, the loop is forcibly broken at
858 runtime if any iteration does actually match an empty substring.
860 10. Include the RunTest script in the distribution.
862 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
863 discrepancies, some of which were old and were also with respect to 5.004. They
864 have now been fixed.
867 Version 1.09 28-Apr-98
868 ----------------------
870 1. A negated single character class followed by a quantifier with a minimum
871 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
872 program crashes, or just wrong answers. This did not apply to negated classes
873 containing more than one character, or to minima other than one.
876 Version 1.08 27-Mar-98
877 ----------------------
879 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
881 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
882 latter must appear before anything that relies on it in the pattern.
885 Version 1.07 16-Feb-98
886 ----------------------
888 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
889 repeat of a potentially empty string).
892 Version 1.06 23-Jan-98
893 ----------------------
895 1. Added Markus Oberhumer's little patches for C++.
897 2. Literal strings longer than 255 characters were broken.
900 Version 1.05 23-Dec-97
901 ----------------------
903 1. Negated character classes containing more than one character were failing if
904 PCRE_CASELESS was set at run time.
907 Version 1.04 19-Dec-97
908 ----------------------
910 1. Corrected the man page, where some "const" qualifiers had been omitted.
912 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
913 input syntax.
915 3. Fixed memory leak which occurred when a regex with back references was
916 matched with an offsets vector that wasn't big enough. The temporary memory
917 that is used in this case wasn't being freed if the match failed.
919 4. Tidied pcretest to ensure it frees memory that it gets.
921 5. Temporary memory was being obtained in the case where the passed offsets
922 vector was exactly big enough.
924 6. Corrected definition of offsetof() from change 5 below.
926 7. I had screwed up change 6 below and broken the rules for the use of
927 setjmp(). Now fixed.
930 Version 1.03 18-Dec-97
931 ----------------------
933 1. A erroneous regex with a missing opening parenthesis was correctly
934 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
935 on some systems.
937 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
938 it was reported that one broken compiler failed on the former because "code" is
939 also an independent variable.
941 3. The erroneous regex a[]b caused an array overrun reference.
943 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
944 fail on data ending with that character. (It was going on too far, and checking
945 the next character, typically a binary zero.) This was specific to the
946 optimized code for single-character negative classes.
948 5. Added a contributed patch from the TIN world which does the following:
950 + Add an undef for memmove, in case the the system defines a macro for it.
952 + Add a definition of offsetof(), in case there isn't one. (I don't know
953 the reason behind this - offsetof() is part of the ANSI standard - but
954 it does no harm).
956 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
957 most of the places where whitespace preceded '#'. I have given up and
958 allowed the remaining 2 cases to be at the margin.
960 + Rename some variables in pcre to eliminate shadowing. This seems very
961 pedantic, but does no harm, of course.
963 6. Moved the call to setjmp() into its own function, to get rid of warnings
964 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
966 7. Constructs such as \d{8,} were compiling into the equivalent of
967 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
968 outcome, but in this particular case used more store than had been allocated,
969 which caused the bug to be discovered because it threw up an internal error.
971 8. The debugging code in both pcre and pcretest for outputting the compiled
972 form of a regex was going wrong in the case of back references followed by
973 curly-bracketed repeats.
976 Version 1.02 12-Dec-97
977 ----------------------
979 1. Typos in pcre.3 and comments in the source fixed.
981 2. Applied a contributed patch to get rid of places where it used to remove
982 'const' from variables, and fixed some signed/unsigned and uninitialized
983 variable warnings.
985 3. Added the "runtest" target to Makefile.
987 4. Set default compiler flag to -O2 rather than just -O.
990 Version 1.01 19-Nov-97
991 ----------------------
993 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
994 like /([ab]*)*/, that is, for classes with more than one character in them.
996 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
997 as /((?>a*))*/ (a PCRE_EXTRA facility).
1000 Version 1.00 18-Nov-97
1001 ----------------------
1003 1. Added compile-time macros to support systems such as SunOS4 which don't have
1004 memmove() or strerror() but have other things that can be used instead.
1006 2. Arranged that "make clean" removes the executables.
1009 Version 0.99 27-Oct-97
1010 ----------------------
1012 1. Fixed bug in code for optimizing classes with only one character. It was
1013 initializing a 32-byte map regardless, which could cause it to run off the end
1014 of the memory it had got.
1016 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1019 Version 0.98 22-Oct-97
1020 ----------------------
1022 1. Fixed bug in code for handling temporary memory usage when there are more
1023 back references than supplied space in the ovector. This could cause segfaults.
1026 Version 0.97 21-Oct-97
1027 ----------------------
1029 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1031 2. Optimized negated single characters not to use a bit map.
1033 3. Brought error texts together as macro definitions; clarified some of them;
1034 fixed one that was wrong - it said "range out of order" when it meant "invalid
1035 escape sequence".
1037 4. Changed some char * arguments to const char *.
1039 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1041 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1042 pcretest.
1045 Version 0.96 16-Oct-97
1046 ----------------------
1048 1. Added a simple "pgrep" utility to the distribution.
1050 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1051 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1052 where "ddd" means "one or more decimal digits".
1054 3. Fixed serious bug. If a pattern had a back reference, but the call to
1055 pcre_exec() didn't supply a large enough ovector to record the related
1056 identifying subpattern, the match always failed. PCRE now remembers the number
1057 of the largest back reference, and gets some temporary memory in which to save
1058 the offsets during matching if necessary, in order to ensure that
1059 backreferences always work.
1061 4. Increased the compatibility with Perl in a number of ways:
1063 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1064 to request this handling. The option can be set at compile or exec time.
1066 (b) $ matches before a terminating newline by default; an option
1067 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
1068 mode). The option can be set at compile or exec time.
1070 (c) The handling of \ followed by a digit other than 0 is now supposed to be
1071 the same as Perl's. If the decimal number it represents is less than 10
1072 or there aren't that many previous left capturing parentheses, an octal
1073 escape is read. Inside a character class, it's always an octal escape,
1074 even if it is a single digit.
1076 (d) An escaped but undefined alphabetic character is taken as a literal,
1077 unless PCRE_EXTRA is set. Currently this just reserves the remaining
1078 escapes.
1080 (e) {0} is now permitted. (The previous item is removed from the compiled
1081 pattern).
1083 5. Changed all the names of code files so that the basic parts are no longer
1084 than 10 characters, and abolished the teeny "globals.c" file.
1086 6. Changed the handling of character classes; they are now done with a 32-byte
1087 bit map always.
1089 7. Added the -d and /D options to pcretest to make it possible to look at the
1090 internals of compilation without having to recompile pcre.
1093 Version 0.95 23-Sep-97
1094 ----------------------
1096 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
1097 \x20 at the start of a run of normal characters. These were being treated as
1098 real characters, instead of the source characters being re-checked.
1101 Version 0.94 18-Sep-97
1102 ----------------------
1104 1. The functions are now thread-safe, with the caveat that the global variables
1105 containing pointers to malloc() and free() or alternative functions are the
1106 same for all threads.
1108 2. Get pcre_study() to generate a bitmap of initial characters for non-
1109 anchored patterns when this is possible, and use it if passed to pcre_exec().
1112 Version 0.93 15-Sep-97
1113 ----------------------
1115 1. /(b)|(:+)/ was computing an incorrect first character.
1117 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
1118 but not actually doing anything yet.
1120 3. Treat "-" characters in classes that cannot be part of ranges as literals,
1121 as Perl does (e.g. [-az] or [az-]).
1123 4. Set the anchored flag if a branch starts with .* or .*? because that tests
1124 all possible positions.
1126 5. Split up into different modules to avoid including unneeded functions in a
1127 compiled binary. However, compile and exec are still in one module. The "study"
1128 function is split off.
1130 6. The character tables are now in a separate module whose source is generated
1131 by an auxiliary program - but can then be edited by hand if required. There are
1132 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
1133 toupper() in the code.
1135 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
1136 make them global. Abolish the function for setting them, as the caller can now
1137 set them directly.
1140 Version 0.92 11-Sep-97
1141 ----------------------
1143 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
1144 (e.g. /a{1,3}/) was broken (I mis-optimized it).
1146 2. Caseless matching was not working in character classes if the characters in
1147 the pattern were in upper case.
1149 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
1151 4. Make PCRE_ANCHORED public and accept as a compile option.
1153 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
1154 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
1155 pass them.
1157 6. Give an error if bad option bits passed at compile or run time.
1159 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
1160 pcretest to cause it to pass that flag.
1162 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
1163 options, and the first character, if set.
1165 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
1168 Version 0.91 10-Sep-97
1169 ----------------------
1171 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
1172 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
1174 2. PCRE was looping on encountering an indefinitely repeated back reference to
1175 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
1176 Perl does - treats the match as successful.
1178 ****

  ViewVC Help
Powered by ViewVC 1.1.5