ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 73 - (show annotations)
Sat Feb 24 21:40:30 2007 UTC (8 years, 9 months ago) by nigel
File size: 64875 byte(s)
Load pcre-4.5 into code/trunk.
1 ChangeLog for PCRE
2 ------------------
4 Version 4.5 01-Dec-03
5 ---------------------
7 1. There has been some re-arrangement of the code for the match() function so
8 that it can be compiled in a version that does not call itself recursively.
9 Instead, it keeps those local variables that need separate instances for
10 each "recursion" in a frame on the heap, and gets/frees frames whenever it
11 needs to "recurse". Keeping track of where control must go is done by means
12 of setjmp/longjmp. The whole thing is implemented by a set of macros that
13 hide most of the details from the main code, and operates only if
14 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
15 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
16 operating.
18 To make it easier for callers to provide specially tailored get/free
19 functions for this usage, two new functions, pcre_stack_malloc, and
20 pcre_stack_free, are used. They are always called in strict stacking order,
21 and the size of block requested is always the same.
23 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
24 PCRE has been compiled to use the stack or the heap for recursion. The
25 -C option of pcretest uses this to show which version is compiled.
27 A new data escape \S, is added to pcretest; it causes the amounts of store
28 obtained and freed by both kinds of malloc/free at match time to be added
29 to the output.
31 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
32 what's available on my current Linux desktop machine.
34 3. When matching a UTF-8 string, the test for a valid string at the start has
35 been extended. If start_offset is not zero, PCRE now checks that it points
36 to a byte that is the start of a UTF-8 character. If not, it returns
37 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
38 this is necessary because there may be backward assertions in the pattern.
39 When matching the same subject several times, it may save resources to use
40 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
42 4. The code for checking the validity of UTF-8 strings has been tightened so
43 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
44 containing "overlong sequences".
46 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
47 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
48 should have been "&", but it just so happened that all the cases this let
49 through by mistake were picked up later in the function.
51 6. I had used a variable called "isblank" - this is a C99 function, causing
52 some compilers to warn. To avoid this, I renamed it (as "blankclass").
54 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
55 prompting; (b) run "./pcretest /dev/null" at the start of the test script
56 so the version is shown; (c) stop "make test" echoing "./RunTest".
58 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
60 9. The prototype for memmove() for systems that don't have it was using
61 size_t, but the inclusion of the header that defines size_t was later. I've
62 moved the #includes for the C headers earlier to avoid this.
64 10. Added some adjustments to the code to make it easier to compiler on certain
65 special systems:
67 (a) Some "const" qualifiers were missing.
68 (b) Added the macro EXPORT before all exported functions; by default this
69 is defined to be empty.
70 (c) Changed the dftables auxiliary program (that builds chartables.c) so
71 that it reads its output file name as an argument instead of writing
72 to the standard output and assuming this can be redirected.
74 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
75 class containing characters with values greater than 255, PCRE compilation
76 went into a loop.
78 12. A recursive reference to a subpattern that was within another subpattern
79 that had a minimum quantifier of zero caused PCRE to crash. For example,
80 (x(y(?2))z)? provoked this bug with a subject that got as far as the
81 recursion. If the recursively-called subpattern itself had a zero repeat,
82 that was OK.
84 13. In pcretest, the buffer for reading a data line was set at 30K, but the
85 buffer into which it was copied (for escape processing) was still set at
86 1024, so long lines caused crashes.
88 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
89 "internal error: code overflow...". This applied to any character class
90 that was followed by a possessive quantifier.
92 15. Modified the Makefile to add libpcre.la as a prerequisite for
93 libpcreposix.la because I was told this is needed for a parallel build to
94 work.
96 16. If a pattern that contained .* following optional items at the start was
97 studied, the wrong optimizing data was generated, leading to matching
98 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
99 matching string must start with a or b or c. The correct conclusion for
100 this pattern is that a match can start with any character.
103 Version 4.4 13-Aug-03
104 ---------------------
106 1. In UTF-8 mode, a character class containing characters with values between
107 127 and 255 was not handled correctly if the compiled pattern was studied.
108 In fixing this, I have also improved the studying algorithm for such
109 classes (slightly).
111 2. Three internal functions had redundant arguments passed to them. Removal
112 might give a very teeny performance improvement.
114 3. Documentation bug: the value of the capture_top field in a callout is *one
115 more than* the number of the hightest numbered captured substring.
117 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
118 in incorrectly linking with a previously installed version. They now link
119 explicitly with libpcre.la.
121 5. configure.in no longer needs to recognize Cygwin specially.
123 6. A problem in pcre.in for Windows platforms is fixed.
125 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
126 pcretest, it used to include the size of the study block as part of its
127 output. Unfortunately, the structure contains a field that has a different
128 size on different hardware architectures. This meant that the tests that
129 showed this size failed. As the block is currently always of a fixed size,
130 this information isn't actually particularly useful in pcretest output, so
131 I have just removed it.
133 8. Three pre-processor statements accidentally did not start in column 1.
134 Sadly, there are *still* compilers around that complain, even though
135 standard C has not required this for well over a decade. Sigh.
137 9. In pcretest, the code for checking callouts passed small integers in the
138 callout_data field, which is a void * field. However, some picky compilers
139 complained about the casts involved for this on 64-bit systems. Now
140 pcretest passes the address of the small integer instead, which should get
141 rid of the warnings.
143 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
144 both compile and run time, and gives an error if an invalid UTF-8 sequence
145 is found. There is a option for disabling this check in cases where the
146 string is known to be correct and/or the maximum performance is wanted.
148 11. In response to a bug report, I changed one line in Makefile.in from
150 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
151 to
152 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
154 to look similar to other lines, but I have no way of telling whether this
155 is the right thing to do, as I do not use Windows. No doubt I'll get told
156 if it's wrong...
159 Version 4.3 21-May-03
160 ---------------------
162 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
163 Makefile.
165 2. Some refactoring to improve the quality of the code:
167 (i) The utf8_table... variables are now declared "const".
169 (ii) The code for \cx, which used the "case flipping" table to upper case
170 lower case letters, now just substracts 32. This is ASCII-specific,
171 but the whole concept of \cx is ASCII-specific, so it seems
172 reasonable.
174 (iii) PCRE was using its character types table to recognize decimal and
175 hexadecimal digits in the pattern. This is silly, because it handles
176 only 0-9, a-f, and A-F, but the character types table is locale-
177 specific, which means strange things might happen. A private
178 table is now used for this - though it costs 256 bytes, a table is
179 much faster than multiple explicit tests. Of course, the standard
180 character types table is still used for matching digits in subject
181 strings against \d.
183 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
184 ending in _t are). So I've renamed it as ESC_tee.
186 3. The first argument for regexec() in the POSIX wrapper should have been
187 defined as "const".
189 4. Changed pcretest to use malloc() for its buffers so that they can be
190 Electric Fenced for debugging.
192 5. There were several places in the code where, in UTF-8 mode, PCRE would try
193 to read one or more bytes before the start of the subject string. Often this
194 had no effect on PCRE's behaviour, but in some circumstances it could
195 provoke a segmentation fault.
197 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
198 to try to read one or more bytes before the start of the subject string.
200 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
201 UTF-8 support could misbehave in various ways if the subject string
202 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
203 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
204 back over UTF-8 characters.)
207 Version 4.2 14-Apr-03
208 ---------------------
210 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
212 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
213 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
214 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
215 [WIN_PREFIX]: new variable, "cyg" for Cygwin
216 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
218 Note: automatic setting of the BUILD variables is not yet working
219 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
220 compile-time but not at link-time
221 [LINK]: use for linking executables only
222 make different versions for Windows and non-Windows
223 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
224 libraries
225 [LINK_FOR_BUILD]: new variable
226 [OBJEXT]: use throughout
227 [EXEEXT]: use throughout
228 <winshared>: new target
229 <wininstall>: new target
230 <dftables.o>: use native compiler
231 <dftables>: use native linker
232 <install>: handle Windows platform correctly
233 <clean>: ditto
234 <check>: ditto
235 copy DLL to top builddir before testing
237 As part of these changes, -no-undefined was removed again. This was reported
238 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
239 in any case.
241 3. Some tidies to get rid of compiler warnings:
243 . In the match_data structure, match_limit was an unsigned long int, whereas
244 match_call_count was an int. I've made them both unsigned long ints.
246 . In pcretest the fact that a const uschar * doesn't automatically cast to
247 a void * provoked a warning.
249 . Turning on some more compiler warnings threw up some "shadow" variables
250 and a few more missing casts.
252 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
253 option, a class that contained a single character with a value between 128
254 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
256 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
257 option, a class that contained several characters, but with at least one
258 whose value was between 128 and 255 caused PCRE to crash.
261 Version 4.1 12-Mar-03
262 ---------------------
264 1. Compiling with gcc -pedantic found a couple of places where casts were
265 needed, and a string in dftables.c that was longer than standard compilers are
266 required to support.
268 2. Compiling with Sun's compiler found a few more places where the code could
269 be tidied up in order to avoid warnings.
271 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
272 first of these names is deprecated in the latest Autoconf in favour of the name
273 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
274 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
275 analogy I have changed it to CFLAGS_FOR_BUILD.
277 4. Added -no-undefined to the linking command in the Makefile, because this is
278 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
279 linking step for the pcreposix library.
281 5. PCRE was failing to diagnose the case of two named groups with the same
282 name.
284 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
285 literal character that is needed in the subject for a match, and scans along to
286 ensure that it is present before embarking on the full matching process. This
287 saves time in cases of nested unlimited repeats that are never going to match.
288 Problem: the scan can take a lot of time if the subject is very long (e.g.
289 megabytes), thus penalizing straightforward matches. It is now done only if the
290 amount of subject to be scanned is less than 1000 bytes.
292 7. A lesser problem with the same optimization is that it was recording the
293 first character of an anchored pattern as "needed", thus provoking a search
294 right along the subject, even when the first match of the pattern was going to
295 fail. The "needed" character is now not set for anchored patterns, unless it
296 follows something in the pattern that is of non-fixed length. Thus, it still
297 fulfils its original purpose of finding quick non-matches in cases of nested
298 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
301 Version 4.0 17-Feb-03
302 ---------------------
304 1. If a comment in an extended regex that started immediately after a meta-item
305 extended to the end of string, PCRE compiled incorrect data. This could lead to
306 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
308 2. Moved to autoconf 2.53 and libtool 1.4.2.
310 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
311 the special perltest8 script is no longer needed - all the tests can be run
312 from a single perltest script.
314 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
315 by \s. It has now been removed in PCRE. This means it isn't recognized as
316 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
317 class [:space:] *does* include VT, thereby creating a mess.
319 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
320 space and tab.
322 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
323 its new features into the main test script, reducing the number of scripts.
325 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
326 were backward compatible, and made the (?i) apply to the whole pattern, as if
327 /i were given. Now it behaves more logically, and applies the option setting
328 only to what follows. PCRE has been changed to follow suit. However, if it
329 finds options settings right at the start of the pattern, it extracts them into
330 the global options, as before. Thus, they show up in the info data.
332 8. Added support for the \Q...\E escape sequence. Characters in between are
333 treated as literals. This is slightly different from Perl in that $ and @ are
334 also handled as literals inside the quotes. In Perl, they will cause variable
335 interpolation. Note the following examples:
337 Pattern PCRE matches Perl matches
339 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
340 \Qabc\$xyz\E abc\$xyz abc\$xyz
341 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
343 For compatibility with Perl, \Q...\E sequences are recognized inside character
344 classes as well as outside them.
346 9. Re-organized 3 code statements in pcretest to avoid "overflow in
347 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
348 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
349 signed/unsigned warnings.
351 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
352 option for pcretest, so I've replaced it by a simple function that does just
353 that job.
355 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
356 "pcregrep -".
358 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
359 Java package. This provides some syntactic sugar for simple cases of what my
360 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
361 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
362 item, you can use this simplified notation. Note that only makes sense with
363 greedy quantifiers. Consequently, the use of the possessive quantifier forces
364 greediness, whatever the setting of the PCRE_UNGREEDY option.
366 13. A change of greediness default within a pattern was not taking effect at
367 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
368 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
369 was abstracted outside.
371 14. PCRE now supports the \G assertion. It is true when the current matching
372 position is at the start point of the match. This differs from \A when the
373 starting offset is non-zero. Used with the /g option of pcretest (or similar
374 code), it works in the same way as it does for Perl's /g option. If all
375 alternatives of a regex begin with \G, the expression is anchored to the start
376 match position, and the "anchored" flag is set in the compiled expression.
378 15. Some bugs concerning the handling of certain option changes within patterns
379 have been fixed. These applied to options other than (?ims). For example,
380 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
381 been the other way round. Some of this was related to change 7 above.
383 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
384 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
385 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
386 POSIX classes only within a class (e.g. /[[:alpha:]]/).
388 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
389 mode. Unlike ".", it always matches newline, whatever the setting of
390 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
391 assertions. Perl allows it, but it doesn't (in general) work because it can't
392 calculate the length of the lookbehind. At least, that's the case for Perl
393 5.8.0 - I've been told they are going to document that it doesn't work in
394 future.
396 18. Added an error diagnosis for escapes that PCRE does not support: these are
397 \L, \l, \N, \P, \p, \U, \u, and \X.
399 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
400 reading past the end of the pattern in cases such as /[abcd/.
402 20. PCRE was getting more memory than necessary for patterns with classes that
403 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
405 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
406 compiling PCRE for use with Virtual Pascal.
408 22. Small fix to the Makefile to make it work properly if the build is done
409 outside the source tree.
411 23. Added a new extension: a condition to go with recursion. If a conditional
412 subpattern starts with (?(R) the "true" branch is used if recursion has
413 happened, whereas the "false" branch is used only at the top level.
415 24. When there was a very long string of literal characters (over 255 bytes
416 without UTF support, over 250 bytes with UTF support), the computation of how
417 much memory was required could be incorrect, leading to segfaults or other
418 strange effects.
420 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
421 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
422 there was a subsequent back reference to those brackets. This meant that, for
423 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
424 possible to check for precisely this case. All we can do is abandon the
425 optimization if .* occurs inside capturing brackets when there are any back
426 references whatsoever. (See below for a better fix that came later.)
428 26. The handling of the optimization for finding the first character of a
429 non-anchored pattern, and for finding a character that is required later in the
430 match were failing in some cases. This didn't break the matching; it just
431 failed to optimize when it could. The way this is done has been re-implemented.
433 27. Fixed typo in error message for invalid (?R item (it said "(?p").
435 28. Added a new feature that provides some of the functionality that Perl
436 provides with (?{...}). The facility is termed a "callout". The way it is done
437 in PCRE is for the caller to provide an optional function, by setting
438 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
439 global variable. By default it is unset, which disables all calling out. To get
440 the function called, the regex must include (?C) at appropriate points. This
441 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
442 This provides a means of identifying different callout points. When PCRE
443 reaches such a point in the regex, if pcre_callout has been set, the external
444 function is called. It is provided with data in a structure called
445 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
446 matching continues; if it returns a non-zero value, the match at the current
447 point fails. However, backtracking will occur if possible. [This was changed
448 later and other features added - see item 49 below.]
450 29. pcretest is upgraded to test the callout functionality. It provides a
451 callout function that displays information. By default, it shows the start of
452 the match and the current position in the text. There are some new data escapes
453 to vary what happens:
455 \C+ in addition, show current contents of captured substrings
456 \C- do not supply a callout function
457 \C!n return 1 when callout number n is reached
458 \C!n!m return 1 when callout number n is reached for the mth time
460 30. If pcregrep was called with the -l option and just a single file name, it
461 output "<stdin>" if a match was found, instead of the file name.
463 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
464 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
465 pcre_exec(). This saves a malloc/free per call. The default value of
466 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
467 when configuring.
469 32. The default maximum size of a compiled pattern is 64K. There have been a
470 few cases of people hitting this limit. The code now uses macros to handle the
471 storing of links as offsets within the compiled pattern. It defaults to 2-byte
472 links, but this can be changed to 3 or 4 bytes by --with-link-size when
473 configuring. Tests 2 and 5 work only with 2-byte links because they output
474 debugging information about compiled patterns.
476 33. Internal code re-arrangements:
478 (a) Moved the debugging function for printing out a compiled regex into
479 its own source file (printint.c) and used #include to pull it into
480 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
481 separate copies.
483 (b) Defined the list of op-code names for debugging as a macro in
484 internal.h so that it is next to the definition of the opcodes.
486 (c) Defined a table of op-code lengths for simpler skipping along compiled
487 code. This is again a macro in internal.h so that it is next to the
488 definition of the opcodes.
490 34. Added support for recursive calls to individual subpatterns, along the
491 lines of Robin Houston's patch (but implemented somewhat differently).
493 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
494 allow it to read and process whole directories in Win32. This code was
495 contributed by Lionel Fourquaux; it has not been tested by me.
497 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
498 used to name a group. Names consist of alphanumerics and underscores, and must
499 be unique. Back references use the syntax (?P=name) and recursive calls use
500 (?P>name) which is a PCRE extension to the Python extension. Groups still have
501 numbers. The function pcre_fullinfo() can be used after compilation to extract
502 a name/number map. There are three relevant calls:
504 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
505 PCRE_INFO_NAMECOUNT yields the number of entries
506 PCRE_INFO_NAMETABLE yields a pointer to the map.
508 The map is a vector of fixed-size entries. The size of each entry depends on
509 the length of the longest name used. The first two bytes of each entry are the
510 group number, most significant byte first. There follows the corresponding
511 name, zero terminated. The names are in alphabetical order.
513 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
514 case instead of 255. Making it the same both with and without UTF-8 support
515 means that the same test output works with both.
517 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
518 calling malloc() with a zero argument.
520 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
521 optimization. I've improved things by keeping a bitmap of backreferences with
522 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
523 fact referenced, the optimization can be applied. It is unlikely that a
524 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
525 the match to follow \n) will appear inside brackets with a number greater than
526 31, but if it does, any back reference > 31 suppresses the optimization.
528 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
529 of disabling numbered capturing parentheses. Any opening parenthesis that is
530 not followed by ? behaves as if it were followed by ?: but named parentheses
531 can still be used for capturing (and they will acquire numbers in the usual
532 way).
534 41. Redesigned the return codes from the match() function into yes/no/error so
535 that errors can be passed back from deep inside the nested calls. A malloc
536 failure while inside a recursive subpattern call now causes the
537 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
539 42. It is now possible to set a limit on the number of times the match()
540 function is called in a call to pcre_exec(). This facility makes it possible to
541 limit the amount of recursion and backtracking, though not in a directly
542 obvious way, because the match() function is used in a number of different
543 circumstances. The count starts from zero for each position in the subject
544 string (for non-anchored patterns). The default limit is, for compatibility, a
545 large number, namely 10 000 000. You can change this in two ways:
547 (a) When configuring PCRE before making, you can use --with-match-limit=n
548 to set a default value for the compiled library.
550 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
551 a different value is set. See 45 below.
553 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
555 43. Added a new function pcre_config(int, void *) to enable run-time extraction
556 of things that can be changed at compile time. The first argument specifies
557 what is wanted and the second points to where the information is to be placed.
558 The current list of available information is:
562 The output is an integer that is set to one if UTF-8 support is available;
563 otherwise it is set to zero.
567 The output is an integer that it set to the value of the code that is used for
568 newline. It is either LF (10) or CR (13).
572 The output is an integer that contains the number of bytes used for internal
573 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
577 The output is an integer that contains the threshold above which the POSIX
578 interface uses malloc() for output vectors. See item 31 above.
582 The output is an unsigned integer that contains the default limit of the number
583 of match() calls in a pcre_exec() execution. See 42 above.
585 44. pcretest has been upgraded by the addition of the -C option. This causes it
586 to extract all the available output from the new pcre_config() function, and to
587 output it. The program then exits immediately.
589 45. A need has arisen to pass over additional data with calls to pcre_exec() in
590 order to support additional features. One way would have been to define
591 pcre_exec2() (for example) with extra arguments, but this would not have been
592 extensible, and would also have required all calls to the original function to
593 be mapped to the new one. Instead, I have chosen to extend the mechanism that
594 is used for passing in "extra" data from pcre_study().
596 The pcre_extra structure is now exposed and defined in pcre.h. It currently
597 contains the following fields:
599 flags a bitmap indicating which of the following fields are set
600 study_data opaque data from pcre_study()
601 match_limit a way of specifying a limit on match() calls for a specific
602 call to pcre_exec()
603 callout_data data for callouts (see 49 below)
605 The flag bits are also defined in pcre.h, and are
611 The pcre_study() function now returns one of these new pcre_extra blocks, with
612 the actual study data pointed to by the study_data field, and the
613 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
614 before. That is, this change is entirely upwards-compatible and requires no
615 change to existing code.
617 If you want to pass in additional data to pcre_exec(), you can either place it
618 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
619 block.
621 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
622 data string contains the escape sequence \M, pcretest calls pcre_exec() several
623 times with different match limits, until it finds the minimum value needed for
624 pcre_exec() to complete. The value is then output. This can be instructive; for
625 most simple matches the number is quite small, but for pathological cases it
626 gets very large very quickly.
628 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
629 returns the size of the data block pointed to by the study_data field in a
630 pcre_extra block, that is, the value that was passed as the argument to
631 pcre_malloc() when PCRE was getting memory in which to place the information
632 created by pcre_study(). The fourth argument should point to a size_t variable.
633 pcretest has been extended so that this information is shown after a successful
634 pcre_study() call when information about the compiled regex is being displayed.
636 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
637 because what follows is always an absolute path. (Later: it turns out that this
638 is more than cosmetic for MinGW, because it doesn't like empty path
639 components.)
641 49. Some changes have been made to the callout feature (see 28 above):
643 (i) A callout function now has three choices for what it returns:
645 0 => success, carry on matching
646 > 0 => failure at this point, but backtrack if possible
647 < 0 => serious error, return this value from pcre_exec()
649 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
650 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
651 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
652 use by callout functions. It will never be used by PCRE itself.
654 (ii) The pcre_extra structure (see 45 above) has a void * field called
655 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
656 pcre_callout_block structure has a field of the same name. The contents of
657 the field passed in the pcre_extra structure are passed to the callout
658 function in the corresponding field in the callout block. This makes it
659 easier to use the same callout-containing regex from multiple threads. For
660 testing, the pcretest program has a new data escape
662 \C*n pass the number n (may be negative) as callout_data
664 If the callout function in pcretest receives a non-zero value as
665 callout_data, it returns that value.
667 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
668 there were some redundant $(CFLAGS) in commands that are now specified as
669 $(LINK), which already includes $(CFLAGS).
671 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
672 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
673 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
674 one-byte characters throughout. Note that case-insensitive matching applies
675 only to characters whose values are less than 256. PCRE doesn't support the
676 notion of cases for higher-valued characters.
678 (i) A character class whose characters are all within 0-255 is handled as
679 a bit map, and the map is inverted for negative classes. Previously, a
680 character > 255 always failed to match such a class; however it should
681 match if the class was a negative one (e.g. [^ab]). This has been fixed.
683 (ii) A negated character class with a single character < 255 is coded as
684 "not this character" (OP_NOT). This wasn't working properly when the test
685 character was multibyte, either singly or repeated.
687 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
688 mode, for example: \x{100}{2,3}.
690 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
691 singly or repeated) now correctly test multibyte characters. However,
692 PCRE doesn't recognize any characters with values greater than 255 as
693 digits, spaces, or word characters. Such characters always match \D, \S,
694 and \W, and never match \d, \s, or \w.
696 (v) Classes may now contain characters and character ranges with values
697 greater than 255. For example: [ab\x{100}-\x{400}].
699 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
700 PCRE in UTF-8 mode.
702 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
703 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
704 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
705 value.)
707 53. The single man page has become too large. I have therefore split it up into
708 a number of separate man pages. These also give rise to individual HTML pages;
709 these are now put in a separate directory, and there is an index.html page that
710 lists them all. Some hyperlinking between the pages has been installed.
712 54. Added convenience functions for handling named capturing parentheses.
714 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
715 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
716 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
717 are faulted.
719 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
720 calling configure. These values are used when compiling the dftables.c program
721 which is run to generate the source of the default character tables. They
722 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
723 you will need to set these values.
725 57. Updated the building process for Windows DLL, as provided by Fred Cox.
728 Version 3.9 02-Jan-02
729 ---------------------
731 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
733 2. If --disable-static was given, the building process failed when trying to
734 build pcretest and pcregrep. (For some reason it was using libtool to compile
735 them, which is not right, as they aren't part of the library.)
738 Version 3.8 18-Dec-01
739 ---------------------
741 1. The experimental UTF-8 code was completely screwed up. It was packing the
742 bytes in the wrong order. How dumb can you get?
745 Version 3.7 29-Oct-01
746 ---------------------
748 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
749 This caused pcretest, when used on the test data, to segfault. Unfortunately,
750 this didn't happen under Solaris 8, where I normally test things.
752 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
753 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
754 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
755 this fix an hour or so after the initial 3.7 release.)
758 Version 3.6 23-Oct-01
759 ---------------------
761 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
762 offsets passed as NULL with zero offset count.
764 2. The config.guess and config.sub files had not been updated when I moved to
765 the latest autoconf.
768 Version 3.5 15-Aug-01
769 ---------------------
771 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
772 had been forgotten.
774 2. By using declared but undefined structures, we can avoid using "void"
775 definitions in pcre.h while keeping the internal definitions of the structures
776 private.
778 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
779 user point of view, this means that both static and shared libraries are built
780 by default, but this can be individually controlled. More of the work of
781 handling this static/shared cases is now inside libtool instead of PCRE's make
782 file.
784 4. The pcretest utility is now installed along with pcregrep because it is
785 useful for users (to test regexs) and by doing this, it automatically gets
786 relinked by libtool. The documentation has been turned into a man page, so
787 there are now .1, .txt, and .html versions in /doc.
789 5. Upgrades to pcregrep:
790 (i) Added long-form option names like gnu grep.
791 (ii) Added --help to list all options with an explanatory phrase.
792 (iii) Added -r, --recursive to recurse into sub-directories.
793 (iv) Added -f, --file to read patterns from a file.
795 6. pcre_exec() was referring to its "code" argument before testing that
796 argument for NULL (and giving an error if it was NULL).
798 7. Upgraded Makefile.in to allow for compiling in a different directory from
799 the source directory.
801 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
802 options bits, the pointer it was passed was to an int instead of to an unsigned
803 long int. This mattered only on 64-bit systems.
805 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
806 generated) instead of pcre.in, which it its source. Also made the same change
807 in several of the .c files.
809 10. A new release of gcc defines printf() as a macro, which broke pcretest
810 because it had an ifdef in the middle of a string argument for printf(). Fixed
811 by using separate calls to printf().
813 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
814 script, to force use of CR or LF instead of \n in the source. On non-Unix
815 systems, the value can be set in config.h.
817 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
818 absolute limit. Changed the text of the error message to make this clear, and
819 likewise updated the man page.
821 13. The limit of 99 on the number of capturing subpatterns has been removed.
822 The new limit is 65535, which I hope will not be a "real" limit.
825 Version 3.4 22-Aug-00
826 ---------------------
828 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
830 2. Diagnose condition (?(0) as an error instead of crashing on matching.
833 Version 3.3 01-Aug-00
834 ---------------------
836 1. If an octal character was given, but the value was greater than \377, it
837 was not getting masked to the least significant bits, as documented. This could
838 lead to crashes in some systems.
840 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
841 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
843 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
844 These just pass their arguments on to (pcre_free)(), but they are provided
845 because some uses of PCRE bind it to non-C systems that can call its functions,
846 but cannot call free() or pcre_free() directly.
848 4. Add "make test" as a synonym for "make check". Corrected some comments in
849 the Makefile.
851 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
852 Makefile.
854 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
855 command called pgrep for grepping around the active processes.
857 7. Added the beginnings of support for UTF-8 character strings.
859 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
860 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
861 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
862 out for the ar command.)
865 Version 3.2 12-May-00
866 ---------------------
868 This is purely a bug fixing release.
870 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
871 of ZA. This was just one example of several cases that could provoke this bug,
872 which was introduced by change 9 of version 2.00. The code for breaking
873 infinite loops after an iteration that matches an empty string was't working
874 correctly.
876 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
877 when matched against abbab (for example). After matching an empty string, it
878 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
879 caused it to match further down the string than it should.
881 3. The code contained an inclusion of sys/types.h. It isn't clear why this
882 was there because it doesn't seem to be needed, and it causes trouble on some
883 systems, as it is not a Standard C header. It has been removed.
885 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
886 were reported on the Macintosh. The changes were from
888 while ((c = *(++ptr)) != 0 && c != '\n');
889 to
890 while ((c = *(++ptr)) != 0 && c != '\n') ;
892 Totally extraordinary, but if that's what it takes...
894 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
895 available. Added HAVE_BCOPY and an autoconf test for it; if neither
896 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
897 assumes the way PCRE uses memmove() (always moving upwards).
899 6. PCRE is being used in one environment where strchr() is not available. There
900 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
901 faster code anyway.
904 Version 3.1 09-Feb-00
905 ---------------------
907 The only change in this release is the fixing of some bugs in Makefile.in for
908 the "install" target:
910 (1) It was failing to install pcreposix.h.
912 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
915 Version 3.0 01-Feb-00
916 ---------------------
918 1. Add support for the /+ modifier to perltest (to output $` like it does in
919 pcretest).
921 2. Add support for the /g modifier to perltest.
923 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
924 matches null strings.
926 4. Fix perltest so that it doesn't do unwanted things when fed an empty
927 pattern. Perl treats empty patterns specially - it reuses the most recent
928 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
929 effect.
931 5. The POSIX interface was broken in that it was just handing over the POSIX
932 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
933 required a bigger vector, with some working space on the end. This means that
934 the POSIX wrapper now has to get and free some memory, and copy the results.
936 6. Added some simple autoconf support, placing the test data and the
937 documentation in separate directories, re-organizing some of the
938 information files, and making it build pcre-config (a GNU standard). Also added
939 libtool support for building PCRE as a shared library, which is now the
940 default.
942 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
943 09 are not valid octal constants. Single digits will be used for minor values
944 less than 10.
946 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
947 existing programs that set these in the POSIX interface can use PCRE without
948 modification.
950 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
951 return all that pcre_info() returns, plus additional data. The pcre_info()
952 function is retained for compatibility, but is considered to be obsolete.
954 10. Added experimental recursion feature (?R) to handle one common case that
955 Perl 5.6 will be able to do with (?p{...}).
957 11. Added support for POSIX character classes like [:alpha:], which Perl is
958 adopting.
961 Version 2.08 31-Aug-99
962 ----------------------
964 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
965 trying to match at the startoffset position, but instead was moving forward to
966 the next newline as if a previous match had failed.
968 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
969 and could get into a loop if a null string was matched other than at the start
970 of the subject.
972 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
973 be distinguished at compile time, and for completeness also added PCRE_DATE.
975 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
976 in GnuWin32 environments.
979 Version 2.07 29-Jul-99
980 ----------------------
982 1. The documentation is now supplied in plain text form and HTML as well as in
983 the form of man page sources.
985 2. C++ compilers don't like assigning (void *) values to other pointer types.
986 In particular this affects malloc(). Although there is no problem in Standard
987 C, I've put in casts to keep C++ compilers happy.
989 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
990 should be (const char *).
992 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
993 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
994 However, I haven't made this a standard facility. The documentation doesn't
995 mention it, and the Makefile doesn't support it.
997 5. The Makefile now contains an "install" target, with editable destinations at
998 the top of the file. The pcretest program is not installed.
1000 6. pgrep -V now gives the PCRE version number and date.
1002 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
1003 causing the entire string to be ignored, instead of just the last character.
1005 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
1006 non-matching string, it can take a very, very long time, even for strings of
1007 quite modest length, because of the nested recursion. PCRE now does better in
1008 some of these cases. It does this by remembering the last required literal
1009 character in the pattern, and pre-searching the subject to ensure it is present
1010 before running the real match. In other words, it applies a heuristic to detect
1011 some types of certain failure quickly, and in the above example, if presented
1012 with a string that has no trailing " it gives "no match" very quickly.
1014 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
1015 other alternatives are tried instead.
1018 Version 2.06 09-Jun-99
1019 ----------------------
1021 1. Change pcretest's output for amount of store used to show just the code
1022 space, because the remainder (the data block) varies in size between 32-bit and
1023 64-bit systems.
1025 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
1026 start matching at. This allows lookbehinds to work when searching for multiple
1027 occurrences in a string.
1029 3. Added additional options to pcretest for testing multiple occurrences:
1031 /+ outputs the rest of the string that follows a match
1032 /g loops for multiple occurrences, using the new startoffset argument
1033 /G loops for multiple occurrences by passing an incremented pointer
1035 4. PCRE wasn't doing the "first character" optimization for patterns starting
1036 with \b or \B, though it was doing it for other lookbehind assertions. That is,
1037 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
1038 the letter 'x'. On long subject strings, this gives a significant speed-up.
1041 Version 2.05 21-Apr-99
1042 ----------------------
1044 1. Changed the type of magic_number from int to long int so that it works
1045 properly on 16-bit systems.
1047 2. Fixed a bug which caused patterns starting with .* not to work correctly
1048 when the subject string contained newline characters. PCRE was assuming
1049 anchoring for such patterns in all cases, which is not correct because .* will
1050 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
1051 DOTALL is set at top level; otherwise it knows that patterns starting with .*
1052 must be retried after every newline in the subject.
1055 Version 2.04 18-Feb-99
1056 ----------------------
1058 1. For parenthesized subpatterns with repeats whose minimum was zero, the
1059 computation of the store needed to hold the pattern was incorrect (too large).
1060 If such patterns were nested a few deep, this could multiply and become a real
1061 problem.
1063 2. Added /M option to pcretest to show the memory requirement of a specific
1064 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
1066 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
1067 compiled in such a way that the backtracking after subsequent failure was
1068 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
1069 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
1072 Version 2.03 02-Feb-99
1073 ----------------------
1075 1. Fixed typo and small mistake in man page.
1077 2. Added 4th condition (GPL supersedes if conflict) and created separate
1078 LICENCE file containing the conditions.
1080 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
1081 Perl, that is the internal \ allows the delimiter to be included in the
1082 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
1083 the final delimiter, add \ to the end of the pattern (to test the error).
1085 4. Added the convenience functions for extracting substrings after a successful
1086 match. Updated pcretest to make it able to test these functions.
1089 Version 2.02 14-Jan-99
1090 ----------------------
1092 1. Initialized the working variables associated with each extraction so that
1093 their saving and restoring doesn't refer to uninitialized store.
1095 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
1096 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
1097 fix the problem.
1099 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
1100 calls, and wasn't printing the correct value for compiling calls. Increased the
1101 default value of LOOPREPEAT, and the number of significant figures in the
1102 times.
1104 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
1106 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
1107 a building problem on Windows NT with a FAT file system.
1110 Version 2.01 21-Oct-98
1111 ----------------------
1113 1. Changed the API for pcre_compile() to allow for the provision of a pointer
1114 to character tables built by pcre_maketables() in the current locale. If NULL
1115 is passed, the default tables are used.
1118 Version 2.00 24-Sep-98
1119 ----------------------
1121 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
1122 it any more.
1124 2. Allow quantification of (?>) groups, and make it work correctly.
1126 3. The first character computation wasn't working for (?>) groups.
1128 4. Correct the implementation of \Z (it is permitted to match on the \n at the
1129 end of the subject) and add 5.005's \z, which really does match only at the
1130 very end of the subject.
1132 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
1134 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
1135 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
1136 localized options. All options to pcre_study() were also removed.
1138 7. Add other new features from 5.005:
1140 $(?<= positive lookbehind
1141 $(?<! negative lookbehind
1142 (?imsx-imsx) added the unsetting capability
1143 such a setting is global if at outer level; local otherwise
1144 (?imsx-imsx:) non-capturing groups with option setting
1145 (?(cond)re|re) conditional pattern matching
1147 A backreference to itself in a repeated group matches the previous
1148 captured string.
1150 8. General tidying up of studying (both automatic and via "study")
1151 consequential on the addition of new assertions.
1153 9. As in 5.005, unlimited repeated groups that could match an empty substring
1154 are no longer faulted at compile time. Instead, the loop is forcibly broken at
1155 runtime if any iteration does actually match an empty substring.
1157 10. Include the RunTest script in the distribution.
1159 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
1160 discrepancies, some of which were old and were also with respect to 5.004. They
1161 have now been fixed.
1164 Version 1.09 28-Apr-98
1165 ----------------------
1167 1. A negated single character class followed by a quantifier with a minimum
1168 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
1169 program crashes, or just wrong answers. This did not apply to negated classes
1170 containing more than one character, or to minima other than one.
1173 Version 1.08 27-Mar-98
1174 ----------------------
1176 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
1178 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
1179 latter must appear before anything that relies on it in the pattern.
1182 Version 1.07 16-Feb-98
1183 ----------------------
1185 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
1186 repeat of a potentially empty string).
1189 Version 1.06 23-Jan-98
1190 ----------------------
1192 1. Added Markus Oberhumer's little patches for C++.
1194 2. Literal strings longer than 255 characters were broken.
1197 Version 1.05 23-Dec-97
1198 ----------------------
1200 1. Negated character classes containing more than one character were failing if
1201 PCRE_CASELESS was set at run time.
1204 Version 1.04 19-Dec-97
1205 ----------------------
1207 1. Corrected the man page, where some "const" qualifiers had been omitted.
1209 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
1210 input syntax.
1212 3. Fixed memory leak which occurred when a regex with back references was
1213 matched with an offsets vector that wasn't big enough. The temporary memory
1214 that is used in this case wasn't being freed if the match failed.
1216 4. Tidied pcretest to ensure it frees memory that it gets.
1218 5. Temporary memory was being obtained in the case where the passed offsets
1219 vector was exactly big enough.
1221 6. Corrected definition of offsetof() from change 5 below.
1223 7. I had screwed up change 6 below and broken the rules for the use of
1224 setjmp(). Now fixed.
1227 Version 1.03 18-Dec-97
1228 ----------------------
1230 1. A erroneous regex with a missing opening parenthesis was correctly
1231 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
1232 on some systems.
1234 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
1235 it was reported that one broken compiler failed on the former because "code" is
1236 also an independent variable.
1238 3. The erroneous regex a[]b caused an array overrun reference.
1240 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
1241 fail on data ending with that character. (It was going on too far, and checking
1242 the next character, typically a binary zero.) This was specific to the
1243 optimized code for single-character negative classes.
1245 5. Added a contributed patch from the TIN world which does the following:
1247 + Add an undef for memmove, in case the the system defines a macro for it.
1249 + Add a definition of offsetof(), in case there isn't one. (I don't know
1250 the reason behind this - offsetof() is part of the ANSI standard - but
1251 it does no harm).
1253 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
1254 most of the places where whitespace preceded '#'. I have given up and
1255 allowed the remaining 2 cases to be at the margin.
1257 + Rename some variables in pcre to eliminate shadowing. This seems very
1258 pedantic, but does no harm, of course.
1260 6. Moved the call to setjmp() into its own function, to get rid of warnings
1261 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
1263 7. Constructs such as \d{8,} were compiling into the equivalent of
1264 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
1265 outcome, but in this particular case used more store than had been allocated,
1266 which caused the bug to be discovered because it threw up an internal error.
1268 8. The debugging code in both pcre and pcretest for outputting the compiled
1269 form of a regex was going wrong in the case of back references followed by
1270 curly-bracketed repeats.
1273 Version 1.02 12-Dec-97
1274 ----------------------
1276 1. Typos in pcre.3 and comments in the source fixed.
1278 2. Applied a contributed patch to get rid of places where it used to remove
1279 'const' from variables, and fixed some signed/unsigned and uninitialized
1280 variable warnings.
1282 3. Added the "runtest" target to Makefile.
1284 4. Set default compiler flag to -O2 rather than just -O.
1287 Version 1.01 19-Nov-97
1288 ----------------------
1290 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
1291 like /([ab]*)*/, that is, for classes with more than one character in them.
1293 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
1294 as /((?>a*))*/ (a PCRE_EXTRA facility).
1297 Version 1.00 18-Nov-97
1298 ----------------------
1300 1. Added compile-time macros to support systems such as SunOS4 which don't have
1301 memmove() or strerror() but have other things that can be used instead.
1303 2. Arranged that "make clean" removes the executables.
1306 Version 0.99 27-Oct-97
1307 ----------------------
1309 1. Fixed bug in code for optimizing classes with only one character. It was
1310 initializing a 32-byte map regardless, which could cause it to run off the end
1311 of the memory it had got.
1313 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
1316 Version 0.98 22-Oct-97
1317 ----------------------
1319 1. Fixed bug in code for handling temporary memory usage when there are more
1320 back references than supplied space in the ovector. This could cause segfaults.
1323 Version 0.97 21-Oct-97
1324 ----------------------
1326 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
1328 2. Optimized negated single characters not to use a bit map.
1330 3. Brought error texts together as macro definitions; clarified some of them;
1331 fixed one that was wrong - it said "range out of order" when it meant "invalid
1332 escape sequence".
1334 4. Changed some char * arguments to const char *.
1336 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
1338 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
1339 pcretest.
1342 Version 0.96 16-Oct-97
1343 ----------------------
1345 1. Added a simple "pgrep" utility to the distribution.
1347 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
1348 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
1349 where "ddd" means "one or more decimal digits".
1351 3. Fixed serious bug. If a pattern had a back reference, but the call to
1352 pcre_exec() didn't supply a large enough ovector to record the related
1353 identifying subpattern, the match always failed. PCRE now remembers the number
1354 of the largest back reference, and gets some temporary memory in which to save
1355 the offsets during matching if necessary, in order to ensure that
1356 backreferences always work.
1358 4. Increased the compatibility with Perl in a number of ways:
1360 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
1361 to request this handling. The option can be set at compile or exec time.
1363 (b) $ matches before a terminating newline by default; an option
1364 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
1365 mode). The option can be set at compile or exec time.
1367 (c) The handling of \ followed by a digit other than 0 is now supposed to be
1368 the same as Perl's. If the decimal number it represents is less than 10
1369 or there aren't that many previous left capturing parentheses, an octal
1370 escape is read. Inside a character class, it's always an octal escape,
1371 even if it is a single digit.
1373 (d) An escaped but undefined alphabetic character is taken as a literal,
1374 unless PCRE_EXTRA is set. Currently this just reserves the remaining
1375 escapes.
1377 (e) {0} is now permitted. (The previous item is removed from the compiled
1378 pattern).
1380 5. Changed all the names of code files so that the basic parts are no longer
1381 than 10 characters, and abolished the teeny "globals.c" file.
1383 6. Changed the handling of character classes; they are now done with a 32-byte
1384 bit map always.
1386 7. Added the -d and /D options to pcretest to make it possible to look at the
1387 internals of compilation without having to recompile pcre.
1390 Version 0.95 23-Sep-97
1391 ----------------------
1393 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
1394 \x20 at the start of a run of normal characters. These were being treated as
1395 real characters, instead of the source characters being re-checked.
1398 Version 0.94 18-Sep-97
1399 ----------------------
1401 1. The functions are now thread-safe, with the caveat that the global variables
1402 containing pointers to malloc() and free() or alternative functions are the
1403 same for all threads.
1405 2. Get pcre_study() to generate a bitmap of initial characters for non-
1406 anchored patterns when this is possible, and use it if passed to pcre_exec().
1409 Version 0.93 15-Sep-97
1410 ----------------------
1412 1. /(b)|(:+)/ was computing an incorrect first character.
1414 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
1415 but not actually doing anything yet.
1417 3. Treat "-" characters in classes that cannot be part of ranges as literals,
1418 as Perl does (e.g. [-az] or [az-]).
1420 4. Set the anchored flag if a branch starts with .* or .*? because that tests
1421 all possible positions.
1423 5. Split up into different modules to avoid including unneeded functions in a
1424 compiled binary. However, compile and exec are still in one module. The "study"
1425 function is split off.
1427 6. The character tables are now in a separate module whose source is generated
1428 by an auxiliary program - but can then be edited by hand if required. There are
1429 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
1430 toupper() in the code.
1432 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
1433 make them global. Abolish the function for setting them, as the caller can now
1434 set them directly.
1437 Version 0.92 11-Sep-97
1438 ----------------------
1440 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
1441 (e.g. /a{1,3}/) was broken (I mis-optimized it).
1443 2. Caseless matching was not working in character classes if the characters in
1444 the pattern were in upper case.
1446 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
1448 4. Make PCRE_ANCHORED public and accept as a compile option.
1450 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
1451 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
1452 pass them.
1454 6. Give an error if bad option bits passed at compile or run time.
1456 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
1457 pcretest to cause it to pass that flag.
1459 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
1460 options, and the first character, if set.
1462 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
1465 Version 0.91 10-Sep-97
1466 ----------------------
1468 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
1469 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
1471 2. PCRE was looping on encountering an indefinitely repeated back reference to
1472 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
1473 Perl does - treats the match as successful.
1475 ****

  ViewVC Help
Powered by ViewVC 1.1.5