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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 29 - (show annotations) (download)
Sat Feb 24 21:38:53 2007 UTC (7 years, 7 months ago) by nigel
File size: 14552 byte(s)
Load pcre-2.03 into code/trunk.

1 ChangeLog for PCRE
2 ------------------
3
4
5 Version 2.03 02-Feb-99
6 ----------------------
7
8 1. Fixed typo and small mistake in man page.
9
10 2. Added 4th condition (GPL supersedes) and created separate LICENCE file.
11
12 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
13 Perl, that is the internal \ allows the delimiter to be included in the
14 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
15 the final delimiter, add \ to the end of the pattern (to test the error).
16
17 4. Added the convenience functions for extracting substrings after a successful
18 match. Updated pcretest to make it able to test these functions.
19
20
21 Version 2.02 14-Jan-99
22 ----------------------
23
24 1. Initialized the working variables associated with each extraction so that
25 their saving and restoring doesn't refer to uninitialized store.
26
27 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
28 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
29 fix the problem.
30
31 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
32 calls, and wasn't printing the correct value for compiling calls. Increased the
33 default value of LOOPREPEAT, and the number of significant figures in the
34 times.
35
36 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
37
38 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
39 a building problem on Windows NT with a FAT file system.
40
41
42 Version 2.01 21-Oct-98
43 ----------------------
44
45 1. Changed the API for pcre_compile() to allow for the provision of a pointer
46 to character tables built by pcre_maketables() in the current locale. If NULL
47 is passed, the default tables are used.
48
49
50 Version 2.00 24-Sep-98
51 ----------------------
52
53 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
54 it any more.
55
56 2. Allow quantification of (?>) groups, and make it work correctly.
57
58 3. The first character computation wasn't working for (?>) groups.
59
60 4. Correct the implementation of \Z (it is permitted to match on the \n at the
61 end of the subject) and add 5.005's \z, which really does match only at the
62 very end of the subject.
63
64 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
65
66 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
67 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
68 localized options. All options to pcre_study() were also removed.
69
70 7. Add other new features from 5.005:
71
72 $(?<= positive lookbehind
73 $(?<! negative lookbehind
74 (?imsx-imsx) added the unsetting capability
75 such a setting is global if at outer level; local otherwise
76 (?imsx-imsx:) non-capturing groups with option setting
77 (?(cond)re|re) conditional pattern matching
78
79 A backreference to itself in a repeated group matches the previous
80 captured string.
81
82 8. General tidying up of studying (both automatic and via "study")
83 consequential on the addition of new assertions.
84
85 9. As in 5.005, unlimited repeated groups that could match an empty substring
86 are no longer faulted at compile time. Instead, the loop is forcibly broken at
87 runtime if any iteration does actually match an empty substring.
88
89 10. Include the RunTest script in the distribution.
90
91 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
92 discrepancies, some of which were old and were also with respect to 5.004. They
93 have now been fixed.
94
95
96 Version 1.09 28-Apr-98
97 ----------------------
98
99 1. A negated single character class followed by a quantifier with a minimum
100 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
101 program crashes, or just wrong answers. This did not apply to negated classes
102 containing more than one character, or to minima other than one.
103
104
105 Version 1.08 27-Mar-98
106 ----------------------
107
108 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
109
110 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
111 latter must appear before anything that relies on it in the pattern.
112
113
114 Version 1.07 16-Feb-98
115 ----------------------
116
117 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
118 repeat of a potentially empty string).
119
120
121 Version 1.06 23-Jan-98
122 ----------------------
123
124 1. Added Markus Oberhumer's little patches for C++.
125
126 2. Literal strings longer than 255 characters were broken.
127
128
129 Version 1.05 23-Dec-97
130 ----------------------
131
132 1. Negated character classes containing more than one character were failing if
133 PCRE_CASELESS was set at run time.
134
135
136 Version 1.04 19-Dec-97
137 ----------------------
138
139 1. Corrected the man page, where some "const" qualifiers had been omitted.
140
141 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
142 input syntax.
143
144 3. Fixed memory leak which occurred when a regex with back references was
145 matched with an offsets vector that wasn't big enough. The temporary memory
146 that is used in this case wasn't being freed if the match failed.
147
148 4. Tidied pcretest to ensure it frees memory that it gets.
149
150 5. Temporary memory was being obtained in the case where the passed offsets
151 vector was exactly big enough.
152
153 6. Corrected definition of offsetof() from change 5 below.
154
155 7. I had screwed up change 6 below and broken the rules for the use of
156 setjmp(). Now fixed.
157
158
159 Version 1.03 18-Dec-97
160 ----------------------
161
162 1. A erroneous regex with a missing opening parenthesis was correctly
163 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
164 on some systems.
165
166 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
167 it was reported that one broken compiler failed on the former because "code" is
168 also an independent variable.
169
170 3. The erroneous regex a[]b caused an array overrun reference.
171
172 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
173 fail on data ending with that character. (It was going on too far, and checking
174 the next character, typically a binary zero.) This was specific to the
175 optimized code for single-character negative classes.
176
177 5. Added a contributed patch from the TIN world which does the following:
178
179 + Add an undef for memmove, in case the the system defines a macro for it.
180
181 + Add a definition of offsetof(), in case there isn't one. (I don't know
182 the reason behind this - offsetof() is part of the ANSI standard - but
183 it does no harm).
184
185 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
186 most of the places where whitespace preceded '#'. I have given up and
187 allowed the remaining 2 cases to be at the margin.
188
189 + Rename some variables in pcre to eliminate shadowing. This seems very
190 pedantic, but does no harm, of course.
191
192 6. Moved the call to setjmp() into its own function, to get rid of warnings
193 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
194
195 7. Constructs such as \d{8,} were compiling into the equivalent of
196 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
197 outcome, but in this particular case used more store than had been allocated,
198 which caused the bug to be discovered because it threw up an internal error.
199
200 8. The debugging code in both pcre and pcretest for outputting the compiled
201 form of a regex was going wrong in the case of back references followed by
202 curly-bracketed repeats.
203
204
205 Version 1.02 12-Dec-97
206 ----------------------
207
208 1. Typos in pcre.3 and comments in the source fixed.
209
210 2. Applied a contributed patch to get rid of places where it used to remove
211 'const' from variables, and fixed some signed/unsigned and uninitialized
212 variable warnings.
213
214 3. Added the "runtest" target to Makefile.
215
216 4. Set default compiler flag to -O2 rather than just -O.
217
218
219 Version 1.01 19-Nov-97
220 ----------------------
221
222 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
223 like /([ab]*)*/, that is, for classes with more than one character in them.
224
225 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
226 as /((?>a*))*/ (a PCRE_EXTRA facility).
227
228
229 Version 1.00 18-Nov-97
230 ----------------------
231
232 1. Added compile-time macros to support systems such as SunOS4 which don't have
233 memmove() or strerror() but have other things that can be used instead.
234
235 2. Arranged that "make clean" removes the executables.
236
237
238 Version 0.99 27-Oct-97
239 ----------------------
240
241 1. Fixed bug in code for optimizing classes with only one character. It was
242 initializing a 32-byte map regardless, which could cause it to run off the end
243 of the memory it had got.
244
245 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
246
247
248 Version 0.98 22-Oct-97
249 ----------------------
250
251 1. Fixed bug in code for handling temporary memory usage when there are more
252 back references than supplied space in the ovector. This could cause segfaults.
253
254
255 Version 0.97 21-Oct-97
256 ----------------------
257
258 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
259
260 2. Optimized negated single characters not to use a bit map.
261
262 3. Brought error texts together as macro definitions; clarified some of them;
263 fixed one that was wrong - it said "range out of order" when it meant "invalid
264 escape sequence".
265
266 4. Changed some char * arguments to const char *.
267
268 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
269
270 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
271 pcretest.
272
273
274 Version 0.96 16-Oct-97
275 ----------------------
276
277 1. Added a simple "pgrep" utility to the distribution.
278
279 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
280 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
281 where "ddd" means "one or more decimal digits".
282
283 3. Fixed serious bug. If a pattern had a back reference, but the call to
284 pcre_exec() didn't supply a large enough ovector to record the related
285 identifying subpattern, the match always failed. PCRE now remembers the number
286 of the largest back reference, and gets some temporary memory in which to save
287 the offsets during matching if necessary, in order to ensure that
288 backreferences always work.
289
290 4. Increased the compatibility with Perl in a number of ways:
291
292 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
293 to request this handling. The option can be set at compile or exec time.
294
295 (b) $ matches before a terminating newline by default; an option
296 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
297 mode). The option can be set at compile or exec time.
298
299 (c) The handling of \ followed by a digit other than 0 is now supposed to be
300 the same as Perl's. If the decimal number it represents is less than 10
301 or there aren't that many previous left capturing parentheses, an octal
302 escape is read. Inside a character class, it's always an octal escape,
303 even if it is a single digit.
304
305 (d) An escaped but undefined alphabetic character is taken as a literal,
306 unless PCRE_EXTRA is set. Currently this just reserves the remaining
307 escapes.
308
309 (e) {0} is now permitted. (The previous item is removed from the compiled
310 pattern).
311
312 5. Changed all the names of code files so that the basic parts are no longer
313 than 10 characters, and abolished the teeny "globals.c" file.
314
315 6. Changed the handling of character classes; they are now done with a 32-byte
316 bit map always.
317
318 7. Added the -d and /D options to pcretest to make it possible to look at the
319 internals of compilation without having to recompile pcre.
320
321
322 Version 0.95 23-Sep-97
323 ----------------------
324
325 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
326 \x20 at the start of a run of normal characters. These were being treated as
327 real characters, instead of the source characters being re-checked.
328
329
330 Version 0.94 18-Sep-97
331 ----------------------
332
333 1. The functions are now thread-safe, with the caveat that the global variables
334 containing pointers to malloc() and free() or alternative functions are the
335 same for all threads.
336
337 2. Get pcre_study() to generate a bitmap of initial characters for non-
338 anchored patterns when this is possible, and use it if passed to pcre_exec().
339
340
341 Version 0.93 15-Sep-97
342 ----------------------
343
344 1. /(b)|(:+)/ was computing an incorrect first character.
345
346 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
347 but not actually doing anything yet.
348
349 3. Treat "-" characters in classes that cannot be part of ranges as literals,
350 as Perl does (e.g. [-az] or [az-]).
351
352 4. Set the anchored flag if a branch starts with .* or .*? because that tests
353 all possible positions.
354
355 5. Split up into different modules to avoid including unneeded functions in a
356 compiled binary. However, compile and exec are still in one module. The "study"
357 function is split off.
358
359 6. The character tables are now in a separate module whose source is generated
360 by an auxiliary program - but can then be edited by hand if required. There are
361 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
362 toupper() in the code.
363
364 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
365 make them global. Abolish the function for setting them, as the caller can now
366 set them directly.
367
368
369 Version 0.92 11-Sep-97
370 ----------------------
371
372 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
373 (e.g. /a{1,3}/) was broken (I mis-optimized it).
374
375 2. Caseless matching was not working in character classes if the characters in
376 the pattern were in upper case.
377
378 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
379
380 4. Make PCRE_ANCHORED public and accept as a compile option.
381
382 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
383 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
384 pass them.
385
386 6. Give an error if bad option bits passed at compile or run time.
387
388 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
389 pcretest to cause it to pass that flag.
390
391 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
392 options, and the first character, if set.
393
394 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
395
396
397 Version 0.91 10-Sep-97
398 ----------------------
399
400 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
401 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
402
403 2. PCRE was looping on encountering an indefinitely repeated back reference to
404 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
405 Perl does - treats the match as successful.
406
407 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12