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

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log


Revision 21 - (show annotations) (download)
Sat Feb 24 21:38:37 2007 UTC (7 years, 2 months ago) by nigel
File size: 11125 byte(s)
Load pcre-1.09 into code/trunk.

1 ChangeLog for PCRE
2 ------------------
3
4
5 Version 1.09 28 Apr-98
6 ----------------------
7
8 1. A negated single character class followed by a quantifier with a minimum
9 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
10 program crashes, or just wrong answers. This did not apply to negated classes
11 containing more than one character, or to minima other than one.
12
13
14 Version 1.08 27-Mar-98
15 ----------------------
16
17 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
18
19 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
20 latter must appear before anything that relies on it in the pattern.
21
22
23 Version 1.07 16-Feb-98
24 ----------------------
25
26 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
27 repeat of a potentially empty string).
28
29
30 Version 1.06 23-Jan-98
31 ----------------------
32
33 1. Added Markus Oberhumer's little patches for C++.
34
35 2. Literal strings longer than 255 characters were broken.
36
37
38 Version 1.05 23-Dec-97
39 ----------------------
40
41 1. Negated character classes containing more than one character were failing if
42 PCRE_CASELESS was set at run time.
43
44
45 Version 1.04 19-Dec-97
46 ----------------------
47
48 1. Corrected the man page, where some "const" qualifiers had been omitted.
49
50 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
51 input syntax.
52
53 3. Fixed memory leak which occurred when a regex with back references was
54 matched with an offsets vector that wasn't big enough. The temporary memory
55 that is used in this case wasn't being freed if the match failed.
56
57 4. Tidied pcretest to ensure it frees memory that it gets.
58
59 5. Temporary memory was being obtained in the case where the passed offsets
60 vector was exactly big enough.
61
62 6. Corrected definition of offsetof() from change 5 below.
63
64 7. I had screwed up change 6 below and broken the rules for the use of
65 setjmp(). Now fixed.
66
67
68 Version 1.03 18-Dec-97
69 ----------------------
70
71 1. A erroneous regex with a missing opening parenthesis was correctly
72 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
73 on some systems.
74
75 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
76 it was reported that one broken compiler failed on the former because "code" is
77 also an independent variable.
78
79 3. The erroneous regex a[]b caused an array overrun reference.
80
81 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
82 fail on data ending with that character. (It was going on too far, and checking
83 the next character, typically a binary zero.) This was specific to the
84 optimized code for single-character negative classes.
85
86 5. Added a contributed patch from the TIN world which does the following:
87
88 + Add an undef for memmove, in case the the system defines a macro for it.
89
90 + Add a definition of offsetof(), in case there isn't one. (I don't know
91 the reason behind this - offsetof() is part of the ANSI standard - but
92 it does no harm).
93
94 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
95 most of the places where whitespace preceded '#'. I have given up and
96 allowed the remaining 2 cases to be at the margin.
97
98 + Rename some variables in pcre to eliminate shadowing. This seems very
99 pedantic, but does no harm, of course.
100
101 6. Moved the call to setjmp() into its own function, to get rid of warnings
102 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
103
104 7. Constructs such as \d{8,} were compiling into the equivalent of
105 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
106 outcome, but in this particular case used more store than had been allocated,
107 which caused the bug to be discovered because it threw up an internal error.
108
109 8. The debugging code in both pcre and pcretest for outputting the compiled
110 form of a regex was going wrong in the case of back references followed by
111 curly-bracketed repeats.
112
113
114 Version 1.02 12-Dec-97
115 ----------------------
116
117 1. Typos in pcre.3 and comments in the source fixed.
118
119 2. Applied a contributed patch to get rid of places where it used to remove
120 'const' from variables, and fixed some signed/unsigned and uninitialized
121 variable warnings.
122
123 3. Added the "runtest" target to Makefile.
124
125 4. Set default compiler flag to -O2 rather than just -O.
126
127
128 Version 1.01 19-Nov-97
129 ----------------------
130
131 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
132 like /([ab]*)*/, that is, for classes with more than one character in them.
133
134 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
135 as /((?>a*))*/ (a PCRE_EXTRA facility).
136
137
138 Version 1.00 18-Nov-97
139 ----------------------
140
141 1. Added compile-time macros to support systems such as SunOS4 which don't have
142 memmove() or strerror() but have other things that can be used instead.
143
144 2. Arranged that "make clean" removes the executables.
145
146
147 Version 0.99 27-Oct-97
148 ----------------------
149
150 1. Fixed bug in code for optimizing classes with only one character. It was
151 initializing a 32-byte map regardless, which could cause it to run off the end
152 of the memory it had got.
153
154 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
155
156
157 Version 0.98 22-Oct-97
158 ----------------------
159
160 1. Fixed bug in code for handling temporary memory usage when there are more
161 back references than supplied space in the ovector. This could cause segfaults.
162
163
164 Version 0.97 21-Oct-97
165 ----------------------
166
167 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
168
169 2. Optimized negated single characters not to use a bit map.
170
171 3. Brought error texts together as macro definitions; clarified some of them;
172 fixed one that was wrong - it said "range out of order" when it meant "invalid
173 escape sequence".
174
175 4. Changed some char * arguments to const char *.
176
177 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
178
179 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
180 pcretest.
181
182
183 Version 0.96 16-Oct-97
184 ----------------------
185
186 1. Added a simple "pgrep" utility to the distribution.
187
188 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
189 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
190 where "ddd" means "one or more decimal digits".
191
192 3. Fixed serious bug. If a pattern had a back reference, but the call to
193 pcre_exec() didn't supply a large enough ovector to record the related
194 identifying subpattern, the match always failed. PCRE now remembers the number
195 of the largest back reference, and gets some temporary memory in which to save
196 the offsets during matching if necessary, in order to ensure that
197 backreferences always work.
198
199 4. Increased the compatibility with Perl in a number of ways:
200
201 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
202 to request this handling. The option can be set at compile or exec time.
203
204 (b) $ matches before a terminating newline by default; an option
205 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
206 mode). The option can be set at compile or exec time.
207
208 (c) The handling of \ followed by a digit other than 0 is now supposed to be
209 the same as Perl's. If the decimal number it represents is less than 10
210 or there aren't that many previous left capturing parentheses, an octal
211 escape is read. Inside a character class, it's always an octal escape,
212 even if it is a single digit.
213
214 (d) An escaped but undefined alphabetic character is taken as a literal,
215 unless PCRE_EXTRA is set. Currently this just reserves the remaining
216 escapes.
217
218 (e) {0} is now permitted. (The previous item is removed from the compiled
219 pattern).
220
221 5. Changed all the names of code files so that the basic parts are no longer
222 than 10 characters, and abolished the teeny "globals.c" file.
223
224 6. Changed the handling of character classes; they are now done with a 32-byte
225 bit map always.
226
227 7. Added the -d and /D options to pcretest to make it possible to look at the
228 internals of compilation without having to recompile pcre.
229
230
231 Version 0.95 23-Sep-97
232 ----------------------
233
234 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
235 \x20 at the start of a run of normal characters. These were being treated as
236 real characters, instead of the source characters being re-checked.
237
238
239 Version 0.94 18-Sep-97
240 ----------------------
241
242 1. The functions are now thread-safe, with the caveat that the global variables
243 containing pointers to malloc() and free() or alternative functions are the
244 same for all threads.
245
246 2. Get pcre_study() to generate a bitmap of initial characters for non-
247 anchored patterns when this is possible, and use it if passed to pcre_exec().
248
249
250 Version 0.93 15-Sep-97
251 ----------------------
252
253 1. /(b)|(:+)/ was computing an incorrect first character.
254
255 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
256 but not actually doing anything yet.
257
258 3. Treat "-" characters in classes that cannot be part of ranges as literals,
259 as Perl does (e.g. [-az] or [az-]).
260
261 4. Set the anchored flag if a branch starts with .* or .*? because that tests
262 all possible positions.
263
264 5. Split up into different modules to avoid including unneeded functions in a
265 compiled binary. However, compile and exec are still in one module. The "study"
266 function is split off.
267
268 6. The character tables are now in a separate module whose source is generated
269 by an auxiliary program - but can then be edited by hand if required. There are
270 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
271 toupper() in the code.
272
273 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
274 make them global. Abolish the function for setting them, as the caller can now
275 set them directly.
276
277
278 Version 0.92 11-Sep-97
279 ----------------------
280
281 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
282 (e.g. /a{1,3}/) was broken (I mis-optimized it).
283
284 2. Caseless matching was not working in character classes if the characters in
285 the pattern were in upper case.
286
287 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
288
289 4. Make PCRE_ANCHORED public and accept as a compile option.
290
291 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
292 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
293 pass them.
294
295 6. Give an error if bad option bits passed at compile or run time.
296
297 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
298 pcretest to cause it to pass that flag.
299
300 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
301 options, and the first character, if set.
302
303 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
304
305
306 Version 0.91 10-Sep-97
307 ----------------------
308
309 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
310 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
311
312 2. PCRE was looping on encountering an indefinitely repeated back reference to
313 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
314 Perl does - treats the match as successful.
315
316 ****

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12