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