Commit | Line | Data |
---|---|---|
108003db RGS |
1 | =head1 NAME |
2 | ||
5a2b28ce | 3 | perlreapi - Perl regular expression plugin interface |
108003db RGS |
4 | |
5 | =head1 DESCRIPTION | |
6 | ||
5a2b28ce KW |
7 | As of Perl 5.9.5 there is a new interface for plugging and using |
8 | regular expression engines other than the default one. | |
a0e97681 RGS |
9 | |
10 | Each engine is supposed to provide access to a constant structure of the | |
11 | following format: | |
108003db RGS |
12 | |
13 | typedef struct regexp_engine { | |
02c01adb KW |
14 | REGEXP* (*comp) (pTHX_ |
15 | const SV * const pattern, const U32 flags); | |
16 | I32 (*exec) (pTHX_ | |
17 | REGEXP * const rx, | |
18 | char* stringarg, | |
19 | char* strend, char* strbeg, | |
20 | I32 minend, SV* screamer, | |
2fdbfb4d | 21 | void* data, U32 flags); |
02c01adb KW |
22 | char* (*intuit) (pTHX_ |
23 | REGEXP * const rx, SV *sv, | |
24 | char *strpos, char *strend, U32 flags, | |
2fdbfb4d | 25 | struct re_scream_pos_data_s *data); |
49d7dfbc AB |
26 | SV* (*checkstr) (pTHX_ REGEXP * const rx); |
27 | void (*free) (pTHX_ REGEXP * const rx); | |
02c01adb KW |
28 | void (*numbered_buff_FETCH) (pTHX_ |
29 | REGEXP * const rx, | |
30 | const I32 paren, | |
31 | SV * const sv); | |
32 | void (*numbered_buff_STORE) (pTHX_ | |
33 | REGEXP * const rx, | |
34 | const I32 paren, | |
5a2b28ce | 35 | SV const * const value); |
02c01adb KW |
36 | I32 (*numbered_buff_LENGTH) (pTHX_ |
37 | REGEXP * const rx, | |
38 | const SV * const sv, | |
39 | const I32 paren); | |
40 | SV* (*named_buff) (pTHX_ | |
41 | REGEXP * const rx, | |
42 | SV * const key, | |
43 | SV * const value, | |
44 | U32 flags); | |
45 | SV* (*named_buff_iter) (pTHX_ | |
46 | REGEXP * const rx, | |
47 | const SV * const lastkey, | |
192b9cd1 | 48 | const U32 flags); |
49d7dfbc | 49 | SV* (*qr_package)(pTHX_ REGEXP * const rx); |
108003db | 50 | #ifdef USE_ITHREADS |
49d7dfbc | 51 | void* (*dupe) (pTHX_ REGEXP * const rx, CLONE_PARAMS *param); |
108003db | 52 | #endif |
3c13cae6 DM |
53 | REGEXP* (*op_comp) (...); |
54 | ||
108003db RGS |
55 | |
56 | When a regexp is compiled, its C<engine> field is then set to point at | |
a0e97681 | 57 | the appropriate structure, so that when it needs to be used Perl can find |
108003db RGS |
58 | the right routines to do so. |
59 | ||
60 | In order to install a new regexp handler, C<$^H{regcomp}> is set | |
61 | to an integer which (when casted appropriately) resolves to one of these | |
62 | structures. When compiling, the C<comp> method is executed, and the | |
5a2b28ce | 63 | resulting C<regexp> structure's engine field is expected to point back at |
108003db RGS |
64 | the same structure. |
65 | ||
5a2b28ce | 66 | The pTHX_ symbol in the definition is a macro used by Perl under threading |
108003db RGS |
67 | to provide an extra argument to the routine holding a pointer back to |
68 | the interpreter that is executing the regexp. So under threading all | |
69 | routines get an extra argument. | |
70 | ||
882227b7 | 71 | =head1 Callbacks |
108003db RGS |
72 | |
73 | =head2 comp | |
74 | ||
3ab4a224 | 75 | REGEXP* comp(pTHX_ const SV * const pattern, const U32 flags); |
108003db | 76 | |
3ab4a224 AB |
77 | Compile the pattern stored in C<pattern> using the given C<flags> and |
78 | return a pointer to a prepared C<REGEXP> structure that can perform | |
79 | the match. See L</The REGEXP structure> below for an explanation of | |
80 | the individual fields in the REGEXP struct. | |
81 | ||
82 | The C<pattern> parameter is the scalar that was used as the | |
5a2b28ce KW |
83 | pattern. Previous versions of Perl would pass two C<char*> indicating |
84 | the start and end of the stringified pattern; the following snippet can | |
3ab4a224 AB |
85 | be used to get the old parameters: |
86 | ||
87 | STRLEN plen; | |
88 | char* exp = SvPV(pattern, plen); | |
89 | char* xend = exp + plen; | |
90 | ||
5a2b28ce | 91 | Since any scalar can be passed as a pattern, it's possible to implement |
3ab4a224 AB |
92 | an engine that does something with an array (C<< "ook" =~ [ qw/ eek |
93 | hlagh / ] >>) or with the non-stringified form of a compiled regular | |
5a2b28ce KW |
94 | expression (C<< "ook" =~ qr/eek/ >>). Perl's own engine will always |
95 | stringify everything using the snippet above, but that doesn't mean | |
3ab4a224 | 96 | other engines have to. |
108003db | 97 | |
a0e97681 | 98 | The C<flags> parameter is a bitfield which indicates which of the |
c998b245 | 99 | C<msixp> flags the regex was compiled with. It also contains |
5a2b28ce | 100 | additional info, such as if C<use locale> is in effect. |
108003db RGS |
101 | |
102 | The C<eogc> flags are stripped out before being passed to the comp | |
5a2b28ce KW |
103 | routine. The regex engine does not need to know if any of these |
104 | are set, as those flags should only affect what Perl does with the | |
c998b245 AB |
105 | pattern and its match variables, not how it gets compiled and |
106 | executed. | |
108003db | 107 | |
c998b245 AB |
108 | By the time the comp callback is called, some of these flags have |
109 | already had effect (noted below where applicable). However most of | |
5a2b28ce | 110 | their effect occurs after the comp callback has run, in routines that |
c998b245 | 111 | read the C<< rx->extflags >> field which it populates. |
108003db | 112 | |
c998b245 AB |
113 | In general the flags should be preserved in C<< rx->extflags >> after |
114 | compilation, although the regex engine might want to add or delete | |
5a2b28ce | 115 | some of them to invoke or disable some special behavior in Perl. The |
c998b245 | 116 | flags along with any special behavior they cause are documented below: |
108003db | 117 | |
c998b245 | 118 | The pattern modifiers: |
108003db | 119 | |
c998b245 | 120 | =over 4 |
108003db | 121 | |
c998b245 | 122 | =item C</m> - RXf_PMf_MULTILINE |
108003db | 123 | |
c998b245 AB |
124 | If this is in C<< rx->extflags >> it will be passed to |
125 | C<Perl_fbm_instr> by C<pp_split> which will treat the subject string | |
126 | as a multi-line string. | |
108003db | 127 | |
c998b245 | 128 | =item C</s> - RXf_PMf_SINGLELINE |
108003db | 129 | |
c998b245 | 130 | =item C</i> - RXf_PMf_FOLD |
108003db | 131 | |
c998b245 | 132 | =item C</x> - RXf_PMf_EXTENDED |
108003db | 133 | |
5a2b28ce | 134 | If present on a regex, C<"#"> comments will be handled differently by the |
c998b245 | 135 | tokenizer in some cases. |
108003db | 136 | |
c998b245 | 137 | TODO: Document those cases. |
108003db | 138 | |
c998b245 | 139 | =item C</p> - RXf_PMf_KEEPCOPY |
108003db | 140 | |
e72ec78c KW |
141 | TODO: Document this |
142 | ||
f0f9b3b8 KW |
143 | =item Character set |
144 | ||
145 | The character set semantics are determined by an enum that is contained | |
146 | in this field. This is still experimental and subject to change, but | |
147 | the current interface returns the rules by use of the in-line function | |
148 | C<get_regex_charset(const U32 flags)>. The only currently documented | |
149 | value returned from it is REGEX_LOCALE_CHARSET, which is set if | |
e72ec78c KW |
150 | C<use locale> is in effect. If present in C<< rx->extflags >>, |
151 | C<split> will use the locale dependent definition of whitespace | |
152 | when RXf_SKIPWHITE or RXf_WHITE is in effect. ASCII whitespace | |
96090e4f | 153 | is defined as per L<isSPACE|perlapi/isSPACE>, and by the internal |
e72ec78c | 154 | macros C<is_utf8_space> under UTF-8, and C<isSPACE_LC> under C<use |
c998b245 | 155 | locale>. |
108003db | 156 | |
f0f9b3b8 KW |
157 | =back |
158 | ||
159 | Additional flags: | |
160 | ||
161 | =over 4 | |
162 | ||
0ac6acae AB |
163 | =item RXf_SPLIT |
164 | ||
9f6ecbda FC |
165 | This flag was removed in perl 5.18.0. C<split ' '> is now special-cased |
166 | solely in the parser. RXf_SPLIT is still #defined, so you can test for it. | |
167 | This is how it used to work: | |
168 | ||
0ac6acae | 169 | If C<split> is invoked as C<split ' '> or with no arguments (which |
5a2b28ce | 170 | really means C<split(' ', $_)>, see L<split|perlfunc/split>), Perl will |
0ac6acae | 171 | set this flag. The regex engine can then check for it and set the |
5a2b28ce | 172 | SKIPWHITE and WHITE extflags. To do this, the Perl engine does: |
0ac6acae AB |
173 | |
174 | if (flags & RXf_SPLIT && r->prelen == 1 && r->precomp[0] == ' ') | |
175 | r->extflags |= (RXf_SKIPWHITE|RXf_WHITE); | |
176 | ||
108003db RGS |
177 | =back |
178 | ||
c998b245 AB |
179 | These flags can be set during compilation to enable optimizations in |
180 | the C<split> operator. | |
181 | ||
182 | =over 4 | |
183 | ||
0ac6acae AB |
184 | =item RXf_SKIPWHITE |
185 | ||
186 | If the flag is present in C<< rx->extflags >> C<split> will delete | |
187 | whitespace from the start of the subject string before it's operated | |
5a2b28ce KW |
188 | on. What is considered whitespace depends on if the subject is a |
189 | UTF-8 string and if the C<RXf_PMf_LOCALE> flag is set. | |
0ac6acae | 190 | |
5a2b28ce KW |
191 | If RXf_WHITE is set in addition to this flag, C<split> will behave like |
192 | C<split " "> under the Perl engine. | |
0ac6acae | 193 | |
c998b245 AB |
194 | =item RXf_START_ONLY |
195 | ||
196 | Tells the split operator to split the target string on newlines | |
197 | (C<\n>) without invoking the regex engine. | |
198 | ||
199 | Perl's engine sets this if the pattern is C</^/> (C<plen == 1 && *exp | |
5a2b28ce | 200 | == '^'>), even under C</^/s>; see L<split|perlfunc>. Of course a |
c998b245 AB |
201 | different regex engine might want to use the same optimizations |
202 | with a different syntax. | |
203 | ||
204 | =item RXf_WHITE | |
205 | ||
206 | Tells the split operator to split the target string on whitespace | |
207 | without invoking the regex engine. The definition of whitespace varies | |
5a2b28ce KW |
208 | depending on if the target string is a UTF-8 string and on |
209 | if RXf_PMf_LOCALE is set. | |
c998b245 | 210 | |
0ac6acae | 211 | Perl's engine sets this flag if the pattern is C<\s+>. |
c998b245 | 212 | |
640f820d AB |
213 | =item RXf_NULL |
214 | ||
a0e97681 | 215 | Tells the split operator to split the target string on |
5a2b28ce | 216 | characters. The definition of character varies depending on if |
640f820d AB |
217 | the target string is a UTF-8 string. |
218 | ||
219 | Perl's engine sets this flag on empty patterns, this optimization | |
a0e97681 | 220 | makes C<split //> much faster than it would otherwise be. It's even |
640f820d AB |
221 | faster than C<unpack>. |
222 | ||
c998b245 | 223 | =back |
108003db RGS |
224 | |
225 | =head2 exec | |
226 | ||
49d7dfbc | 227 | I32 exec(pTHX_ REGEXP * const rx, |
108003db RGS |
228 | char *stringarg, char* strend, char* strbeg, |
229 | I32 minend, SV* screamer, | |
230 | void* data, U32 flags); | |
231 | ||
8fd1a950 DM |
232 | Execute a regexp. The arguments are |
233 | ||
234 | =over 4 | |
235 | ||
236 | =item rx | |
237 | ||
238 | The regular expression to execute. | |
239 | ||
240 | =item screamer | |
241 | ||
242 | This strangely-named arg is the SV to be matched against. Note that the | |
243 | actual char array to be matched against is supplied by the arguments | |
244 | described below; the SV is just used to determine UTF8ness, C<pos()> etc. | |
245 | ||
246 | =item strbeg | |
247 | ||
248 | Pointer to the physical start of the string. | |
249 | ||
250 | =item strend | |
251 | ||
252 | Pointer to the character following the physical end of the string (i.e. | |
5a2b28ce | 253 | the C<\0>). |
8fd1a950 DM |
254 | |
255 | =item stringarg | |
256 | ||
257 | Pointer to the position in the string where matching should start; it might | |
258 | not be equal to C<strbeg> (for example in a later iteration of C</.../g>). | |
259 | ||
260 | =item minend | |
261 | ||
262 | Minimum length of string (measured in bytes from C<stringarg>) that must | |
263 | match; if the engine reaches the end of the match but hasn't reached this | |
264 | position in the string, it should fail. | |
265 | ||
266 | =item data | |
267 | ||
268 | Optimisation data; subject to change. | |
269 | ||
270 | =item flags | |
271 | ||
272 | Optimisation flags; subject to change. | |
273 | ||
274 | =back | |
108003db RGS |
275 | |
276 | =head2 intuit | |
277 | ||
49d7dfbc | 278 | char* intuit(pTHX_ REGEXP * const rx, |
108003db | 279 | SV *sv, char *strpos, char *strend, |
49d7dfbc | 280 | const U32 flags, struct re_scream_pos_data_s *data); |
108003db RGS |
281 | |
282 | Find the start position where a regex match should be attempted, | |
5a2b28ce KW |
283 | or possibly if the regex engine should not be run because the |
284 | pattern can't match. This is called, as appropriate, by the core, | |
285 | depending on the values of the C<extflags> member of the C<regexp> | |
108003db RGS |
286 | structure. |
287 | ||
288 | =head2 checkstr | |
289 | ||
49d7dfbc | 290 | SV* checkstr(pTHX_ REGEXP * const rx); |
108003db RGS |
291 | |
292 | Return a SV containing a string that must appear in the pattern. Used | |
293 | by C<split> for optimising matches. | |
294 | ||
295 | =head2 free | |
296 | ||
49d7dfbc | 297 | void free(pTHX_ REGEXP * const rx); |
108003db | 298 | |
5a2b28ce | 299 | Called by Perl when it is freeing a regexp pattern so that the engine |
108003db | 300 | can release any resources pointed to by the C<pprivate> member of the |
5a2b28ce KW |
301 | C<regexp> structure. This is only responsible for freeing private data; |
302 | Perl will handle releasing anything else contained in the C<regexp> structure. | |
108003db | 303 | |
192b9cd1 | 304 | =head2 Numbered capture callbacks |
108003db | 305 | |
192b9cd1 AB |
306 | Called to get/set the value of C<$`>, C<$'>, C<$&> and their named |
307 | equivalents, ${^PREMATCH}, ${^POSTMATCH} and $^{MATCH}, as well as the | |
c27a5cfe | 308 | numbered capture groups (C<$1>, C<$2>, ...). |
49d7dfbc | 309 | |
c149d39e DM |
310 | The C<paren> parameter will be C<1> for C<$1>, C<2> for C<$2> and so |
311 | forth, and have these symbolic values for the special variables: | |
312 | ||
313 | ${^PREMATCH} RX_BUFF_IDX_CARET_PREMATCH | |
314 | ${^POSTMATCH} RX_BUFF_IDX_CARET_POSTMATCH | |
315 | ${^MATCH} RX_BUFF_IDX_CARET_FULLMATCH | |
316 | $` RX_BUFF_IDX_PREMATCH | |
317 | $' RX_BUFF_IDX_POSTMATCH | |
318 | $& RX_BUFF_IDX_FULLMATCH | |
319 | ||
5a2b28ce | 320 | Note that in Perl 5.17.3 and earlier, the last three constants were also |
c149d39e DM |
321 | used for the caret variants of the variables. |
322 | ||
49d7dfbc | 323 | |
192b9cd1 AB |
324 | The names have been chosen by analogy with L<Tie::Scalar> methods |
325 | names with an additional B<LENGTH> callback for efficiency. However | |
326 | named capture variables are currently not tied internally but | |
327 | implemented via magic. | |
328 | ||
329 | =head3 numbered_buff_FETCH | |
330 | ||
331 | void numbered_buff_FETCH(pTHX_ REGEXP * const rx, const I32 paren, | |
332 | SV * const sv); | |
333 | ||
334 | Fetch a specified numbered capture. C<sv> should be set to the scalar | |
335 | to return, the scalar is passed as an argument rather than being | |
5a2b28ce | 336 | returned from the function because when it's called Perl already has a |
192b9cd1 AB |
337 | scalar to store the value, creating another one would be |
338 | redundant. The scalar can be set with C<sv_setsv>, C<sv_setpvn> and | |
339 | friends, see L<perlapi>. | |
49d7dfbc | 340 | |
5a2b28ce | 341 | This callback is where Perl untaints its own capture variables under |
c998b245 | 342 | taint mode (see L<perlsec>). See the C<Perl_reg_numbered_buff_fetch> |
49d7dfbc AB |
343 | function in F<regcomp.c> for how to untaint capture variables if |
344 | that's something you'd like your engine to do as well. | |
108003db | 345 | |
192b9cd1 | 346 | =head3 numbered_buff_STORE |
108003db | 347 | |
02c01adb KW |
348 | void (*numbered_buff_STORE) (pTHX_ |
349 | REGEXP * const rx, | |
350 | const I32 paren, | |
2fdbfb4d | 351 | SV const * const value); |
108003db | 352 | |
192b9cd1 AB |
353 | Set the value of a numbered capture variable. C<value> is the scalar |
354 | that is to be used as the new value. It's up to the engine to make | |
355 | sure this is used as the new value (or reject it). | |
2fdbfb4d AB |
356 | |
357 | Example: | |
358 | ||
359 | if ("ook" =~ /(o*)/) { | |
ccf3535a | 360 | # 'paren' will be '1' and 'value' will be 'ee' |
2fdbfb4d AB |
361 | $1 =~ tr/o/e/; |
362 | } | |
363 | ||
364 | Perl's own engine will croak on any attempt to modify the capture | |
a0e97681 | 365 | variables, to do this in another engine use the following callback |
2fdbfb4d AB |
366 | (copied from C<Perl_reg_numbered_buff_store>): |
367 | ||
368 | void | |
02c01adb KW |
369 | Example_reg_numbered_buff_store(pTHX_ |
370 | REGEXP * const rx, | |
371 | const I32 paren, | |
372 | SV const * const value) | |
2fdbfb4d AB |
373 | { |
374 | PERL_UNUSED_ARG(rx); | |
375 | PERL_UNUSED_ARG(paren); | |
376 | PERL_UNUSED_ARG(value); | |
377 | ||
378 | if (!PL_localizing) | |
379 | Perl_croak(aTHX_ PL_no_modify); | |
380 | } | |
381 | ||
5a2b28ce | 382 | Actually Perl will not I<always> croak in a statement that looks |
2fdbfb4d | 383 | like it would modify a numbered capture variable. This is because the |
5a2b28ce | 384 | STORE callback will not be called if Perl can determine that it |
2fdbfb4d AB |
385 | doesn't have to modify the value. This is exactly how tied variables |
386 | behave in the same situation: | |
387 | ||
388 | package CaptureVar; | |
389 | use base 'Tie::Scalar'; | |
390 | ||
391 | sub TIESCALAR { bless [] } | |
392 | sub FETCH { undef } | |
393 | sub STORE { die "This doesn't get called" } | |
394 | ||
395 | package main; | |
396 | ||
c69ca1d4 | 397 | tie my $sv => "CaptureVar"; |
2fdbfb4d AB |
398 | $sv =~ y/a/b/; |
399 | ||
5a2b28ce | 400 | Because C<$sv> is C<undef> when the C<y///> operator is applied to it, |
2fdbfb4d | 401 | the transliteration won't actually execute and the program won't |
192b9cd1 | 402 | C<die>. This is different to how 5.8 and earlier versions behaved |
5a2b28ce | 403 | since the capture variables were READONLY variables then; now they'll |
192b9cd1 | 404 | just die when assigned to in the default engine. |
2fdbfb4d | 405 | |
192b9cd1 | 406 | =head3 numbered_buff_LENGTH |
2fdbfb4d | 407 | |
02c01adb KW |
408 | I32 numbered_buff_LENGTH (pTHX_ |
409 | REGEXP * const rx, | |
410 | const SV * const sv, | |
2fdbfb4d AB |
411 | const I32 paren); |
412 | ||
413 | Get the C<length> of a capture variable. There's a special callback | |
5a2b28ce KW |
414 | for this so that Perl doesn't have to do a FETCH and run C<length> on |
415 | the result, since the length is (in Perl's case) known from an offset | |
416 | stored in C<< rx->offs >>, this is much more efficient: | |
2fdbfb4d AB |
417 | |
418 | I32 s1 = rx->offs[paren].start; | |
419 | I32 s2 = rx->offs[paren].end; | |
420 | I32 len = t1 - s1; | |
421 | ||
422 | This is a little bit more complex in the case of UTF-8, see what | |
423 | C<Perl_reg_numbered_buff_length> does with | |
424 | L<is_utf8_string_loclen|perlapi/is_utf8_string_loclen>. | |
425 | ||
192b9cd1 AB |
426 | =head2 Named capture callbacks |
427 | ||
5a2b28ce | 428 | Called to get/set the value of C<%+> and C<%->, as well as by some |
192b9cd1 AB |
429 | utility functions in L<re>. |
430 | ||
431 | There are two callbacks, C<named_buff> is called in all the cases the | |
432 | FETCH, STORE, DELETE, CLEAR, EXISTS and SCALAR L<Tie::Hash> callbacks | |
433 | would be on changes to C<%+> and C<%-> and C<named_buff_iter> in the | |
434 | same cases as FIRSTKEY and NEXTKEY. | |
435 | ||
436 | The C<flags> parameter can be used to determine which of these | |
5a2b28ce | 437 | operations the callbacks should respond to. The following flags are |
192b9cd1 AB |
438 | currently defined: |
439 | ||
440 | Which L<Tie::Hash> operation is being performed from the Perl level on | |
441 | C<%+> or C<%+>, if any: | |
442 | ||
f1b875a0 YO |
443 | RXapif_FETCH |
444 | RXapif_STORE | |
445 | RXapif_DELETE | |
446 | RXapif_CLEAR | |
447 | RXapif_EXISTS | |
448 | RXapif_SCALAR | |
449 | RXapif_FIRSTKEY | |
450 | RXapif_NEXTKEY | |
192b9cd1 | 451 | |
5a2b28ce | 452 | If C<%+> or C<%-> is being operated on, if any. |
2fdbfb4d | 453 | |
f1b875a0 YO |
454 | RXapif_ONE /* %+ */ |
455 | RXapif_ALL /* %- */ | |
2fdbfb4d | 456 | |
5a2b28ce | 457 | If this is being called as C<re::regname>, C<re::regnames> or |
c998b245 | 458 | C<re::regnames_count>, if any. The first two will be combined with |
f1b875a0 | 459 | C<RXapif_ONE> or C<RXapif_ALL>. |
192b9cd1 | 460 | |
f1b875a0 YO |
461 | RXapif_REGNAME |
462 | RXapif_REGNAMES | |
463 | RXapif_REGNAMES_COUNT | |
192b9cd1 AB |
464 | |
465 | Internally C<%+> and C<%-> are implemented with a real tied interface | |
466 | via L<Tie::Hash::NamedCapture>. The methods in that package will call | |
467 | back into these functions. However the usage of | |
468 | L<Tie::Hash::NamedCapture> for this purpose might change in future | |
469 | releases. For instance this might be implemented by magic instead | |
470 | (would need an extension to mgvtbl). | |
471 | ||
472 | =head3 named_buff | |
473 | ||
474 | SV* (*named_buff) (pTHX_ REGEXP * const rx, SV * const key, | |
475 | SV * const value, U32 flags); | |
476 | ||
477 | =head3 named_buff_iter | |
478 | ||
02c01adb KW |
479 | SV* (*named_buff_iter) (pTHX_ |
480 | REGEXP * const rx, | |
481 | const SV * const lastkey, | |
192b9cd1 | 482 | const U32 flags); |
108003db | 483 | |
49d7dfbc | 484 | =head2 qr_package |
108003db | 485 | |
49d7dfbc | 486 | SV* qr_package(pTHX_ REGEXP * const rx); |
108003db RGS |
487 | |
488 | The package the qr// magic object is blessed into (as seen by C<ref | |
49d7dfbc | 489 | qr//>). It is recommended that engines change this to their package |
5a2b28ce | 490 | name for identification regardless of if they implement methods |
49d7dfbc AB |
491 | on the object. |
492 | ||
192b9cd1 | 493 | The package this method returns should also have the internal |
d5213412 | 494 | C<Regexp> package in its C<@ISA>. C<< qr//->isa("Regexp") >> should always |
192b9cd1 AB |
495 | be true regardless of what engine is being used. |
496 | ||
497 | Example implementation might be: | |
108003db RGS |
498 | |
499 | SV* | |
192b9cd1 | 500 | Example_qr_package(pTHX_ REGEXP * const rx) |
108003db RGS |
501 | { |
502 | PERL_UNUSED_ARG(rx); | |
503 | return newSVpvs("re::engine::Example"); | |
504 | } | |
505 | ||
506 | Any method calls on an object created with C<qr//> will be dispatched to the | |
507 | package as a normal object. | |
508 | ||
509 | use re::engine::Example; | |
510 | my $re = qr//; | |
511 | $re->meth; # dispatched to re::engine::Example::meth() | |
512 | ||
f7e71195 AB |
513 | To retrieve the C<REGEXP> object from the scalar in an XS function use |
514 | the C<SvRX> macro, see L<"REGEXP Functions" in perlapi|perlapi/REGEXP | |
515 | Functions>. | |
108003db RGS |
516 | |
517 | void meth(SV * rv) | |
518 | PPCODE: | |
f7e71195 | 519 | REGEXP * re = SvRX(sv); |
108003db | 520 | |
108003db RGS |
521 | =head2 dupe |
522 | ||
49d7dfbc | 523 | void* dupe(pTHX_ REGEXP * const rx, CLONE_PARAMS *param); |
108003db RGS |
524 | |
525 | On threaded builds a regexp may need to be duplicated so that the pattern | |
a0e97681 | 526 | can be used by multiple threads. This routine is expected to handle the |
108003db | 527 | duplication of any private data pointed to by the C<pprivate> member of |
5a2b28ce KW |
528 | the C<regexp> structure. It will be called with the preconstructed new |
529 | C<regexp> structure as an argument, the C<pprivate> member will point at | |
a0e97681 | 530 | the B<old> private structure, and it is this routine's responsibility to |
5a2b28ce | 531 | construct a copy and return a pointer to it (which Perl will then use to |
108003db RGS |
532 | overwrite the field as passed to this routine.) |
533 | ||
534 | This allows the engine to dupe its private data but also if necessary | |
535 | modify the final structure if it really must. | |
536 | ||
537 | On unthreaded builds this field doesn't exist. | |
538 | ||
3c13cae6 DM |
539 | =head2 op_comp |
540 | ||
5a2b28ce | 541 | This is private to the Perl core and subject to change. Should be left |
3c13cae6 DM |
542 | null. |
543 | ||
108003db RGS |
544 | =head1 The REGEXP structure |
545 | ||
546 | The REGEXP struct is defined in F<regexp.h>. All regex engines must be able to | |
547 | correctly build such a structure in their L</comp> routine. | |
548 | ||
5a2b28ce | 549 | The REGEXP structure contains all the data that Perl needs to be aware of |
108003db | 550 | to properly work with the regular expression. It includes data about |
5a2b28ce | 551 | optimisations that Perl can use to determine if the regex engine should |
108003db | 552 | really be used, and various other control info that is needed to properly |
5a2b28ce KW |
553 | execute patterns in various contexts, such as if the pattern anchored in |
554 | some way, or what flags were used during the compile, or if the | |
555 | program contains special constructs that Perl needs to be aware of. | |
108003db | 556 | |
882227b7 AB |
557 | In addition it contains two fields that are intended for the private |
558 | use of the regex engine that compiled the pattern. These are the | |
559 | C<intflags> and C<pprivate> members. C<pprivate> is a void pointer to | |
5a2b28ce KW |
560 | an arbitrary structure, whose use and management is the responsibility |
561 | of the compiling engine. Perl will never modify either of these | |
882227b7 | 562 | values. |
108003db RGS |
563 | |
564 | typedef struct regexp { | |
565 | /* what engine created this regexp? */ | |
566 | const struct regexp_engine* engine; | |
567 | ||
568 | /* what re is this a lightweight copy of? */ | |
569 | struct regexp* mother_re; | |
570 | ||
5a2b28ce | 571 | /* Information about the match that the Perl core uses to manage |
02c01adb | 572 | * things */ |
108003db | 573 | U32 extflags; /* Flags used both externally and internally */ |
2d608413 KW |
574 | I32 minlen; /* mininum possible number of chars in */ |
575 | string to match */ | |
576 | I32 minlenret; /* mininum possible number of chars in $& */ | |
108003db RGS |
577 | U32 gofs; /* chars left of pos that we search from */ |
578 | ||
579 | /* substring data about strings that must appear | |
580 | in the final match, used for optimisations */ | |
581 | struct reg_substr_data *substrs; | |
582 | ||
c27a5cfe | 583 | U32 nparens; /* number of capture groups */ |
108003db RGS |
584 | |
585 | /* private engine specific data */ | |
586 | U32 intflags; /* Engine Specific Internal flags */ | |
587 | void *pprivate; /* Data private to the regex engine which | |
588 | created this object. */ | |
589 | ||
02c01adb KW |
590 | /* Data about the last/current match. These are modified during |
591 | * matching*/ | |
82f14494 DM |
592 | U32 lastparen; /* highest close paren matched ($+) */ |
593 | U32 lastcloseparen; /* last close paren matched ($^N) */ | |
108003db | 594 | regexp_paren_pair *swap; /* Swap copy of *offs */ |
02c01adb KW |
595 | regexp_paren_pair *offs; /* Array of offsets for (@-) and |
596 | (@+) */ | |
108003db | 597 | |
02c01adb KW |
598 | char *subbeg; /* saved or original string so \digit works |
599 | forever. */ | |
108003db RGS |
600 | SV_SAVED_COPY /* If non-NULL, SV which is COW from original */ |
601 | I32 sublen; /* Length of string pointed by subbeg */ | |
02c01adb KW |
602 | I32 suboffset; /* byte offset of subbeg from logical start of |
603 | str */ | |
6502e081 | 604 | I32 subcoffset; /* suboffset equiv, but in chars (for @-/@+) */ |
108003db RGS |
605 | |
606 | /* Information about the match that isn't often used */ | |
607 | I32 prelen; /* length of precomp */ | |
608 | const char *precomp; /* pre-compilation regular expression */ | |
609 | ||
108003db RGS |
610 | char *wrapped; /* wrapped version of the pattern */ |
611 | I32 wraplen; /* length of wrapped */ | |
612 | ||
02c01adb KW |
613 | I32 seen_evals; /* number of eval groups in the pattern - for |
614 | security checks */ | |
108003db RGS |
615 | HV *paren_names; /* Optional hash of paren names */ |
616 | ||
617 | /* Refcount of this regexp */ | |
618 | I32 refcnt; /* Refcount of this regexp */ | |
619 | } regexp; | |
620 | ||
621 | The fields are discussed in more detail below: | |
622 | ||
882227b7 | 623 | =head2 C<engine> |
108003db | 624 | |
5a2b28ce | 625 | This field points at a C<regexp_engine> structure which contains pointers |
108003db RGS |
626 | to the subroutines that are to be used for performing a match. It |
627 | is the compiling routine's responsibility to populate this field before | |
628 | returning the regexp object. | |
629 | ||
630 | Internally this is set to C<NULL> unless a custom engine is specified in | |
5a2b28ce | 631 | C<$^H{regcomp}>, Perl's own set of callbacks can be accessed in the struct |
108003db RGS |
632 | pointed to by C<RE_ENGINE_PTR>. |
633 | ||
882227b7 | 634 | =head2 C<mother_re> |
108003db RGS |
635 | |
636 | TODO, see L<http://www.mail-archive.com/perl5-changes@perl.org/msg17328.html> | |
637 | ||
882227b7 | 638 | =head2 C<extflags> |
108003db | 639 | |
5a2b28ce | 640 | This will be used by Perl to see what flags the regexp was compiled |
192b9cd1 | 641 | with, this will normally be set to the value of the flags parameter by |
c998b245 AB |
642 | the L<comp|/comp> callback. See the L<comp|/comp> documentation for |
643 | valid flags. | |
108003db | 644 | |
882227b7 | 645 | =head2 C<minlen> C<minlenret> |
108003db | 646 | |
2d608413 KW |
647 | The minimum string length (in characters) required for the pattern to match. |
648 | This is used to | |
108003db RGS |
649 | prune the search space by not bothering to match any closer to the end of a |
650 | string than would allow a match. For instance there is no point in even | |
651 | starting the regex engine if the minlen is 10 but the string is only 5 | |
652 | characters long. There is no way that the pattern can match. | |
653 | ||
2d608413 KW |
654 | C<minlenret> is the minimum length (in characters) of the string that would |
655 | be found in $& after a match. | |
108003db RGS |
656 | |
657 | The difference between C<minlen> and C<minlenret> can be seen in the | |
658 | following pattern: | |
659 | ||
660 | /ns(?=\d)/ | |
661 | ||
662 | where the C<minlen> would be 3 but C<minlenret> would only be 2 as the \d is | |
663 | required to match but is not actually included in the matched content. This | |
664 | distinction is particularly important as the substitution logic uses the | |
5a2b28ce KW |
665 | C<minlenret> to tell if it can do in-place substitutions (these can |
666 | result in considerable speed-up). | |
108003db | 667 | |
882227b7 | 668 | =head2 C<gofs> |
108003db RGS |
669 | |
670 | Left offset from pos() to start match at. | |
671 | ||
882227b7 | 672 | =head2 C<substrs> |
108003db | 673 | |
192b9cd1 | 674 | Substring data about strings that must appear in the final match. This |
5a2b28ce | 675 | is currently only used internally by Perl's engine, but might be |
c998b245 | 676 | used in the future for all engines for optimisations. |
108003db | 677 | |
1cecf2c0 | 678 | =head2 C<nparens>, C<lastparen>, and C<lastcloseparen> |
108003db RGS |
679 | |
680 | These fields are used to keep track of how many paren groups could be matched | |
681 | in the pattern, which was the last open paren to be entered, and which was | |
682 | the last close paren to be entered. | |
683 | ||
882227b7 | 684 | =head2 C<intflags> |
108003db RGS |
685 | |
686 | The engine's private copy of the flags the pattern was compiled with. Usually | |
192b9cd1 | 687 | this is the same as C<extflags> unless the engine chose to modify one of them. |
108003db | 688 | |
882227b7 | 689 | =head2 C<pprivate> |
108003db | 690 | |
5a2b28ce | 691 | A void* pointing to an engine-defined data structure. The Perl engine uses the |
108003db RGS |
692 | C<regexp_internal> structure (see L<perlreguts/Base Structures>) but a custom |
693 | engine should use something else. | |
694 | ||
882227b7 | 695 | =head2 C<swap> |
108003db | 696 | |
5a2b28ce | 697 | Unused. Left in for compatibility with Perl 5.10.0. |
108003db | 698 | |
882227b7 | 699 | =head2 C<offs> |
108003db RGS |
700 | |
701 | A C<regexp_paren_pair> structure which defines offsets into the string being | |
702 | matched which correspond to the C<$&> and C<$1>, C<$2> etc. captures, the | |
703 | C<regexp_paren_pair> struct is defined as follows: | |
704 | ||
705 | typedef struct regexp_paren_pair { | |
706 | I32 start; | |
707 | I32 end; | |
708 | } regexp_paren_pair; | |
709 | ||
710 | If C<< ->offs[num].start >> or C<< ->offs[num].end >> is C<-1> then that | |
c27a5cfe | 711 | capture group did not match. C<< ->offs[0].start/end >> represents C<$&> (or |
c149d39e | 712 | C<${^MATCH}> under C<//p>) and C<< ->offs[paren].end >> matches C<$$paren> where |
108003db RGS |
713 | C<$paren >= 1>. |
714 | ||
882227b7 | 715 | =head2 C<precomp> C<prelen> |
108003db | 716 | |
192b9cd1 AB |
717 | Used for optimisations. C<precomp> holds a copy of the pattern that |
718 | was compiled and C<prelen> its length. When a new pattern is to be | |
719 | compiled (such as inside a loop) the internal C<regcomp> operator | |
5a2b28ce | 720 | checks if the last compiled C<REGEXP>'s C<precomp> and C<prelen> |
192b9cd1 AB |
721 | are equivalent to the new one, and if so uses the old pattern instead |
722 | of compiling a new one. | |
723 | ||
724 | The relevant snippet from C<Perl_pp_regcomp>: | |
725 | ||
726 | if (!re || !re->precomp || re->prelen != (I32)len || | |
727 | memNE(re->precomp, t, len)) | |
728 | /* Compile a new pattern */ | |
108003db | 729 | |
882227b7 | 730 | =head2 C<paren_names> |
108003db | 731 | |
c27a5cfe | 732 | This is a hash used internally to track named capture groups and their |
108003db RGS |
733 | offsets. The keys are the names of the buffers the values are dualvars, |
734 | with the IV slot holding the number of buffers with the given name and the | |
735 | pv being an embedded array of I32. The values may also be contained | |
736 | independently in the data array in cases where named backreferences are | |
737 | used. | |
738 | ||
c998b245 | 739 | =head2 C<substrs> |
108003db RGS |
740 | |
741 | Holds information on the longest string that must occur at a fixed | |
742 | offset from the start of the pattern, and the longest string that must | |
743 | occur at a floating offset from the start of the pattern. Used to do | |
744 | Fast-Boyer-Moore searches on the string to find out if its worth using | |
745 | the regex engine at all, and if so where in the string to search. | |
746 | ||
6502e081 DM |
747 | =head2 C<subbeg> C<sublen> C<saved_copy> C<suboffset> C<subcoffset> |
748 | ||
749 | Used during the execution phase for managing search and replace patterns, | |
750 | and for providing the text for C<$&>, C<$1> etc. C<subbeg> points to a | |
751 | buffer (either the original string, or a copy in the case of | |
752 | C<RX_MATCH_COPIED(rx)>), and C<sublen> is the length of the buffer. The | |
753 | C<RX_OFFS> start and end indices index into this buffer. | |
754 | ||
755 | In the presence of the C<REXEC_COPY_STR> flag, but with the addition of | |
756 | the C<REXEC_COPY_SKIP_PRE> or C<REXEC_COPY_SKIP_POST> flags, an engine | |
757 | can choose not to copy the full buffer (although it must still do so in | |
758 | the presence of C<RXf_PMf_KEEPCOPY> or the relevant bits being set in | |
759 | C<PL_sawampersand>). In this case, it may set C<suboffset> to indicate the | |
760 | number of bytes from the logical start of the buffer to the physical start | |
761 | (i.e. C<subbeg>). It should also set C<subcoffset>, the number of | |
762 | characters in the offset. The latter is needed to support C<@-> and C<@+> | |
763 | which work in characters, not bytes. | |
108003db | 764 | |
882227b7 | 765 | =head2 C<wrapped> C<wraplen> |
108003db | 766 | |
5a2b28ce | 767 | Stores the string C<qr//> stringifies to. The Perl engine for example |
ed215d3c | 768 | stores C<(?^:eek)> in the case of C<qr/eek/>. |
108003db | 769 | |
c998b245 AB |
770 | When using a custom engine that doesn't support the C<(?:)> construct |
771 | for inline modifiers, it's probably best to have C<qr//> stringify to | |
772 | the supplied pattern, note that this will create undesired patterns in | |
773 | cases such as: | |
108003db RGS |
774 | |
775 | my $x = qr/a|b/; # "a|b" | |
192b9cd1 | 776 | my $y = qr/c/i; # "c" |
108003db RGS |
777 | my $z = qr/$x$y/; # "a|bc" |
778 | ||
192b9cd1 AB |
779 | There's no solution for this problem other than making the custom |
780 | engine understand a construct like C<(?:)>. | |
108003db | 781 | |
882227b7 | 782 | =head2 C<seen_evals> |
108003db RGS |
783 | |
784 | This stores the number of eval groups in the pattern. This is used for security | |
785 | purposes when embedding compiled regexes into larger patterns with C<qr//>. | |
786 | ||
882227b7 | 787 | =head2 C<refcnt> |
108003db | 788 | |
5a2b28ce | 789 | The number of times the structure is referenced. When this falls to 0, the |
108003db RGS |
790 | regexp is automatically freed by a call to pregfree. This should be set to 1 in |
791 | each engine's L</comp> routine. | |
792 | ||
108003db RGS |
793 | =head1 HISTORY |
794 | ||
795 | Originally part of L<perlreguts>. | |
796 | ||
797 | =head1 AUTHORS | |
798 | ||
799 | Originally written by Yves Orton, expanded by E<AElig>var ArnfjE<ouml>rE<eth> | |
800 | Bjarmason. | |
801 | ||
802 | =head1 LICENSE | |
803 | ||
804 | Copyright 2006 Yves Orton and 2007 E<AElig>var ArnfjE<ouml>rE<eth> Bjarmason. | |
805 | ||
806 | This program is free software; you can redistribute it and/or modify it under | |
807 | the same terms as Perl itself. | |
808 | ||
809 | =cut |