usr.bin/perl/regexec.c

54  *	2. The origin of this software must not be misrepresented, either
57  *	3. Altered versions must be plainly marked as such, and must not
138 /* like HOPMAYBE3 but backwards. lim must be +ve. Returns NULL on overshoot */
151 /* lim must be +ve. Returns NULL on overshoot */
160  * off must be >=0; args should be vars rather than expressions */
178  * VERBS must NOT be included. JUMPABLE is used to determine  if we can ignore a
199   Search for mandatory following text node; for lookahead, the text must
561     /* Compare non-UTF-8 using Unicode (Latin1) semantics.  s2 must already be
855 /* minend:    end of match must be >= minend bytes after stringarg. */
1081              * start of the regex) substr must also be anchored relative
1101                      * Unless it's a multiline match, the lengths must match
1280         SV* must;
1293          * This means that the regex origin must lie somewhere
1297          * So the fixed  substr must lie somewhere between
1337          * to CHR_SVLEN(must) - !!SvTAIL(must) + prog->float_end_shift
1375         must = utf8_target ? other->utf8_substr : other->substr;
1376         assert(SvPOK(must));
1379             char *to   = last + SvCUR(must) - (SvTAIL(must)!=0);
1395                     must,
1409                 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
1413                 quoted, RE_SV_TAIL(must));
1532         /* if we get here, the check substr must have been float,
1980  * occurrence, and there is a further condition that must be matched besides
2018  * match ascii, but it must advance based on UTF-8.   The other three handle
2022  * must be different.  Find the "wordness" of the character just prior to this
2045  * at entry to the loop, and to get to the IF_FAIL branch, tmp must equal
2297     /* We know what class it must start with. The case statements below have
3420                             /* we must be accepting here */
3656 /* minend:    end of match must be >= minend bytes after stringarg. */
3668     SSize_t minlen;		/* must match at least this many chars */
3834         /* SAVEFREESV, not sv_mortalcopy, as this SV must last until after
3982         /* it must be a one character string (XXXX Except is_utf8_pat?) */
4029         SV *must;
4042                 must = prog->anchored_utf8;
4050                 must = prog->anchored_substr;
4058                 must = prog->float_utf8;
4066                 must = prog->float_substr;
4076                   -(SSize_t)(CHR_SVLEN(must)
4077                          - (SvTAIL(must) != 0) + back_min), strbeg);
4085            check_substr==must. */
4090                                   (unsigned char*)strend, must,
4125                 SvPVX_const(must), RE_SV_DUMPLEN(must), 30);
4127                               ((must == prog->anchored_substr || must == prog->anchored_utf8)
4129                 quoted, RE_SV_TAIL(must));
4695      * if there is anything in the node that can't, the match must fail */
4719         if (utf8_pat) { /* Here, must have been EXACTFL, so look at the
4978             /* In /iaa nodes, neither or both must be ASCII to be a legal fold
5078              * single zero bit.  If a string matches those, masked, it must be
7146                                       must be UTF-8. */
8224                  * embedded qr//s), then we must ensure that during any
8752 must always point to the *current* CURLYX block, the rules are:
9469                     /* A{m,n} must come at the end of the string, there's
9537                          * means those bytes must match identically, so can use
10058         /* we have successfully completed a subexpression, but we must now
10553         /* we know the first byte must be the FLAGS field */
11336     /* free all slabs above current one - this must be the last action
11498 Additionally all decimal digits must come from the same consecutive sequence of
11508 such as Arabic, have more than one set of digits.  All digits must come from
11581     /* All code points in 0..255 are either Common or Latin, so must be a
11686         /* For the character's script to be Unknown, it must be the first
11873          * know that for this to be a digit, it must be in the range of +0..+9