diff options
Diffstat (limited to 'contrib/perl5/regexp.h')
| -rw-r--r-- | contrib/perl5/regexp.h | 131 |
1 files changed, 68 insertions, 63 deletions
diff --git a/contrib/perl5/regexp.h b/contrib/perl5/regexp.h index fbc9237..ca0e9ed 100644 --- a/contrib/perl5/regexp.h +++ b/contrib/perl5/regexp.h @@ -17,78 +17,62 @@ struct regnode { typedef struct regnode regnode; -struct reg_data { - U32 count; - U8 *what; - void* data[1]; -}; - -struct reg_substr_datum { - I32 min_offset; - I32 max_offset; - SV *substr; -}; - -struct reg_substr_data { - struct reg_substr_datum data[3]; /* Actual array */ -}; +struct reg_substr_data; typedef struct regexp { - I32 refcnt; - char **startp; - char **endp; + I32 *startp; + I32 *endp; regnode *regstclass; + struct reg_substr_data *substrs; + char *precomp; /* pre-compilation regular expression */ + struct reg_data *data; /* Additional data. */ + char *subbeg; /* saved or original string + so \digit works forever. */ + I32 sublen; /* Length of string pointed by subbeg */ + I32 refcnt; I32 minlen; /* mininum possible length of $& */ I32 prelen; /* length of precomp */ U32 nparens; /* number of parentheses */ U32 lastparen; /* last paren matched */ - char *precomp; /* pre-compilation regular expression */ - char *subbase; /* saved string so \digit works forever */ - char *subbeg; /* same, but not responsible for allocation */ - char *subend; /* end of subbase */ - U16 naughty; /* how exponential is this pattern? */ - U16 reganch; /* Internal use only + + U32 reganch; /* Internal use only + Tainted information used by regexec? */ -#if 0 - SV *anchored_substr; /* Substring at fixed position wrt start. */ - I32 anchored_offset; /* Position of it. */ - SV *float_substr; /* Substring at variable position wrt start. */ - I32 float_min_offset; /* Minimal position of it. */ - I32 float_max_offset; /* Maximal position of it. */ - SV *check_substr; /* Substring to check before matching. */ - I32 check_offset_min; /* Offset of the above. */ - I32 check_offset_max; /* Offset of the above. */ -#else - struct reg_substr_data *substrs; -#endif - struct reg_data *data; /* Additional data. */ regnode program[1]; /* Unwarranted chumminess with compiler. */ } regexp; -#define anchored_substr substrs->data[0].substr -#define anchored_offset substrs->data[0].min_offset -#define float_substr substrs->data[1].substr -#define float_min_offset substrs->data[1].min_offset -#define float_max_offset substrs->data[1].max_offset -#define check_substr substrs->data[2].substr -#define check_offset_min substrs->data[2].min_offset -#define check_offset_max substrs->data[2].max_offset - -#define ROPT_ANCH (ROPT_ANCH_BOL|ROPT_ANCH_MBOL|ROPT_ANCH_GPOS) -#define ROPT_ANCH_SINGLE (ROPT_ANCH_BOL|ROPT_ANCH_GPOS) -#define ROPT_ANCH_BOL 1 -#define ROPT_ANCH_MBOL 2 -#define ROPT_ANCH_GPOS 4 -#define ROPT_SKIP 8 -#define ROPT_IMPLICIT 0x10 /* Converted .* to ^.* */ -#define ROPT_NOSCAN 0x20 /* Check-string always at start. */ -#define ROPT_GPOS_SEEN 0x40 -#define ROPT_CHECK_ALL 0x80 -#define ROPT_LOOKBEHIND_SEEN 0x100 -#define ROPT_EVAL_SEEN 0x200 -#define ROPT_TAINTED_SEEN 0x400 +#define ROPT_ANCH (ROPT_ANCH_BOL|ROPT_ANCH_MBOL|ROPT_ANCH_GPOS|ROPT_ANCH_SBOL) +#define ROPT_ANCH_SINGLE (ROPT_ANCH_SBOL|ROPT_ANCH_GPOS) +#define ROPT_ANCH_BOL 0x00001 +#define ROPT_ANCH_MBOL 0x00002 +#define ROPT_ANCH_SBOL 0x00004 +#define ROPT_ANCH_GPOS 0x00008 +#define ROPT_SKIP 0x00010 +#define ROPT_IMPLICIT 0x00020 /* Converted .* to ^.* */ +#define ROPT_NOSCAN 0x00040 /* Check-string always at start. */ +#define ROPT_GPOS_SEEN 0x00080 +#define ROPT_CHECK_ALL 0x00100 +#define ROPT_LOOKBEHIND_SEEN 0x00200 +#define ROPT_EVAL_SEEN 0x00400 + /* 0xf800 of reganch is used by PMf_COMPILETIME */ +#define ROPT_UTF8 0x10000 +#define ROPT_NAUGHTY 0x20000 /* how exponential is this pattern? */ +#define ROPT_COPY_DONE 0x40000 /* subbeg is a copy of the string */ +#define ROPT_TAINTED_SEEN 0x80000 + +#define RE_USE_INTUIT_NOML 0x0100000 /* Best to intuit before matching */ +#define RE_USE_INTUIT_ML 0x0200000 +#define REINT_AUTORITATIVE_NOML 0x0400000 /* Can trust a positive answer */ +#define REINT_AUTORITATIVE_ML 0x0800000 +#define REINT_ONCE_NOML 0x1000000 /* Intuit can succed once only. */ +#define REINT_ONCE_ML 0x2000000 +#define RE_INTUIT_ONECHAR 0x4000000 +#define RE_INTUIT_TAIL 0x8000000 + +#define RE_USE_INTUIT (RE_USE_INTUIT_NOML|RE_USE_INTUIT_ML) +#define REINT_AUTORITATIVE (REINT_AUTORITATIVE_NOML|REINT_AUTORITATIVE_ML) +#define REINT_ONCE (REINT_ONCE_NOML|REINT_ONCE_ML) + #define RX_MATCH_TAINTED(prog) ((prog)->reganch & ROPT_TAINTED_SEEN) #define RX_MATCH_TAINTED_on(prog) ((prog)->reganch |= ROPT_TAINTED_SEEN) #define RX_MATCH_TAINTED_off(prog) ((prog)->reganch &= ~ROPT_TAINTED_SEEN) @@ -96,8 +80,29 @@ typedef struct regexp { ? RX_MATCH_TAINTED_on(prog) \ : RX_MATCH_TAINTED_off(prog)) -#define REXEC_COPY_STR 1 /* Need to copy the string. */ -#define REXEC_CHECKED 2 /* check_substr already checked. */ +#define RX_MATCH_COPIED(prog) ((prog)->reganch & ROPT_COPY_DONE) +#define RX_MATCH_COPIED_on(prog) ((prog)->reganch |= ROPT_COPY_DONE) +#define RX_MATCH_COPIED_off(prog) ((prog)->reganch &= ~ROPT_COPY_DONE) +#define RX_MATCH_COPIED_set(prog,t) ((t) \ + ? RX_MATCH_COPIED_on(prog) \ + : RX_MATCH_COPIED_off(prog)) + +#define REXEC_COPY_STR 0x01 /* Need to copy the string. */ +#define REXEC_CHECKED 0x02 /* check_substr already checked. */ +#define REXEC_SCREAM 0x04 /* use scream table. */ +#define REXEC_IGNOREPOS 0x08 /* \G matches at start. */ +#define REXEC_NOT_FIRST 0x10 /* This is another iteration of //g. */ +#define REXEC_ML 0x20 /* $* was set. */ + +#define ReREFCNT_inc(re) ((void)(re && re->refcnt++), re) +#define ReREFCNT_dec(re) CALLREGFREE(aTHX_ re) + +#define FBMcf_TAIL_DOLLAR 1 +#define FBMcf_TAIL_DOLLARM 2 +#define FBMcf_TAIL_Z 4 +#define FBMcf_TAIL_z 8 +#define FBMcf_TAIL (FBMcf_TAIL_DOLLAR|FBMcf_TAIL_DOLLARM|FBMcf_TAIL_Z|FBMcf_TAIL_z) + +#define FBMrf_MULTILINE 1 -#define ReREFCNT_inc(re) ((re && re->refcnt++), re) -#define ReREFCNT_dec(re) pregfree(re) +struct re_scream_pos_data_s; |
