X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=blobdiff_plain;f=regexp.h;h=33e7c2091ec7a27c78e79170e88a748aa6a32518;hb=c944cfb99a453bf0989f7e0433237ee9661e575b;hp=abc744c38da88106b9017d901e80341b5c2e2dd7;hpb=3c6f95d1b67124c82147911f846eb1993443a39a;p=p5sagit%2Fp5-mst-13.2.git diff --git a/regexp.h b/regexp.h index abc744c..33e7c20 100644 --- a/regexp.h +++ b/regexp.h @@ -1,7 +1,7 @@ /* regexp.h * * Copyright (C) 1993, 1994, 1996, 1997, 1999, 2000, 2001, 2003, - * by Larry Wall and others + * 2005, 2006 by Larry Wall and others * * You may distribute under the terms of either the GNU General Public * License or the Artistic License, as specified in the README file. @@ -15,8 +15,8 @@ * not the System V one. */ #ifndef PLUGGABLE_RE_EXTENSION -/* we don't want to include this stuff if we are inside Nicholas' - * pluggable regex engine code */ +/* we don't want to include this stuff if we are inside of + an external regex engine based on the core one - like re 'debug'*/ struct regnode { U8 flags; @@ -31,22 +31,36 @@ struct reg_substr_data; struct reg_data; struct regexp_engine; - -typedef struct regexp_paren_ofs { - I32 *startp; - I32 *endp; -} regexp_paren_ofs; +struct regexp; + +struct reg_substr_datum { + I32 min_offset; + I32 max_offset; + SV *substr; /* non-utf8 variant */ + SV *utf8_substr; /* utf8 variant */ + I32 end_shift; +}; +struct reg_substr_data { + struct reg_substr_datum data[3]; /* Actual array */ +}; #ifdef PERL_OLD_COPY_ON_WRITE #define SV_SAVED_COPY SV *saved_copy; /* If non-NULL, SV which is COW from original */ #else #define SV_SAVED_COPY #endif + +typedef struct regexp_paren_pair { + I32 start; + I32 end; +} regexp_paren_pair; + /* this is ordered such that the most commonly used fields are at the start of the struct */ typedef struct regexp { /* what engine created this regexp? */ const struct regexp_engine* engine; + struct regexp* mother_re; /* what re is this a lightweight copy of? */ /* Information about the match that the perl core uses to manage things */ U32 extflags; /* Flags used both externally and internally */ @@ -65,17 +79,21 @@ typedef struct regexp { /* Data about the last/current match. These are modified during matching*/ U32 lastparen; /* last open paren matched */ U32 lastcloseparen; /* last close paren matched */ - I32 *startp; /* Array of offsets from start of string (@-) */ - I32 *endp; /* Array of offsets from start of string (@+) */ + regexp_paren_pair *swap; /* Swap copy of *offs */ + regexp_paren_pair *offs; /* Array of offsets for (@-) and (@+) */ + char *subbeg; /* saved or original string so \digit works forever. */ - I32 sublen; /* Length of string pointed by subbeg */ SV_SAVED_COPY /* If non-NULL, SV which is COW from original */ + I32 sublen; /* Length of string pointed by subbeg */ /* Information about the match that isn't often used */ - char *precomp; /* pre-compilation regular expression */ I32 prelen; /* length of precomp */ + const char *precomp; /* pre-compilation regular expression */ + /* wrapped can't be const char*, as it is returned by sv_2pv_flags */ + char *wrapped; /* wrapped version of the pattern */ + I32 wraplen; /* length of wrapped */ I32 seen_evals; /* number of eval groups in the pattern - for security checks */ HV *paren_names; /* Optional hash of paren names */ @@ -83,34 +101,18 @@ typedef struct regexp { I32 refcnt; /* Refcount of this regexp */ } regexp; - -typedef struct regexp_internal { - regexp_paren_ofs *swap; /* Swap copy of *startp / *endp */ - U32 *offsets; /* offset annotations 20001228 MJD - data about mapping the program to the - string*/ - regnode *regstclass; /* Optional startclass as identified or constructed - by the optimiser */ - struct reg_data *data; /* Additional miscellaneous data used by the program. - Used to make it easier to clone and free arbitrary - data that the regops need. Often the ARG field of - a regop is an index into this structure */ - regnode program[1]; /* Unwarranted chumminess with compiler. */ -} regexp_internal; - -#define RXi_SET(x,y) (x)->pprivate = (void*)(y) -#define RXi_GET(x) ((regexp_internal *)((x)->pprivate)) -#define RXi_GET_DECL(r,ri) regexp_internal *ri = RXi_GET(r) - +/* used for high speed searches */ typedef struct re_scream_pos_data_s { char **scream_olds; /* match pos */ I32 *scream_pos; /* Internal iterator of scream. */ } re_scream_pos_data; +/* regexp_engine structure. This is the dispatch table for regexes. + * Any regex engine implementation must be able to build one of these. + */ typedef struct regexp_engine { - regexp* (*comp) (pTHX_ char* exp, char* xend, PMOP* pm); - regexp* (*compile) (pTHX_ char *exp, char *xend, PMOP *pm); + REGEXP* (*comp) (pTHX_ const SV * const pattern, const U32 flags); I32 (*exec) (pTHX_ regexp* prog, char* stringarg, char* strend, char* strbeg, I32 minend, SV* screamer, void* data, U32 flags); @@ -119,24 +121,20 @@ typedef struct regexp_engine { struct re_scream_pos_data_s *data); SV* (*checkstr) (pTHX_ regexp *prog); void (*free) (pTHX_ struct regexp* r); + SV* (*numbered_buff_get) (pTHX_ const REGEXP * const rx, I32 paren, SV* usesv); + SV* (*named_buff_get)(pTHX_ const REGEXP * const rx, SV* namesv, U32 flags); + SV* (*qr_pkg)(pTHX_ const REGEXP * const rx); #ifdef USE_ITHREADS void* (*dupe) (pTHX_ const regexp *r, CLONE_PARAMS *param); #endif } regexp_engine; -/* - * Flags stored in regexp->intflags - * These are used only internally to the regexp engine - */ -#define PREGf_SKIP 0x00000001 -#define PREGf_IMPLICIT 0x00000002 /* Converted .* to ^.* */ -#define PREGf_NAUGHTY 0x00000004 /* how exponential is this pattern? */ -#define PREGf_VERBARG_SEEN 0x00000008 -#define PREGf_CUTGROUP_SEEN 0x00000010 - - /* Flags stored in regexp->extflags * These are used by code external to the regexp engine + * + * Note that flags starting with RXf_PMf_ have exact equivalents + * stored in op_pmflags and which are defined in op.h, they are defined + * numerically here only for clarity. */ /* Anchor and GPOS related stuff */ @@ -146,24 +144,66 @@ typedef struct regexp_engine { #define RXf_ANCH_GPOS 0x00000008 #define RXf_GPOS_SEEN 0x00000010 #define RXf_GPOS_FLOAT 0x00000020 -/* five bits here */ +/* two bits here */ #define RXf_ANCH (RXf_ANCH_BOL|RXf_ANCH_MBOL|RXf_ANCH_GPOS|RXf_ANCH_SBOL) #define RXf_GPOS_CHECK (RXf_GPOS_SEEN|RXf_ANCH_GPOS) -#define RXf_ANCH_SINGLE (RXf_ANCH_SBOL|RXf_ANCH_GPOS) -/* - * 0xF800 of extflags is used by PMf_COMPILETIME - * These are the regex equivelent of the PMf_xyz stuff defined - * in op.h +#define RXf_ANCH_SINGLE (RXf_ANCH_SBOL|RXf_ANCH_GPOS) + +/* Flags indicating special patterns */ +#define RXf_SKIPWHITE 0x00000100 /* Pattern is for a split / / */ +#define RXf_START_ONLY 0x00000200 /* Pattern is /^/ */ +#define RXf_WHITE 0x00000400 /* Pattern is /\s+/ */ + +/* 0x1F800 of extflags is used by (RXf_)PMf_COMPILETIME */ +#define RXf_PMf_LOCALE 0x00000800 /* use locale */ +#define RXf_PMf_MULTILINE 0x00001000 /* /m */ +#define RXf_PMf_SINGLELINE 0x00002000 /* /s */ +#define RXf_PMf_FOLD 0x00004000 /* /i */ +#define RXf_PMf_EXTENDED 0x00008000 /* /x */ +#define RXf_PMf_KEEPCOPY 0x00010000 /* /k */ +/* these flags are transfered from the PMOP->op_pmflags member during compilation */ +#define RXf_PMf_STD_PMMOD (RXf_PMf_MULTILINE|RXf_PMf_SINGLELINE|RXf_PMf_FOLD|RXf_PMf_EXTENDED) +#define RXf_PMf_COMPILETIME (RXf_PMf_MULTILINE|RXf_PMf_SINGLELINE|RXf_PMf_LOCALE|RXf_PMf_FOLD|RXf_PMf_EXTENDED|RXf_PMf_KEEPCOPY) + +#define CASE_STD_PMMOD_FLAGS_PARSE_SET(pmfl) \ + case IGNORE_PAT_MOD: *(pmfl) |= RXf_PMf_FOLD; break; \ + case MULTILINE_PAT_MOD: *(pmfl) |= RXf_PMf_MULTILINE; break; \ + case SINGLE_PAT_MOD: *(pmfl) |= RXf_PMf_SINGLELINE; break; \ + case XTENDED_PAT_MOD: *(pmfl) |= RXf_PMf_EXTENDED; break + +/* chars and strings used as regex pattern modifiers + * Singlular is a 'c'har, plural is a "string" + * + * NOTE, KEEPCOPY was originally 'k', but was changed to 'p' for preserve + * for compatibility reasons with Regexp::Common which highjacked (?k:...) + * for its own uses. So 'k' is out as well. */ -#define RXf_PMf_LOCALE 0x00000800 -#define RXf_PMf_MULTILINE 0x00001000 -#define RXf_PMf_SINGLELINE 0x00002000 -#define RXf_PMf_FOLD 0x00004000 -#define RXf_PMf_EXTENDED 0x00008000 -#define RXf_PMf_COMPILETIME (RXf_PMf_MULTILINE|RXf_PMf_SINGLELINE|RXf_PMf_LOCALE|RXf_PMf_FOLD|RXf_PMf_EXTENDED) +#define EXEC_PAT_MOD 'e' +#define KEEPCOPY_PAT_MOD 'p' +#define ONCE_PAT_MOD 'o' +#define GLOBAL_PAT_MOD 'g' +#define CONTINUE_PAT_MOD 'c' +#define MULTILINE_PAT_MOD 'm' +#define SINGLE_PAT_MOD 's' +#define IGNORE_PAT_MOD 'i' +#define XTENDED_PAT_MOD 'x' + +#define ONCE_PAT_MODS "o" +#define KEEPCOPY_PAT_MODS "p" +#define EXEC_PAT_MODS "e" +#define LOOP_PAT_MODS "gc" + +#define STD_PAT_MODS "msix" + +#define INT_PAT_MODS STD_PAT_MODS KEEPCOPY_PAT_MODS + +#define EXT_PAT_MODS ONCE_PAT_MODS KEEPCOPY_PAT_MODS +#define QR_PAT_MODS STD_PAT_MODS EXT_PAT_MODS +#define M_PAT_MODS QR_PAT_MODS LOOP_PAT_MODS +#define S_PAT_MODS M_PAT_MODS EXEC_PAT_MODS + /* What we have seen */ -/* one bit here */ #define RXf_LOOKBEHIND_SEEN 0x00020000 #define RXf_EVAL_SEEN 0x00040000 #define RXf_CANY_SEEN 0x00080000 @@ -186,7 +226,7 @@ typedef struct regexp_engine { /* Copy and tainted info */ #define RXf_COPY_DONE 0x10000000 #define RXf_TAINTED_SEEN 0x20000000 -/* two bits here */ +#define RXf_TAINTED 0x80000000 /* this pattern is tainted */ #define RX_HAS_CUTGROUP(prog) ((prog)->intflags & PREGf_CUTGROUP_SEEN) @@ -271,7 +311,9 @@ typedef struct { /* structures for holding and saving the state maintained by regmatch() */ -#define MAX_RECURSE_EVAL_NOCHANGE_DEPTH 50 +#ifndef MAX_RECURSE_EVAL_NOCHANGE_DEPTH +#define MAX_RECURSE_EVAL_NOCHANGE_DEPTH 1000 +#endif typedef I32 CHECKPOINT; @@ -288,16 +330,41 @@ typedef struct regmatch_state { struct regmatch_state *prev_yes_state; } yes; - struct { + /* branchlike members */ + /* this is a fake union member that matches the first elements + * of each member that needs to behave like a branch */ + struct { /* this first element must match u.yes */ struct regmatch_state *prev_yes_state; - reg_trie_accepted *accept_buff; + U32 lastparen; + CHECKPOINT cp; + + } branchlike; + + struct { + /* the first elements must match u.branchlike */ + struct regmatch_state *prev_yes_state; + U32 lastparen; + CHECKPOINT cp; + + regnode *next_branch; /* next branch node */ + } branch; + + struct { + /* the first elements must match u.branchlike */ + struct regmatch_state *prev_yes_state; + U32 lastparen; + CHECKPOINT cp; + + reg_trie_accepted *accept_buff; /* accepting states we have seen */ U32 accepted; /* how many accepting states we have seen */ U16 *jump; /* positive offsets from me */ regnode *B; /* node following the trie */ regnode *me; /* Which node am I - needed for jump tries*/ } trie; + /* special types - these members are used to store state for special + regops like eval, if/then, lookaround and the markpoint state */ struct { /* this first element must match u.yes */ struct regmatch_state *prev_yes_state; @@ -309,20 +376,42 @@ typedef struct regmatch_state { inner and outer rexen */ CHECKPOINT cp; /* remember current savestack indexes */ CHECKPOINT lastcp; - regnode *B; /* the node following us */ U32 close_paren; /* which close bracket is our end */ + regnode *B; /* the node following us */ } eval; struct { /* this first element must match u.yes */ struct regmatch_state *prev_yes_state; + I32 wanted; + I32 logical; /* saved copy of 'logical' var */ + regnode *me; /* the IFMATCH/SUSPEND/UNLESSM node */ + } ifmatch; /* and SUSPEND/UNLESSM */ + + struct { + /* this first element must match u.yes */ + struct regmatch_state *prev_yes_state; + struct regmatch_state *prev_mark; + SV* mark_name; + char *mark_loc; + } mark; + + struct { + int val; + } keeper; + + /* quantifiers - these members are used for storing state for + for the regops used to implement quantifiers */ + struct { + /* this first element must match u.yes */ + struct regmatch_state *prev_yes_state; struct regmatch_state *prev_curlyx; /* previous cur_curlyx */ + regnode *A, *B; /* the nodes corresponding to /A*B/ */ CHECKPOINT cp; /* remember current savestack index */ bool minmod; int parenfloor;/* how far back to strip paren data */ int min; /* the minimal number of A's to match */ int max; /* the maximal number of A's to match */ - regnode *A, *B; /* the nodes corresponding to /A*B/ */ /* these two are modified by WHILEM */ int count; /* how many instances of A we've matched */ @@ -343,14 +432,6 @@ typedef struct regmatch_state { struct { /* this first element must match u.yes */ struct regmatch_state *prev_yes_state; - U32 lastparen; - regnode *next_branch; /* next branch node */ - CHECKPOINT cp; - } branch; - - struct { - /* this first element must match u.yes */ - struct regmatch_state *prev_yes_state; I32 c1, c2; /* case fold search */ CHECKPOINT cp; I32 alen; /* length of first-matched A string */ @@ -371,21 +452,6 @@ typedef struct regmatch_state { regnode *A, *B; /* the nodes corresponding to /A*B/ */ } curly; /* and CURLYN/PLUS/STAR */ - struct { - /* this first element must match u.yes */ - struct regmatch_state *prev_yes_state; - I32 wanted; - I32 logical; /* saved copy of 'logical' var */ - regnode *me; /* the IFMATCH/SUSPEND/UNLESSM node */ - } ifmatch; /* and SUSPEND/UNLESSM */ - - struct { - /* this first element must match u.yes */ - struct regmatch_state *prev_yes_state; - struct regmatch_state *prev_mark; - SV* mark_name; - char *mark_loc; - } mark; } u; } regmatch_state; @@ -405,8 +471,7 @@ typedef struct regmatch_slab { #define PL_bostr PL_reg_state.re_state_bostr #define PL_reginput PL_reg_state.re_state_reginput #define PL_regeol PL_reg_state.re_state_regeol -#define PL_regstartp PL_reg_state.re_state_regstartp -#define PL_regendp PL_reg_state.re_state_regendp +#define PL_regoffs PL_reg_state.re_state_regoffs #define PL_reglastparen PL_reg_state.re_state_reglastparen #define PL_reglastcloseparen PL_reg_state.re_state_reglastcloseparen #define PL_reg_start_tmp PL_reg_state.re_state_reg_start_tmp @@ -429,28 +494,27 @@ typedef struct regmatch_slab { struct re_save_state { U32 re_state_reg_flags; /* from regexec.c */ + U32 re_state_reg_start_tmpl; /* from regexec.c */ + I32 re_state_reg_eval_set; /* from regexec.c */ + bool re_state_reg_match_utf8; /* from regexec.c */ char *re_state_bostr; char *re_state_reginput; /* String-input pointer. */ char *re_state_regeol; /* End of input, for $ check. */ - I32 *re_state_regstartp; /* Pointer to startp array. */ - I32 *re_state_regendp; /* Ditto for endp. */ + regexp_paren_pair *re_state_regoffs; /* Pointer to start/end pairs */ U32 *re_state_reglastparen; /* Similarly for lastparen. */ U32 *re_state_reglastcloseparen; /* Similarly for lastcloseparen. */ char **re_state_reg_start_tmp; /* from regexec.c */ - U32 re_state_reg_start_tmpl; /* from regexec.c */ - I32 re_state_reg_eval_set; /* from regexec.c */ - bool re_state_reg_match_utf8; /* from regexec.c */ MAGIC *re_state_reg_magic; /* from regexec.c */ - I32 re_state_reg_oldpos; /* from regexec.c */ PMOP *re_state_reg_oldcurpm; /* from regexec.c */ PMOP *re_state_reg_curpm; /* from regexec.c */ char *re_state_reg_oldsaved; /* old saved substr during match */ STRLEN re_state_reg_oldsavedlen; /* old length of saved substr during match */ + STRLEN re_state_reg_poscache_size; /* size of pos cache of WHILEM */ + I32 re_state_reg_oldpos; /* from regexec.c */ I32 re_state_reg_maxiter; /* max wait until caching pos */ I32 re_state_reg_leftiter; /* wait until caching pos */ - char *re_state_reg_poscache; /* cache of pos of WHILEM */ - STRLEN re_state_reg_poscache_size; /* size of pos cache of WHILEM */ U32 re_state_regsize; /* from regexec.c */ + char *re_state_reg_poscache; /* cache of pos of WHILEM */ char *re_state_reg_starttry; /* from regexec.c */ #ifdef PERL_OLD_COPY_ON_WRITE SV *re_state_nrs; /* was placeholder: unused since 5.8.0 (5.7.2 patch #12027 for bug ID 20010815.012). Used to save rx->saved_copy */ @@ -459,6 +523,7 @@ struct re_save_state { #define SAVESTACK_ALLOC_FOR_RE_SAVE_STATE \ (1 + ((sizeof(struct re_save_state) - 1) / sizeof(*PL_savestack))) + /* * Local variables: * c-indentation-style: bsd