X-Git-Url: https://perl5.git.perl.org/perl5.git/blobdiff_plain/f65e70f560c80556143956eb9980619a9dfbe2f4..e99e4210c1b38bfe638e230c77e29f80d75265c5:/regexp.h

diff --git a/regexp.h b/regexp.h
index bd3aeb6..928a374 100644
--- a/regexp.h
+++ b/regexp.h
@@ -36,11 +36,11 @@ struct regexp_engine;
 struct regexp;
 
 struct reg_substr_datum {
-    I32 min_offset;
-    I32 max_offset;
+    SSize_t min_offset;
+    SSize_t max_offset;
     SV *substr;		/* non-utf8 variant */
     SV *utf8_substr;	/* utf8 variant */
-    I32 end_shift;
+    SSize_t end_shift;
 };
 struct reg_substr_data {
     struct reg_substr_datum data[3];	/* Actual array */
@@ -55,15 +55,15 @@ struct reg_substr_data {
 /* offsets within a string of a particular /(.)/ capture */
 
 typedef struct regexp_paren_pair {
-    I32 start;
-    I32 end;
+    SSize_t start;
+    SSize_t end;
     /* 'start_tmp' records a new opening position before the matching end
      * has been found, so that the old start and end values are still
      * valid, e.g.
      *	  "abc" =~ /(.(?{print "[$1]"}))+/
      *outputs [][a][b]
      * This field is not part of the API.  */
-    I32 start_tmp;
+    SSize_t start_tmp;
 } regexp_paren_pair;
 
 #if defined(PERL_IN_REGCOMP_C) || defined(PERL_IN_UTF8_C)
@@ -104,9 +104,9 @@ struct reg_code_block {
 	/* Information about the match that the perl core uses to */	\
 	/* manage things */						\
 	U32 extflags;	/* Flags used both externally and internally */	\
-	I32 minlen;	/* mininum possible number of chars in string to match */\
-	I32 minlenret;	/* mininum possible number of chars in $& */		\
-	U32 gofs;	/* chars left of pos that we search from */	\
+	SSize_t minlen;	/* mininum possible number of chars in string to match */\
+	SSize_t minlenret; /* mininum possible number of chars in $& */		\
+	STRLEN gofs;	/* chars left of pos that we search from */	\
 	/* substring data about strings that must appear in the */	\
 	/* final match, used for optimisations */			\
 	struct reg_substr_data *substrs;				\
@@ -124,9 +124,9 @@ struct reg_code_block {
 	/* saved or original string so \digit works forever. */		\
 	char *subbeg;							\
 	SV_SAVED_COPY	/* If non-NULL, SV which is COW from original */\
-	I32 sublen;	/* Length of string pointed by subbeg */	\
-	I32 suboffset;	/* byte offset of subbeg from logical start of str */ \
-	I32 subcoffset;	/* suboffset equiv, but in chars (for @-/@+) */ \
+	SSize_t sublen;	/* Length of string pointed by subbeg */	\
+	SSize_t suboffset; /* byte offset of subbeg from logical start of str */ \
+	SSize_t subcoffset; /* suboffset equiv, but in chars (for @-/@+) */ \
 	/* Information about the match that isn't often used */		\
 	/* offset from wrapped to the start of precomp */		\
 	PERL_BITFIELD32 pre_prefix:4;					\
@@ -146,7 +146,7 @@ typedef struct regexp {
 typedef struct re_scream_pos_data_s
 {
     char **scream_olds;		/* match pos */
-    I32 *scream_pos;		/* Internal iterator of scream. */
+    SSize_t *scream_pos;	/* Internal iterator of scream. */
 } re_scream_pos_data;
 
 /* regexp_engine structure. This is the dispatch table for regexes.
@@ -155,7 +155,7 @@ typedef struct re_scream_pos_data_s
 typedef struct regexp_engine {
     REGEXP* (*comp) (pTHX_ SV * const pattern, U32 flags);
     I32     (*exec) (pTHX_ REGEXP * const rx, char* stringarg, char* strend,
-                     char* strbeg, I32 minend, SV* screamer,
+                     char* strbeg, SSize_t minend, SV* sv,
                      void* data, U32 flags);
     char*   (*intuit) (pTHX_
                         REGEXP * const rx,
@@ -501,6 +501,10 @@ get_regex_charset_name(const U32 flags, STRLEN* const lenp)
 #define RX_LASTPAREN(prog)	(ReANY(prog)->lastparen)
 #define RX_LASTCLOSEPAREN(prog)	(ReANY(prog)->lastcloseparen)
 #define RX_SAVED_COPY(prog)	(ReANY(prog)->saved_copy)
+/* last match was zero-length */
+#define RX_ZERO_LEN(prog) \
+        (RX_OFFS(prog)[0].start + (SSize_t)RX_GOFS(prog) \
+          == RX_OFFS(prog)[0].end)
 
 #endif /* PLUGGABLE_RE_EXTENSION */
 
@@ -534,16 +538,24 @@ get_regex_charset_name(const U32 flags, STRLEN* const lenp)
 /* Whether the pattern stored at RX_WRAPPED is in UTF-8  */
 #define RX_UTF8(prog)			SvUTF8(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. */
-                                    /* under REXEC_COPY_STR, it's ok for the
-                                     * engine (modulo PL_sawamperand etc)
-                                     * to skip copying ... */
-#define REXEC_COPY_SKIP_PRE  0x20   /* ...the $` part of the string, or */
-#define REXEC_COPY_SKIP_POST 0x40   /* ...the $' part of the string */
+
+/* bits in flags arg of Perl_regexec_flags() */
+
+#define REXEC_COPY_STR  0x01    /* Need to copy the string for captures. */
+#define REXEC_CHECKED   0x02    /* re_intuit_start() already called. */
+#define REXEC_SCREAM    0x04    /* currently unused. */
+#define REXEC_IGNOREPOS 0x08    /* use stringarg, not pos(), for \G match */
+#define REXEC_NOT_FIRST 0x10    /* This is another iteration of //g:
+                                   no need to copy string again */
+
+                                     /* under REXEC_COPY_STR, it's ok for the
+                                        engine (modulo PL_sawamperand etc)
+                                        to skip copying: ... */
+#define REXEC_COPY_SKIP_PRE  0x20    /* ...the $` part of the string, or */
+#define REXEC_COPY_SKIP_POST 0x40    /* ...the $' part of the string */
+#define REXEC_FAIL_ON_UNDERFLOW 0x80 /* fail the match if $& would start before
+                                        the start pos (so s/.\G// would fail
+                                        on second iteration */
 
 #if defined(__GNUC__) && !defined(PERL_GCC_BRACE_GROUPS_FORBIDDEN)
 #  define ReREFCNT_inc(re)						\
@@ -594,7 +606,8 @@ typedef struct {
     STRLEN  suboffset;  /* saved suboffset  field from rex */
     STRLEN  subcoffset; /* saved subcoffset field from rex */
     MAGIC   *pos_magic; /* pos() magic attached to $_ */
-    I32     pos;        /* the original value of pos() in pos_magic */
+    SSize_t pos;        /* the original value of pos() in pos_magic */
+    U8      pos_flags;  /* flags to be restored; currently only MGf_BYTES*/
 } regmatch_info_aux_eval;
 
 
@@ -618,13 +631,13 @@ typedef struct {
  * regmatch_state union. */
 
 typedef struct {
-    REGEXP *prog;
+    REGEXP *prog;        /* the regex being executed */
     const char * strbeg; /* real start of string */
-    char *strend;   /* one byte beyond last char of match string */
-    char *till;
-    SV *sv;
-    char *ganch;
-    char *cutpoint;
+    char *strend;        /* one byte beyond last char of match string */
+    char *till;          /* matches shorter than this fail (see minlen arg) */
+    SV *sv;              /* the SV string currently being matched */
+    char *ganch;         /* position of \G anchor */
+    char *cutpoint;      /* (*COMMIT) position (if any) */
     regmatch_info_aux      *info_aux; /* extra fields that need cleanup */
     regmatch_info_aux_eval *info_aux_eval; /* extra saved state for (?{}) */
     I32  poscache_maxiter; /* how many whilems todo before S-L cache kicks in */
@@ -828,14 +841,7 @@ typedef struct regmatch_slab {
     struct regmatch_slab *prev, *next;
 } regmatch_slab;
 
-#define PL_reg_starttry		PL_reg_state.re_state_reg_starttry
-
-struct re_save_state {
-    char *re_state_reg_starttry;	/* from regexec.c */
-};
 
-#define SAVESTACK_ALLOC_FOR_RE_SAVE_STATE \
-	(1 + ((sizeof(struct re_save_state) - 1) / sizeof(*PL_savestack)))
 
 /*
  * Local variables: