Re: [PATCH] New regex syntax omnibus

[perl5.git] / regcomp.h

diff --git a/regcomp.h b/regcomp.h
index 3213fc8..f64168a 100644 (file)
--- a/regcomp.h
+++ b/regcomp.h
@@ -15,6 +15,12 @@ typedef OP OP_4tree;                 /* Will be redefined later. */
 #define PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION 1
 #define PERL_ENABLE_POSITIVE_ASSERTION_STUDY 1
 #define PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS 0
+/* Unless the next line is uncommented it is illegal to combine lazy 
+   matching with possessive matching. Frankly it doesn't make much sense 
+   to allow it as X*?+ matches nothing, X+?+ matches a single char only, 
+   and X{min,max}?+ matches min times only.
+ */
+/* #define REG_ALLOW_MINMOD_SUSPEND */
 
 /*
  * The "internal use only" fields in regexp.h are present to pass info from
@@ -86,6 +92,8 @@ struct regnode_string {
     char string[1];
 };
 
+/* Argument bearing node - workhorse, 
+   arg1 is often for the data field */
 struct regnode_1 {
     U8 flags;
     U8  type;
@@ -93,6 +101,16 @@ struct regnode_1 {
     U32 arg1;
 };
 
+/* Similar to a regnode_1 but with an extra signed argument */
+struct regnode_2L {
+    U8 flags;
+    U8  type;
+    U16 next_off;
+    U32 arg1;
+    I32 arg2;
+};
+
+/* 'Two field' -- Two 16 bit unsigned args */
 struct regnode_2 {
     U8 flags;
     U8  type;
@@ -101,6 +119,7 @@ struct regnode_2 {
     U16 arg2;
 };
 
+
 #define ANYOF_BITMAP_SIZE      32      /* 256 b/(8 b/B) */
 #define ANYOF_CLASSBITMAP_SIZE  4      /* up to 32 (8*4) named classes */
 
@@ -154,10 +173,12 @@ struct regnode_charclass_class {  /* has [[:blah:]] classes */
 #define ARG(p) ARG_VALUE(ARG_LOC(p))
 #define ARG1(p) ARG_VALUE(ARG1_LOC(p))
 #define ARG2(p) ARG_VALUE(ARG2_LOC(p))
+#define ARG2L(p) ARG_VALUE(ARG2L_LOC(p))
 
 #define ARG_SET(p, val) ARG__SET(ARG_LOC(p), (val))
 #define ARG1_SET(p, val) ARG__SET(ARG1_LOC(p), (val))
 #define ARG2_SET(p, val) ARG__SET(ARG2_LOC(p), (val))
+#define ARG2L_SET(p, val) ARG__SET(ARG2L_LOC(p), (val))
 
 #undef NEXT_OFF
 #undef NODE_ALIGN
@@ -190,7 +211,7 @@ struct regnode_charclass_class {    /* has [[:blah:]] classes */
 #define        ARG_LOC(p)      (((struct regnode_1 *)p)->arg1)
 #define        ARG1_LOC(p)     (((struct regnode_2 *)p)->arg1)
 #define        ARG2_LOC(p)     (((struct regnode_2 *)p)->arg2)
-
+#define ARG2L_LOC(p)   (((struct regnode_2L *)p)->arg2)
 
 #define NODE_STEP_REGNODE      1       /* sizeof(regnode)/sizeof(regnode) */
 #define EXTRA_STEP_2ARGS       EXTRA_SIZE(struct regnode_2)
@@ -328,6 +349,10 @@ struct regnode_charclass_class {   /* has [[:blah:]] classes */
 #define REG_SEEN_EVAL          0x00000008
 #define REG_SEEN_CANY          0x00000010
 #define REG_SEEN_SANY          REG_SEEN_CANY /* src bckwrd cmpt */
+#define REG_SEEN_RECURSE        0x00000020
+#define REG_TOP_LEVEL_BRANCHES  0x00000040
+#define REG_SEEN_VERBARG        0x00000080
+#define REG_SEEN_CUTGROUP       0x00000100
 
 START_EXTERN_C
 
@@ -374,6 +399,7 @@ EXTCONST regexp_engine PL_core_reg_engine = {
         Perl_re_intuit_start,
         Perl_re_intuit_string, 
         Perl_pregfree, 
+        Perl_reg_stringify,
 #if defined(USE_ITHREADS)        
         Perl_regdupe 
 #endif        
@@ -396,6 +422,7 @@ END_EXTERN_C
  *       in the character class
  *   t - trie struct
  *   T - aho-trie struct
+ *   S - sv for named capture lookup
  * 20010712 mjd@plover.com
  * (Remember to update re_dup() and pregfree() if you add any items.)
  */
@@ -495,7 +522,8 @@ struct _reg_trie_data {
                                         for the node following a given word. */
     U16                    *nextword;       /* optional 1 indexed array to support linked list
                                         of duplicate wordnums */
-    U32             laststate;       /* Build only */
+    U32             statecount;      /* Build only - number of states in the states array 
+                                        (including the unused zero state) */
     U32             wordcount;       /* Build only */
 #ifdef DEBUGGING
     STRLEN          charcount;       /* Build only */
@@ -599,6 +627,8 @@ re.pm, especially to the documentation.
 #define RE_DEBUG_EXTRA_OFFDEBUG    0x040000
 #define RE_DEBUG_EXTRA_STATE       0x080000
 #define RE_DEBUG_EXTRA_OPTIMISE    0x100000
+/* combined */
+#define RE_DEBUG_EXTRA_STACK       0x280000
 
 #define RE_DEBUG_FLAG(x) (re_debug_flags & x)
 /* Compile */
@@ -632,6 +662,8 @@ re.pm, especially to the documentation.
     if (re_debug_flags & RE_DEBUG_EXTRA_OFFSETS) x  )
 #define DEBUG_STATE_r(x) DEBUG_r( \
     if (re_debug_flags & RE_DEBUG_EXTRA_STATE) x )
+#define DEBUG_STACK_r(x) DEBUG_r( \
+    if (re_debug_flags & RE_DEBUG_EXTRA_STACK) x )
 #define DEBUG_OPTIMISE_MORE_r(x) DEBUG_r( \
     if ((RE_DEBUG_EXTRA_OPTIMISE|RE_DEBUG_COMPILE_OPTIMISE) == \
          (re_debug_flags & (RE_DEBUG_EXTRA_OPTIMISE|RE_DEBUG_COMPILE_OPTIMISE)) ) x )