regcomp.c - extend REF to hold the paren it needs to regcppush

this way we can avoid pushing every buffer, we only need to push
the nestroot of the ref.

7 files changed, 88 insertions, 83 deletions

diff --git a/pod/perldebguts.pod b/pod/perldebguts.pod
index 90f5fb66ae..52aef07396 100644
--- a/pod/perldebguts.pod
+++ b/pod/perldebguts.pod
@@ -773,25 +773,25 @@ will be lost.
  SROPEN           none       Same as OPEN, but for script run
  SRCLOSE          none       Close preceding SROPEN
 
- REF              num 1      Match some already matched string
- REFF             num 1      Match already matched string, using /di
+ REF              num 2      Match some already matched string
+ REFF             num 2      Match already matched string, using /di
                              rules.
- REFFL            num 1      Match already matched string, using /li
+ REFFL            num 2      Match already matched string, using /li
                              rules.
- REFFU            num 1      Match already matched string, usng /ui.
- REFFA            num 1      Match already matched string, using /aai
+ REFFU            num 2      Match already matched string, usng /ui.
+ REFFA            num 2      Match already matched string, using /aai
                              rules.
 
  # Named references.  Code in regcomp.c assumes that these all are after
  # the numbered references
- REFN             no-sv 1    Match some already matched string
- REFFN            no-sv 1    Match already matched string, using /di
+ REFN             no-sv 2    Match some already matched string
+ REFFN            no-sv 2    Match already matched string, using /di
                              rules.
- REFFLN           no-sv 1    Match already matched string, using /li
+ REFFLN           no-sv 2    Match already matched string, using /li
                              rules.
- REFFUN           num 1      Match already matched string, using /ui
+ REFFUN           num 2      Match already matched string, using /ui
                              rules.
- REFFAN           num 1      Match already matched string, using /aai
+ REFFAN           num 2      Match already matched string, using /aai
                              rules.
 
  # Support for long RE
diff --git a/regcomp.c b/regcomp.c
index da8b3e32f1..94428fad9c 100644
--- a/regcomp.c
+++ b/regcomp.c
@@ -2754,7 +2754,7 @@ S_handle_named_backref(pTHX_ RExC_state_t *pRExC_state,
         SvREFCNT_inc_simple_void_NN(sv_dat);
     }
     RExC_sawback = 1;
-    ret = reg1node(pRExC_state,
+    ret = reg2node(pRExC_state,
                    ((! FOLD)
                      ? REFN
                      : (ASCII_FOLD_RESTRICTED)
@@ -2764,7 +2764,9 @@ S_handle_named_backref(pTHX_ RExC_state_t *pRExC_state,
                          : (LOC)
                            ? REFFLN
                            : REFFN),
-                    num);
+                    num, RExC_nestroot);
+    if (RExC_nestroot && num >= (U32)RExC_nestroot)
+        REGNODE_p(ret)->flags = VOLATILE_REF;
     *flagp |= HASWIDTH;
 
     nextchar(pRExC_state);
@@ -6028,7 +6030,7 @@ S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
                     }
                 }
                 RExC_sawback = 1;
-                ret = reg1node(pRExC_state,
+                ret = reg2node(pRExC_state,
                                ((! FOLD)
                                  ? REF
                                  : (ASCII_FOLD_RESTRICTED)
@@ -6038,7 +6040,7 @@ S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
                                      : (LOC)
                                        ? REFFL
                                        : REFF),
-                                num);
+                                num, RExC_nestroot);
                 if (RExC_nestroot && num >= RExC_nestroot)
                     REGNODE_p(ret)->flags = VOLATILE_REF;
                 if (OP(REGNODE_p(ret)) == REFF) {
diff --git a/regcomp.sym b/regcomp.sym
index 1c0af0cd53..2c0f4a0501 100644
--- a/regcomp.sym
+++ b/regcomp.sym
@@ -233,22 +233,22 @@ CLOSE       CLOSE,      num 1     ; Close corresponding OPEN of #n.
 SROPEN      SROPEN,     none      ; Same as OPEN, but for script run
 SRCLOSE     SRCLOSE,    none      ; Close preceding SROPEN
 
-REF         REF,        num 1 V   ; Match some already matched string
-REFF        REF,        num 1 V   ; Match already matched string, using /di rules.
-REFFL       REF,        num 1 V   ; Match already matched string, using /li rules.
+REF         REF,        num 2 V   ; Match some already matched string
+REFF        REF,        num 2 V   ; Match already matched string, using /di rules.
+REFFL       REF,        num 2 V   ; Match already matched string, using /li rules.
 # N?REFF[AU] could have been implemented using the FLAGS field of the
 # regnode, but by having a separate node type, we can use the existing switch
 # statement to avoid some tests
-REFFU       REF,        num 1 V   ; Match already matched string, usng /ui.
-REFFA       REF,        num 1 V   ; Match already matched string, using /aai rules.
+REFFU       REF,        num 2 V   ; Match already matched string, usng /ui.
+REFFA       REF,        num 2 V   ; Match already matched string, using /aai rules.
 
 #*Named references.  Code in regcomp.c assumes that these all are after
 #*the numbered references
-REFN        REF,        no-sv 1 V ; Match some already matched string
-REFFN       REF,        no-sv 1 V ; Match already matched string, using /di rules.
-REFFLN      REF,        no-sv 1 V ; Match already matched string, using /li rules.
-REFFUN      REF,        num   1 V ; Match already matched string, using /ui rules.
-REFFAN      REF,        num   1 V ; Match already matched string, using /aai rules.
+REFN        REF,        no-sv 2 V ; Match some already matched string
+REFFN       REF,        no-sv 2 V ; Match already matched string, using /di rules.
+REFFLN      REF,        no-sv 2 V ; Match already matched string, using /li rules.
+REFFUN      REF,        num   2 V ; Match already matched string, using /ui rules.
+REFFAN      REF,        num   2 V ; Match already matched string, using /aai rules.
 
 #*Support for long RE
 LONGJMP     LONGJMP,    off 1 . 1 ; Jump far away.
diff --git a/regcomp_debug.c b/regcomp_debug.c
index d02dbfd4cb..8c5065bacd 100644
--- a/regcomp_debug.c
+++ b/regcomp_debug.c
@@ -489,9 +489,9 @@ Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o, const regmatch_
              || k == GROUPP || op == ACCEPT)
     {
         AV *name_list= NULL;
-        U32 parno= (op == ACCEPT)              ? (U32)ARG2i(o) :
-                   (op == OPEN || op == CLOSE) ? (U32)PARNO(o) :
-                                                 (U32)ARG1u(o);
+        U32 parno= (op == ACCEPT)              ? ARG2u(o) :
+                   (op == OPEN || op == CLOSE) ? PARNO(o) :
+                                                 ARG1u(o);
         if ( RXp_PAREN_NAMES(prog) ) {
             name_list= MUTABLE_AV(progi->data->data[progi->name_list_idx]);
         } else if ( pRExC_state ) {
@@ -542,6 +542,9 @@ Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o, const regmatch_
                 Perl_sv_catpvf(aTHX_ sv, "/%" UVuf, (UV)parno);     /* Parenth number */
 
         }
+        if ( k == REF ) {
+            Perl_sv_catpvf(aTHX_ sv, " <%" IVdf ">", (IV)ARG2i(o));
+        }
         if ( k == REF && reginfo) {
             U32 n = ARG1u(o);  /* which paren pair */
             I32 ln = RXp_OFFS_START(prog,n);
diff --git a/regexec.c b/regexec.c
index 3a8e160478..eb1fe4e06a 100644
--- a/regexec.c
+++ b/regexec.c
@@ -8068,7 +8068,7 @@ S_regmatch(pTHX_ regmatch_info *reginfo, char *startpos, regnode *prog)
         }
         ref_yes:
             if (scan->flags) { /* == VOLATILE_REF but only other value is 0 */
-                ST.cp = regcppush(rex, 0, maxopenparen);
+                ST.cp = regcppush(rex, ARG2u(scan) - 1, maxopenparen);
                 REGCP_SET(ST.lastcp);
                 PUSH_STATE_GOTO(REF_next, next, locinput, loceol,
                                 script_run_begin);
@@ -8105,7 +8105,7 @@ S_regmatch(pTHX_ regmatch_info *reginfo, char *startpos, regnode *prog)
             U32 arg;
 
         case GOSUB: /*    /(...(?1))/   /(...(?&foo))/   */
-            arg= (U32)ARG1u(scan);
+            arg = ARG1u(scan);
             if (cur_eval && cur_eval->locinput == locinput) {
                 if ( ++nochange_depth > max_nochange_depth )
                     Perl_croak(aTHX_
@@ -8620,7 +8620,7 @@ S_regmatch(pTHX_ regmatch_info *reginfo, char *startpos, regnode *prog)
             is_accepted = true;
             if (scan->flags)
                 sv_yes_mark = MUTABLE_SV(rexi->data->data[ ARG1u( scan ) ]);
-            utmp = (U32)ARG2i(scan);
+            utmp = ARG2u(scan);
 
             if ( utmp ) {
                 regnode *cursor;
diff --git a/regnodes.h b/regnodes.h
index a675874378..c4c7ba6a45 100644
--- a/regnodes.h
+++ b/regnodes.h
@@ -94,16 +94,16 @@ typedef struct regnode                           tregnode_POSIXL;
 typedef struct regnode                           tregnode_POSIXU;
 typedef struct regnode_1                         tregnode_PRUNE;
 typedef struct regnode                           tregnode_PSEUDO;
-typedef struct regnode_1                         tregnode_REF;
-typedef struct regnode_1                         tregnode_REFF;
-typedef struct regnode_1                         tregnode_REFFA;
-typedef struct regnode_1                         tregnode_REFFAN;
-typedef struct regnode_1                         tregnode_REFFL;
-typedef struct regnode_1                         tregnode_REFFLN;
-typedef struct regnode_1                         tregnode_REFFN;
-typedef struct regnode_1                         tregnode_REFFU;
-typedef struct regnode_1                         tregnode_REFFUN;
-typedef struct regnode_1                         tregnode_REFN;
+typedef struct regnode_2                         tregnode_REF;
+typedef struct regnode_2                         tregnode_REFF;
+typedef struct regnode_2                         tregnode_REFFA;
+typedef struct regnode_2                         tregnode_REFFAN;
+typedef struct regnode_2                         tregnode_REFFL;
+typedef struct regnode_2                         tregnode_REFFLN;
+typedef struct regnode_2                         tregnode_REFFN;
+typedef struct regnode_2                         tregnode_REFFU;
+typedef struct regnode_2                         tregnode_REFFUN;
+typedef struct regnode_2                         tregnode_REFN;
 typedef struct regnode_p                         tregnode_REGEX_SET;
 typedef struct regnode                           tregnode_REG_ANY;
 typedef struct regnode_1                         tregnode_RENUM;
diff --git a/t/re/pat_advanced.t b/t/re/pat_advanced.t
index d836525c15..d64bd1b94b 100644
--- a/t/re/pat_advanced.t
+++ b/t/re/pat_advanced.t
@@ -2590,10 +2590,10 @@ Starting parse and generation
 <\g{c}>         |   6|            brnc   
                 |    |              piec   
                 |    |                atom   
-<>              |   8|            tail~ OPEN1 'b' (4) -> REFN
-                |    |            Setting close paren #1 to 8
-                |  10|          lsbr~ tying lastbr REFN (6) to ender CLOSE1 'b' (8) offset 2
-                |    |            tail~ REFN (6) -> CLOSE
+<>              |   9|            tail~ OPEN1 'b' (4) -> REFN
+                |    |            Setting close paren #1 to 9
+                |  11|          lsbr~ tying lastbr REFN <1> (6) to ender CLOSE1 'b' (9) offset 3
+                |    |            tail~ REFN <1> (6) -> CLOSE
 Unmatched ( in regex; marked by <-- HERE in m/(?{a})( <-- HERE ?<b>\g{c}/ at - line 1.
 Freeing REx: "(?{a})(?<b>\g{c}"
 EOF_DEBUG_OUT
@@ -2618,35 +2618,35 @@ Starting parse and generation
 <\g{c})(?<c>>...|   3|            brnc   
                 |    |              piec   
                 |    |                atom   
-<)(?<c>x)(?&b)> |   5|            tail~ OPEN1 'b' (1) -> REFN
-                |   7|          lsbr~ tying lastbr REFN (3) to ender CLOSE1 'b' (5) offset 2
-                |    |            tail~ REFN (3) -> CLOSE
+<)(?<c>x)(?&b)> |   6|            tail~ OPEN1 'b' (1) -> REFN
+                |   8|          lsbr~ tying lastbr REFN <1> (3) to ender CLOSE1 'b' (6) offset 3
+                |    |            tail~ REFN <1> (3) -> CLOSE
 <(?<c>x)(?&b)>  |    |      piec   
                 |    |        atom   
 <?<c>x)(?&b)>   |    |          reg    
-<x)(?&b)>       |   9|            brnc   
+<x)(?&b)>       |  10|            brnc
                 |    |              piec   
                 |    |                atom   
-<)(?&b)>        |  11|            tail~ OPEN2 'c' (7) -> EXACT
-                |  13|          lsbr~ tying lastbr EXACT <x> (9) to ender CLOSE2 'c' (11) offset 2
-                |    |            tail~ EXACT <x> (9) -> CLOSE
+<)(?&b)>        |  12|            tail~ OPEN2 'c' (8) -> EXACT
+                |  14|          lsbr~ tying lastbr EXACT <x> (10) to ender CLOSE2 'c' (12) offset 2
+                |    |            tail~ EXACT <x> (10) -> CLOSE
 <(?&b)>         |    |      tail~ OPEN1 'b' (1)  
-                |    |          ~ REFN (3)
-                |    |          ~ CLOSE1 'b' (5) -> OPEN
+                |    |          ~ REFN <1> (3)
+                |    |          ~ CLOSE1 'b' (6) -> OPEN
                 |    |      piec   
                 |    |        atom   
 <?&b)>          |    |          reg    
-<>              |  16|      tail~ OPEN2 'c' (7)  
-                |    |          ~ EXACT <x> (9)  
-                |    |          ~ CLOSE2 'c' (11) -> GOSUB
-                |  17|  lsbr~ tying lastbr OPEN1 'b' (1) to ender END (16) offset 15
+<>              |  17|      tail~ OPEN2 'c' (8)
+                |    |          ~ EXACT <x> (10)
+                |    |          ~ CLOSE2 'c' (12) -> GOSUB
+                |  18|  lsbr~ tying lastbr OPEN1 'b' (1) to ender END (17) offset 16
                 |    |    tail~ OPEN1 'b' (1)  
-                |    |        ~ REFN (3)
-                |    |        ~ CLOSE1 'b' (5)  
-                |    |        ~ OPEN2 'c' (7)  
-                |    |        ~ EXACT <x> (9)  
-                |    |        ~ CLOSE2 'c' (11)  
-                |    |        ~ GOSUB1[+0:13] 'b' (13) -> END
+                |    |        ~ REFN <1> (3)
+                |    |        ~ CLOSE1 'b' (6)
+                |    |        ~ OPEN2 'c' (8)
+                |    |        ~ EXACT <x> (10)
+                |    |        ~ CLOSE2 'c' (12)
+                |    |        ~ GOSUB1[+0:14] 'b' (14) -> END
 Need to redo parse
 Freeing REx: "(?<b>\g{c})(?<c>x)(?&b)"
 Starting parse and generation
@@ -2658,36 +2658,36 @@ Starting parse and generation
 <\g{c})(?<c>>...|   3|            brnc   
                 |    |              piec   
                 |    |                atom   
-<)(?<c>x)(?&b)> |   5|            tail~ OPEN1 'b' (1) -> REFN
-                |   7|          lsbr~ tying lastbr REFN2 'c' (3) to ender CLOSE1 'b' (5) offset 2
-                |    |            tail~ REFN2 'c' (3) -> CLOSE
+<)(?<c>x)(?&b)> |   6|            tail~ OPEN1 'b' (1) -> REFN
+                |   8|          lsbr~ tying lastbr REFN2 'c' <1> (3) to ender CLOSE1 'b' (6) offset 3
+                |    |            tail~ REFN2 'c' <1> (3) -> CLOSE
 <(?<c>x)(?&b)>  |    |      piec   
                 |    |        atom   
 <?<c>x)(?&b)>   |    |          reg    
-<x)(?&b)>       |   9|            brnc   
+<x)(?&b)>       |  10|            brnc
                 |    |              piec   
                 |    |                atom   
-<)(?&b)>        |  11|            tail~ OPEN2 'c' (7) -> EXACT
-                |  13|          lsbr~ tying lastbr EXACT <x> (9) to ender CLOSE2 'c' (11) offset 2
-                |    |            tail~ EXACT <x> (9) -> CLOSE
+<)(?&b)>        |  12|            tail~ OPEN2 'c' (8) -> EXACT
+                |  14|          lsbr~ tying lastbr EXACT <x> (10) to ender CLOSE2 'c' (12) offset 2
+                |    |            tail~ EXACT <x> (10) -> CLOSE
 <(?&b)>         |    |      tail~ OPEN1 'b' (1)  
-                |    |          ~ REFN2 'c' (3)
-                |    |          ~ CLOSE1 'b' (5) -> OPEN
+                |    |          ~ REFN2 'c' <1> (3)
+                |    |          ~ CLOSE1 'b' (6) -> OPEN
                 |    |      piec   
                 |    |        atom   
 <?&b)>          |    |          reg    
-<>              |  16|      tail~ OPEN2 'c' (7)  
-                |    |          ~ EXACT <x> (9)  
-                |    |          ~ CLOSE2 'c' (11) -> GOSUB
-                |  17|  lsbr~ tying lastbr OPEN1 'b' (1) to ender END (16) offset 15
+<>              |  17|      tail~ OPEN2 'c' (8)
+                |    |          ~ EXACT <x> (10)
+                |    |          ~ CLOSE2 'c' (12) -> GOSUB
+                |  18|  lsbr~ tying lastbr OPEN1 'b' (1) to ender END (17) offset 16
                 |    |    tail~ OPEN1 'b' (1)  
-                |    |        ~ REFN2 'c' (3)
-                |    |        ~ CLOSE1 'b' (5)  
-                |    |        ~ OPEN2 'c' (7)  
-                |    |        ~ EXACT <x> (9)  
-                |    |        ~ CLOSE2 'c' (11)  
-                |    |        ~ GOSUB1[+0:13] 'b' (13) -> END
-Required size 16 nodes
+                |    |        ~ REFN2 'c' <1> (3)
+                |    |        ~ CLOSE1 'b' (6)
+                |    |        ~ OPEN2 'c' (8)
+                |    |        ~ EXACT <x> (10)
+                |    |        ~ CLOSE2 'c' (12)
+                |    |        ~ GOSUB1[+0:14] 'b' (14) -> END
+Required size 17 nodes
 first at 3
 Freeing REx: "(?<b>\g{c})(?<c>x)(?&b)"
 EOF_DEBUG_OUT