From c3c43fd39c715827de5cc74846ba977c5b1d2181 Mon Sep 17 00:00:00 2001
From: Kaz Kylheku <kaz@kylheku.com>
Date: Fri, 20 Apr 2012 00:47:46 -0700
Subject: * parser.y (regtoken): New nonterminal symbol. (regterm): REGTOKEN
 production factored out to regtoken. (regclass): Reverted prior commmit's
 changes. (regclassterm): Reverted prior commit, removing REGTOKEN production
 for character classes, and introduced a regtoken production. So now the
 keyword symbols are part of the character class abstract syntax. (regtoken):
 New production rule.

* regex.c (regex_space_chars): Converted to internal linkage.
(char_set_compile): Handle token keywords in character class
abstract syntax.

* regex.h (regex_space_chars): External declaration removed.
---
 parser.y | 56 +++++++++++++++++++++-----------------------------------
 1 file changed, 21 insertions(+), 35 deletions(-)

(limited to 'parser.y')

diff --git a/parser.y b/parser.y
index c57bc901..46a0052b 100644
--- a/parser.y
+++ b/parser.y
@@ -90,7 +90,7 @@ static val parsed_spec;
 %type <val> repeat_clause repeat_parts_opt o_line
 %type <val> o_elems_opt o_elems o_elem o_var rep_elem rep_parts_opt
 %type <val> regex lisp_regex regexpr regbranch
-%type <val> regterm regclass regclassterm regrange
+%type <val> regterm regtoken regclass regclassterm regrange
 %type <val> strlit chrlit quasilit quasi_items quasi_item litchars
 %type <chr> regchar
 %type <lineno> '(' '['
@@ -796,48 +796,20 @@ regterm : regterm '*'           { $$ = list(zeroplus_s, $1, nao); }
         | ']'                   { $$ = chr(']'); }
         | '-'                   { $$ = chr('-'); }
         | REGCHAR               { $$ = chr($1); }
-        | REGTOKEN              { switch ($1)
-                                  { case 's':
-                                      $$ = space_k; break;
-                                    case 'S':
-                                      $$ = cspace_k; break;
-                                    case 'd':
-                                      $$ = digit_k; break;
-                                    case 'D':
-                                      $$ = cdigit_k; break;
-                                    case 'w':
-                                      $$ = word_char_k; break;
-                                    case 'W':
-                                      $$ = cword_char_k; break; }}
+        | regtoken              { $$ = $1; }
         | '(' regexpr ')'       { $$ = $2; }
         | '(' error             { $$ = nil;
                                   yybadtoken(yychar,
                                              lit("regex subexpression")); }
         ;
 
-regclass : regclassterm                 { $$ = $1; }
-         | regclassterm regclass        { $$ = nappend2($1, $2); }
+regclass : regclassterm                 { $$ = cons($1, nil); }
+         | regclassterm regclass        { $$ = cons($1, $2); }
          ;
 
-regclassterm : regrange         { $$ = cons($1, nil); }
-             | regchar          { $$ = cons(chr($1), nil); }
-             | REGTOKEN         { switch ($1)
-                                  { case 's':
-                                      $$ = regex_space_chars;
-                                      break;
-                                    case 'd':
-                                      $$ = cons(cons(chr('0'), chr('9')), nil);
-                                      break;
-                                    case 'w':
-                                      $$ = list(cons(chr('A'), chr('Z')),
-                                                cons(chr('a'), chr('z')),
-                                                chr('_'), nao);
-                                      break;
-                                    default:
-                                      yyerrorf(lit("complemented token "
-                                                   "\\~a not allowed "
-                                                   "in regex character class"),
-                                               chr($1), nao); } }
+regclassterm : regrange         { $$ = $1; }
+             | regchar          { $$ = chr($1); }
+             | regtoken         { $$ = $1; }
              ;
 
 regrange : regchar '-' regchar  { $$ = cons(chr($1), chr($3)); }
@@ -856,6 +828,20 @@ regchar : '?'                   { $$ = '?'; }
         | REGCHAR               { $$ = $1; }
         ;
 
+regtoken : REGTOKEN             { switch ($1)
+                                  { case 's':
+                                      $$ = space_k; break;
+                                    case 'S':
+                                      $$ = cspace_k; break;
+                                    case 'd':
+                                      $$ = digit_k; break;
+                                    case 'D':
+                                      $$ = cdigit_k; break;
+                                    case 'w':
+                                      $$ = word_char_k; break;
+                                    case 'W':
+                                      $$ = cword_char_k; break; }}
+
 newl : '\n'
      | error '\n'       { yyerror("newline expected after directive");
                           yyerrok; }
-- 
cgit v1.2.3