Update.

	* posix/regcomp.c (build_word_op): Rename like...
	(build_charclass_op): ...this.  Accept two extra parameters,
	CLASS_NAME and EXTRA.  Add EXTRA to the result, not only _.
	(peek_token): accept \s and \S as OP_SPACE and OP_NOTSPACE.
	(parse_expression): replace build_word_op with
	build_charclass_op, add new arguments, accept OP_SPACE
	and OP_NOTSPACE.
	* posix/regex_internal.h (re_token_type_t): Add OP_SPACE
	and OP_NOTSPACE.

diff --git a/ChangeLog b/ChangeLog
index 0a6301cfbafc3c22240c27d3f5e0cc735ed935e4..9fc0a6576c9aa25e7bf377eea2db85b927e013a5 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,15 @@
 2003-09-20  Paolo Bonzini  <bonzini@gnu.org>
 
+       * posix/regcomp.c (build_word_op): Rename like...
+       (build_charclass_op): ...this.  Accept two extra parameters,
+       CLASS_NAME and EXTRA.  Add EXTRA to the result, not only _.
+       (peek_token): accept \s and \S as OP_SPACE and OP_NOTSPACE.
+       (parse_expression): replace build_word_op with
+       build_charclass_op, add new arguments, accept OP_SPACE
+       and OP_NOTSPACE.
+       * posix/regex_internal.h (re_token_type_t): Add OP_SPACE
+       and OP_NOTSPACE.
+
        * posix/regcomp.c (peek_token): Don't look back for ( or |
        to check whether to treat a caret as special.  It fails
        for the (extended) regex \(^.

diff --git a/posix/regcomp.c b/posix/regcomp.c
index 4682ca6c1b69586169dc7ded92cbaa0bf7714d8c..3d54f9936362873e089bf3141f1ca48497afe9df 100644 (file)
--- a/posix/regcomp.c
+++ b/posix/regcomp.c
@@ -118,8 +118,10 @@ static reg_errcode_t build_charclass (RE_TRANSLATE_TYPE trans,
                                      const unsigned char *class_name,
                                      reg_syntax_t syntax);
 #endif /* not RE_ENABLE_I18N */
-static bin_tree_t *build_word_op (re_dfa_t *dfa, RE_TRANSLATE_TYPE trans,
-                                 int not, reg_errcode_t *err);
+static bin_tree_t *build_charclass_op (re_dfa_t *dfa, RE_TRANSLATE_TYPE trans,
+                                      const unsigned char *class_name, 
+                                      const unsigned char *extra, int not,
+                                      reg_errcode_t *err);
 static void free_bin_tree (bin_tree_t *tree);
 static bin_tree_t *create_tree (bin_tree_t *left, bin_tree_t *right,
                                re_token_type_t type, int index);
@@ -1561,6 +1563,14 @@ peek_token (token, input, syntax)
          if (!(syntax & RE_NO_GNU_OPS))
            token->type = OP_NOTWORD;
          break;
+       case 's':
+         if (!(syntax & RE_NO_GNU_OPS))
+           token->type = OP_SPACE;
+         break;
+       case 'S':
+         if (!(syntax & RE_NO_GNU_OPS))
+           token->type = OP_NOTSPACE;
+         break;
        case '`':
          if (!(syntax & RE_NO_GNU_OPS))
            {
@@ -2076,12 +2086,22 @@ parse_expression (regexp, preg, token, syntax, nest, err)
        dfa->has_mb_node = 1;
       break;
     case OP_WORD:
-      tree = build_word_op (dfa, regexp->trans, 0, err);
+      tree = build_charclass_op (dfa, regexp->trans, "alnum", "_", 0, err);
       if (BE (*err != REG_NOERROR && tree == NULL, 0))
        return NULL;
       break;
     case OP_NOTWORD:
-      tree = build_word_op (dfa, regexp->trans, 1, err);
+      tree = build_charclass_op (dfa, regexp->trans, "alnum", "_", 1, err);
+      if (BE (*err != REG_NOERROR && tree == NULL, 0))
+       return NULL;
+      break;
+    case OP_SPACE:
+      tree = build_charclass_op (dfa, regexp->trans, "space", "", 0, err);
+      if (BE (*err != REG_NOERROR && tree == NULL, 0))
+       return NULL;
+      break;
+    case OP_NOTSPACE:
+      tree = build_charclass_op (dfa, regexp->trans, "space", "", 1, err);
       if (BE (*err != REG_NOERROR && tree == NULL, 0))
        return NULL;
       break;
@@ -3284,9 +3304,11 @@ build_charclass (trans, sbcset, class_name, syntax)
 }
 
 static bin_tree_t *
-build_word_op (dfa, trans, not, err)
+build_charclass_op (dfa, trans, class_name, extra, not, err)
      re_dfa_t *dfa;
      RE_TRANSLATE_TYPE trans;
+     const unsigned char *class_name;
+     const unsigned char *extra;
      int not;
      reg_errcode_t *err;
 {
@@ -3340,7 +3362,7 @@ build_word_op (dfa, trans, not, err)
 #ifdef RE_ENABLE_I18N
                         mbcset, &alloc,
 #endif /* RE_ENABLE_I18N */
-                        (const unsigned char *) "alnum", 0);
+                        class_name, 0);
 
   if (BE (ret != REG_NOERROR, 0))
     {
@@ -3352,7 +3374,8 @@ build_word_op (dfa, trans, not, err)
       return NULL;
     }
   /* \w match '_' also.  */
-  bitset_set (sbcset, '_');
+  for (; *extra; extra++)
+    bitset_set (sbcset, *extra);
 
   /* If it is non-matching list.  */
 #ifdef RE_ENABLE_I18N

diff --git a/posix/regex_internal.h b/posix/regex_internal.h
index 9dd3fe55b30d85aea4a099e3a610f7e04000e2b3..b2fe766f780b0c581de638fee37e85f7b1ebc712 100644 (file)
--- a/posix/regex_internal.h
+++ b/posix/regex_internal.h
@@ -182,6 +182,8 @@ typedef enum
   OP_CLOSE_CHAR_CLASS,
   OP_WORD,
   OP_NOTWORD,
+  OP_SPACE,
+  OP_NOTSPACE,
   BACK_SLASH,
 
   /* Tree type, these are used only by tree. */