Towards working facets

[yaz-moved-to-github.git] / src / ccltoken.c

diff --git a/src/ccltoken.c b/src/ccltoken.c
index a073712..6d201f4 100644 (file)
--- a/src/ccltoken.c
+++ b/src/ccltoken.c
@@ -1,16 +1,18 @@
 /* This file is part of the YAZ toolkit.
- * Copyright (C) 1995-2009 Index Data
+ * Copyright (C) 1995-2013 Index Data
  * See the file LICENSE for details.
  */
-/** 
+/**
  * \file ccltoken.c
  * \brief Implements CCL lexical analyzer (scanner)
  */
+#if HAVE_CONFIG_H
+#include <config.h>
+#endif
 
 #include <string.h>
 #include <stdlib.h>
-#include <ctype.h>
-
+#include <yaz/yaz-iconv.h>
 #include "cclp.h"
 
 /*
@@ -101,7 +103,7 @@ struct ccl_token *ccl_parser_tokenize(CCL_parser cclp, const char *command)
         case '%':
         case '!':
             last->kind = CCL_TOK_PROX;
-            while (isdigit(*cp))
+            while (yaz_isdigit(*cp))
             {
                 ++ last->len;
                 cp++;
@@ -121,29 +123,39 @@ struct ccl_token *ccl_parser_tokenize(CCL_parser cclp, const char *command)
             else
                 last->kind = CCL_TOK_REL;
             break;
-        case '\"':
+        default:
+            --cp;
+            --last->len;
+
             last->kind = CCL_TOK_TERM;
             last->name = (const char *) cp;
-            last->len = 0;
-            while (*cp && *cp != '\"')
+            while (*cp && !strchr("(),%!><= \t\n\r", *cp))
             {
-                cp++;
-                ++ last->len;
-            }
-            if (*cp == '\"')
-                cp++;
-            break;
-        default:
-            if (!strchr("(),%!><= \t\n\r", cp[-1]))
-            {
-                while (*cp && !strchr("(),%!><= \t\n\r", *cp))
+                if (*cp == '\\' && cp[1])
                 {
                     cp++;
                     ++ last->len;
                 }
+                else if (*cp == '"')
+                {
+                    while (*cp)
+                    {
+                        cp++;
+                        ++ last->len;
+                        if (*cp == '\\' && cp[1])
+                        {
+                            cp++;
+                            ++ last->len;
+                        }
+                        else if (*cp == '"')
+                            break;
+                    }
+                }
+                if (!*cp)
+                    break;
+                cp++;
+                ++ last->len;
             }
-            last->kind = CCL_TOK_TERM;
-
             aliases = ccl_qual_search_special(cclp->bibset, "and");
             if (!aliases)
                 aliases = cclp->ccl_token_and;
@@ -190,7 +202,7 @@ struct ccl_token *ccl_token_add(struct ccl_token *at)
     n->ws_prefix_len = 0;
     return n;
 }
-    
+
 /*
  * ccl_token_del: delete CCL tokens
  */
@@ -198,7 +210,7 @@ void ccl_token_del(struct ccl_token *list)
 {
     struct ccl_token *list1;
 
-    while (list) 
+    while (list)
     {
         list1 = list->next;
         xfree(list);
@@ -277,6 +289,7 @@ int ccl_parser_get_error(CCL_parser cclp, int *pos)
 /*
  * Local variables:
  * c-basic-offset: 4
+ * c-file-style: "Stroustrup"
  * indent-tabs-mode: nil
  * End:
  * vim: shiftwidth=4 tabstop=8 expandtab