* Europagate, 1995
*
* $Log: ccltoken.c,v $
- * Revision 1.7 1997-09-01 08:48:12 adam
+ * Revision 1.17 2001-10-03 23:54:41 adam
+ * Fixes for numeric ranges (date=1980-1990).
+ *
+ * Revision 1.16 2001/03/07 13:24:40 adam
+ * Member and_not in Z_Operator is kept for backwards compatibility.
+ * Added support for definition of CCL operators in field spec file.
+ *
+ * Revision 1.15 2000/05/01 09:36:50 adam
+ * Range operator only treated in ordered ranges so that minus (-) can be
+ * used for, say, the and-not operator.
+ *
+ * Revision 1.14 2000/03/14 09:06:11 adam
+ * Added POSIX threads support for frontend server.
+ *
+ * Revision 1.13 2000/02/08 10:39:53 adam
+ * Added a few functions to set name of operands, etc.
+ *
+ * Revision 1.12 2000/01/31 13:15:21 adam
+ * Removed uses of assert(3). Cleanup of ODR. CCL parser update so
+ * that some characters are not surrounded by spaces in resulting term.
+ * ILL-code updates.
+ *
+ * Revision 1.11 1999/11/30 13:47:11 adam
+ * Improved installation. Moved header files to include/yaz.
+ *
+ * Revision 1.10 1998/07/07 15:49:41 adam
+ * Added braces to avoid warning.
+ *
+ * Revision 1.9 1998/02/11 11:53:33 adam
+ * Changed code so that it compiles as C++.
+ *
+ * Revision 1.8 1997/09/29 08:56:38 adam
+ * Changed CCL parser to be thread safe. New type, CCL_parser, declared
+ * and a create/destructers ccl_parser_create/ccl_parser/destory has
+ * been added.
+ *
+ * Revision 1.7 1997/09/01 08:48:12 adam
* New windows NT/95 port using MSV5.0. Only a few changes made
* to avoid warnings.
*
*
*/
-#include <stdio.h>
#include <string.h>
#include <stdlib.h>
-#include <assert.h>
-#include <ccl.h>
-
-const char *ccl_token_and = "and";
-const char *ccl_token_or = "or";
-const char *ccl_token_not = "not andnot";
-const char *ccl_token_set = "set";
-int ccl_case_sensitive = 1;
+#include <yaz/ccl.h>
/*
* token_cmp: Compare token with keyword(s)
* return: 1 if token string matches one of the keywords in list;
* 0 otherwise.
*/
-static int token_cmp (const char *kw, struct ccl_token *token)
+static int token_cmp (CCL_parser cclp, const char *kw, struct ccl_token *token)
{
const char *cp1 = kw;
const char *cp2;
+ const char *aliases;
+ int case_sensitive = cclp->ccl_case_sensitive;
+
+ aliases = ccl_qual_search_special(cclp->bibset, "case");
+ if (aliases)
+ case_sensitive = atoi(aliases);
if (!kw)
return 0;
while ((cp2 = strchr (cp1, ' ')))
{
if (token->len == (size_t) (cp2-cp1))
- if (ccl_case_sensitive)
+ {
+ if (case_sensitive)
{
if (!memcmp (cp1, token->name, token->len))
return 1;
if (!ccl_memicmp (cp1, token->name, token->len))
return 1;
}
+ }
cp1 = cp2+1;
}
- if (ccl_case_sensitive)
+ if (case_sensitive)
return token->len == strlen(cp1)
&& !memcmp (cp1, token->name, token->len);
return token->len == strlen(cp1) &&
}
if (!first)
{
- first = last = malloc (sizeof (*first));
- assert (first);
+ first = last = (struct ccl_token *)malloc (sizeof (*first));
+ ccl_assert (first);
last->prev = NULL;
}
else
{
- last->next = malloc (sizeof(*first));
- assert (last->next);
+ last->next = (struct ccl_token *)malloc (sizeof(*first));
+ ccl_assert (last->next);
last->next->prev = last;
last = last->next;
}
return first;
}
+
/*
* ccl_tokenize: tokenize CCL command string.
* return: CCL token list.
*/
-struct ccl_token *ccl_tokenize (const char *command)
+struct ccl_token *ccl_parser_tokenize (CCL_parser cclp, const char *command)
{
+ const char *aliases;
const char *cp = command;
struct ccl_token *first = NULL;
struct ccl_token *last = NULL;
}
if (!first)
{
- first = last = malloc (sizeof (*first));
- assert (first);
+ first = last = (struct ccl_token *)malloc (sizeof (*first));
+ ccl_assert (first);
last->prev = NULL;
}
else
{
- last->next = malloc (sizeof(*first));
- assert (last->next);
+ last->next = (struct ccl_token *)malloc (sizeof(*first));
+ ccl_assert (last->next);
last->next->prev = last;
last = last->next;
}
else
last->kind = CCL_TOK_REL;
break;
- case '-':
- last->kind = CCL_TOK_MINUS;
- break;
case '\"':
last->kind = CCL_TOK_TERM;
last->name = cp;
cp++;
break;
default:
- while (*cp && !strchr ("(),%!><=- \t\n\r", *cp))
+ if (!strchr ("(),%!><=- \t\n\r", cp[-1]))
{
- cp++;
- ++ last->len;
+ while (*cp && !strchr ("(),%!><=- \t\n\r", *cp))
+ {
+ cp++;
+ ++ last->len;
+ }
}
- if (token_cmp (ccl_token_and, last))
+ last->kind = CCL_TOK_TERM;
+
+ aliases = ccl_qual_search_special(cclp->bibset, "and");
+ if (!aliases)
+ aliases = cclp->ccl_token_and;
+ if (token_cmp (cclp, aliases, last))
last->kind = CCL_TOK_AND;
- else if (token_cmp (ccl_token_or, last))
+
+ aliases = ccl_qual_search_special(cclp->bibset, "or");
+ if (!aliases)
+ aliases = cclp->ccl_token_or;
+ if (token_cmp (cclp, aliases, last))
last->kind = CCL_TOK_OR;
- else if (token_cmp (ccl_token_not, last))
+
+ aliases = ccl_qual_search_special(cclp->bibset, "not");
+ if (!aliases)
+ aliases = cclp->ccl_token_not;
+ if (token_cmp (cclp, aliases, last))
last->kind = CCL_TOK_NOT;
- else if (token_cmp (ccl_token_set, last))
+
+ aliases = ccl_qual_search_special(cclp->bibset, "set");
+ if (!aliases)
+ aliases = cclp->ccl_token_set;
+
+ if (token_cmp (cclp, aliases, last))
last->kind = CCL_TOK_SET;
- else
- last->kind = CCL_TOK_TERM;
}
}
return first;
}
+struct ccl_token *ccl_tokenize (const char *command)
+{
+ CCL_parser cclp = ccl_parser_create ();
+ struct ccl_token *list;
+
+ list = ccl_parser_tokenize (cclp, command);
+
+ ccl_parser_destroy (cclp);
+ return list;
+}
+
/*
* ccl_token_del: delete CCL tokens
*/
list = list1;
}
}
+
+char *ccl_strdup (const char *str)
+{
+ int len = strlen(str);
+ char *p = (char*) malloc (len+1);
+ strcpy (p, str);
+ return p;
+}
+
+CCL_parser ccl_parser_create (void)
+{
+ CCL_parser p = (CCL_parser)malloc (sizeof(*p));
+ if (!p)
+ return p;
+ p->look_token = NULL;
+ p->error_code = 0;
+ p->error_pos = NULL;
+ p->bibset = NULL;
+
+ p->ccl_token_and = ccl_strdup("and");
+ p->ccl_token_or = ccl_strdup("or");
+ p->ccl_token_not = ccl_strdup("not andnot");
+ p->ccl_token_set = ccl_strdup("set");
+ p->ccl_case_sensitive = 1;
+
+ return p;
+}
+
+void ccl_parser_destroy (CCL_parser p)
+{
+ if (!p)
+ return;
+ free (p->ccl_token_and);
+ free (p->ccl_token_or);
+ free (p->ccl_token_not);
+ free (p->ccl_token_set);
+ free (p);
+}
+
+void ccl_parser_set_op_and (CCL_parser p, const char *op)
+{
+ if (p && op)
+ {
+ if (p->ccl_token_and)
+ free (p->ccl_token_and);
+ p->ccl_token_and = ccl_strdup (op);
+ }
+}
+
+void ccl_parser_set_op_or (CCL_parser p, const char *op)
+{
+ if (p && op)
+ {
+ if (p->ccl_token_or)
+ free (p->ccl_token_or);
+ p->ccl_token_or = ccl_strdup (op);
+ }
+}
+void ccl_parser_set_op_not (CCL_parser p, const char *op)
+{
+ if (p && op)
+ {
+ if (p->ccl_token_not)
+ free (p->ccl_token_not);
+ p->ccl_token_not = ccl_strdup (op);
+ }
+}
+void ccl_parser_set_op_set (CCL_parser p, const char *op)
+{
+ if (p && op)
+ {
+ if (p->ccl_token_set)
+ free (p->ccl_token_set);
+ p->ccl_token_set = ccl_strdup (op);
+ }
+}
+
+void ccl_parser_set_case (CCL_parser p, int case_sensitivity_flag)
+{
+ if (p)
+ p->ccl_case_sensitive = case_sensitivity_flag;
+}