* Sebastian Hammer, Adam Dickmeiss
*
* $Log: zrpn.c,v $
- * Revision 1.93 1999-06-17 14:38:40 adam
+ * Revision 1.95 1999-09-07 07:19:21 adam
+ * Work on character mapping. Implemented replace rules.
+ *
+ * Revision 1.94 1999/07/20 13:59:18 adam
+ * Fixed bug that occurred when phrases had 0 hits.
+ *
+ * Revision 1.93 1999/06/17 14:38:40 adam
* Bug fix: Scan SEGV'ed when getting unknown use attribute.
*
* Revision 1.92 1999/05/26 07:49:13 adam
while (*src)
{
const char *cp = zebra_maps_output (zh->zebra_maps, reg_type, &src);
- while (*cp)
- *dst++ = *cp++;
+ if (!cp)
+ *dst++ = *src++;
+ else
+ while (*cp)
+ *dst++ = *cp++;
}
*dst = '\0';
}
return *s0;
}
+/* term_100: handle term, where trunc=none (no operators at all) */
static int term_100 (ZebraMaps zebra_maps, int reg_type,
const char **src, char *dst, int space_split,
char *dst_term)
return i;
}
+/* term_101: handle term, where trunc=Process # */
static int term_101 (ZebraMaps zebra_maps, int reg_type,
const char **src, char *dst, int space_split,
char *dst_term)
return i;
}
-
+/* term_103: handle term, where trunc=re-2 (regular expressions) */
static int term_103 (ZebraMaps zebra_maps, int reg_type, const char **src,
char *dst, int *errors, int space_split,
char *dst_term)
return i;
}
+/* term_103: handle term, where trunc=re-1 (regular expressions) */
static int term_102 (ZebraMaps zebra_maps, int reg_type, const char **src,
char *dst, int space_split, char *dst_term)
{
dst_term);
}
+
+/* term_104: handle term, where trunc=Process # and ! */
+static int term_104 (ZebraMaps zebra_maps, int reg_type,
+ const char **src, char *dst, int space_split,
+ char *dst_term)
+{
+ const char *s0, *s1;
+ const char **map;
+ int i = 0;
+ int j = 0;
+
+ if (!term_pre (zebra_maps, reg_type, src, "#!", "#!"))
+ return 0;
+ s0 = *src;
+ while (*s0)
+ {
+ if (*s0 == '#')
+ {
+ dst[i++] = '.';
+ dst[i++] = '*';
+ dst_term[j++] = *s0++;
+ }
+ else if (*s0 == '!')
+ {
+ dst[i++] = '.';
+ dst_term[j++] = *s0++;
+ }
+ {
+ s1 = s0;
+ map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
+ if (space_split && **map == *CHR_SPACE)
+ break;
+ while (s1 < s0)
+ {
+ if (!isalnum (*s1))
+ dst[i++] = '\\';
+ dst_term[j++] = *s1;
+ dst[i++] = *s1++;
+ }
+ }
+ }
+ dst[i] = '\0';
+ dst_term[j++] = '\0';
+ *src = s0;
+ return i;
+}
+
+
/* gen_regular_rel - generate regular expression from relation
* val: border value (inclusive)
* islt: 1 if <=; 0 if >=.
dst[dst_p] = '\0';
if (islt)
{
- for (i=1; i<pos; i++)
- strcat (dst, "[0-9]?");
+ /* match everything less than 10^(pos-1) */
+ strcat (dst, "0*");
+ for (i=1; i<pos; i++)
+ strcat (dst, "[0-9]?");
}
else
{
+ /* match everything greater than 10^pos */
for (i = 0; i <= pos; i++)
strcat (dst, "[0-9]");
strcat (dst, "[0-9]*");
logf (LOG_WARN, "dict_lookup_grep err, trunc=eregular: %d",
r);
break;
+ case 104: /* process # and ! in term */
+ term_dict[j++] = '(';
+ if (!term_104 (zh->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ return 0;
+ strcat (term_dict, ")");
+ r = dict_lookup_grep (zh->dict, term_dict, 0, grep_info,
+ &max_pos, 0, grep_handle);
+ if (r)
+ logf (LOG_WARN, "dict_lookup_grep err, trunc=#/!: %d", r);
+ break;
}
}
*term_sub = termp;
more = (int *) xmalloc (sizeof(*more)*rset_no);
buf = (struct it_key **) xmalloc (sizeof(*buf)*rset_no);
+ *prox_term = '\0';
for (i = 0; i<rset_no; i++)
{
int j;
- buf[i] = (struct it_key *) xmalloc (sizeof(**buf));
- rsfd[i] = rset_open (rset[i], RSETF_READ);
- if (!(more[i] = rset_read (rset[i], rsfd[i], buf[i], &term_index)))
- break;
for (j = 0; j<rset[i]->no_rset_terms; j++)
{
const char *nflags = rset[i]->rset_terms[j]->flags;
char *term = rset[i]->rset_terms[j]->name;
int lterm = strlen(term);
- if (length_prox_term)
- prox_term[length_prox_term++] = ' ';
- strcpy (prox_term + length_prox_term, term);
- length_prox_term += lterm;
+ if (lterm + length_prox_term < sizeof(prox_term)-1)
+ {
+ if (length_prox_term)
+ prox_term[length_prox_term++] = ' ';
+ strcpy (prox_term + length_prox_term, term);
+ length_prox_term += lterm;
+ }
if (min_nn > rset[i]->rset_terms[j]->nn)
min_nn = rset[i]->rset_terms[j]->nn;
flags = nflags;
}
}
+ for (i = 0; i<rset_no; i++)
+ {
+ buf[i] = (struct it_key *) xmalloc (sizeof(**buf));
+ rsfd[i] = rset_open (rset[i], RSETF_READ);
+ if (!(more[i] = rset_read (rset[i], rsfd[i], buf[i], &term_index)))
+ break;
+ }
if (i != rset_no)
{
rset_null_parms parms;
xfree (buf[i]);
--i;
}
- parms.rset_term = rset_term_create (prox_term, -1, flags);
+ parms.rset_term = rset_term_create (prox_term, length_prox_term,
+ flags);
parms.rset_term->nn = 0;
result = rset_create (rset_kind_null, &parms);
}
rset_temp_parms parms;
RSFD rsfd_result;
- parms.rset_term = rset_term_create (prox_term, -1, flags);
+ parms.rset_term = rset_term_create (prox_term, length_prox_term,
+ flags);
parms.rset_term->nn = min_nn;
parms.key_size = sizeof (struct it_key);
parms.temp_path = res_get (zh->res, "setTmpDir");
return result;
}
+
+char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char *termz, NMEM stream, unsigned reg_id)
+{
+ WRBUF wrbuf = 0;
+ AttrType truncation;
+ int truncation_value;
+ char *ex_list = 0;
+
+ attr_init (&truncation, zapt, 5);
+ truncation_value = attr_find (&truncation, NULL);
+
+ switch (truncation_value)
+ {
+ default:
+ ex_list = "";
+ break;
+ case 101:
+ ex_list = "#";
+ break;
+ case 102:
+ case 103:
+ ex_list = 0;
+ break;
+ case 104:
+ ex_list = "!#";
+ break;
+ }
+ if (ex_list)
+ wrbuf = zebra_replace(zh->zebra_maps, reg_id, ex_list,
+ termz, strlen(termz));
+ if (!wrbuf)
+ return nmem_strdup(stream, termz);
+ else
+ {
+ char *buf = (char*) nmem_malloc (stream, wrbuf_len(wrbuf)+1);
+ memcpy (buf, wrbuf_buf(wrbuf), wrbuf_len(wrbuf));
+ buf[wrbuf_len(wrbuf)] = '\0';
+ return buf;
+ }
+}
+
static RSET rpn_search_APT_phrase (ZebraHandle zh,
Z_AttributesPlusTerm *zapt,
- const char *termz,
+ const char *termz_org,
oid_value attributeSet,
NMEM stream,
int reg_type, int complete_flag,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
- const char *termp = termz;
RSET rset[60], result;
int i, r, rset_no = 0;
struct grep_info grep_info;
+ char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
+ const char *termp = termz;
#ifdef TERM_COUNT
grep_info.term_no = 0;
static RSET rpn_search_APT_or_list (ZebraHandle zh,
Z_AttributesPlusTerm *zapt,
- const char *termz,
+ const char *termz_org,
oid_value attributeSet,
NMEM stream,
int reg_type, int complete_flag,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
- const char *termp = termz;
RSET rset[60], result;
int i, r, rset_no = 0;
struct grep_info grep_info;
-
+ char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
+ const char *termp = termz;
#ifdef TERM_COUNT
grep_info.term_no = 0;
#endif
static RSET rpn_search_APT_and_list (ZebraHandle zh,
Z_AttributesPlusTerm *zapt,
- const char *termz,
+ const char *termz_org,
oid_value attributeSet,
NMEM stream,
int reg_type, int complete_flag,
int num_bases, char **basenames)
{
char term_dst[IT_MAX_WORD+1];
- const char *termp = termz;
RSET rset[60], result;
int i, r, rset_no = 0;
struct grep_info grep_info;
+ char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type);
+ const char *termp = termz;
#ifdef TERM_COUNT
grep_info.term_no = 0;
len_prefix = strlen(scan_info->prefix);
if (memcmp (name, scan_info->prefix, len_prefix))
return 1;
- if (pos > 0)
- idx = scan_info->after - pos + scan_info->before;
+ if (pos > 0) idx = scan_info->after - pos + scan_info->before;
else
idx = - pos - 1;
scan_info->list[idx].term = (char *)