+typedef struct {
+ int type;
+ int major;
+ int minor;
+ Z_AttributesPlusTerm *zapt;
+} AttrType;
+
+static int attr_find (AttrType *src, oid_value *attributeSetP)
+{
+ int num_attributes;
+
+#ifdef ASN_COMPILED
+ num_attributes = src->zapt->attributes->num_attributes;
+#else
+ num_attributes = src->zapt->num_attributes;
+#endif
+ while (src->major < num_attributes)
+ {
+ Z_AttributeElement *element;
+
+#ifdef ASN_COMPILED
+ element = src->zapt->attributes->attributes[src->major];
+#else
+ element = src->zapt->attributeList[src->major];
+#endif
+ if (src->type == *element->attributeType)
+ {
+ switch (element->which)
+ {
+ case Z_AttributeValue_numeric:
+ ++(src->major);
+ if (element->attributeSet && attributeSetP)
+ {
+ oident *attrset;
+
+ attrset = oid_getentbyoid (element->attributeSet);
+ *attributeSetP = attrset->value;
+ }
+ return *element->value.numeric;
+ break;
+ case Z_AttributeValue_complex:
+ if (src->minor >= element->value.complex->num_list ||
+ element->value.complex->list[src->minor]->which !=
+ Z_StringOrNumeric_numeric)
+ break;
+ ++(src->minor);
+ if (element->attributeSet && attributeSetP)
+ {
+ oident *attrset;
+
+ attrset = oid_getentbyoid (element->attributeSet);
+ *attributeSetP = attrset->value;
+ }
+ return *element->value.complex->list[src->minor-1]->u.numeric;
+ default:
+ assert (0);
+ }
+ }
+ ++(src->major);
+ }
+ return -1;
+}
+
+static void attr_init (AttrType *src, Z_AttributesPlusTerm *zapt,
+ int type)
+{
+ src->zapt = zapt;
+ src->type = type;
+ src->major = 0;
+ src->minor = 0;
+}
+
+#define TERM_COUNT
+
+struct grep_info {
+#ifdef TERM_COUNT
+ int *term_no;
+#endif
+ ISAM_P *isam_p_buf;
+ int isam_p_size;
+ int isam_p_indx;
+ ZebraHandle zh;
+ int reg_type;
+};
+
+static void term_untrans (ZebraHandle zh, int reg_type,
+ char *dst, const char *src)
+{
+ while (*src)
+ {
+ const char *cp = zebra_maps_output (zh->zebra_maps, reg_type, &src);
+ while (*cp)
+ *dst++ = *cp++;
+ }
+ *dst = '\0';
+}
+
+static void add_isam_p (const char *name, const char *info,
+ struct grep_info *p)
+{
+ char term_tmp[1024];
+ if (p->isam_p_indx == p->isam_p_size)
+ {
+ ISAM_P *new_isam_p_buf;
+#ifdef TERM_COUNT
+ int *new_term_no;
+#endif
+ p->isam_p_size = 2*p->isam_p_size + 100;
+ new_isam_p_buf = xmalloc (sizeof(*new_isam_p_buf) *
+ p->isam_p_size);
+ if (p->isam_p_buf)
+ {
+ memcpy (new_isam_p_buf, p->isam_p_buf,
+ p->isam_p_indx * sizeof(*p->isam_p_buf));
+ xfree (p->isam_p_buf);
+ }
+ p->isam_p_buf = new_isam_p_buf;
+
+#ifdef TERM_COUNT
+ new_term_no = xmalloc (sizeof(*new_term_no) *
+ p->isam_p_size);
+ if (p->term_no)
+ {
+ memcpy (new_term_no, p->isam_p_buf,
+ p->isam_p_indx * sizeof(*p->term_no));
+ xfree (p->term_no);
+ }
+ p->term_no = new_term_no;
+#endif
+ }
+ assert (*info == sizeof(*p->isam_p_buf));
+ memcpy (p->isam_p_buf + p->isam_p_indx, info+1, sizeof(*p->isam_p_buf));
+
+#if 0
+ term_untrans (p->zh, p->reg_type, term_tmp, name+2);
+ logf (LOG_DEBUG, "grep: %s", term_tmp);
+#endif
+ (p->isam_p_indx)++;
+}
+
+static int grep_handle (char *name, const char *info, void *p)
+{
+ add_isam_p (name, info, p);
+ return 0;
+}
+
+static int term_pre (ZebraMaps zebra_maps, int reg_type, const char **src,
+ const char *ct1, const char *ct2)
+{
+ const char *s1, *s0 = *src;
+ const char **map;
+
+ /* skip white space */
+ while (*s0)
+ {
+ if (ct1 && strchr (ct1, *s0))
+ break;
+ if (ct2 && strchr (ct2, *s0))
+ break;
+ s1 = s0;
+ map = zebra_maps_input (zebra_maps, reg_type, &s1, strlen(s1));
+ if (**map != *CHR_SPACE)
+ break;
+ s0 = s1;
+ }
+ *src = s0;
+ return *s0;
+}
+
+static int term_100 (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, NULL, NULL))
+ return 0;
+ s0 = *src;
+ while (*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) && *s1 != '-')
+ dst[i++] = '\\';
+ dst_term[j++] = *s1;
+ dst[i++] = *s1++;
+ }
+ }
+ dst[i] = '\0';
+ dst_term[j] = '\0';
+ *src = s0;
+ return i;
+}
+
+static int term_101 (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
+ {
+ 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;
+}
+
+
+static int term_103 (ZebraMaps zebra_maps, int reg_type, const char **src,
+ char *dst, int *errors, int space_split,
+ char *dst_term)
+{
+ int i = 0;
+ int j = 0;
+ const char *s0, *s1;
+ const char **map;
+
+ if (!term_pre (zebra_maps, reg_type, src, "^\\()[].*+?|", "("))
+ return 0;
+ s0 = *src;
+ if (errors && *s0 == '+' && s0[1] && s0[2] == '+' && s0[3] &&
+ isdigit (s0[1]))
+ {
+ *errors = s0[1] - '0';
+ s0 += 3;
+ if (*errors > 3)
+ *errors = 3;
+ }
+ while (*s0)
+ {
+ if (strchr ("^\\()[].*+?|-", *s0))
+ {
+ dst_term[j++] = *s0;
+ dst[i++] = *s0++;
+ }
+ else
+ {
+ s1 = s0;
+ map = zebra_maps_input (zebra_maps, reg_type, &s0, strlen(s0));
+ if (**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;
+}
+
+static int term_102 (ZebraMaps zebra_maps, int reg_type, const char **src,
+ char *dst, int space_split, char *dst_term)
+{
+ return term_103 (zebra_maps, reg_type, src, dst, NULL, space_split,
+ dst_term);
+}
+
+/* gen_regular_rel - generate regular expression from relation
+ * val: border value (inclusive)
+ * islt: 1 if <=; 0 if >=.
+ */
+static void gen_regular_rel (char *dst, int val, int islt)
+{
+ int dst_p;
+ int w, d, i;
+ int pos = 0;
+ char numstr[20];
+
+ logf (LOG_DEBUG, "gen_regular_rel. val=%d, islt=%d", val, islt);
+ if (val >= 0)
+ {
+ if (islt)
+ strcpy (dst, "(-[0-9]+|(");
+ else
+ strcpy (dst, "((");
+ }
+ else
+ {
+ if (!islt)
+ {
+ strcpy (dst, "([0-9]+|-(");
+ dst_p = strlen (dst);
+ islt = 1;
+ }
+ else
+ {
+ strcpy (dst, "((-");
+ islt = 0;
+ }
+ val = -val;
+ }
+ dst_p = strlen (dst);
+ sprintf (numstr, "%d", val);
+ for (w = strlen(numstr); --w >= 0; pos++)
+ {
+ d = numstr[w];
+ if (pos > 0)
+ {
+ if (islt)
+ {
+ if (d == '0')
+ continue;
+ d--;
+ }
+ else
+ {
+ if (d == '9')
+ continue;
+ d++;
+ }
+ }
+
+ strcpy (dst + dst_p, numstr);
+ dst_p = strlen(dst) - pos - 1;
+
+ if (islt)
+ {
+ if (d != '0')
+ {
+ dst[dst_p++] = '[';
+ dst[dst_p++] = '0';
+ dst[dst_p++] = '-';
+ dst[dst_p++] = d;
+ dst[dst_p++] = ']';
+ }
+ else
+ dst[dst_p++] = d;
+ }
+ else
+ {
+ if (d != '9')
+ {
+ dst[dst_p++] = '[';
+ dst[dst_p++] = d;
+ dst[dst_p++] = '-';
+ dst[dst_p++] = '9';
+ dst[dst_p++] = ']';
+ }
+ else
+ dst[dst_p++] = d;
+ }
+ for (i = 0; i<pos; i++)
+ {
+ dst[dst_p++] = '[';
+ dst[dst_p++] = '0';
+ dst[dst_p++] = '-';
+ dst[dst_p++] = '9';
+ dst[dst_p++] = ']';
+ }
+ dst[dst_p++] = '|';
+ }
+ dst[dst_p] = '\0';
+ if (islt)
+ {
+ for (i=1; i<pos; i++)
+ strcat (dst, "[0-9]?");
+ }
+ else
+ {
+ for (i = 0; i <= pos; i++)
+ strcat (dst, "[0-9]");
+ strcat (dst, "[0-9]*");
+ }
+ strcat (dst, "))");
+}
+
+static int relational_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ char *term_dict,
+ oid_value attributeSet,
+ struct grep_info *grep_info,
+ int *max_pos,
+ int reg_type,
+ char *term_dst)
+{
+ AttrType relation;
+ int relation_value;
+ int term_value;
+ int r;
+ char *term_tmp = term_dict + strlen(term_dict);
+
+ attr_init (&relation, zapt, 2);
+ relation_value = attr_find (&relation, NULL);
+
+ logf (LOG_DEBUG, "relation value=%d", relation_value);
+ switch (relation_value)
+ {
+ case 1:
+ if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1,
+ term_dst))
+ return 0;
+ term_value = atoi (term_tmp);
+ logf (LOG_DEBUG, "Relation <");
+ gen_regular_rel (term_tmp, term_value-1, 1);
+ break;
+ case 2:
+ if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1,
+ term_dst))
+ return 0;
+ term_value = atoi (term_tmp);
+ logf (LOG_DEBUG, "Relation <=");
+ gen_regular_rel (term_tmp, term_value, 1);
+ break;
+ case 4:
+ if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1,
+ term_dst))
+ return 0;
+ term_value = atoi (term_tmp);
+ logf (LOG_DEBUG, "Relation >=");
+ gen_regular_rel (term_tmp, term_value, 0);
+ break;
+ case 5:
+ if (!term_100 (zh->zebra_maps, reg_type, term_sub, term_tmp, 1,
+ term_dst))
+ return 0;
+ term_value = atoi (term_tmp);
+ logf (LOG_DEBUG, "Relation >");
+ gen_regular_rel (term_tmp, term_value+1, 0);
+ break;
+ default:
+ return 0;
+ }
+ logf (LOG_DEBUG, "dict_lookup_grep: %s", term_tmp);
+ r = dict_lookup_grep (zh->dict, term_dict, 0, grep_info, max_pos,
+ 0, grep_handle);
+ if (r)
+ logf (LOG_WARN, "dict_lookup_grep fail, rel=gt: %d", r);
+ logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
+ return 1;
+}
+
+static int field_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ const char **term_sub,
+ oid_value attributeSet, struct grep_info *grep_info,
+ int reg_type, int complete_flag,
+ int num_bases, char **basenames,
+ char *term_dst)
+{
+ char term_dict[2*IT_MAX_WORD+2];
+ int j, r, base_no;
+ AttrType truncation;
+ int truncation_value;
+ AttrType use;
+ int use_value;
+ oid_value curAttributeSet = attributeSet;
+ const char *termp;
+ struct rpn_char_map_info rcmi;
+ int space_split = complete_flag ? 0 : 1;
+
+ rpn_char_map_prepare (zh, reg_type, &rcmi);
+ attr_init (&use, zapt, 1);
+ use_value = attr_find (&use, &curAttributeSet);
+ logf (LOG_DEBUG, "field_term, use value %d", use_value);
+ attr_init (&truncation, zapt, 5);
+ truncation_value = attr_find (&truncation, NULL);
+ logf (LOG_DEBUG, "truncation value %d", truncation_value);
+
+ if (use_value == -1)
+ use_value = 1016;
+
+ for (base_no = 0; base_no < num_bases; base_no++)
+ {
+ attent attp;
+ data1_local_attribute *local_attr;
+ int max_pos, prefix_len = 0;
+
+ termp = *term_sub;
+ if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value)))
+ {
+ logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d",
+ curAttributeSet, use_value, r);
+ if (r == -1)
+ zh->errCode = 114;
+ else
+ zh->errCode = 121;
+ return -1;
+ }
+ if (zebraExplain_curDatabase (zh->zei, basenames[base_no]))
+ {
+ zh->errCode = 109; /* Database unavailable */
+ zh->errString = basenames[base_no];
+ return -1;
+ }
+ for (local_attr = attp.local_attributes; local_attr;
+ local_attr = local_attr->next)
+ {
+ int ord;
+ char ord_buf[32];
+ int i, ord_len;
+
+ ord = zebraExplain_lookupSU (zh->zei, attp.attset_ordinal,
+ local_attr->local);
+ if (ord < 0)
+ continue;
+ if (prefix_len)
+ term_dict[prefix_len++] = '|';
+ else
+ term_dict[prefix_len++] = '(';
+
+ ord_len = key_SU_code (ord, ord_buf);
+ for (i = 0; i<ord_len; i++)
+ {
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = ord_buf[i];
+ }
+ }
+ if (!prefix_len)
+ {
+ zh->errCode = 114;
+ return -1;
+ }
+ term_dict[prefix_len++] = ')';
+ term_dict[prefix_len++] = 1;
+ term_dict[prefix_len++] = reg_type;
+ logf (LOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]);
+ term_dict[prefix_len] = '\0';
+ if (!relational_term (zh, zapt, &termp, term_dict,
+ attributeSet, grep_info, &max_pos, reg_type,
+ term_dst))
+ {
+ j = prefix_len;
+ switch (truncation_value)
+ {
+ case -1: /* not specified */
+ case 100: /* do not truncate */
+ term_dict[j++] = '(';
+ if (!term_100 (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=none:%d", r);
+ break;
+ case 1: /* right truncation */
+ term_dict[j++] = '(';
+ if (!term_100 (zh->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ return 0;
+ strcat (term_dict, ".*)");
+ dict_lookup_grep (zh->dict, term_dict, 0, grep_info,
+ &max_pos, 0, grep_handle);
+ break;
+ case 2: /* left truncation */
+ case 3: /* left&right truncation */
+ zh->errCode = 120;
+ return -1;
+ case 101: /* process # in term */
+ term_dict[j++] = '(';
+ if (!term_101 (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;
+ case 102: /* Regexp-1 */
+ term_dict[j++] = '(';
+ if (!term_102 (zh->zebra_maps, reg_type,
+ &termp, term_dict + j, space_split, term_dst))
+ return 0;
+ strcat (term_dict, ")");
+ logf (LOG_DEBUG, "Regexp-1 tolerance=%d", r);
+ 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=regular: %d",
+ r);
+ break;
+ case 103: /* Regexp-2 */
+ r = 1;
+ term_dict[j++] = '(';
+ if (!term_103 (zh->zebra_maps, reg_type,
+ &termp, term_dict + j, &r, space_split, term_dst))
+ return 0;
+ strcat (term_dict, ")");
+ logf (LOG_DEBUG, "Regexp-2 tolerance=%d", r);
+ r = dict_lookup_grep (zh->dict, term_dict, r, grep_info,
+ &max_pos, 2, grep_handle);
+ if (r)
+ logf (LOG_WARN, "dict_lookup_grep err, trunc=eregular: %d",
+ r);
+ break;
+ }
+ }
+ }
+ *term_sub = termp;
+ logf (LOG_DEBUG, "%d positions", grep_info->isam_p_indx);
+ return 1;
+}
+
+static void trans_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
+ char *termz)