1 /* $Id: rpnscan.c,v 1.22 2007-11-13 13:41:51 adam Exp $
2 Copyright (C) 1995-2007
5 This file is part of the Zebra server.
7 Zebra is free software; you can redistribute it and/or modify it under
8 the terms of the GNU General Public License as published by the Free
9 Software Foundation; either version 2, or (at your option) any later
12 Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
13 WARRANTY; without even the implied warranty of MERCHANTABILITY or
14 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
17 You should have received a copy of the GNU General Public License
18 along with this program; if not, write to the Free Software
19 Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
33 #include <yaz/diagbib1.h>
35 #include <zebra_xpath.h>
36 #include <yaz/wrbuf.h>
40 #include <yaz/oid_db.h>
42 #define RPN_MAX_ORDS 32
44 /* convert APT SCAN term to internal cmap */
45 static ZEBRA_RES trans_scan_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt,
46 char *termz, zebra_map_t zm)
48 char termz0[IT_MAX_WORD];
50 if (zapt_term_to_utf8(zh, zapt, termz0) == ZEBRA_FAIL)
51 return ZEBRA_FAIL; /* error */
55 const char *cp = (const char *) termz0;
56 const char *cp_end = cp + strlen(cp);
59 const char *space_map = NULL;
62 while ((len = (cp_end - cp)) > 0)
64 map = zebra_maps_input(zm, &cp, len, 0);
65 if (**map == *CHR_SPACE)
70 for (src = space_map; *src; src++)
73 for (src = *map; *src; src++)
82 static void count_set(ZebraHandle zh, RSET rset, zint *count, zint approx_limit)
88 yaz_log(YLOG_DEBUG, "count_set");
90 rset->hits_limit = approx_limit;
93 rfd = rset_open(rset, RSETF_READ);
94 while (rset_read(rfd, &key,0 /* never mind terms */))
96 if (key.mem[0] != psysno)
99 if (rfd->counted_items >= rset->hits_limit)
104 *count = rset->hits_count;
107 static void get_first_snippet_from_rset(ZebraHandle zh,
108 RSET rset, zebra_snippets *snippets,
114 size_t sysno_mem_index = 0;
116 if (zh->m_staticrank)
119 yaz_log(YLOG_DEBUG, "get_first_snippet_from_rset");
121 rfd = rset_open(rset, RSETF_READ);
123 while (rset_read(rfd, &key, &termid))
125 if (key.mem[sysno_mem_index] != *sysno)
129 *sysno = key.mem[sysno_mem_index];
134 for (ol = termid->ol; ol; ol = ol->next)
136 zebra_snippets_append(snippets, key.mem[key.len-1], 0,
137 ol->ord, termid->name);
144 struct scan2_info_entry {
152 static int scan_handle2(char *name, const char *info, int pos, void *client)
155 struct scan2_info_entry *scan_info = (struct scan2_info_entry *) client;
157 if (scan_info->pos_to_save != pos)
160 len_prefix = strlen(scan_info->prefix);
161 if (memcmp(name, scan_info->prefix, len_prefix))
164 /* skip special terms such as first-in-field specials */
165 if (name[len_prefix] < CHR_BASE_CHAR)
168 wrbuf_rewind(scan_info->term);
169 wrbuf_puts(scan_info->term, name+len_prefix);
171 assert(*info == sizeof(ISAM_P));
172 memcpy(&scan_info->isam_p, info+1, sizeof(ISAM_P));
177 static int scan_save_set(ZebraHandle zh, ODR stream, NMEM nmem,
178 struct rset_key_control *kc,
179 Z_AttributesPlusTerm *zapt,
182 const char *index_type,
183 struct scan2_info_entry *ar, int ord_no,
184 ZebraScanEntry *glist, int pos)
188 zint approx_limit = zh->approx_limit;
189 AttrType global_hits_limit_attr;
191 attr_init_APT(&global_hits_limit_attr, zapt, 12);
193 l = attr_find(&global_hits_limit_attr, NULL);
197 for (i = 0; i < ord_no; i++)
199 if (ar[i].isam_p && strcmp(wrbuf_cstr(ar[i].term), term) == 0)
201 struct ord_list *ol = ord_list_create(nmem);
204 ol = ord_list_append(nmem, ol, ar[i].ord);
208 zh, &ar[i].isam_p, 1,
209 wrbuf_buf(ar[i].term), wrbuf_len(ar[i].term),
210 NULL, 1, zapt->term->which, nmem,
211 kc, kc->scope, ol, index_type,
212 0 /* hits_limit_value */,
213 0 /* term_ref_id_str */);
222 rset = rset_create_or(nmem, kc, kc->scope, 0 /* termid */,
231 /* merge with limit_set if given */
236 rsets[1] = rset_dup(limit_set);
238 rset = rset_create_and(nmem, kc, kc->scope, 2, rsets);
241 count_set(zh, rset, &count, approx_limit);
247 zebra_snippets *rec_snippets = zebra_snippets_create();
248 zebra_snippets *hit_snippets = zebra_snippets_create();
251 glist[pos].display_term = 0;
253 get_first_snippet_from_rset(zh, rset, hit_snippets, &sysno);
255 code = zebra_get_rec_snippets(zh, sysno, rec_snippets);
259 const struct zebra_snippet_word *w =
260 zebra_snippets_lookup(rec_snippets, hit_snippets);
263 glist[pos].display_term = odr_strdup(stream, w->term);
266 if (!glist[pos].term)
267 zebra_term_untrans_iconv(zh, stream->mem, index_type,
268 &glist[pos].term, term);
269 glist[pos].occurrences = count;
270 zebra_snippets_destroy(rec_snippets);
271 zebra_snippets_destroy(hit_snippets);
282 static ZEBRA_RES rpn_scan_norm(ZebraHandle zh, ODR stream, NMEM nmem,
283 struct rset_key_control *kc,
284 Z_AttributesPlusTerm *zapt,
285 int *position, int *num_entries,
286 ZebraScanEntry **list,
287 int *is_partial, RSET limit_set,
288 const char *index_type,
289 int ord_no, int *ords)
291 struct scan2_info_entry *ar = nmem_malloc(nmem, sizeof(*ar) * ord_no);
292 struct rpn_char_map_info rcmi;
293 zebra_map_t zm = zebra_map_get_or_add(zh->reg->zebra_maps, index_type);
298 ZebraScanEntry *glist = (ZebraScanEntry *)
299 odr_malloc(stream, *num_entries * sizeof(*glist));
302 if (*position > *num_entries+1)
309 rpn_char_map_prepare(zh->reg, zm, &rcmi);
311 for (i = 0; i < ord_no; i++)
312 ar[i].term = wrbuf_alloc();
314 for (i = 0; i < ord_no; i++)
316 char termz[IT_MAX_WORD+20];
319 prefix_len = key_SU_encode(ords[i], termz);
320 termz[prefix_len] = 0;
321 strcpy(ar[i].prefix, termz);
323 if (trans_scan_term(zh, zapt, termz+prefix_len, zm) ==
326 for (i = 0; i < ord_no; i++)
327 wrbuf_destroy(ar[i].term);
330 wrbuf_rewind(ar[i].term);
331 wrbuf_puts(ar[i].term, termz + prefix_len);
335 /** deal with terms before position .. */
336 /* the glist index starts at zero (unlike scan positions */
337 for (pos = *position-2; pos >= 0; )
341 /* scan on all maximum terms */
342 for (i = 0; i < ord_no; i++)
344 if (ar[i].isam_p == 0)
346 char termz[IT_MAX_WORD+20];
350 ar[i].pos_to_save = -1;
352 strcpy(termz, ar[i].prefix);
353 strcat(termz, wrbuf_cstr(ar[i].term));
354 dict_scan(zh->reg->dict, termz, &before, &after,
358 /* get maximum after scan */
359 for (i = 0; i < ord_no; i++)
362 && (hi == 0 || strcmp(wrbuf_cstr(ar[i].term), hi) > 0))
363 hi = wrbuf_cstr(ar[i].term);
367 if (scan_save_set(zh, stream, nmem, kc, zapt, limit_set, hi,
368 index_type, ar, ord_no, glist,
369 (pos >= 0 && pos < *num_entries) ? pos : -1))
372 /* see if we got all terms before.. */
376 /* did not get all terms; adjust the real position and reduce
378 yaz_log(YLOG_LOG, "before terms dif=%d", dif);
384 for (i = 0; i < ord_no; i++)
386 char termz[IT_MAX_WORD+20];
389 prefix_len = key_SU_encode(ords[i], termz);
390 termz[prefix_len] = 0;
391 strcpy(ar[i].prefix, termz);
393 if (trans_scan_term(zh, zapt, termz+prefix_len, zm) ==
396 wrbuf_rewind(ar[i].term);
397 wrbuf_puts(ar[i].term, termz + prefix_len);
402 after_pos = 1; /* immediate term first.. */
403 for (pos = *position-1; pos < *num_entries; )
407 /* scan on all minimum terms */
408 for (i = 0; i < ord_no; i++)
410 if (ar[i].isam_p == 0)
412 char termz[IT_MAX_WORD+20];
414 int after = after_pos;
416 ar[i].pos_to_save = 1;
418 strcpy(termz, ar[i].prefix);
419 strcat(termz, wrbuf_cstr(ar[i].term));
420 dict_scan(zh->reg->dict, termz, &before, &after,
424 after_pos = 2; /* next round we grab following term */
426 /* get minimum after scan */
427 for (i = 0; i < ord_no; i++)
430 && (lo == 0 || strcmp(wrbuf_cstr(ar[i].term), lo) < 0))
431 lo = wrbuf_cstr(ar[i].term);
435 if (scan_save_set(zh, stream, nmem, kc, zapt, limit_set, lo,
436 index_type, ar, ord_no, glist,
437 (pos >= 0 && pos < *num_entries) ? pos : -1))
441 if (pos != *num_entries)
452 for (i = 0; i < ord_no; i++)
453 wrbuf_destroy(ar[i].term);
458 struct scan1_info_entry {
464 struct scan1_info_entry *list;
470 ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
471 const Odr_oid *attributeset,
472 int num_bases, char **basenames,
473 int *position, int *num_entries, ZebraScanEntry **list,
474 int *is_partial, const char *set_name)
477 int ords[RPN_MAX_ORDS], ord_no = 0;
479 const char *index_type;
480 char *search_type = NULL;
486 struct rset_key_control *kc = 0;
493 attributeset = yaz_oid_attset_bib_1;
497 /* see if there is a @attr 8=set */
499 int termset_value_numeric;
500 const char *termset_value_string = 0;
501 attr_init_APT(&termset, zapt, 8);
502 termset_value_numeric =
503 attr_find_ex(&termset, NULL, &termset_value_string);
504 if (termset_value_numeric != -1)
506 if (termset_value_numeric != -2)
509 sprintf(resname, "%d", termset_value_numeric);
510 set_name = odr_strdup(stream, resname);
513 set_name = odr_strdup(stream, termset_value_string);
519 limit_set = resultSetRef(zh, set_name);
524 YAZ_BIB1_SPECIFIED_RESULT_SET_DOES_NOT_EXIST,
530 yaz_log(YLOG_DEBUG, "position = %d, num = %d",
531 *position, *num_entries);
533 if (zebra_maps_attr(zh->reg->zebra_maps, zapt, &index_type, &search_type,
534 rank_type, &complete_flag, &sort_flag))
537 zebra_setError(zh, YAZ_BIB1_UNSUPP_ATTRIBUTE_TYPE, 0);
540 if (num_bases > RPN_MAX_ORDS)
542 zebra_setError(zh, YAZ_BIB1_TOO_MANY_DATABASES_SPECIFIED, 0);
547 return rpn_facet(zh, stream, zapt, attributeset, position, num_entries,
548 list, is_partial, set_name);
550 for (base_no = 0; base_no < num_bases; base_no++)
554 if (zebraExplain_curDatabase(zh->reg->zei, basenames[base_no]))
556 zebra_setError(zh, YAZ_BIB1_DATABASE_UNAVAILABLE,
561 if (zebra_apt_get_ord(zh, zapt, index_type, 0, attributeset, &ord)
564 ords[ord_no++] = ord;
568 *num_entries = 0; /* zebra_apt_get_ord should set error reason */
571 if (*num_entries < 1)
576 nmem = nmem_create();
577 kc = zebra_key_control_create(zh);
579 res = rpn_scan_norm(zh, stream, nmem, kc, zapt, position, num_entries,
580 list, is_partial, limit_set,
581 index_type, ord_no, ords);
590 * indent-tabs-mode: nil
592 * vim: shiftwidth=4 tabstop=8 expandtab