1 /* This file is part of Pazpar2.
2 Copyright (C) 2006-2009 Index Data
4 Pazpar2 is free software; you can redistribute it and/or modify it under
5 the terms of the GNU General Public License as published by the Free
6 Software Foundation; either version 2, or (at your option) any later
9 Pazpar2 is distributed in the hope that it will be useful, but WITHOUT ANY
10 WARRANTY; without even the implied warranty of MERCHANTABILITY or
11 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, write to the Free Software
16 Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
26 #include <yaz/yaz-util.h>
30 #include "jenkins_hash.h"
34 struct reclist_bucket **hashtable;
38 struct record_cluster **flatlist;
46 static struct reclist_sortparms *qsort_sortparms = 0; /* thread pr */
50 struct record_cluster *record;
51 struct reclist_bucket *next;
55 struct reclist_sortparms *
56 reclist_sortparms_insert_field_id(NMEM nmem,
57 struct reclist_sortparms **sortparms,
59 enum conf_sortkey_type type,
62 struct reclist_sortparms * tmp_rlsp = 0;
65 if(!sortparms || field_id < 0)
68 // construct new reclist_sortparms
69 tmp_rlsp = nmem_malloc(nmem, sizeof(struct reclist_sortparms));
70 tmp_rlsp->offset = field_id;
71 tmp_rlsp->type = type;
72 tmp_rlsp->increasing = increasing;
75 // insert in *sortparms place, moving *sortparms one down the list
76 tmp_rlsp->next = *sortparms;
77 *sortparms = tmp_rlsp;
84 struct reclist_sortparms *
85 reclist_sortparms_insert(NMEM nmem,
86 struct reclist_sortparms **sortparms,
87 struct conf_service * service,
93 if (!sortparms || !service || !name)
96 field_id = conf_service_sortkey_field_id(service, name);
101 return reclist_sortparms_insert_field_id(nmem, sortparms, field_id,
102 service->sortkeys[field_id].type,
108 struct reclist_sortparms *reclist_parse_sortparms(NMEM nmem, const char *parms,
109 struct conf_service *service)
111 struct reclist_sortparms *res = 0;
112 struct reclist_sortparms **rp = &res;
114 if (strlen(parms) > 256)
124 enum conf_sortkey_type type;
125 struct reclist_sortparms *new;
127 if (!(cpp = strchr(parms, ',')))
128 cpp = parms + strlen(parms);
129 strncpy(parm, parms, cpp - parms);
130 parm[cpp-parms] = '\0';
132 if ((pp = strchr(parm, ':')))
134 increasing = pp[1] == '1' ? 1 : 0;
139 if (!strcmp(parm, "relevance"))
141 type = Metadata_sortkey_relevance;
145 for (i = 0; i < service->num_sortkeys; i++)
147 struct conf_sortkey *sk = &service->sortkeys[i];
148 if (!strcmp(sk->name, parm))
151 if (type == Metadata_sortkey_skiparticle)
152 type = Metadata_sortkey_string;
156 if (i >= service->num_sortkeys)
158 yaz_log(YLOG_FATAL, "Bad sortkey: %s", parm);
164 new = *rp = nmem_malloc(nmem, sizeof(struct reclist_sortparms));
166 new->offset = offset;
168 new->increasing = increasing;
176 static int reclist_cmp(const void *p1, const void *p2)
178 struct record_cluster *r1 = (*(struct record_cluster**) p1);
179 struct record_cluster *r2 = (*(struct record_cluster**) p2);
180 struct reclist_sortparms *s;
183 for (s = qsort_sortparms; s && res == 0; s = s->next)
185 union data_types *ut1 = r1->sortkeys[s->offset];
186 union data_types *ut2 = r2->sortkeys[s->offset];
191 case Metadata_sortkey_relevance:
192 res = r2->relevance - r1->relevance;
194 case Metadata_sortkey_string:
195 s1 = ut1 ? ut1->text.sort : "";
196 s2 = ut2 ? ut2->text.sort : "";
197 res = strcmp(s2, s1);
204 case Metadata_sortkey_numeric:
208 res = ut1->number.min - ut2->number.min;
210 res = ut2->number.max - ut1->number.max;
212 else if (ut1 && !ut2)
214 else if (!ut1 && ut2)
220 yaz_log(YLOG_WARN, "Bad sort type: %d", s->type);
227 void reclist_sort(struct reclist *l, struct reclist_sortparms *parms)
229 qsort_sortparms = parms;
230 qsort(l->flatlist, l->num_records,
231 sizeof(struct record_cluster*), reclist_cmp);
235 struct record_cluster *reclist_read_record(struct reclist *l)
237 if (l && l->pointer < l->num_records)
238 return l->flatlist[l->pointer++];
243 void reclist_rewind(struct reclist *l)
249 struct reclist *reclist_create(NMEM nmem, int numrecs)
255 while (hashsize < numrecs)
257 res = nmem_malloc(nmem, sizeof(struct reclist));
259 = nmem_malloc(nmem, hashsize * sizeof(struct reclist_bucket*));
260 memset(res->hashtable, 0, hashsize * sizeof(struct reclist_bucket*));
261 res->hashtable_size = hashsize;
263 res->hashmask = hashsize - 1; // Creates a bitmask
265 res->num_records = 0;
266 res->flatlist = nmem_malloc(nmem, numrecs * sizeof(struct record_cluster*));
267 res->flatlist_size = numrecs;
272 int reclist_get_num_records(struct reclist *l)
275 return l->num_records;
279 struct record_cluster *reclist_get_cluster(struct reclist *l, int i)
281 return l->flatlist[i];
284 // Insert a record. Return record cluster (newly formed or pre-existing)
285 struct record_cluster *reclist_insert( struct reclist *l,
286 struct conf_service *service,
287 struct record *record,
288 char *merge_key, int *total)
291 struct reclist_bucket **p;
292 struct record_cluster *cluster = 0;
300 bucket = jenkins_hash((unsigned char*) merge_key) & l->hashmask;
302 for (p = &l->hashtable[bucket]; *p; p = &(*p)->next)
304 // We found a matching record. Merge them
305 if (!strcmp(merge_key, (*p)->record->merge_key))
307 struct record_cluster *existing = (*p)->record;
308 record->next = existing->records;
309 existing->records = record;
314 if (!cluster && l->num_records < l->flatlist_size)
316 struct reclist_bucket *new =
317 nmem_malloc(l->nmem, sizeof(struct reclist_bucket));
318 struct record_cluster *newc =
319 nmem_malloc(l->nmem, sizeof(struct record_cluster));
324 newc->records = record;
325 newc->merge_key = merge_key;
327 newc->term_frequency_vec = 0;
328 newc->recid = merge_key;
330 newc->metadata = nmem_malloc(l->nmem,
331 sizeof(struct record_metadata*) * service->num_metadata);
332 memset(newc->metadata, 0,
333 sizeof(struct record_metadata*) * service->num_metadata);
334 newc->sortkeys = nmem_malloc(l->nmem,
335 sizeof(struct record_metadata*) * service->num_sortkeys);
336 memset(newc->sortkeys, 0,
337 sizeof(union data_types*) * service->num_sortkeys);
340 l->flatlist[l->num_records++] = newc;
350 * c-file-style: "Stroustrup"
351 * indent-tabs-mode: nil
353 * vim: shiftwidth=4 tabstop=8 expandtab