1 /* $Id: rank1.c,v 1.19 2004-10-28 10:37:15 heikki Exp $
2 Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003
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 Zebra; see the file LICENSE.zebra. If not, write to the
19 Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
37 struct rank_class_info {
41 struct rank_term_info {
51 struct rank_set_info {
55 struct rank_term_info *entries;
59 static int log2_int (unsigned g)
68 * create: Creates/Initialises this rank handler. This routine is
69 * called exactly once. The routine returns the class_handle.
71 static void *create (ZebraHandle zh)
73 struct rank_class_info *ci =
74 (struct rank_class_info *) xmalloc (sizeof(*ci));
76 yaz_log (LOG_DEBUG, "rank-1 create");
81 * destroy: Destroys this rank handler. This routine is called
82 * when the handler is no longer needed - i.e. when the server
83 * dies. The class_handle was previously returned by create.
85 static void destroy (struct zebra_register *reg, void *class_handle)
87 struct rank_class_info *ci = (struct rank_class_info *) class_handle;
89 yaz_log (LOG_DEBUG, "rank-1 destroy");
95 * begin: Prepares beginning of "real" ranking. Called once for
96 * each result set. The returned handle is a "set handle" and
97 * will be used in each of the handlers below.
99 static void *begin (struct zebra_register *reg,
100 void *class_handle, RSET rset, NMEM nmem,
101 TERMID *terms, int numterms)
103 struct rank_set_info *si =
104 (struct rank_set_info *) nmem_malloc (nmem,sizeof(*si));
108 yaz_log (LOG_LOG, "rank-1 begin");
110 si->no_entries = numterms;
111 si->no_rank_entries = 0;
113 si->entries = (struct rank_term_info *)
114 nmem_malloc (si->nmem, sizeof(*si->entries)*numterms);
115 for (i = 0; i < numterms; i++)
117 zint g = rset_count(terms[i]->rset);
119 yaz_log(LOG_LOG, "i=%d flags=%s '%s'", i,
120 terms[i]->flags, terms[i]->name );
122 if (!strncmp (terms[i]->flags, "rank,", 5))
124 const char *cp = strstr(terms[i]->flags+4, ",w=");
125 si->entries[i].rank_flag = 1;
127 si->entries[i].rank_weight = atoi (cp+3);
129 si->entries[i].rank_weight = 34;
131 yaz_log (LOG_LOG, " i=%d weight=%d g="ZINT_FORMAT, i,
132 si->entries[i].rank_weight, g);
134 (si->no_rank_entries)++;
137 si->entries[i].rank_flag = 0;
138 si->entries[i].local_occur = 0; /* FIXME */
139 si->entries[i].global_occur = g;
140 si->entries[i].global_inv = 32 - log2_int (g);
141 yaz_log (LOG_DEBUG, " global_inv = %d g = " ZINT_FORMAT,
142 (int) (32-log2_int (g)), g);
143 si->entries[i].term=terms[i];
144 si->entries[i].term_index=i;
145 terms[i]->rankpriv=&(si->entries[i]);
151 * end: Terminates ranking process. Called after a result set
154 static void end (struct zebra_register *reg, void *set_handle)
156 yaz_log (LOG_DEBUG, "rank-1 end");
157 /* no need to free anything, they are in nmems */
162 * add: Called for each word occurence in a result set. This routine
163 * should be as fast as possible. This routine should "incrementally"
166 static void add (void *set_handle, int seqno, TERMID term)
168 struct rank_set_info *si = (struct rank_set_info *) set_handle;
169 struct rank_term_info *ti= (struct rank_term_info *) term->rankpriv;
174 yaz_log (LOG_LOG, "rank-1 add seqno=%d term=%s", seqno, term->name);
176 si->last_pos = seqno;
181 * calc: Called for each document in a result. This handler should
182 * produce a score based on previous call(s) to the add handler. The
183 * score should be between 0 and 1000. If score cannot be obtained
184 * -1 should be returned.
186 static int calc (void *set_handle, zint sysno)
188 int i, lo, divisor, score = 0;
189 struct rank_set_info *si = (struct rank_set_info *) set_handle;
191 if (!si->no_rank_entries)
195 yaz_log(LOG_LOG, "calc");
197 for (i = 0; i < si->no_entries; i++)
200 yaz_log(LOG_LOG, "calc: i=%d rank_flag=%d lo=%d",
201 i, si->entries[i].rank_flag, si->entries[i].local_occur);
203 if (si->entries[i].rank_flag && (lo = si->entries[i].local_occur))
204 score += (8+log2_int (lo)) * si->entries[i].global_inv *
205 si->entries[i].rank_weight;
207 divisor = si->no_rank_entries * (8+log2_int (si->last_pos/si->no_entries));
208 score = score / divisor;
210 yaz_log (LOG_LOG, "calc sysno=" ZINT_FORMAT " score=%d", sysno, score);
214 for (i = 0; i < si->no_entries; i++)
215 si->entries[i].local_occur = 0;
220 * Pseudo-meta code with sequence of calls as they occur in a
221 * server. Handlers are prefixed by --:
237 static struct rank_control rank_control = {
247 struct rank_control *rank1_class = &rank_control;