-/*
- * Copyright (C) 1998-1999, Index Data
- * All rights reserved.
- * Sebastian Hammer, Adam Dickmeiss
- *
- * $Log: rank1.c,v $
- * Revision 1.9 2002-04-11 11:39:59 heikki
- * Removed to logf calls from tight inside loops
- *
- * Revision 1.8 2002/04/04 14:14:13 adam
- * Multiple registers (alpha early)
- *
- * Revision 1.7 2001/11/14 22:06:27 adam
- * Rank-weight may be controlled via query.
- *
- * Revision 1.6 2000/03/15 15:00:30 adam
- * First work on threaded version.
- *
- * Revision 1.5 1999/05/26 07:49:13 adam
- * C++ compilation.
- *
- * Revision 1.4 1999/02/02 14:51:01 adam
- * Updated WIN32 code specific sections. Changed header.
- *
- * Revision 1.3 1998/06/12 12:21:53 adam
- * Fixed memory-leak.
- *
- * Revision 1.2 1998/03/05 13:03:29 adam
- * Improved ranking.
- *
- * Revision 1.1 1998/03/05 08:45:12 adam
- * New result set model and modular ranking system. Moved towards
- * descent server API. System information stored as "SGML" records.
- *
- */
+/* $Id: rank1.c,v 1.30 2006-05-10 08:13:22 adam Exp $
+ Copyright (C) 1995-2005
+ Index Data ApS
+
+This file is part of the Zebra server.
+
+Zebra is free software; you can redistribute it and/or modify it under
+the terms of the GNU General Public License as published by the Free
+Software Foundation; either version 2, or (at your option) any later
+version.
+
+Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
+WARRANTY; without even the implied warranty of MERCHANTABILITY or
+FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+for more details.
+
+You should have received a copy of the GNU General Public License
+along with Zebra; see the file LICENSE.zebra. If not, write to the
+Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
+02111-1307, USA.
+*/
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
#include <io.h>
-#else
+#endif
+#if HAVE_UNISTD_H
#include <unistd.h>
#endif
#include "index.h"
+#include "rank.h"
+
+static int log_level = 0;
+static int log_initialized = 0;
struct rank_class_info {
int dummy;
struct rank_term_info {
int local_occur;
- int global_occur;
+ zint global_occur;
int global_inv;
int rank_flag;
int rank_weight;
+ TERMID term;
+ int term_index;
};
struct rank_set_info {
int no_entries;
int no_rank_entries;
struct rank_term_info *entries;
+ NMEM nmem;
};
-static int log2_int (unsigned g)
+static int log2_int (zint g)
{
int n = 0;
+ if (g < 0)
+ return 0;
while ((g = g>>1))
n++;
return n;
* create: Creates/Initialises this rank handler. This routine is
* called exactly once. The routine returns the class_handle.
*/
-static void *create (struct zebra_register *reg)
+static void *create (ZebraHandle zh)
{
- struct rank_class_info *ci = (struct rank_class_info *)
- xmalloc (sizeof(*ci));
+ struct rank_class_info *ci =
+ (struct rank_class_info *) xmalloc (sizeof(*ci));
- logf (LOG_DEBUG, "rank-1 create");
+ if (!log_initialized)
+ {
+ log_level = yaz_log_module_level("rank1");
+ log_initialized = 1;
+ }
+ yaz_log(log_level, "rank-1 create");
return ci;
}
{
struct rank_class_info *ci = (struct rank_class_info *) class_handle;
- logf (LOG_DEBUG, "rank-1 destroy");
+ yaz_log(log_level, "rank-1 destroy");
xfree (ci);
}
-/*
+/**
* begin: Prepares beginning of "real" ranking. Called once for
* each result set. The returned handle is a "set handle" and
* will be used in each of the handlers below.
*/
-static void *begin (struct zebra_register *reg, void *class_handle, RSET rset)
+static void *begin (struct zebra_register *reg,
+ void *class_handle, RSET rset, NMEM nmem,
+ TERMID *terms, int numterms)
{
- struct rank_set_info *si = (struct rank_set_info *) xmalloc (sizeof(*si));
+ struct rank_set_info *si =
+ (struct rank_set_info *) nmem_malloc (nmem,sizeof(*si));
int i;
- logf (LOG_DEBUG, "rank-1 begin");
- si->no_entries = rset->no_rset_terms;
+ yaz_log(log_level, "rank-1 begin");
+ si->no_entries = numterms;
si->no_rank_entries = 0;
+ si->nmem=nmem;
si->entries = (struct rank_term_info *)
- xmalloc (sizeof(*si->entries)*si->no_entries);
- for (i = 0; i < si->no_entries; i++)
+ nmem_malloc (si->nmem, sizeof(*si->entries)*numterms);
+ for (i = 0; i < numterms; i++)
{
- int g = rset->rset_terms[i]->nn;
- if (!strncmp (rset->rset_terms[i]->flags, "rank,", 5))
+ zint g = rset_count(terms[i]->rset);
+ yaz_log(log_level, "i=%d flags=%s '%s'", i,
+ terms[i]->flags, terms[i]->name );
+ if (!strncmp (terms[i]->flags, "rank,", 5))
{
- yaz_log (LOG_LOG, "%s", rset->rset_terms[i]->flags);
+ const char *cp = strstr(terms[i]->flags+4, ",w=");
si->entries[i].rank_flag = 1;
- si->entries[i].rank_weight = atoi (rset->rset_terms[i]->flags+5);
- yaz_log (LOG_LOG, "i=%d weight=%d", i, si->entries[i].rank_weight);
+ if (cp)
+ si->entries[i].rank_weight = atoi (cp+3);
+ else
+ si->entries[i].rank_weight = 34; /* sqrroot of 1000 */
+ yaz_log(log_level, " i=%d weight=%d g="ZINT_FORMAT, i,
+ si->entries[i].rank_weight, g);
(si->no_rank_entries)++;
}
else
si->entries[i].rank_flag = 0;
- si->entries[i].local_occur = 0;
+ si->entries[i].local_occur = 0; /* FIXME */
si->entries[i].global_occur = g;
si->entries[i].global_inv = 32 - log2_int (g);
- logf (LOG_DEBUG, "-------- %d ------", 32 - log2_int (g));
+ yaz_log(log_level, " global_inv = %d g = " ZINT_FORMAT,
+ (int) (32-log2_int (g)), g);
+ si->entries[i].term = terms[i];
+ si->entries[i].term_index=i;
+ terms[i]->rankpriv = &(si->entries[i]);
}
return si;
}
*/
static void end (struct zebra_register *reg, void *set_handle)
{
- struct rank_set_info *si = (struct rank_set_info *) set_handle;
- logf (LOG_DEBUG, "rank-1 end");
- xfree (si->entries);
- xfree (si);
+ yaz_log(log_level, "rank-1 end");
+ /* no need to free anything, they are in nmems */
}
-/*
+
+/**
* add: Called for each word occurence in a result set. This routine
* should be as fast as possible. This routine should "incrementally"
* update the score.
*/
-static void add (void *set_handle, int seqno, int term_index)
+static void add (void *set_handle, int seqno, TERMID term)
{
struct rank_set_info *si = (struct rank_set_info *) set_handle;
- /*logf (LOG_DEBUG, "rank-1 add seqno=%d term_index=%d", seqno, term_index);*/
+ struct rank_term_info *ti;
+ assert(si);
+ if (!term)
+ {
+ yaz_log(log_level, "rank-1 add NULL term");
+ return;
+ }
+ ti= (struct rank_term_info *) term->rankpriv;
+ assert(ti);
si->last_pos = seqno;
- si->entries[term_index].local_occur++;
+ ti->local_occur++;
+ yaz_log(log_level, "rank-1 add seqno=%d term=%s count=%d",
+ seqno, term->name,ti->local_occur);
}
/*
* score should be between 0 and 1000. If score cannot be obtained
* -1 should be returned.
*/
-static int calc (void *set_handle, int sysno)
+static int calc (void *set_handle, zint sysno, zint staticrank,
+ int *stop_flag)
{
int i, lo, divisor, score = 0;
struct rank_set_info *si = (struct rank_set_info *) set_handle;
- if (!si->no_rank_entries)
- return -1;
+ if (!si->no_rank_entries)
+ return -1; /* ranking not enabled for any terms */
for (i = 0; i < si->no_entries; i++)
+ {
+ yaz_log(log_level, "calc: i=%d rank_flag=%d lo=%d",
+ i, si->entries[i].rank_flag, si->entries[i].local_occur);
if (si->entries[i].rank_flag && (lo = si->entries[i].local_occur))
score += (8+log2_int (lo)) * si->entries[i].global_inv *
si->entries[i].rank_weight;
+ }
divisor = si->no_rank_entries * (8+log2_int (si->last_pos/si->no_entries));
score = score / divisor;
- yaz_log (LOG_LOG, "sysno=%d score=%d", sysno, score);
+ yaz_log(log_level, "calc sysno=" ZINT_FORMAT " score=%d", sysno, score);
if (score > 1000)
score = 1000;
+ /* reset the counts for the next term */
for (i = 0; i < si->no_entries; i++)
si->entries[i].local_occur = 0;
return score;
add,
};
-struct rank_control *rank1_class = &rank_control;
+struct rank_control *rank_1_class = &rank_control;
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+