projects
/
idzebra-moved-to-github.git
/ blobdiff
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
shortlog
|
log
|
commit
|
commitdiff
|
tree
raw
|
inline
| side by side
Added code which maintains number of term occurrences and document
[idzebra-moved-to-github.git]
/
index
/
rank1.c
diff --git
a/index/rank1.c
b/index/rank1.c
index
581e78a
..
d9153b5
100644
(file)
--- a/
index/rank1.c
+++ b/
index/rank1.c
@@
-1,6
+1,6
@@
-/* $Id: rank1.c,v 1.18 2004-10-26 15:32:11 heikki Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003
- Index Data Aps
+/* $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.
This file is part of the Zebra server.
@@
-20,19
+20,20
@@
Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
02111-1307, USA.
*/
02111-1307, USA.
*/
-
-
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
#include <io.h>
#include <stdio.h>
#include <assert.h>
#ifdef WIN32
#include <io.h>
-#else
+#endif
+#if HAVE_UNISTD_H
#include <unistd.h>
#endif
#include <unistd.h>
#endif
-#define DEBUG_RANK 1
-
#include "index.h"
#include "index.h"
+#include "rank.h"
+
+static int log_level = 0;
+static int log_initialized = 0;
struct rank_class_info {
int dummy;
struct rank_class_info {
int dummy;
@@
-56,9
+57,11
@@
struct rank_set_info {
NMEM nmem;
};
NMEM nmem;
};
-static int log2_int (unsigned g)
+static int log2_int (zint g)
{
int n = 0;
{
int n = 0;
+ if (g < 0)
+ return 0;
while ((g = g>>1))
n++;
return n;
while ((g = g>>1))
n++;
return n;
@@
-73,7
+76,12
@@
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));
- yaz_log (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;
}
return ci;
}
@@
-86,7
+94,7
@@
static void destroy (struct zebra_register *reg, void *class_handle)
{
struct rank_class_info *ci = (struct rank_class_info *) class_handle;
{
struct rank_class_info *ci = (struct rank_class_info *) class_handle;
- yaz_log (LOG_DEBUG, "rank-1 destroy");
+ yaz_log(log_level, "rank-1 destroy");
xfree (ci);
}
xfree (ci);
}
@@
-104,9
+112,7
@@
static void *begin (struct zebra_register *reg,
(struct rank_set_info *) nmem_malloc (nmem,sizeof(*si));
int i;
(struct rank_set_info *) nmem_malloc (nmem,sizeof(*si));
int i;
-#if DEBUG_RANK
- yaz_log (LOG_LOG, "rank-1 begin");
-#endif
+ yaz_log(log_level, "rank-1 begin");
si->no_entries = numterms;
si->no_rank_entries = 0;
si->nmem=nmem;
si->no_entries = numterms;
si->no_rank_entries = 0;
si->nmem=nmem;
@@
-115,10
+121,8
@@
static void *begin (struct zebra_register *reg,
for (i = 0; i < numterms; i++)
{
zint g = rset_count(terms[i]->rset);
for (i = 0; i < numterms; i++)
{
zint g = rset_count(terms[i]->rset);
-#if DEBUG_RANK
- yaz_log(LOG_LOG, "i=%d flags=%s '%s'", i,
+ yaz_log(log_level, "i=%d flags=%s '%s'", i,
terms[i]->flags, terms[i]->name );
terms[i]->flags, terms[i]->name );
-#endif
if (!strncmp (terms[i]->flags, "rank,", 5))
{
const char *cp = strstr(terms[i]->flags+4, ",w=");
if (!strncmp (terms[i]->flags, "rank,", 5))
{
const char *cp = strstr(terms[i]->flags+4, ",w=");
@@
-126,11
+130,9
@@
static void *begin (struct zebra_register *reg,
if (cp)
si->entries[i].rank_weight = atoi (cp+3);
else
if (cp)
si->entries[i].rank_weight = atoi (cp+3);
else
- si->entries[i].rank_weight = 34;
-#if DEBUG_RANK
- yaz_log (LOG_LOG, " i=%d weight=%d g="ZINT_FORMAT, i,
+ 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->entries[i].rank_weight, g);
-#endif
(si->no_rank_entries)++;
}
else
(si->no_rank_entries)++;
}
else
@@
-138,11
+140,11
@@
static void *begin (struct zebra_register *reg,
si->entries[i].local_occur = 0; /* FIXME */
si->entries[i].global_occur = g;
si->entries[i].global_inv = 32 - log2_int (g);
si->entries[i].local_occur = 0; /* FIXME */
si->entries[i].global_occur = g;
si->entries[i].global_inv = 32 - log2_int (g);
- yaz_log (LOG_DEBUG, " global_inv = %d g = " ZINT_FORMAT,
+ yaz_log(log_level, " global_inv = %d g = " ZINT_FORMAT,
(int) (32-log2_int (g)), g);
(int) (32-log2_int (g)), g);
- si->entries[i].term=terms[i];
+ si->entries[i].term = terms[i];
si->entries[i].term_index=i;
si->entries[i].term_index=i;
- terms[i]->rankpriv=&(si->entries[i]);
+ terms[i]->rankpriv = &(si->entries[i]);
}
return si;
}
}
return si;
}
@@
-153,7
+155,7
@@
static void *begin (struct zebra_register *reg,
*/
static void end (struct zebra_register *reg, void *set_handle)
{
*/
static void end (struct zebra_register *reg, void *set_handle)
{
- yaz_log (LOG_DEBUG, "rank-1 end");
+ yaz_log(log_level, "rank-1 end");
/* no need to free anything, they are in nmems */
}
/* no need to free anything, they are in nmems */
}
@@
-166,15
+168,19
@@
static void end (struct zebra_register *reg, void *set_handle)
static void add (void *set_handle, int seqno, TERMID term)
{
struct rank_set_info *si = (struct rank_set_info *) set_handle;
static void add (void *set_handle, int seqno, TERMID term)
{
struct rank_set_info *si = (struct rank_set_info *) set_handle;
- struct rank_term_info *ti= (struct rank_term_info *) term->rankpriv;
+ struct rank_term_info *ti;
assert(si);
assert(si);
- assert(term);
+ if (!term)
+ {
+ yaz_log(log_level, "rank-1 add NULL term");
+ return;
+ }
+ ti= (struct rank_term_info *) term->rankpriv;
assert(ti);
assert(ti);
-#if DEBUG_RANK
- yaz_log (LOG_LOG, "rank-1 add seqno=%d term=%s", seqno, term->name);
-#endif
si->last_pos = seqno;
ti->local_occur++;
si->last_pos = seqno;
ti->local_occur++;
+ yaz_log(log_level, "rank-1 add seqno=%d term=%s count=%d",
+ seqno, term->name,ti->local_occur);
}
/*
}
/*
@@
-183,34
+189,29
@@
static void add (void *set_handle, int seqno, TERMID term)
* score should be between 0 and 1000. If score cannot be obtained
* -1 should be returned.
*/
* score should be between 0 and 1000. If score cannot be obtained
* -1 should be returned.
*/
-static int calc (void *set_handle, zint 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;
{
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 */
-#if DEBUG_RANK
- yaz_log(LOG_LOG, "calc");
-#endif
for (i = 0; i < si->no_entries; i++)
{
for (i = 0; i < si->no_entries; i++)
{
-#if DEBUG_RANK
- yaz_log(LOG_LOG, "calc: i=%d rank_flag=%d lo=%d",
+ yaz_log(log_level, "calc: i=%d rank_flag=%d lo=%d",
i, si->entries[i].rank_flag, si->entries[i].local_occur);
i, si->entries[i].rank_flag, si->entries[i].local_occur);
-#endif
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;
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;
-#if DEBUG_RANK
- yaz_log (LOG_LOG, "calc sysno=" ZINT_FORMAT " score=%d", sysno, score);
-#endif
+ yaz_log(log_level, "calc sysno=" ZINT_FORMAT " score=%d", sysno, score);
if (score > 1000)
score = 1000;
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;
for (i = 0; i < si->no_entries; i++)
si->entries[i].local_occur = 0;
return score;
@@
-244,4
+245,12
@@
static struct rank_control rank_control = {
add,
};
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
+ */
+