X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Frank1.c;h=f233fe34c935b90d6583263bebfabd007bd43751;hb=eb8857cfa2ceaa2635f14888b95d6f71b8def045;hp=3369c7925ab997305667b946141dee458823798a;hpb=ce3907338568fce46c5751e7e1091a5ad1c8e291;p=idzebra-moved-to-github.git diff --git a/index/rank1.c b/index/rank1.c index 3369c79..f233fe3 100644 --- a/index/rank1.c +++ b/index/rank1.c @@ -1,24 +1,38 @@ -/* - * Copyright (C) 1998, Index Data I/S - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: rank1.c,v $ - * 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 1.12 2003-01-13 22:45:22 adam Exp $ + Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002 + 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 #include -#ifdef WINDOWS +#ifdef WIN32 #include #else #include #endif -#include "zserver.h" +#define DEBUG_RANK 0 + +#include "index.h" struct rank_class_info { int dummy; @@ -29,6 +43,7 @@ struct rank_term_info { int global_occur; int global_inv; int rank_flag; + int rank_weight; }; struct rank_set_info { @@ -50,11 +65,12 @@ static int log2_int (unsigned g) * create: Creates/Initialises this rank handler. This routine is * called exactly once. The routine returns the class_handle. */ -static void *create (ZebraHandle zh) +static void *create (struct zebra_register *reg) { - struct rank_class_info *ci = xmalloc (sizeof(*ci)); + struct rank_class_info *ci = (struct rank_class_info *) + xmalloc (sizeof(*ci)); - logf (LOG_DEBUG, "rank-1 create"); + yaz_log (LOG_DEBUG, "rank-1 create"); return ci; } @@ -63,11 +79,11 @@ static void *create (ZebraHandle zh) * when the handler is no longer needed - i.e. when the server * dies. The class_handle was previously returned by create. */ -static void destroy (ZebraHandle zh, void *class_handle) +static void destroy (struct zebra_register *reg, void *class_handle) { - struct rank_class_info *ci = class_handle; + struct rank_class_info *ci = (struct rank_class_info *) class_handle; - logf (LOG_DEBUG, "rank-1 destroy"); + yaz_log (LOG_DEBUG, "rank-1 destroy"); xfree (ci); } @@ -77,21 +93,31 @@ static void destroy (ZebraHandle zh, void *class_handle) * each result set. The returned handle is a "set handle" and * will be used in each of the handlers below. */ -static void *begin (ZebraHandle zh, void *class_handle, RSET rset) +static void *begin (struct zebra_register *reg, void *class_handle, RSET rset) { - struct rank_set_info *si = xmalloc (sizeof(*si)); + struct rank_set_info *si = (struct rank_set_info *) xmalloc (sizeof(*si)); int i; - logf (LOG_DEBUG, "rank-1 begin"); +#if DEBUG_RANK + yaz_log (LOG_LOG, "rank-1 begin"); +#endif si->no_entries = rset->no_rset_terms; si->no_rank_entries = 0; - si->entries = xmalloc (sizeof(*si->entries)*si->no_entries); + si->entries = (struct rank_term_info *) + xmalloc (sizeof(*si->entries)*si->no_entries); for (i = 0; i < si->no_entries; i++) { int g = rset->rset_terms[i]->nn; - if (!strcmp (rset->rset_terms[i]->flags, "rank")) +#if DEBUG_RANK + yaz_log(LOG_LOG, "i=%d flags=%s", i, rset->rset_terms[i]->flags); +#endif + if (!strncmp (rset->rset_terms[i]->flags, "rank,", 5)) { si->entries[i].rank_flag = 1; + si->entries[i].rank_weight = atoi (rset->rset_terms[i]->flags+5); +#if DEBUG_RANK + yaz_log (LOG_LOG, " weight=%d", i, si->entries[i].rank_weight); +#endif (si->no_rank_entries)++; } else @@ -99,7 +125,7 @@ static void *begin (ZebraHandle zh, void *class_handle, RSET rset) si->entries[i].local_occur = 0; 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_DEBUG, "-------- %d ------", 32 - log2_int (g)); } return si; } @@ -108,10 +134,11 @@ static void *begin (ZebraHandle zh, void *class_handle, RSET rset) * end: Terminates ranking process. Called after a result set * has been ranked. */ -static void end (ZebraHandle zh, void *set_handle) +static void end (struct zebra_register *reg, void *set_handle) { - struct rank_set_info *si = set_handle; - logf (LOG_DEBUG, "rank-1 end"); + struct rank_set_info *si = (struct rank_set_info *) set_handle; + yaz_log (LOG_DEBUG, "rank-1 end"); + xfree (si->entries); xfree (si); } @@ -122,8 +149,10 @@ static void end (ZebraHandle zh, void *set_handle) */ static void add (void *set_handle, int seqno, int term_index) { - struct rank_set_info *si = set_handle; - logf (LOG_DEBUG, "rank-1 add seqno=%d term_index=%d", seqno, term_index); + struct rank_set_info *si = (struct rank_set_info *) set_handle; +#if DEBUG_RANK + yaz_log (LOG_LOG, "rank-1 add seqno=%d term_index=%d", seqno, term_index); +#endif si->last_pos = seqno; si->entries[term_index].local_occur++; } @@ -136,19 +165,32 @@ static void add (void *set_handle, int seqno, int term_index) */ static int calc (void *set_handle, int sysno) { - int i, lu, score = 0; - struct rank_set_info *si = set_handle; - - logf (LOG_DEBUG, "rank-1 calc sysno=%d", sysno); + 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 DEBUG_RANK + yaz_log(LOG_LOG, "calc"); +#endif for (i = 0; i < si->no_entries; i++) - if (si->entries[i].rank_flag && (lu = si->entries[i].local_occur)) - score += (2+log2_int (lu)) * si->entries[i].global_inv; - logf (LOG_DEBUG, " dividend=%d", 60*score); - logf (LOG_DEBUG, " divisor=%d", si->no_rank_entries * log2_int (4+si->last_pos)); - score = (60 * score)/(si->no_rank_entries * log2_int (4+si->last_pos)); + { +#if DEBUG_RANK + yaz_log(LOG_LOG, "i=%d rank_flag=%d lo=%d", + 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 DEBUG_RANK + yaz_log (LOG_LOG, "sysno=%d score=%d", sysno, score); +#endif + if (score > 1000) + score = 1000; for (i = 0; i < si->no_entries; i++) si->entries[i].local_occur = 0; return score;