X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fretrieve.c;h=fcedc3b235eb062402378631fe44f04760cd238c;hb=73dc84a6886c4d2c3f8c82b888a0728e34a5b51b;hp=2eb556ab05bd8b5858853b95934ec71750cde87a;hpb=c41c84a497ae744aa825a90f144c85b54f1cd4bb;p=idzebra-moved-to-github.git diff --git a/index/retrieve.c b/index/retrieve.c index 2eb556a..fcedc3b 100644 --- a/index/retrieve.c +++ b/index/retrieve.c @@ -1,51 +1,24 @@ -/* - * Copyright (C) 1995-1999, Index Data - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: retrieve.c,v $ - * Revision 1.13 2000-03-20 19:08:36 adam - * Added remote record import using Z39.50 extended services and Segment - * Requests. - * - * Revision 1.12 2000/03/15 15:00:30 adam - * First work on threaded version. - * - * Revision 1.11 1999/10/29 10:00:00 adam - * Fixed minor bug where database name wasn't set in zebra_record_fetch. - * - * Revision 1.10 1999/05/26 07:49:13 adam - * C++ compilation. - * - * Revision 1.9 1999/05/20 12:57:18 adam - * Implemented TCL filter. Updated recctrl system. - * - * Revision 1.8 1999/03/09 16:27:49 adam - * More work on SDRKit integration. - * - * Revision 1.7 1999/03/02 16:15:43 quinn - * Added "tagsysno" and "tagrank" directives to zebra.cfg. - * - * Revision 1.6 1999/02/18 15:01:25 adam - * Minor changes. - * - * Revision 1.5 1999/02/17 11:29:56 adam - * Fixed in record_fetch. Minor updates to API. - * - * Revision 1.4 1999/02/02 14:51:07 adam - * Updated WIN32 code specific sections. Changed header. - * - * Revision 1.3 1998/10/28 10:54:40 adam - * SDRKit integration. - * - * Revision 1.2 1998/10/16 08:14:33 adam - * Updated record control system. - * - * Revision 1.1 1998/03/05 08:45:13 adam - * New result set model and modular ranking system. Moved towards - * descent server API. System information stored as "SGML" records. - * - */ +/* $Id: retrieve.c,v 1.34 2005-08-19 14:05:29 marc 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 #include @@ -54,20 +27,13 @@ #ifdef WIN32 #include #include -#else -#include #endif - -#include -#include "zserver.h" - -#ifndef ZEBRASDR -#define ZEBRASDR 0 +#if HAVE_UNISTD_H +#include #endif -#if ZEBRASDR -#include "zebrasdr.h" -#endif +#include "index.h" +#include int zebra_record_ext_read (void *fh, char *buf, size_t count) { @@ -105,7 +71,7 @@ int zebra_record_int_read (void *fh, char *buf, size_t count) int l = fc->record_int_len - fc->record_int_pos; if (l <= 0) return 0; - l = (l < (int) count) ? l : count; + l = (l < (int) count) ? l : (int) count; memcpy (buf, fc->record_int_buf + fc->record_int_pos, l); fc->record_int_pos += l; return l; @@ -117,28 +83,43 @@ void zebra_record_int_end (void *fh, off_t off) fc->offset_end = off; } -int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, +int zebra_record_fetch (ZebraHandle zh, SYSNO sysno, int score, + zebra_snippets *hit_snippet, ODR stream, oid_value input_format, Z_RecordComposition *comp, oid_value *output_format, char **rec_bufp, - int *rec_lenp, char **basenamep) + int *rec_lenp, char **basenamep, + char **addinfo) { Record rec; char *fname, *file_type, *basename; RecType rt; struct recRetrieveCtrl retrieveCtrl; - char subType[128]; struct zebra_fetch_control fc; RecordAttr *recordAttr; void *clientData; + int raw_mode = 0; - rec = rec_get (zh->service->records, sysno); + *basenamep = 0; + *addinfo = 0; + if (comp && comp->which == Z_RecordComp_simple && + comp->u.simple->which == Z_ElementSetNames_generic && + !strcmp (comp->u.simple->u.generic, "_sysno_")) + { + char rec_str[60]; + sprintf(rec_str, ZINT_FORMAT, sysno); + *output_format = VAL_SUTRS; + *rec_lenp = strlen(rec_str); + *rec_bufp = odr_strdup(stream, rec_str); + return 0; + } + rec = rec_get (zh->reg->records, sysno); if (!rec) { - logf (LOG_DEBUG, "rec_get fail on sysno=%d", sysno); - *basenamep = 0; + yaz_log (YLOG_DEBUG, "rec_get fail on sysno=" ZINT_FORMAT, sysno); + *basenamep = 0; return 14; } - recordAttr = rec_init_attr (zh->service->zei, rec); + recordAttr = rec_init_attr (zh->reg->zei, rec); file_type = rec->info[recInfo_fileType]; fname = rec->info[recInfo_filename]; @@ -146,15 +127,22 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, *basenamep = (char *) odr_malloc (stream, strlen(basename)+1); strcpy (*basenamep, basename); - if (!(rt = recType_byName (zh->service->recTypes, - file_type, subType, &clientData))) + if (comp && comp->which == Z_RecordComp_simple && + comp->u.simple->which == Z_ElementSetNames_generic && + !strcmp (comp->u.simple->u.generic, "R")) + { + raw_mode = 1; + } + if (!(rt = recType_byName (zh->reg->recTypes, zh->res, + file_type, &clientData))) { - logf (LOG_WARN, "Retrieve: Cannot handle type %s", file_type); + yaz_log (YLOG_WARN, "Retrieve: Cannot handle type %s", file_type); return 14; } - logf (LOG_DEBUG, "retrieve localno=%d score=%d", sysno, score); + yaz_log (YLOG_DEBUG, "retrieve localno=" ZINT_FORMAT " score=%d", sysno,score); retrieveCtrl.fh = &fc; fc.fd = -1; + retrieveCtrl.fname = fname; if (rec->size[recInfo_storeData] > 0) { retrieveCtrl.readf = zebra_record_int_read; @@ -163,14 +151,34 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, fc.record_int_len = rec->size[recInfo_storeData]; fc.record_int_buf = rec->info[recInfo_storeData]; fc.record_int_pos = 0; - logf (LOG_DEBUG, "Internal retrieve. %d bytes", fc.record_int_len); + yaz_log (YLOG_DEBUG, "Internal retrieve. %d bytes", fc.record_int_len); + if (raw_mode) + { + *output_format = VAL_SUTRS; + *rec_lenp = rec->size[recInfo_storeData]; + *rec_bufp = (char *) odr_malloc(stream, *rec_lenp); + memcpy(*rec_bufp, rec->info[recInfo_storeData], *rec_lenp); + rec_rm (&rec); + return 0; + } } else { - if ((fc.fd = open (fname, O_BINARY|O_RDONLY)) == -1) + char full_rep[1024]; + + if (zh->path_reg && !yaz_is_abspath (fname)) { - logf (LOG_WARN|LOG_ERRNO, "Retrieve fail; missing file: %s", - fname); + strcpy (full_rep, zh->path_reg); + strcat (full_rep, "/"); + strcat (full_rep, fname); + } + else + strcpy (full_rep, fname); + + if ((fc.fd = open (full_rep, O_BINARY|O_RDONLY)) == -1) + { + yaz_log (YLOG_WARN|YLOG_ERRNO, "Retrieve fail; missing file: %s", + full_rep); rec_rm (&rec); return 14; } @@ -181,18 +189,69 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, retrieveCtrl.tellf = zebra_record_ext_tell; zebra_record_ext_seek (retrieveCtrl.fh, 0); + if (raw_mode) + { + *output_format = VAL_SUTRS; + *rec_lenp = recordAttr->recordSize; + *rec_bufp = (char *) odr_malloc(stream, *rec_lenp); + zebra_record_ext_read(&fc, *rec_bufp, *rec_lenp); + rec_rm (&rec); + close (fc.fd); + return 0; + } } - retrieveCtrl.subType = subType; retrieveCtrl.localno = sysno; retrieveCtrl.score = score; retrieveCtrl.recordSize = recordAttr->recordSize; retrieveCtrl.odr = stream; retrieveCtrl.input_format = retrieveCtrl.output_format = input_format; retrieveCtrl.comp = comp; + retrieveCtrl.encoding = zh->record_encoding; retrieveCtrl.diagnostic = 0; - retrieveCtrl.dh = zh->service->dh; - retrieveCtrl.res = zh->service->res; - (*rt->retrieve)(clientData, &retrieveCtrl); + retrieveCtrl.addinfo = 0; + retrieveCtrl.dh = zh->reg->dh; + retrieveCtrl.res = zh->res; + retrieveCtrl.rec_buf = 0; + retrieveCtrl.rec_len = -1; + retrieveCtrl.hit_snippet = hit_snippet; + retrieveCtrl.doc_snippet = zebra_snippets_create(); + + if (0) + { + /* snippets code */ + struct recKeys reckeys; + zebra_snippets *snippet; + reckeys.buf = rec->info[recInfo_delKeys]; + reckeys.buf_used = rec->size[recInfo_delKeys]; + + zebra_snippets_rec_keys(zh, &reckeys, retrieveCtrl.doc_snippet); + +#if 0 + /* for debugging purposes */ + yaz_log(YLOG_LOG, "DOC SNIPPET:"); + zebra_snippets_log(retrieveCtrl.doc_snippet, YLOG_LOG); + yaz_log(YLOG_LOG, "HIT SNIPPET:"); + zebra_snippets_log(retrieveCtrl.hit_snippet, YLOG_LOG); +#endif + snippet = zebra_snippets_window(retrieveCtrl.doc_snippet, + retrieveCtrl.hit_snippet, + 10); +#if 0 + /* for debugging purposes */ + yaz_log(YLOG_LOG, "WINDOW SNIPPET:"); + zebra_snippets_log(snippet, YLOG_LOG); +#endif + (*rt->retrieve)(clientData, &retrieveCtrl); + + zebra_snippets_destroy(snippet); + } + else + { + (*rt->retrieve)(clientData, &retrieveCtrl); + } + + zebra_snippets_destroy(retrieveCtrl.doc_snippet); + *output_format = retrieveCtrl.output_format; *rec_bufp = (char *) retrieveCtrl.rec_buf; *rec_lenp = retrieveCtrl.rec_len; @@ -200,5 +259,6 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, close (fc.fd); rec_rm (&rec); + *addinfo = retrieveCtrl.addinfo; return retrieveCtrl.diagnostic; }