X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fretrieve.c;h=d60a98988f494726aa8cf95341190f3939395b1a;hb=593927cb1897c1e3163c284448eff7fee6ddad51;hp=cf0787c01614d05936c92a4973a255bd26620960;hpb=47ea1fc957c7b97bb30a26698f072109cae275e4;p=idzebra-moved-to-github.git diff --git a/index/retrieve.c b/index/retrieve.c index cf0787c..d60a989 100644 --- a/index/retrieve.c +++ b/index/retrieve.c @@ -1,69 +1,24 @@ -/* - * Copyright (C) 1995-1999, Index Data - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: retrieve.c,v $ - * Revision 1.19 2002-07-25 13:06:43 adam - * Character set negotiation updates - * - * Revision 1.18 2002/07/02 20:20:09 adam - * idzebra:{filename,score,size,localnumber} tags for XML - * - * Revision 1.17 2002/05/03 13:49:04 adam - * play with shellsort - * - * Revision 1.16 2002/04/04 20:50:37 adam - * Multi register works with record paths and data1 profile path - * - * Revision 1.15 2002/04/04 14:14:13 adam - * Multiple registers (alpha early) - * - * Revision 1.14 2001/01/22 11:41:41 adam - * Added support for raw retrieval (element set name "R"). - * - * 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 1.36 2005-10-28 07:25:30 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 #include @@ -72,7 +27,8 @@ #ifdef WIN32 #include #include -#else +#endif +#if HAVE_UNISTD_H #include #endif @@ -115,7 +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; @@ -127,25 +83,40 @@ 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; + *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->reg->zei, rec); @@ -157,18 +128,18 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, strcpy (*basenamep, basename); if (comp && comp->which == Z_RecordComp_simple && - comp->u.simple->which == Z_ElementSetNames_generic) + comp->u.simple->which == Z_ElementSetNames_generic && + !strcmp (comp->u.simple->u.generic, "R")) { - if (!strcmp (comp->u.simple->u.generic, "R")) - file_type = "text"; + raw_mode = 1; } - if (!(rt = recType_byName (zh->reg->recTypes, - file_type, subType, &clientData))) + 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; @@ -180,7 +151,16 @@ 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 { @@ -194,11 +174,10 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, } else strcpy (full_rep, fname); - if ((fc.fd = open (full_rep, O_BINARY|O_RDONLY)) == -1) { - logf (LOG_WARN|LOG_ERRNO, "Retrieve fail; missing file: %s", + yaz_log (YLOG_WARN|YLOG_ERRNO, "Retrieve fail; missing file: %s", full_rep); rec_rm (&rec); return 14; @@ -210,8 +189,17 @@ 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; @@ -220,12 +208,62 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, retrieveCtrl.comp = comp; retrieveCtrl.encoding = zh->record_encoding; retrieveCtrl.diagnostic = 0; + 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(); - (*rt->retrieve)(clientData, &retrieveCtrl); + if (1) + { + /* snippets code */ + zebra_snippets *snippet; + +#if NEW_REC_KEYS + zebra_rec_keys_t reckeys = zebra_rec_keys_open(); + + zebra_rec_keys_set_buf(reckeys, + rec->info[recInfo_delKeys], + rec->size[recInfo_delKeys], + 0); + zebra_snippets_rec_keys(zh, reckeys, retrieveCtrl.doc_snippet); + zebra_rec_keys_close(reckeys); +#else + struct recKeys reckeys; + reckeys.buf = rec->info[recInfo_delKeys]; + reckeys.buf_used = rec->size[recInfo_delKeys]; + zebra_snippets_rec_keys(zh, &reckeys, retrieveCtrl.doc_snippet); +#endif + + +#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; @@ -233,5 +271,6 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, close (fc.fd); rec_rm (&rec); + *addinfo = retrieveCtrl.addinfo; return retrieveCtrl.diagnostic; }