X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fretrieve.c;h=909c41f92b832f6fe76bfd444e0019ec4e448484;hb=2e4e9c6def27f1e1463dcb6f205fab6a98054f38;hp=43661f279eca6441560c3e502c746914502bc334;hpb=6e044d5a523bb4363d07af757ebf1bcfe0ad6c04;p=idzebra-moved-to-github.git diff --git a/index/retrieve.c b/index/retrieve.c index 43661f2..909c41f 100644 --- a/index/retrieve.c +++ b/index/retrieve.c @@ -1,26 +1,24 @@ -/* - * Copyright (C) 1995-1999, Index Data - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: retrieve.c,v $ - * 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 1.22 2004-08-04 08:35:23 adam Exp $ + Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003,2004 + 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 @@ -33,58 +31,42 @@ #include #endif -#include -#include "zserver.h" - -#ifndef ZEBRASDR -#define ZEBRASDR 0 -#endif +#include "index.h" +#include -#if ZEBRASDR -#include "zebrasdr.h" -#endif - -struct fetch_control { - int record_offset; - int record_int_pos; - char *record_int_buf; - int record_int_len; - int fd; -}; - -static int record_ext_read (void *fh, char *buf, size_t count) +int zebra_record_ext_read (void *fh, char *buf, size_t count) { - struct fetch_control *fc = fh; + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; return read (fc->fd, buf, count); } -static off_t record_ext_seek (void *fh, off_t offset) +off_t zebra_record_ext_seek (void *fh, off_t offset) { - struct fetch_control *fc = fh; + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; return lseek (fc->fd, offset + fc->record_offset, SEEK_SET); } -static off_t record_ext_tell (void *fh) +off_t zebra_record_ext_tell (void *fh) { - struct fetch_control *fc = fh; + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; return lseek (fc->fd, 0, SEEK_CUR) - fc->record_offset; } -static off_t record_int_seek (void *fh, off_t offset) +off_t zebra_record_int_seek (void *fh, off_t offset) { - struct fetch_control *fc = fh; + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; return (off_t) (fc->record_int_pos = offset); } -static off_t record_int_tell (void *fh) +off_t zebra_record_int_tell (void *fh) { - struct fetch_control *fc = fh; + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; return (off_t) fc->record_int_pos; } -static int record_int_read (void *fh, char *buf, size_t count) +int zebra_record_int_read (void *fh, char *buf, size_t count) { - struct fetch_control *fc = fh; + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; int l = fc->record_int_len - fc->record_int_pos; if (l <= 0) return 0; @@ -94,7 +76,13 @@ static int record_int_read (void *fh, char *buf, size_t count) return l; } -int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, +void zebra_record_int_end (void *fh, off_t off) +{ + struct zebra_fetch_control *fc = (struct zebra_fetch_control *) fh; + fc->offset_end = off; +} + +int zebra_record_fetch (ZebraHandle zh, SYSNO sysno, int score, ODR stream, oid_value input_format, Z_RecordComposition *comp, oid_value *output_format, char **rec_bufp, int *rec_lenp, char **basenamep) @@ -104,122 +92,98 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, RecType rt; struct recRetrieveCtrl retrieveCtrl; char subType[128]; - struct fetch_control fc; + struct zebra_fetch_control fc; RecordAttr *recordAttr; + void *clientData; + int raw_mode = 0; - rec = rec_get (zh->records, sysno); + rec = rec_get (zh->reg->records, sysno); if (!rec) { - logf (LOG_DEBUG, "rec_get fail on sysno=%d", sysno); + logf (LOG_DEBUG, "rec_get fail on sysno=" ZINT_FORMAT, sysno); + *basenamep = 0; return 14; } - recordAttr = rec_init_attr (zh->zei, rec); + recordAttr = rec_init_attr (zh->reg->zei, rec); file_type = rec->info[recInfo_fileType]; fname = rec->info[recInfo_filename]; basename = rec->info[recInfo_databaseName]; - *basenamep = odr_malloc (stream, strlen(basename)+1); + *basenamep = (char *) odr_malloc (stream, strlen(basename)+1); strcpy (*basenamep, basename); - if (!(rt = recType_byName (zh->recTypes, file_type, subType))) + if (comp && comp->which == Z_RecordComp_simple && + comp->u.simple->which == Z_ElementSetNames_generic) + { + if (!strcmp (comp->u.simple->u.generic, "R")) + raw_mode = 1; + } + if (!(rt = recType_byName (zh->reg->recTypes, + file_type, subType, &clientData))) { logf (LOG_WARN, "Retrieve: Cannot handle type %s", file_type); return 14; } - logf (LOG_DEBUG, "retrieve localno=%d score=%d", sysno, score); + logf (LOG_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 = record_int_read; - retrieveCtrl.seekf = record_int_seek; - retrieveCtrl.tellf = record_int_tell; + retrieveCtrl.readf = zebra_record_int_read; + retrieveCtrl.seekf = zebra_record_int_seek; + retrieveCtrl.tellf = zebra_record_int_tell; 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); - } -#if ZEBRASDR - else if (*fname == '%') - { - ZebraSdrHandle h; - int segment = 0, r; - char *cp, xname[128]; - unsigned char *buf; - - logf (LOG_DEBUG, "SDR"); - strcpy (xname, fname+1); - if ((cp = strrchr (xname, '.'))) - { - *cp++ = 0; - segment = atoi(cp); - } - h = zebraSdr_open (xname); - if (!h) + if (raw_mode) { - logf (LOG_WARN, "sdr open %s", xname); - return 0; - } - if (zebraSdr_segment (h, &segment) < 0) - { - logf (LOG_WARN, "zebraSdr_segment fail segment=%d", - segment); + *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 14; - } - r = zebraSdr_read (h, &buf); - if (r < 1) - { - logf (LOG_WARN, "zebraSdr_read fail segment=%d", - segment); - rec_rm (&rec); - return 14; - } - zebraSdr_close (h); - - fc.record_int_len = recordAttr->recordSize; - fc.record_int_buf = buf + recordAttr->recordOffset; - fc.record_int_pos = 0; - - logf (LOG_LOG, "segment = %d len=%d off=%d", - segment, - recordAttr->recordSize, - recordAttr->recordOffset); - if (fc.record_int_len > 180) - { - logf (LOG_LOG, "%0.70s", fc.record_int_buf); - logf (LOG_LOG, "%0.70s", fc.record_int_buf + - (fc.record_int_len - 70)); + return 0; } - else - logf (LOG_LOG, "%0.*s", - fc.record_int_len, fc.record_int_buf); - - /* the following two lines makes rec_rm delete buf */ - rec->size[recInfo_storeData] = r; - rec->info[recInfo_storeData] = buf; - - retrieveCtrl.readf = record_int_read; - retrieveCtrl.seekf = record_int_seek; - retrieveCtrl.tellf = record_int_tell; } -#endif else { - if ((fc.fd = open (fname, O_BINARY|O_RDONLY)) == -1) + char full_rep[1024]; + + if (zh->path_reg && !yaz_is_abspath (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) { logf (LOG_WARN|LOG_ERRNO, "Retrieve fail; missing file: %s", - fname); + full_rep); rec_rm (&rec); return 14; } fc.record_offset = recordAttr->recordOffset; - retrieveCtrl.readf = record_ext_read; - retrieveCtrl.seekf = record_ext_seek; - retrieveCtrl.tellf = record_ext_tell; + retrieveCtrl.readf = zebra_record_ext_read; + retrieveCtrl.seekf = zebra_record_ext_seek; + retrieveCtrl.tellf = zebra_record_ext_tell; - record_ext_seek (retrieveCtrl.fh, 0); + 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; @@ -228,13 +192,17 @@ int zebra_record_fetch (ZebraHandle zh, int sysno, int score, ODR stream, 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->dh; - (*rt->retrieve)(&retrieveCtrl); + retrieveCtrl.dh = zh->reg->dh; + retrieveCtrl.res = zh->res; + retrieveCtrl.rec_buf = 0; + retrieveCtrl.rec_len = -1; + + (*rt->retrieve)(clientData, &retrieveCtrl); *output_format = retrieveCtrl.output_format; + *rec_bufp = (char *) retrieveCtrl.rec_buf; *rec_lenp = retrieveCtrl.rec_len; - *rec_bufp = odr_malloc (stream, *rec_lenp); - memcpy (*rec_bufp, retrieveCtrl.rec_buf, *rec_lenp); if (fc.fd != -1) close (fc.fd); rec_rm (&rec);