X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Frectext.c;h=3b14203a8910dbe5774b8f14b0d4f6d67e73b7e8;hb=b09c1d0aa414cb43fb667be54a29b3cc7279ea85;hp=7c7c9c0338a4cc3caf33e588371497f76d420d7f;hpb=3d01abbe1238c82d62e4c01eb7cac760d3d53377;p=idzebra-moved-to-github.git diff --git a/recctrl/rectext.c b/recctrl/rectext.c index 7c7c9c0..3b14203 100644 --- a/recctrl/rectext.c +++ b/recctrl/rectext.c @@ -1,61 +1,54 @@ -/* - * Copyright (C) 1994-1995, Index Data I/S - * All rights reserved. - * Sebastian Hammer, Adam Dickmeiss - * - * $Log: rectext.c,v $ - * Revision 1.5 1997-10-27 14:33:06 adam - * Moved towards generic character mapping depending on "structure" - * field in abstract syntax file. Fixed a few memory leaks. Fixed - * bug with negative integers when doing searches with relational - * operators. - * - * Revision 1.4 1996/11/04 14:09:16 adam - * Minor changes. - * - * Revision 1.3 1996/11/01 09:00:33 adam - * This simple "text" format now supports element specs B and M. - * - * Revision 1.2 1996/10/29 14:02:45 adam - * Uses buffered read to speed up things. - * - * Revision 1.1 1996/10/11 10:57:28 adam - * New module recctrl. Used to manage records (extract/retrieval). - * - * Revision 1.7 1996/01/17 14:57:55 adam - * Prototype changed for reader functions in extract/retrieve. File - * is identified by 'void *' instead of 'int. - * - * Revision 1.6 1995/10/10 13:59:24 adam - * Function rset_open changed its wflag parameter to general flags. - * - * Revision 1.5 1995/10/02 16:24:39 adam - * Use attribute actually used in search requests. - * - * Revision 1.4 1995/10/02 15:42:55 adam - * Extract uses file descriptors instead of FILE pointers. - * - * Revision 1.3 1995/09/28 09:19:45 adam - * xfree/xmalloc used everywhere. - * Extract/retrieve method seems to work for text records. - * - * Revision 1.2 1995/09/15 14:45:21 adam - * Retrieve control. - * Work on truncation. - * - * Revision 1.1 1995/09/14 07:48:25 adam - * Record control management. - * - */ +/* $Id: rectext.c,v 1.24 2005-03-05 09:19:16 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 #include #include -#include "rectext.h" +#include + +struct text_info { + char *sep; +}; + +static void *text_init (Res res, RecType recType) +{ + struct text_info *tinfo = (struct text_info *) xmalloc(sizeof(*tinfo)); + tinfo->sep = 0; + return tinfo; +} + +static void text_config(void *clientData, Res res, const char *args) +{ + +} -static void text_init (void) +static void text_destroy (void *clientData) { + struct text_info *tinfo = clientData; + xfree (tinfo->sep); + xfree (tinfo); } struct buf_info { @@ -67,16 +60,16 @@ struct buf_info { struct buf_info *buf_open (struct recExtractCtrl *p) { - struct buf_info *fi = xmalloc (sizeof(*fi)); + struct buf_info *fi = (struct buf_info *) xmalloc (sizeof(*fi)); fi->p = p; - fi->buf = xmalloc (4096); + fi->buf = (char *) xmalloc (4096); fi->offset = 1; fi->max = 1; return fi; } -int buf_read (struct buf_info *fi, char *dst) +int buf_read (struct text_info *tinfo, struct buf_info *fi, char *dst) { if (fi->offset >= fi->max) { @@ -88,6 +81,12 @@ int buf_read (struct buf_info *fi, char *dst) return 0; } *dst = fi->buf[(fi->offset)++]; + if (tinfo->sep && *dst == *tinfo->sep) + { + off_t off = (*fi->p->tellf)(fi->p->fh); + (*fi->p->endf)(fi->p->fh, off - (fi->max - fi->offset)); + return 0; + } return 1; } @@ -97,46 +96,50 @@ void buf_close (struct buf_info *fi) xfree (fi); } -static int text_extract (struct recExtractCtrl *p) +static int text_extract (void *clientData, struct recExtractCtrl *p) { - char w[256]; + struct text_info *tinfo = clientData; + char w[512]; RecWord recWord; - int r, seqno = 1; + int r; struct buf_info *fi = buf_open (p); - (*p->init)(&recWord); +#if 0 + yaz_log(YLOG_LOG, "text_extract off=%ld", + (long) (*fi->p->tellf)(fi->p->fh)); +#endif + xfree(tinfo->sep); + tinfo->sep = 0; + (*p->init)(p, &recWord); recWord.reg_type = 'w'; do { int i = 0; - r = buf_read (fi, w); - while (r > 0 && i < 255 && isalnum(w[i])) + r = buf_read (tinfo, fi, w); + while (r > 0 && i < 511 && w[i] != '\n' && w[i] != '\r') { i++; - r = buf_read (fi, w + i); - } + r = buf_read (tinfo, fi, w + i); + } if (i) { - int j; - for (j = 0; jadd)(&recWord); + recWord.term_buf = w; + recWord.term_len = i; + (*p->tokenAdd)(&recWord); } } while (r > 0); buf_close (fi); - return 0; + return RECCTRL_EXTRACT_OK; } -static int text_retrieve (struct recRetrieveCtrl *p) +static int text_retrieve (void *clientData, struct recRetrieveCtrl *p) { int r, text_ptr = 0; static char *text_buf = NULL; static int text_size = 0; - int start_flag = 1; + int make_header = 1; + int make_body = 1; const char *elementSetName = NULL; int no_lines = 0; @@ -144,6 +147,21 @@ static int text_retrieve (struct recRetrieveCtrl *p) p->comp->u.simple->which == Z_ElementSetNames_generic) elementSetName = p->comp->u.simple->u.generic; + if (elementSetName) + { + /* don't make header for the R(aw) element set name */ + if (!strcmp(elementSetName, "R")) + { + make_header = 0; + make_body = 1; + } + /* only make header for the H(eader) element set name */ + else if (!strcmp(elementSetName, "H")) + { + make_header = 1; + make_body = 0; + } + } while (1) { if (text_ptr + 4096 >= text_size) @@ -151,7 +169,7 @@ static int text_retrieve (struct recRetrieveCtrl *p) char *nb; text_size = 2*text_size + 8192; - nb = xmalloc (text_size); + nb = (char *) xmalloc (text_size); if (text_buf) { memcpy (nb, text_buf, text_ptr); @@ -159,17 +177,25 @@ static int text_retrieve (struct recRetrieveCtrl *p) } text_buf = nb; } - if (start_flag) + if (make_header && text_ptr == 0) { - start_flag = 0; if (p->score >= 0) { sprintf (text_buf, "Rank: %d\n", p->score); text_ptr = strlen(text_buf); } - sprintf (text_buf + text_ptr, "Local Number: %d\n", p->localno); + sprintf (text_buf + text_ptr, "Local Number: " ZINT_FORMAT "\n", + p->localno); text_ptr = strlen(text_buf); + if (p->fname) + { + sprintf (text_buf + text_ptr, "Filename: %s\n", p->fname); + text_ptr = strlen(text_buf); + } + strcpy(text_buf+text_ptr++, "\n"); } + if (!make_body) + break; r = (*p->readf)(p->fh, text_buf + text_ptr, 4096); if (r <= 0) break; @@ -205,8 +231,20 @@ static int text_retrieve (struct recRetrieveCtrl *p) static struct recType text_type = { "text", text_init, + text_config, + text_destroy, text_extract, text_retrieve }; -RecType recTypeText = &text_type; +RecType +#ifdef IDZEBRA_STATIC_TEXT +idzebra_filter_text +#else +idzebra_filter +#endif + +[] = { + &text_type, + 0, +};