X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Fmarcread.c;h=0a125650d55d4fc101b6b738596bef3e45ba6024;hb=131e8143a9b8da294d582f0793833679101a2672;hp=c8a12e2fffc77a067827048454be528e1005406f;hpb=4478d785b7769691261005c98063b98a5a5971b3;p=idzebra-moved-to-github.git

diff --git a/index/marcread.c b/index/marcread.c
index c8a12e2..0a12565 100644
--- a/index/marcread.c
+++ b/index/marcread.c
@@ -1,5 +1,5 @@
-/* $Id: marcread.c,v 1.2 2006-08-14 10:40:15 adam Exp $
-   Copyright (C) 1995-2006
+/* $Id: marcread.c,v 1.9 2007-03-19 21:57:25 adam Exp $
+   Copyright (C) 1995-2007
    Index Data ApS
 
 This file is part of the Zebra server.
@@ -58,7 +58,7 @@ static data1_node *grs_read_iso2709 (struct grs_read_info *p, int marc_xml)
     char *absynName;
     data1_marctab *marctab;
 
-    if ((*p->readf)(p->fh, buf, 5) != 5)
+    if (p->stream->readf(p->stream, buf, 5) != 5)
         return NULL;
     while (*buf < '0' || *buf > '9')
     {
@@ -69,7 +69,7 @@ static data1_node *grs_read_iso2709 (struct grs_read_info *p, int marc_xml)
 	for (i = 0; i<4; i++)
 	    buf[i] = buf[i+1];
 
-	if ((*p->readf)(p->fh, buf+4, 1) != 1)
+	if (p->stream->readf(p->stream, buf+4, 1) != 1)
 	    return NULL;
     }
     record_length = atoi_n (buf, 5);
@@ -78,21 +78,35 @@ static data1_node *grs_read_iso2709 (struct grs_read_info *p, int marc_xml)
         yaz_log (YLOG_WARN, "MARC record length < 25, is %d", record_length);
         return NULL;
     }
-    /* read remaining part - attempt to read one byte furhter... */
-    read_bytes = (*p->readf)(p->fh, buf+5, record_length-4);
+
+    read_bytes = p->stream->readf(p->stream, buf+5, record_length-5);
     if (read_bytes < record_length-5)
     {
         yaz_log (YLOG_WARN, "Couldn't read whole MARC record");
         return NULL;
     }
-    if (read_bytes == record_length - 4)
+    /* skip until we meet a record separator */
+    while (buf[record_length-1] != ISO2709_RS)
     {
-        off_t cur_offset = (*p->tellf)(p->fh);
-	if (cur_offset <= 27)
-	    return NULL;
-	if (p->endf)
-	    (*p->endf)(p->fh, cur_offset - 1);
+        if (record_length > sizeof(buf)-2)
+            break;
+        read_bytes = p->stream->readf(p->stream, buf+record_length, 1);
+        if (read_bytes != 1)
+            break;
+        record_length++;
     }
+    /* read one byte ahead to see if there is more ... */
+    read_bytes = p->stream->readf(p->stream, buf+record_length, 1);
+    if (read_bytes == 1)
+    {
+        off_t cur_offset = p->stream->tellf(p->stream);
+	if (p->stream->endf)
+        {
+            off_t end_offset = cur_offset - 1;
+	    p->stream->endf(p->stream, &end_offset);
+        }
+    }
+
     absynName = mi->type;
     res_root = data1_mk_root (p->dh, p->mem, absynName);
     if (!res_root)
@@ -113,7 +127,7 @@ static data1_node *grs_read_iso2709 (struct grs_read_info *p, int marc_xml)
     else
 	res_top = data1_mk_tag (p->dh, p->mem, absynName, 0, res_root);
 
-    if ((marctab = data1_absyn_getmarctab(p->dh, res_root->u.root.absyn)))
+    if ((marctab = data1_absyn_getmarctab(p->dh, res_root)))
     {
 	memcpy(marctab->leader, buf, 24);
         memcpy(marctab->implementation_codes, buf+6, 4);
@@ -703,7 +717,7 @@ static int is_empty(char *s)
 static void parse_data1_tree(struct grs_read_info *p, const char *mc_stmnt,
 			     data1_node *root)
 {
-    data1_marctab *marctab = data1_absyn_getmarctab(p->dh, root->u.root.absyn);
+    data1_marctab *marctab = data1_absyn_getmarctab(p->dh, root);
     data1_node *top = root->child;
     data1_node *field;
     mc_context *c;
@@ -733,9 +747,12 @@ static void parse_data1_tree(struct grs_read_info *p, const char *mc_stmnt,
 	yaz_log(YLOG_LOG,"parse_data1_tree(): try LEADER from {%d} to {%d} positions",
 	    pf->interval.start, pf->interval.end);
 #endif	
-	new = data1_mk_tag_n(p->dh, p->mem, mc_stmnt, strlen(mc_stmnt), 0, top);
-	data1_mk_text_n(p->dh, p->mem, marctab->leader+pf->interval.start,
-	    pf->interval.end-pf->interval.start+1, new);
+        if (marctab)
+        {
+            new = data1_mk_tag_n(p->dh, p->mem, mc_stmnt, strlen(mc_stmnt), 0, top);
+            data1_mk_text_n(p->dh, p->mem, marctab->leader+pf->interval.start,
+                            pf->interval.end-pf->interval.start+1, new);
+        }
     }
     else
     {
@@ -755,14 +772,15 @@ static void parse_data1_tree(struct grs_read_info *p, const char *mc_stmnt,
 
 		field = cat_field(p, pf, buf, field);
 		
-		pb = wrbuf_buf(buf);
+		wrbuf_cstr(buf);
+                pb = wrbuf_buf(buf);
 		for (pb = strtok(pb, "\n"); pb; pb = strtok(NULL, "\n"))
 		{
-			if (!is_empty(pb))
-			{
-		    		new = data1_mk_tag_n(p->dh, p->mem, mc_stmnt, strlen(mc_stmnt), 0, top);
-		    		data1_mk_text_n(p->dh, p->mem, pb, strlen(pb), new);
-			}
+                    if (!is_empty(pb))
+                    {
+                        new = data1_mk_tag_n(p->dh, p->mem, mc_stmnt, strlen(mc_stmnt), 0, top);
+                        data1_mk_text_n(p->dh, p->mem, pb, strlen(pb), new);
+                    }
 		}
 	    }
 	    else
@@ -773,7 +791,7 @@ static void parse_data1_tree(struct grs_read_info *p, const char *mc_stmnt,
     }
     mc_destroy_field(pf);
     mc_destroy_context(c);
-    wrbuf_free(buf, 1);
+    wrbuf_destroy(buf);
 }
 
 data1_node *grs_read_marcxml(struct grs_read_info *p)
@@ -784,7 +802,7 @@ data1_node *grs_read_marcxml(struct grs_read_info *p)
     if (!root)
 	return 0;
 	
-    for (e = data1_absyn_getelements(p->dh, root->u.root.absyn); e; e=e->next)
+    for (e = data1_absyn_getelements(p->dh, root); e; e=e->next)
     {
 	data1_tag *tag = e->tag;
 	
@@ -803,7 +821,7 @@ data1_node *grs_read_marc(struct grs_read_info *p)
     if (!root)
 	return 0;
 	
-    for (e = data1_absyn_getelements(p->dh, root->u.root.absyn); e; e=e->next)
+    for (e = data1_absyn_getelements(p->dh, root); e; e=e->next)
     {
 	data1_tag *tag = e->tag;