X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=retrieval%2Fd1_read.c;h=0af749251df77c3602e5943a91ce0708c260e2ee;hb=3ee667935c4cac82445ecd05c8ae421888dfd40f;hp=a2753415569c2eacc237b9493bc07bba2c16f907;hpb=43801085d371c69455730182f72c4e53d8c8218f;p=yaz-moved-to-github.git

diff --git a/retrieval/d1_read.c b/retrieval/d1_read.c
index a275341..0af7492 100644
--- a/retrieval/d1_read.c
+++ b/retrieval/d1_read.c
@@ -1,10 +1,25 @@
 /*
- * Copyright (c) 1995-1999, Index Data.
+ * Copyright (c) 1995-2001, Index Data.
  * See the file LICENSE for details.
  * Sebastian Hammer, Adam Dickmeiss
  *
  * $Log: d1_read.c,v $
- * Revision 1.33  2000-11-29 14:22:47  adam
+ * Revision 1.38  2001-03-27 23:06:21  adam
+ * Quotes and slashes may occur within attributes.
+ *
+ * Revision 1.37  2001/02/28 09:00:06  adam
+ * Fixed problem with stack overflow for very nested records.
+ *
+ * Revision 1.36  2001/02/21 13:46:53  adam
+ * C++ fixes.
+ *
+ * Revision 1.35  2000/12/05 14:44:25  adam
+ * Readers skips <! ...> sections.
+ *
+ * Revision 1.34  2000/12/05 10:06:23  adam
+ * Added support for null-data rules like <tag/>.
+ *
+ * Revision 1.33  2000/11/29 14:22:47  adam
  * Implemented XML/SGML attributes for data1 so that d1_read reads them
  * and d1_write generates proper attributes for XML/SGML records. Added
  * register locking for threaded version.
@@ -342,7 +357,7 @@ data1_xattr *data1_read_xattr (data1_handle dh, NMEM m,
 	    c = (*get_byte)(fh);
 	if (!c  || c == '>' || c == '/')
 	    break;
-	*pp = p = nmem_malloc (m, sizeof(*p));
+	*pp = p = (data1_xattr *) nmem_malloc (m, sizeof(*p));
 	p->next = 0;
 	pp = &p->next;
 	p->value = 0;
@@ -355,25 +370,36 @@ data1_xattr *data1_read_xattr (data1_handle dh, NMEM m,
 	}
 	wrbuf_putc (wrbuf, '\0');
 	len = wrbuf_len(wrbuf);
-	p->name = nmem_malloc (m, len);
+	p->name = (char*) nmem_malloc (m, len);
 	strcpy (p->name, wrbuf_buf(wrbuf));
 	if (c == '=')
 	{
 	    c = (*get_byte)(fh);
 	    if (c == '"')
-		c = (*get_byte)(fh);	
-	    wrbuf_rewind(wrbuf);
-	    while (c && c != '"' && c != '>' && c != '/')
 	    {
-		wrbuf_putc (wrbuf, c);
-		c = (*get_byte)(fh);
+		c = (*get_byte)(fh);	
+		wrbuf_rewind(wrbuf);
+		while (c && c != '"')
+		{
+		    wrbuf_putc (wrbuf, c);
+		    c = (*get_byte)(fh);
+	        }
+	        if (c)
+		    c = (*get_byte)(fh);	
 	    }
+	    else
+	    {
+	        wrbuf_rewind(wrbuf);
+	        while (c && c != '>' && c != '/')
+	        {
+		    wrbuf_putc (wrbuf, c);
+		    c = (*get_byte)(fh);
+	        }
+            }
 	    wrbuf_putc (wrbuf, '\0');
 	    len = wrbuf_len(wrbuf);
-	    p->value = nmem_malloc (m, len);
+	    p->value = (char*) nmem_malloc (m, len);
 	    strcpy (p->value, wrbuf_buf(wrbuf));
-	    if (c == '"')
-		c = (*get_byte)(fh);	
 	}
     }
     *ch = c;
@@ -416,10 +442,30 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m,
 #endif
 	    char tag[64];
 	    char args[256];
-	    size_t i;
-	    for (i = 0; (c=(*get_byte)(fh)) && c != '>' && !d1_isspace(c);)
+	    int null_tag = 0;
+	    int end_tag = 0;
+	    size_t i = 0;
+
+	    c = (*get_byte)(fh);
+	    if (c == '/')
+	    {
+		end_tag = 1;
+		c = (*get_byte)(fh);
+	    }
+	    else if (c == '!')  /* tags/comments that we don't deal with yet */
+	    {
+		while (c && c != '>')
+		    c = (*get_byte)(fh);
+		if (c)
+		    c = (*get_byte)(fh);
+		continue;
+	    }
+	    while (c && c != '>' && c != '/' && !d1_isspace(c))
+	    {
 		if (i < (sizeof(tag)-1))
 		    tag[i++] = c;
+		c = (*get_byte)(fh);
+	    }
 	    tag[i] = '\0';
 #if DATA1_USING_XATTR
 	    xattr = data1_read_xattr (dh, m, get_byte, fh, wrbuf, &c);
@@ -427,11 +473,16 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m,
 #else
 	    while (d1_isspace(c))
 		c = (*get_byte)(fh);
-	    for (i = 0; c && c != '>'; c = (*get_byte)(fh))
+	    for (i = 0; c && c != '>' && c != '/'; c = (*get_byte)(fh))
 		if (i < (sizeof(args)-1))
 		    args[i++] = c;
 	    args[i] = '\0';
 #endif
+	    if (c == '/')
+	    {    /* <tag attrs/> or <tag/> */
+		null_tag = 1;
+		c = (*get_byte)(fh);
+	    }
 	    if (c != '>')
 	    {
 		yaz_log(LOG_WARN, "d1: %d: Malformed tag", line);
@@ -441,9 +492,9 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m,
 		c = (*get_byte)(fh);
 
 	    /* End tag? */
-	    if (*tag == '/')       
+	    if (end_tag)       
 	    {
-		if (tag[1] == '\0')
+		if (*tag == '\0')
 		    --level;        /* </> */
 		else
 		{                   /* </tag> */
@@ -452,9 +503,9 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m,
 		    {
 			parent = d1_stack[--i];
 			if ((parent->which == DATA1N_root &&
-			     !strcmp(tag+1, parent->u.root.type)) ||
+			     !strcmp(tag, parent->u.root.type)) ||
 			    (parent->which == DATA1N_tag &&
-			     !strcmp(tag+1, parent->u.tag.tag)))
+			     !strcmp(tag, parent->u.tag.tag)))
 			{
 			    level = i;
 			    break;
@@ -561,7 +612,9 @@ data1_node *data1_read_nodex (data1_handle dh, NMEM m,
 	    else if (parent)
 		parent->child = res;
 	    d1_stack[level] = res;
-	    d1_stack[++level] = 0;
+	    d1_stack[level+1] = 0;
+	    if (level < 250 && !null_tag)
+		++level;
 	}
 	else /* != '<'... this is a body of text */
 	{