X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=retrieval%2Fd1_write.c;h=565af5e0b35fc56891d9a634ed5a07b1849c7ea6;hb=9dffa786b23759414b476e152f590f6aa0e8aa3a;hp=66e8b26be9d76d73671db2d7ff3e44bae3839ec7;hpb=8916c5245235fa1229f8a23ba7ebb1500e98b818;p=yaz-moved-to-github.git

diff --git a/retrieval/d1_write.c b/retrieval/d1_write.c
index 66e8b26..565af5e 100644
--- a/retrieval/d1_write.c
+++ b/retrieval/d1_write.c
@@ -1,141 +1,240 @@
 /*
- * Copyright (c) 1995, Index Data.
+ * Copyright (c) 1995-2002, Index Data.
  * See the file LICENSE for details.
  * Sebastian Hammer, Adam Dickmeiss
  *
- * $Log: d1_write.c,v $
- * Revision 1.2  1995-12-13 17:14:27  quinn
- * *** empty log message ***
- *
- * Revision 1.1  1995/12/13  15:38:43  quinn
- * Added SGML-output filter.
- *
- *
+ * $Id: d1_write.c,v 1.17 2002-08-23 14:25:07 adam Exp $
  */
 
 #include <string.h>
-#include <ctype.h>
 
-#include <data1.h>
-#include <wrbuf.h>
+#include <yaz/data1.h>
+#include <yaz/wrbuf.h>
 
 #define IDSGML_MARGIN 75
 
-static int wordlen(char *b)
+#define PRETTY_FORMAT 0
+
+static int wordlen(char *b, int max)
 {
     int l = 0;
 
-    while (*b && !isspace(*b))
+    while (l < max && !d1_isspace(*b))
 	l++, b++;
     return l;
 }
 
-static int nodetoidsgml(data1_node *n, int select, WRBUF b, int col)
+static void indent (WRBUF b, int col)
+{
+    int i;
+    for (i = 0; i<col; i++)
+        wrbuf_putc (b, ' ');
+}
+
+static void wrbuf_write_cdata(WRBUF b, const char *msg, int len)
+{
+    int i;
+
+    for (i = 0; i < len; i++)
+    {
+        switch (msg[i])
+        {
+        case '"':
+            wrbuf_puts (b, "&quot;");
+            break;
+        case '\'':
+            wrbuf_puts (b, "&apos;");
+            break;
+        case '>':
+            wrbuf_puts (b, "&gt;");
+            break;
+        case '<':
+            wrbuf_puts (b, "&lt;");
+            break;
+        case '&':
+            wrbuf_puts (b, "&amp;");
+            break;
+        default:
+            wrbuf_putc(b, msg[i]);
+        }
+    }
+}
+
+static void wrbuf_put_cdata(WRBUF b, const char *msg)
+{
+    wrbuf_write_cdata (b, msg, strlen(msg));
+}
+
+static int nodetoidsgml(data1_node *n, int select, WRBUF b, int col,
+                        int pretty_format)
 {
     data1_node *c;
-    char line[1024];
 
     for (c = n->child; c; c = c->next)
     {
 	char *tag;
 
-	if (c->which == DATA1N_tag)
+        if (c->which == DATA1N_preprocess)
+        {
+            data1_xattr *p;
+
+            if (pretty_format)
+                indent (b, col);
+	    wrbuf_puts (b, "<?");
+            wrbuf_put_cdata (b, c->u.preprocess.target);
+            for (p = c->u.preprocess.attributes; p; p = p->next)
+            {
+                wrbuf_putc (b, ' ');
+                wrbuf_put_cdata (b, p->name);
+                wrbuf_putc (b, '=');
+                wrbuf_putc (b, '"');
+                wrbuf_put_cdata (b, p->value);
+                wrbuf_putc (b, '"');
+            }
+            if (c->child)
+                wrbuf_puts(b, " ");
+            if (nodetoidsgml(c, select, b, (col > 40) ? 40 : col+2,
+                             pretty_format) < 0)
+                return -1;
+            wrbuf_puts (b, "?>\n");
+        }
+        else if (c->which == DATA1N_tag)
 	{
 	    if (select && c->u.tag.node_selected)
 		continue;
-	    if (c->u.tag.element && c->u.tag.element->tag)
-		tag = c->u.tag.element->tag->names->name; /* first name */
+            tag = c->u.tag.tag;
+	    if (!data1_matchstr(tag, "wellknown")) /* skip wellknown */
+	    {
+		if (nodetoidsgml(c, select, b, col, pretty_format) < 0)
+		    return -1;
+	    }
 	    else
-		tag = c->u.tag.tag; /* local string tag */
-	    if (data1_matchstr(tag, "wellknown")) /* skip wellknown */
 	    {
-		sprintf(line, "<%s>\n", tag);
-		wrbuf_write(b, line, strlen(line));
-		col = 0;
+		data1_xattr *p;
+
+                if (pretty_format)
+                    indent (b, col);
+		wrbuf_puts (b, "<");	
+		wrbuf_put_cdata (b, tag);
+		for (p = c->u.tag.attributes; p; p = p->next)
+		{
+		    wrbuf_putc (b, ' ');
+		    wrbuf_put_cdata (b, p->name);
+		    wrbuf_putc (b, '=');
+		    wrbuf_putc (b, '"');
+		    wrbuf_put_cdata (b, p->value);
+		    wrbuf_putc (b, '"');
+		}
+		wrbuf_puts(b, ">");
+                if (pretty_format)
+                    wrbuf_puts(b, "\n");
+		if (nodetoidsgml(c, select, b, (col > 40) ? 40 : col+2,
+                                 pretty_format) < 0)
+		    return -1;
+                if (pretty_format)
+                    indent (b, col);
+		wrbuf_puts(b, "</");
+		wrbuf_put_cdata(b, tag);
+		wrbuf_puts(b, ">");
+                if (pretty_format)
+                    wrbuf_puts (b, "\n");
 	    }
-	    if (nodetoidsgml(c, select, b, col) < 0)
-		return -1;
-	    wrbuf_write(b, "</>\n", 4);
-	    col = 0;
 	}
-	else if (c->which == DATA1N_data)
+	else if (c->which == DATA1N_data || c->which == DATA1N_comment)
 	{
 	    char *p = c->u.data.data;
 	    int l = c->u.data.len;
 	    int first = 1;
+	    int lcol = col;
 
-	    if (c->u.data.what == DATA1I_text)
-	    {
-		while (l)
-		{
-		    int wlen;
-
-		    while (l && isspace(*p))
-			p++, l--;
-		    if (!l)
-			break;
-		    /* break if we'll cross margin and word is not too long */
-		    if (col + (wlen = wordlen(p)) > IDSGML_MARGIN && wlen <
-			IDSGML_MARGIN)
-		    {
-			sprintf(line, "\n");
-			col = 0;
-			wrbuf_write(b, line, strlen(line));
-			first = 1;
-		    }
-		    if (!first)
-		    {
-			wrbuf_putc(b, ' ');
-			col++;
-		    }
-		    while (l && !isspace(*p))
-		    {
-#if 0
-			if (col > NTOBUF_MARGIN)
-			{
-			    wrbuf_putc(b, '=');
-			    wrbuf_putc(b, '\n');
-			    sprintf(line, "%*s", indent * NTOBUF_INDENT, "");
-			    wrbuf_write(b, line, strlen(line));
-			    col = indent * NTOBUF_INDENT;
-			}
-#endif
-			wrbuf_putc(b, *p);
-			p++;
-			l--;
-			col++;
-		    }
-		    first = 0;
-		}
-		wrbuf_write(b, "\n", 1);
-		col = 0;
-	    }
-	    else if (c->u.data.what == DATA1I_num)
+            if (pretty_format && !c->u.data.formatted_text)
+                indent (b, col);
+            if (c->which == DATA1N_comment)
+                wrbuf_puts (b, "<!--");
+	    switch (c->u.data.what)
 	    {
-		wrbuf_putc(b, ' ');
-		wrbuf_write(b, c->u.data.data, c->u.data.len);
+	    case DATA1I_text:
+                if (!pretty_format || c->u.data.formatted_text)
+                {
+                    wrbuf_write_cdata (b, p, l);
+                }
+                else
+                {
+                    while (l)
+                    {
+                        int wlen;
+                        
+                        while (l && d1_isspace(*p))
+                            p++, l--;
+                        if (!l)
+                            break;
+                        /* break if we cross margin and word is not too long */
+                        if (lcol + (wlen = wordlen(p, l)) > IDSGML_MARGIN &&
+                            wlen < IDSGML_MARGIN)
+                        {
+                            wrbuf_puts (b, "\n");
+                            indent (b, col);
+                            lcol = col;
+                            first = 1;
+                        }
+                        if (!first)
+                        {
+                            wrbuf_putc(b, ' ');
+                            lcol++;
+                        }
+                        while (l && !d1_isspace(*p))
+                        {
+                            wrbuf_putc(b, *p);
+                            p++;
+                            l--;
+                            lcol++;
+                        }
+                        first = 0;
+                    }
+                    wrbuf_puts(b, "\n");
+                }
+		break;
+	    case DATA1I_num:
+		wrbuf_write_cdata(b, c->u.data.data, c->u.data.len);
+                if (pretty_format)
+                    wrbuf_puts(b, "\n");
+		break;
+	    case DATA1I_oid:
+		wrbuf_write_cdata(b, c->u.data.data, c->u.data.len);
+                if (pretty_format)
+                    wrbuf_puts(b, "\n");
 	    }
+            if (c->which == DATA1N_comment)
+            {
+                wrbuf_puts(b, "-->");
+                if (pretty_format)
+                    wrbuf_puts(b, "\n");
+            }
 	}
     }
     return 0;
 }
 
-char *data1_nodetoidsgml(data1_node *n, int select, int *len)
+char *data1_nodetoidsgml (data1_handle dh, data1_node *n, int select, int *len)
 {
-    static WRBUF b = 0;
-    char line[1024];
-
-    if (!b)
-	b = wrbuf_alloc();
-    else
-	wrbuf_rewind(b);
+    WRBUF b = data1_get_wrbuf (dh);
+    
+    wrbuf_rewind(b);
     
-    sprintf(line, "<%s>\n", n->u.root.type);
-    wrbuf_write(b, line, strlen(line));
-    if (nodetoidsgml(n, select, b, 0))
+    if (!data1_is_xmlmode (dh))
+    {
+        wrbuf_puts (b, "<");
+        wrbuf_puts (b, n->u.root.type);
+        wrbuf_puts (b, ">\n");
+    }
+    if (nodetoidsgml(n, select, b, 0, 0 /* no pretty format */))
 	return 0;
-    sprintf(line, "</%s>\n", n->u.root.type);
-    wrbuf_write(b, line, strlen(line));
+    if (!data1_is_xmlmode (dh))
+    {
+        wrbuf_puts (b, "</");
+        wrbuf_puts (b, n->u.root.type);
+        wrbuf_puts (b, ">\n");
+    }
     *len = wrbuf_len(b);
     return wrbuf_buf(b);
 }