Add basic support for RSS parsing for application/rss+xml content types

This means the RSS source will be highlighted, but by default the HTML
renderer will be used for the default rendering.

src/document/dom/renderer.c

@@ -811,6 +811,7 @@ render_dom_document(struct cache_entry *cached, struct document *document,
 	struct dom_renderer renderer;
 	struct conv_table *convert_table;
 	struct sgml_parser *parser;
+	enum sgml_document_type doctype;
 
 	assert(document->options.plain);
 
@@ -824,7 +825,13 @@ render_dom_document(struct cache_entry *cached, struct document *document,
 
 	document->bgcolor = document->options.default_bg;
 
-	parser = init_sgml_parser(SGML_PARSER_STREAM, SGML_DOCTYPE_HTML,
+	if (cached->content_type
+	    && !strlcasecmp("application/rss+xml", 19, cached->content_type, -1))
+		doctype = SGML_DOCTYPE_RSS;
+	else
+		doctype = SGML_DOCTYPE_HTML;
+
+	parser = init_sgml_parser(SGML_PARSER_STREAM, doctype,
 				  &renderer, cached->uri,
 				  dom_source_renderer_push_callbacks,
 				  dom_source_renderer_pop_callbacks);

src/document/renderer.c

@@ -19,6 +19,7 @@
 #include "document/html/renderer.h"
 #include "document/plain/renderer.h"
 #include "document/renderer.h"
+#include "document/rss/renderer.h"
 #include "document/view.h"
 #include "ecmascript/ecmascript.h"
 #include "encoding/encoding.h"
@@ -241,7 +242,8 @@ render_encoded_document(struct cache_entry *cached, struct document *document)
 	if (document->options.plain) {
 #ifdef CONFIG_DOM
 		if (cached->content_type
-		    && !strlcasecmp("text/html", 9, cached->content_type, -1))
+		    && (!strlcasecmp("text/html", 9, cached->content_type, -1)
+		        || !strlcasecmp("application/rss+xml", 19, cached->content_type, -1)))
 			render_dom_document(cached, document, &buffer);
 		else
 #endif

src/document/sgml/Makefile

@@ -1,7 +1,7 @@
 top_builddir=../../..
 include $(top_builddir)/Makefile.config
 
-SUBDIRS = html
+SUBDIRS = html rss
 OBJS	= sgml.o parser.o scanner.o
 
 include $(top_srcdir)/Makefile.lib

src/document/sgml/rss/attribute.inc

@@ -0,0 +1,3 @@
+/* RSS attributes */
+
+RSS_(ATTRIBUTE, ISPERMALINK,		0),

src/document/sgml/rss/element.inc

@@ -0,0 +1,10 @@
+/* RSS elements */
+
+RSS_(ELEMENT, CHANNEL,		0),
+RSS_(ELEMENT, ITEM,		0),
+RSS_(ELEMENT, TITLE,		0),
+RSS_(ELEMENT, AUTHOR,		0),
+RSS_(ELEMENT, PUBDATE,		0),
+RSS_(ELEMENT, GUID,		0),
+RSS_(ELEMENT, LINK,		0),
+RSS_(ELEMENT, DESCRIPTION,	0),

src/document/sgml/rss/rss.c

@@ -0,0 +1,35 @@
+/* SGML node handling */
+
+#ifdef HAVE_CONFIG_H
+#include "config.h"
+#endif
+
+#include <stdlib.h>
+#include <string.h>
+
+#include "elinks.h"
+
+#include "document/sgml/rss/rss.h"
+#include "document/sgml/sgml.h"
+
+
+#define RSS_(node, name, id)	SGML_NODE_INFO(RSS, node, name, id)
+
+static struct sgml_node_info rss_attributes[RSS_ATTRIBUTES] = {
+	SGML_NODE_HEAD(RSS, ATTRIBUTE),
+
+#include "document/sgml/rss/attribute.inc"
+};
+
+static struct sgml_node_info rss_elements[RSS_ELEMENTS] = {
+	SGML_NODE_HEAD(RSS, ELEMENT),
+
+#include "document/sgml/rss/element.inc"
+};
+
+
+struct sgml_info sgml_rss_info = {
+	SGML_DOCTYPE_RSS,
+	rss_attributes,
+	rss_elements,
+};

src/document/sgml/rss/rss.h

@@ -0,0 +1,28 @@
+#ifndef EL__DOCUMENT_SGML_RSS_RSS_H
+#define EL__DOCUMENT_SGML_RSS_RSS_H
+
+#include "document/sgml/sgml.h"
+
+extern struct sgml_info sgml_rss_info;
+
+#define RSS_(node, name, flags)	SGML_NODE_INFO_TYPE(RSS, node, name)
+
+enum rss_element_type {
+	RSS_ELEMENT_UNKNOWN,
+
+#include "document/sgml/rss/element.inc"
+
+	RSS_ELEMENTS,
+};
+
+enum rss_attribute_type {
+	RSS_ATTRIBUTE_UNKNOWN,
+
+#include "document/sgml/rss/attribute.inc"
+
+	RSS_ATTRIBUTES,
+};
+
+#undef	RSS_
+
+#endif

src/document/sgml/sgml.c

@@ -17,6 +17,7 @@
 /* Backend includes: */
 
 #include "document/sgml/html/html.h"
+#include "document/sgml/rss/rss.h"
 
 
 int
@@ -30,6 +31,7 @@ sgml_info_strcmp(const void *key_, const void *node_)
 
 struct sgml_info *sgml_info[SGML_DOCTYPES] = {
 	&sgml_html_info,
+	&sgml_rss_info,
 };
 
 struct sgml_info *

src/document/sgml/sgml.h

@@ -77,6 +77,7 @@ get_sgml_node_info(struct sgml_node_info list[], struct dom_node *node)
 
 enum sgml_document_type {
 	SGML_DOCTYPE_HTML,
+	SGML_DOCTYPE_RSS,
 
 	SGML_DOCTYPES,
 };

src/session/download.c

@@ -1146,6 +1146,7 @@ struct {
 } static known_types[] = {
 	{ "text/html",			0 },
 	{ "application/xhtml+xml",	0 }, /* RFC 3236 */
+	{ "application/rss+xml",	0 },
 	{ "text/plain",			1 },
 	{ NULL,				1 },
 };