static void
parse_channel_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *value;
value = xml_get_attribute (cur, "resource");
if (!xmlStrcmp (BAD_CAST "errorReportsTo", cur->name))
metadata_list_set (&(ctxt->subscription->metadata), "errorReportsTo", value);
else if (!xmlStrcmp (BAD_CAST "generatorAgent", cur->name))
metadata_list_set (&(ctxt->subscription->metadata), "feedgeneratorUri", value);
g_free (value);
}
static void
parse_item_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *tmp, *thumbnail, *imgsrc;
if (!xmlStrcmp (BAD_CAST"thumbnail", cur->name) ||
!xmlStrcmp (BAD_CAST"thumb", cur->name)) {
tmp = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1);
if (tmp) {
if (g_utf8_strlen (tmp, -1) > 0)
g_hash_table_insert (ctxt->item->tmpdata, "photo:thumbnail", tmp);
else
g_free (tmp);
}
} else if (!xmlStrcmp (BAD_CAST"imgsrc", cur->name)) {
tmp = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1);
if (tmp) {
if (g_utf8_strlen (tmp, -1) > 0)
g_hash_table_insert (ctxt->item->tmpdata, "photo:imgsrc", tmp);
else
g_free (tmp);
}
}
thumbnail = g_hash_table_lookup (ctxt->item->tmpdata, "photo:thumbnail");
imgsrc = g_hash_table_lookup (ctxt->item->tmpdata, "photo:imgsrc");
if (!thumbnail) {
/* we do nothing */
} else {
tmp = g_strdup_printf ("%s,%s", thumbnail, imgsrc?imgsrc:"");
metadata_list_set (&(ctxt->item->metadata), "photo", tmp);
g_free (tmp);
}
}
void
subscription_set_homepage (subscriptionPtr subscription, const gchar *newHtmlUrl)
{
gchar *htmlUrl = NULL;
if (newHtmlUrl) {
if (strstr (newHtmlUrl, "://")) {
/* absolute URI can be used directly */
htmlUrl = g_strchomp (g_strdup (newHtmlUrl));
} else {
/* relative URI part needs to be expanded */
gchar *tmp, *source;
source = g_strdup (subscription_get_source (subscription));
tmp = strrchr (source, '/');
if (tmp)
*(tmp+1) = '\0';
htmlUrl = common_build_url (newHtmlUrl, source);
g_free (source);
}
metadata_list_set (&subscription->metadata, "homepage", htmlUrl);
g_free (htmlUrl);
}
}
static void
atom10_parse_feed_subtitle (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *subtitle;
subtitle = atom10_parse_text_construct (cur, TRUE);
if (subtitle) {
metadata_list_set (&ctxt->subscription->metadata, "description", subtitle);
g_free (subtitle);
}
}
static void
atom10_parse_feed_logo (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *logoUrl;
logoUrl = atom10_parse_text_construct (cur, FALSE);
if (logoUrl) {
metadata_list_set (&ctxt->subscription->metadata, "imageUrl", logoUrl);
g_free (logoUrl);
}
}
static void
parse_item_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *uri = NULL;
if (!xmlStrcmp (BAD_CAST"commentRss", cur->name) || !xmlStrcmp (BAD_CAST"commentRSS", cur->name))
uri = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
if (uri) {
metadata_list_set (&(ctxt->item->metadata), "commentFeedUri", uri);
g_free (uri);
}
}
static void
ns_itunes_parse_channel_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *tmp;
const gchar *old;
if (!xmlStrcmp (cur->name, BAD_CAST"summary") || !xmlStrcmp (cur->name, BAD_CAST"subtitle")) {
tmp = xhtml_extract (cur, 0, NULL);
old = metadata_list_get (ctxt->subscription->metadata, "description");
if (!old || strlen (old) < strlen (tmp))
metadata_list_set (&ctxt->subscription->metadata, "description", tmp);
g_free (tmp);
}
}
static void
parse_item_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *date, *source, *sourceURL, *tmp;
gboolean sourceTag = FALSE;
if (!xmlStrcmp (BAD_CAST "source", cur->name)) {
sourceTag = TRUE;
g_hash_table_insert (ctxt->item->tmpdata, g_strdup ("ag:source"), (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1));
} else if (!xmlStrcmp (BAD_CAST "sourceURL", cur->name)) {
sourceTag = TRUE;
g_hash_table_insert (ctxt->item->tmpdata, g_strdup ("ag:sourceURL"), (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1));
}
if (sourceTag) {
source = g_hash_table_lookup (ctxt->item->tmpdata, "ag:source");
sourceURL = g_hash_table_lookup (ctxt->item->tmpdata, "ag:sourceURL");
if (source && sourceURL)
tmp = g_strdup_printf ("<a href=\"%s\">%s</a>", sourceURL, source);
else if (!source)
tmp = g_strdup_printf ("<a href=\"%s\">%s</a>", sourceURL, sourceURL);
else
tmp = g_strdup (source);
metadata_list_set (&(ctxt->item->metadata), "agSource", tmp);
} else if (!xmlStrcmp (BAD_CAST "timestamp", cur->name)) {
if (NULL != (tmp = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1))) {
date = date_format (date_parse_ISO8601 (tmp), _("%b %d %H:%M"));
metadata_list_set (&(ctxt->item->metadata), "agTimestamp", date);
g_free (date);
g_free (tmp);
}
}
}
static void
ttrss_source_merge_feed (ttrssSourcePtr source, const gchar *url, const gchar *title, gint64 id)
{
nodePtr node;
gchar *tmp;
/* check if node to be merged already exists */
node = feedlist_find_node (source->root, NODE_BY_URL, url);
if (!node) {
debug2 (DEBUG_UPDATE, "adding %s (%s)", title, url);
node = node_new (feed_get_node_type ());
node_set_title (node, title);
node_set_data (node, feed_new ());
node_set_subscription (node, subscription_new (url, NULL, NULL));
node->subscription->type = &ttrssSourceFeedSubscriptionType;
/* Save tt-rss feed id which we need to fetch items... */
tmp = g_strdup_printf ("%" G_GINT64_FORMAT, id);
metadata_list_set (&node->subscription->metadata, "ttrss-feed-id", tmp);
g_free (tmp);
node_set_parent (node, source->root, -1);
feedlist_node_imported (node);
/**
* @todo mark the ones as read immediately after this is done
* the feed as retrieved by this has the read and unread
* status inherently.
*/
subscription_update (node->subscription, FEED_REQ_RESET_TITLE | FEED_REQ_PRIORITY_HIGH);
subscription_update_favicon (node->subscription);
/* Important: we must not loose the feed id! */
db_subscription_update (node->subscription);
}
debug2 (DEBUG_UPDATE, "updating folder for %s (%s)", title, url);
ttrss_source_update_folder (source, node);
}
/* simple function to retrieve an OPML document and
parse and output all depth 1 outline tags as
HTML into a buffer */
static void
ns_blogChannel_download_request_cb (const struct updateResult * const result, gpointer user_data, guint32 flags)
{
struct requestData *requestData = user_data;
xmlDocPtr doc = NULL;
xmlNodePtr cur;
GString *buffer = NULL;
g_assert (NULL != requestData);
if (result->data) {
buffer = g_string_new (NULL);
while (1) {
doc = xmlRecoverMemory (result->data, result->size);
if (NULL == doc)
break;
if (NULL == (cur = xmlDocGetRootElement (doc)))
break;
if (!xmlStrcmp (cur->name, BAD_CAST"opml") ||
!xmlStrcmp (cur->name, BAD_CAST"oml") ||
!xmlStrcmp (cur->name, BAD_CAST"outlineDocument")) {
/* nothing */
} else
break;
cur = cur->xmlChildrenNode;
while (cur) {
if (!xmlStrcmp (cur->name, BAD_CAST"body")) {
/* process all <outline> tags */
cur = cur->xmlChildrenNode;
while (cur) {
if (!xmlStrcmp (cur->name, BAD_CAST"outline")) {
GString *tmp = getOutlineContents (cur);
g_string_append_printf (buffer, "%s<br />", tmp->str);
g_string_free (tmp, TRUE);
}
cur = cur->next;
}
break;
}
cur = cur->next;
}
break;
}
if (doc)
xmlFreeDoc (doc);
}
if (buffer) {
switch (requestData->tag) {
case TAG_BLOGROLL:
g_string_prepend (buffer, BLOGROLL_START);
g_string_append (buffer, BLOGROLL_END);
g_hash_table_insert (requestData->ctxt->tmpdata, g_strdup ("bC:blogRoll"), buffer->str);
break;
case TAG_MYSUBSCRIPTIONS:
g_string_prepend (buffer, MYSUBSCR_START);
g_string_append (buffer, MYSUBSCR_END);
g_hash_table_insert (requestData->ctxt->tmpdata, g_strdup ("bC:mySubscriptions"), buffer->str);
break;
default:
g_error ("wrong requestData->tag value");
break;
}
g_string_free (buffer, FALSE);
buffer = g_string_new (NULL);
g_string_append (buffer, g_hash_table_lookup (requestData->ctxt->tmpdata, "bC:blink"));
g_string_append (buffer, g_hash_table_lookup (requestData->ctxt->tmpdata, "bC:blogRoll"));
g_string_append (buffer, g_hash_table_lookup (requestData->ctxt->tmpdata, "bC:mySubscriptions"));
metadata_list_set (&(requestData->ctxt->subscription->metadata), "blogChannel", buffer->str);
g_string_free (buffer, TRUE);
}
g_list_free (requestData->ctxt->items);
feed_free_parser_ctxt (requestData->ctxt);
g_free (requestData);
}
/* reads a PIE feed URL and returns a new channel structure (even if
the feed could not be read) */
static void pie_parse(feedParserCtxtPtr ctxt, xmlNodePtr cur) {
gchar *tmp2, *tmp = NULL, *tmp3;
NsHandler *nsh;
parseChannelTagFunc pf;
while(TRUE) {
if(xmlStrcmp(cur->name, BAD_CAST"feed")) {
g_string_append(ctxt->feed->parseErrors, "<p>Could not find Atom/Echo/PIE header!</p>");
break;
}
/* parse feed contents */
cur = cur->xmlChildrenNode;
while(cur) {
if(!cur->name || cur->type != XML_ELEMENT_NODE) {
cur = cur->next;
continue;
}
/* check namespace of this tag */
if(cur->ns) {
if((cur->ns->href && (nsh = (NsHandler *)g_hash_table_lookup(ns_pie_ns_uri_table, (gpointer)cur->ns->href))) ||
(cur->ns->prefix && (nsh = (NsHandler *)g_hash_table_lookup(pie_nstable, (gpointer)cur->ns->prefix)))) {
pf = nsh->parseChannelTag;
if(pf)
(*pf)(ctxt, cur);
cur = cur->next;
continue;
} else {
/*g_print("unsupported namespace \"%s\"\n", cur->ns->prefix);*/
}
} /* explicitly no following else !!! */
if(!xmlStrcmp(cur->name, BAD_CAST"title")) {
tmp = unhtmlize(pie_parse_content_construct(cur));
if(tmp) {
if(ctxt->title)
g_free(ctxt->title);
ctxt->title = tmp;
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"link")) {
tmp = xml_get_attribute (cur, "href");
if(tmp) {
/* 0.3 link : rel, type and href attribute */
tmp2 = xml_get_attribute (cur, "rel");
if(tmp2 && g_str_equal(tmp2, "alternate"))
subscription_set_homepage (ctxt->subscription, tmp);
/* else
FIXME: Maybe do something with other links? */
g_free(tmp2);
g_free(tmp);
} else {
/* 0.2 link : element content is the link, or non-alternate link in 0.3 */
tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1);
if(tmp) {
subscription_set_homepage (ctxt->subscription, tmp);
g_free(tmp);
}
}
/* parse feed author */
} else if(!xmlStrcmp(cur->name, BAD_CAST"author")) {
/* parse feed author */
tmp = parseAuthor(cur);
if(tmp) {
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, "author", tmp);
g_free(tmp);
}
} else if (!xmlStrcmp (cur->name, BAD_CAST"tagline")) {
tmp = pie_parse_content_construct (cur);
if (tmp) {
metadata_list_set (&ctxt->subscription->metadata, "description", tmp);
g_free (tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"generator")) {
tmp = unhtmlize((gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1));
if(tmp && tmp[0] != '\0') {
tmp2 = xml_get_attribute (cur, "version");
if(tmp2) {
tmp3 = g_strdup_printf("%s %s", tmp, tmp2);
g_free(tmp);
g_free(tmp2);
tmp = tmp3;
}
tmp2 = xml_get_attribute (cur, "url");
if(tmp2) {
tmp3 = g_strdup_printf("<a href=\"%s\">%s</a>", tmp2, tmp);
g_free(tmp2);
g_free(tmp);
tmp = tmp3;
}
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, "feedgenerator", tmp);
}
g_free(tmp);
} else if(!xmlStrcmp(cur->name, BAD_CAST"copyright")) {
tmp = pie_parse_content_construct(cur);
if(tmp) {
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, "copyright", tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"modified")) { /* Modified was last used in IETF draft 02) */
tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1);
if(tmp) {
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, "pubDate", tmp);
ctxt->feed->time = date_parse_ISO8601 (tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"updated")) { /* Updated was added in IETF draft 03 */
tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1);
if(tmp) {
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, "pubDate", tmp);
ctxt->feed->time = date_parse_ISO8601 (tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"contributor")) {
tmp = parseAuthor(cur);
if(tmp) {
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, "contributor", tmp);
g_free(tmp);
}
} else if((!xmlStrcmp(cur->name, BAD_CAST"entry"))) {
ctxt->item = parseEntry(ctxt, cur);
if(ctxt->item) {
if(0 == ctxt->item->time)
ctxt->item->time = ctxt->feed->time;
ctxt->items = g_list_append(ctxt->items, ctxt->item);
}
}
cur = cur->next;
}
break;
}
}
/* Note: this function is called for both item and feed context */
static gchar *
atom10_parse_link (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *href, *alternate = NULL;
href = xml_get_ns_attribute (cur, "href", NULL);
if (href) {
xmlChar *baseURL = xmlNodeGetBase (cur->doc, cur);
gchar *url, *relation, *type, *escTitle = NULL, *title;
const gchar *feedURL = subscription_get_homepage (ctxt->subscription);
if (!baseURL && feedURL && feedURL[0] != '|' && strstr (feedURL, "://"))
baseURL = xmlStrdup (BAD_CAST (feedURL));
url = (gchar *)common_build_url (href, (gchar *)baseURL);
type = xml_get_ns_attribute (cur, "type", NULL);
relation = xml_get_ns_attribute (cur, "rel", NULL);
title = xml_get_ns_attribute (cur, "title", NULL);
if (title)
escTitle = g_markup_escape_text (title, -1);
if (!xmlHasNsProp (cur, BAD_CAST"rel", NULL) || !relation || g_str_equal (relation, BAD_CAST"alternate"))
alternate = g_strdup (url);
else if (g_str_equal (relation, "replies")) {
if (!type || g_str_equal (type, BAD_CAST"application/atom+xml")) {
gchar *commentUri = (gchar *)common_build_url ((gchar *)url, subscription_get_homepage (ctxt->subscription));
if (ctxt->item)
metadata_list_set (&ctxt->item->metadata, "commentFeedUri", commentUri);
g_free (commentUri);
}
} else if (g_str_equal (relation, "enclosure")) {
if (ctxt->item) {
gsize length = 0;
gchar *lengthStr = xml_get_ns_attribute (cur, "length", NULL);
if (lengthStr)
length = atol (lengthStr);
g_free (lengthStr);
gchar *encStr = enclosure_values_to_string (url, type, length, FALSE /* not yet downloaded */);
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "enclosure", encStr);
ctxt->item->hasEnclosure = TRUE;
g_free (encStr);
}
} else if (g_str_equal (relation, "related") || g_str_equal (relation, "via")) {
if (ctxt->item)
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, relation, url);
} else {
/* g_warning ("Unhandled Atom link with unexpected relation \"%s\"\n", relation); */
}
xmlFree (title);
xmlFree (baseURL);
g_free (escTitle);
g_free (url);
g_free(relation);
g_free(type);
g_free(href);
} else {
/* FIXME: @href is required, this document is not valid Atom */;
}
return alternate;
}
/* method to parse standard tags for the channel element */
static void parseCDFChannel(feedParserCtxtPtr ctxt, xmlNodePtr cur, CDFChannelPtr cp) {
gchar *tmp, *tmp2, *tmp3;
cur = cur->xmlChildrenNode;
while(cur) {
if(!cur->name || cur->type != XML_ELEMENT_NODE) {
cur = cur->next;
continue;
}
if((!xmlStrcasecmp(cur->name, BAD_CAST"logo"))) {
tmp = (gchar *)xmlGetProp(cur, BAD_CAST"HREF");
if(tmp) {
tmp = (gchar *)xmlGetProp(cur, BAD_CAST"href");
metadata_list_set (&ctxt->subscription->metadata, "imageUrl", tmp);
g_free(tmp);
}
} else if((!xmlStrcasecmp(cur->name, BAD_CAST"a"))) {
xmlChar *value = xmlGetProp(cur, BAD_CAST"HREF");
if(value) {
subscription_set_homepage (ctxt->subscription, (gchar *)value);
xmlFree(value);
}
} else if((!xmlStrcasecmp(cur->name, BAD_CAST"item"))) {
ctxt->item = parseCDFItem(ctxt, cur, cp);
if(ctxt->item) {
if(0 == ctxt->item->time)
ctxt->item->time = cp->time;
ctxt->items = g_list_append(ctxt->items, ctxt->item);
}
} else if(!xmlStrcasecmp(cur->name, BAD_CAST "title")) {
tmp = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, TRUE);
if(tmp) {
tmp = unhtmlize(tmp);
if(ctxt->title)
g_free(ctxt->title);
ctxt->title = tmp;
}
} else if (!xmlStrcasecmp(cur->name, BAD_CAST "abstract")) {
tmp = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, TRUE);
if (tmp) {
metadata_list_set (&ctxt->subscription->metadata, "description", tmp);
xmlFree (tmp);
}
} else {
tmp = g_ascii_strdown((gchar *)cur->name, -1);
tmp2 = g_hash_table_lookup(channelHash, tmp);
if(tmp2) {
tmp3 = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, TRUE);
if(tmp3) {
ctxt->subscription->metadata = metadata_list_append(ctxt->subscription->metadata, tmp2, tmp3);
g_free(tmp3);
}
}
g_free(tmp);
}
cur = cur->next;
}
}
