/* generic tag parsing (used for RSS and Atom) */
static void
parse_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur, gboolean isFeedTag)
{
int i, j;
gchar *date, *value, *tmp;
const gchar *mapping;
gboolean isNotEmpty;
if (!isFeedTag) {
/* special handling for the ISO 8601 date item tags */
if (!xmlStrcmp (BAD_CAST "date", cur->name)) {
if (NULL != (date = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1))) {
i = date_parse_ISO8601 (date);
ctxt->item->time = i;
g_free (date);
}
return;
}
/* special handling for item titles */
if(!xmlStrcmp (BAD_CAST "title", cur->name)) {
value = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1);
if(value) {
item_set_title(ctxt->item, value);
g_free(value);
}
return;
}
}
/* compare with each possible tag name */
for (i = 0; taglist[i] != NULL; i++) {
if (!xmlStrcmp ((const xmlChar *)taglist[i], cur->name)) {
value = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1);
if (value) {
/* check if value consist of whitespaces only */
for (j = 0, tmp = value, isNotEmpty = FALSE; j < g_utf8_strlen (value, -1); j++) {
if (!g_unichar_isspace (*tmp)) {
isNotEmpty = TRUE;
break;
}
tmp = g_utf8_next_char (tmp);
}
if (isNotEmpty) {
if (isFeedTag) {
if (NULL != (mapping = mapToFeedMetadata[i]))
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, mapping, value);
} else {
if (NULL != (mapping = mapToItemMetadata[i]))
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, mapping, value);
}
}
g_free (value);
}
return;
}
}
}
static void
theoldreader_source_merge_feed (TheOldReaderSourcePtr source, const gchar *url, const gchar *title, const gchar *id, nodePtr folder)
{
nodePtr node;
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 = source->root->source->type->feedSubscriptionType;
/* Save TheOldReader feed id which we need to fetch items... */
node->subscription->metadata = metadata_list_append (node->subscription->metadata, "theoldreader-feed-id", id);
db_subscription_update (node->subscription);
node_set_parent (node, folder?folder: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);
} else {
node_source_update_folder (node, folder);
}
}
static void
atom10_parse_feed_generator (xmlNodePtr cur, feedParserCtxtPtr ctxt, itemPtr ip, struct atom10ParserState *state)
{
gchar *ret, *version, *tmp = NULL, *uri;
ret = unhtmlize ((gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1));
if (ret && ret[0] != '\0') {
version = xml_get_ns_attribute (cur, "version", NULL);
if (version) {
tmp = g_strdup_printf ("%s %s", ret, version);
g_free (ret);
g_free (version);
ret = tmp;
}
uri = xml_get_ns_attribute (cur, "uri", NULL);
if (uri) {
tmp = g_markup_printf_escaped ("<a href=\"%s\">%s</a>", uri, ret);
g_free (uri);
g_free (ret);
ret = tmp;
}
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, "feedgenerator", tmp);
}
g_free (ret);
}
static void
atom10_parse_feed_contributor (xmlNodePtr cur, feedParserCtxtPtr ctxt, itemPtr ip, struct atom10ParserState *state)
{
/* parse feed contributors */
gchar *contributer = atom10_parse_person_construct (cur);
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, "contributor", contributer);
g_free (contributer);
}
static void
atom10_parse_feed_author (xmlNodePtr cur, feedParserCtxtPtr ctxt, itemPtr ip, struct atom10ParserState *state)
{
/* parse feed author */
gchar *author = atom10_parse_person_construct (cur);
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, "author", author);
g_free (author);
/* FIXME: make item parsing use this author if not specified elsewhere */
}
static GSList *
db_metadata_list_append (GSList *metadata, const char *key, const char *value)
{
if (metadata_is_type_registered (key))
metadata = metadata_list_append (metadata, key, value);
else
debug1 (DEBUG_DB, "Trying to load unregistered metadata type %s from DB.", key);
return metadata;
}
static void
atom10_parse_entry_contributor (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *contributor;
contributor = atom10_parse_person_construct (cur);
if (contributor) {
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "contributor", contributor);
g_free (contributor);
}
}
static void
atom10_parse_feed_rights (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *rights;
rights = atom10_parse_text_construct (cur, FALSE);
if (rights) {
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, "copyright", rights);
g_free (rights);
}
}
static void
atom10_parse_feed_updated (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *timestamp;
timestamp = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
if (timestamp) {
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, "contentUpdateDate", timestamp);
ctxt->feed->time = date_parse_ISO8601 (timestamp);
g_free (timestamp);
}
}
static void
atom10_parse_entry_published (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *datestr;
datestr = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
if (datestr) {
ctxt->item->time = date_parse_ISO8601 (datestr);
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "pubDate", datestr);
g_free (datestr);
}
}
static void
atom10_parse_feed_icon (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *icon_uri;
icon_uri = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
if (icon_uri) {
debug1 (DEBUG_PARSING, "icon URI found in atom feed: %s", icon_uri);
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata,
"icon", icon_uri);
}
}
static void
ns_itunes_parse_item_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *tmp;
if (!xmlStrcmp(cur->name, BAD_CAST"author")) {
tmp = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
if (tmp) {
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "author", tmp);
g_free (tmp);
}
}
if (!xmlStrcmp (cur->name, BAD_CAST"summary")) {
tmp = xhtml_extract (cur, 0, NULL);
item_set_description (ctxt->item, tmp);
g_free (tmp);
}
if (!xmlStrcmp(cur->name, BAD_CAST"keywords")) {
gchar *keyword = tmp = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
gchar *allocated = tmp;
/* parse comma separated list and strip leading spaces... */
while (tmp) {
tmp = strchr (tmp, ',');
if (tmp) {
*tmp = 0;
tmp++;
}
while (g_unichar_isspace (*keyword)) {
keyword = g_utf8_next_char (keyword);
}
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "category", keyword);
keyword = tmp;
}
g_free (allocated);
}
}
static void
atom10_parse_entry_updated (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *datestr;
datestr = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
/* if pubDate is already set, don't overwrite it */
if (datestr && !metadata_list_get(ctxt->item->metadata, "pubDate")) {
ctxt->item->time = date_parse_ISO8601 (datestr);
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "contentUpdateDate", datestr);
}
g_free (datestr);
}
static void
atom10_parse_feed_category (xmlNodePtr cur, feedParserCtxtPtr ctxt, itemPtr ip, struct atom10ParserState *state)
{
gchar *label = NULL;
label = xml_get_ns_attribute (cur, "label", NULL);
if (!label)
label = xml_get_ns_attribute (cur, "term", NULL);
if (label) {
gchar *escaped = g_markup_escape_text (label, -1);
ctxt->subscription->metadata = metadata_list_append (ctxt->subscription->metadata, "category", escaped);
g_free (escaped);
xmlFree (label);
}
}
static void
atom10_parse_entry_category (xmlNodePtr cur, feedParserCtxtPtr ctxt, struct atom10ParserState *state)
{
gchar *category = NULL;
category = xml_get_ns_attribute (cur, "label", NULL);
if (!category)
category = xml_get_ns_attribute (cur, "term", NULL);
if (category) {
gchar *escaped = g_markup_escape_text (category, -1);
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "category", escaped);
g_free (escaped);
xmlFree (category);
}
}
GSList *
metadata_list_copy (GSList *list)
{
GSList *copy = NULL;
GSList *iter2, *iter = list;
struct pair *p;
while (iter) {
p = (struct pair*)iter->data;
iter2 = p->data;
while (iter2) {
copy = metadata_list_append (copy, p->strid, iter2->data);
iter2 = iter2->next;
}
iter = iter->next;
}
return copy;
}
static void
parse_item_tag (feedParserCtxtPtr ctxt, xmlNodePtr cur)
{
gchar *tmp;
/* We ignore the "ping" tag */
if (xmlStrcmp (cur->name, BAD_CAST"about"))
return;
/* RSS 1.0 */
tmp = xml_get_attribute (cur, "about");
/* RSS 2.0 */
if (!tmp)
tmp = (gchar *)xmlNodeListGetString (cur->doc, cur->xmlChildrenNode, 1);
if (tmp) {
ctxt->item->metadata = metadata_list_append (ctxt->item->metadata, "related", tmp);
g_free (tmp);
}
}
static void
inoreader_source_merge_feed (InoreaderSourcePtr source, const gchar *url, const gchar *title, const gchar *id)
{
nodePtr node;
GSList *iter;
/* check if node to be merged already exists */
iter = source->root->children;
while (iter) {
node = (nodePtr)iter->data;
if (g_str_equal (node->subscription->source, url))
return;
iter = g_slist_next (iter);
}
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 = &inoreaderSourceFeedSubscriptionType;
/* Save Inoreader feed id which we need to fetch items... */
node->subscription->metadata = metadata_list_append (node->subscription->metadata, "inoreader-feed-id", id);
db_subscription_update (node->subscription);
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);
}
/* method to parse standard tags for each item element */
itemPtr parseCDFItem(feedParserCtxtPtr ctxt, xmlNodePtr cur, CDFChannelPtr cp) {
gchar *tmp = NULL, *tmp2, *tmp3;
if(CDFToMetadataMapping == NULL) {
CDFToMetadataMapping = g_hash_table_new(g_str_hash, g_str_equal);
g_hash_table_insert(CDFToMetadataMapping, "author", "author");
g_hash_table_insert(CDFToMetadataMapping, "category", "category");
}
ctxt->item = item_new();
/* save the item link */
if(!(tmp = (gchar *)xmlGetProp(cur, BAD_CAST"href")))
tmp = (gchar *)xmlGetProp(cur, BAD_CAST"HREF");
if(tmp) {
item_set_source(ctxt->item, tmp);
g_free(tmp);
}
cur = cur->xmlChildrenNode;
while(cur) {
if(!cur->name || cur->type != XML_ELEMENT_NODE) {
cur = cur->next;
continue;
}
/* save first link to a channel image */
if(NULL != (tmp = g_ascii_strdown((gchar *)cur->name, -1))) {
if(NULL != (tmp2 = g_hash_table_lookup(CDFToMetadataMapping, tmp))) {
if(NULL != (tmp3 = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, TRUE))) {
ctxt->item->metadata = metadata_list_append(ctxt->item->metadata, tmp2, tmp3);
g_free(tmp3);
}
}
g_free(tmp);
}
if((!xmlStrcasecmp(cur->name, BAD_CAST"logo"))) {
if(!(tmp = (gchar *)xmlGetProp(cur, BAD_CAST"href")))
tmp = (gchar *)xmlGetProp(cur, BAD_CAST"HREF");
if(tmp) {
ctxt->item->metadata = metadata_list_append(ctxt->item->metadata, "imageUrl", tmp);
g_free(tmp);
}
} else if((!xmlStrcasecmp(cur->name, BAD_CAST"title"))) {
if(NULL != (tmp = unhtmlize((gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1)))) {
item_set_title(ctxt->item, tmp);
g_free(tmp);
}
} else if((!xmlStrcasecmp(cur->name, BAD_CAST"abstract"))) {
if(NULL != (tmp = (gchar *)xmlNodeListGetString(cur->doc, cur->xmlChildrenNode, 1))) {
item_set_description(ctxt->item, tmp);
g_free(tmp);
}
} else if((!xmlStrcasecmp(cur->name, BAD_CAST"a"))) {
if(!(tmp = (gchar *)xmlGetProp(cur, BAD_CAST"href")))
tmp = (gchar *)xmlGetProp(cur, BAD_CAST"HREF");
if(tmp) {
item_set_source(ctxt->item, tmp);
g_free(tmp);
}
}
cur = cur->next;
}
ctxt->item->readStatus = FALSE;
return ctxt->item;
}
/* 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;
}
}
GList *
json_api_get_items (const gchar *json, const gchar *root, jsonApiMapping *mapping, jsonApiItemCallbackFunc callback)
{
GList *items = NULL;
JsonParser *parser = json_parser_new ();
if (json_parser_load_from_data (parser, json, -1, NULL)) {
JsonArray *array = json_node_get_array (json_get_node (json_parser_get_root (parser), root));
GList *elements = json_array_get_elements (array);
GList *iter = elements;
debug1 (DEBUG_PARSING, "JSON API: found items root node \"%s\"", root);
while (iter) {
JsonNode *node = (JsonNode *)iter->data;
itemPtr item = item_new ();
/* Parse default feeds */
item_set_id (item, json_api_get_string (node, mapping->id));
item_set_title (item, json_api_get_string (node, mapping->title));
item_set_source (item, json_api_get_string (node, mapping->link));
item->time = json_api_get_int (node, mapping->updated);
item->readStatus = json_api_get_bool (node, mapping->read);
item->flagStatus = json_api_get_bool (node, mapping->flag);
if (mapping->negateRead)
item->readStatus = !item->readStatus;
/* Handling encoded content */
const gchar *content;
gchar *xhtml;
content = json_api_get_string (node, mapping->description);
if (mapping->xhtml) {
xhtml = xhtml_extract_from_string (content, NULL);
item_set_description (item, xhtml);
xmlFree (xhtml);
} else {
item_set_description (item, content);
}
/* Optional meta data */
const gchar *tmp = json_api_get_string (node, mapping->author);
if (tmp)
item->metadata = metadata_list_append (item->metadata, "author", tmp);
items = g_list_append (items, (gpointer)item);
/* Allow optional item callback to process stuff */
if (callback)
(*callback)(node, item);
iter = g_list_next (iter);
}
g_list_free (elements);
g_object_unref (parser);
} else {
debug1 (DEBUG_PARSING, "Could not parse JSON \"%s\"", json);
}
return items;
}
/* method to parse standard tags for each item element */
itemPtr parseEntry(feedParserCtxtPtr ctxt, xmlNodePtr cur) {
xmlChar *xtmp;
gchar *tmp2, *tmp;
NsHandler *nsh;
parseItemTagFunc pf;
g_assert(NULL != cur);
ctxt->item = item_new();
cur = cur->xmlChildrenNode;
while(cur) {
if(!cur->name) {
g_warning("invalid XML: parser returns NULL value -> tag ignored!");
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)))) {
if(NULL != (pf = nsh->parseItemTag))
(*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")) {
if(NULL != (tmp = unhtmlize(pie_parse_content_construct(cur)))) {
item_set_title(ctxt->item, tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"link")) {
if(NULL != (tmp2 = xml_get_attribute(cur, "href"))) {
/* 0.3 link : rel, type and href attribute */
xtmp = xmlGetProp(cur, BAD_CAST"rel");
if(xtmp != NULL && !xmlStrcmp(xtmp, BAD_CAST"alternate"))
item_set_source(ctxt->item, tmp2);
/* else
FIXME: Maybe do something with other links? */
xmlFree(xtmp);
g_free(tmp2);
} else {
/* 0.2 link : element content is the link, or non-alternate link in 0.3 */
if(NULL != (tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1))) {
item_set_source(ctxt->item, tmp);
g_free(tmp);
}
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"author")) {
/* parse feed author */
tmp = parseAuthor(cur);
ctxt->item->metadata = metadata_list_append(ctxt->item->metadata, "author", tmp);
g_free(tmp);
} else if(!xmlStrcmp(cur->name, BAD_CAST"contributor")) {
/* parse feed contributors */
tmp = parseAuthor(cur);
ctxt->item->metadata = metadata_list_append(ctxt->item->metadata, "contributor", tmp);
g_free(tmp);
} else if(!xmlStrcmp(cur->name, BAD_CAST"id")) {
if(NULL != (tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1))) {
item_set_id(ctxt->item, tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"issued")) {
/* FIXME: is <modified> or <issued> or <created> the time tag we want to display? */
if(NULL != (tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1))) {
ctxt->item->time = date_parse_ISO8601 (tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"content")) {
/* <content> support */
if(NULL != (tmp = pie_parse_content_construct(cur))) {
item_set_description(ctxt->item, tmp);
g_free(tmp);
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"summary")) {
/* <summary> can be used for short text descriptions, if there is no
<content> description we show the <summary> content */
if(!item_get_description(ctxt->item)) {
if(NULL != (tmp = pie_parse_content_construct(cur))) {
item_set_description(ctxt->item, tmp);
g_free(tmp);
}
}
} else if(!xmlStrcmp(cur->name, BAD_CAST"copyright")) {
if(NULL != (tmp = (gchar *)xmlNodeListGetString(ctxt->doc, cur->xmlChildrenNode, 1))) {
ctxt->item->metadata = metadata_list_append(ctxt->item->metadata, "copyright", tmp);
g_free(tmp);
}
}
cur = cur->next;
}
/* after parsing we fill the infos into the itemPtr structure */
ctxt->item->readStatus = FALSE;
return ctxt->item;
}
/* 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;
}
}
