1
0
Fork 0
mirror of https://github.com/miniflux/v2.git synced 2025-06-27 16:36:00 +00:00

feat(rss): add workaround for RSS item title with HTML content

This commit is contained in:
Frédéric Guillot 2025-02-14 21:15:45 -08:00
parent 502db64b28
commit a3ce03cc9d
3 changed files with 32 additions and 3 deletions

View file

@ -173,13 +173,13 @@ func findFeedAuthor(rssChannel *RSSChannel) string {
}
func findEntryTitle(rssItem *RSSItem) string {
title := rssItem.Title
title := sanitizer.StripTags(rssItem.Title.Inner)
if rssItem.DublinCoreTitle != "" {
title = rssItem.DublinCoreTitle
}
return html.UnescapeString(strings.TrimSpace(title))
return html.UnescapeString(html.UnescapeString(strings.TrimSpace(title)))
}
func findEntryURL(rssItem *RSSItem) string {

View file

@ -1023,6 +1023,30 @@ func TestParseEntryTitleWithWhitespaces(t *testing.T) {
}
}
func TestParseEntryTitleWithInnerHTML(t *testing.T) {
data := `<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0">
<channel>
<title>Example</title>
<link>http://example.org</link>
<item>
<title>Test: <b>bold</b></title>
<link>http://www.example.org/entries/1</link>
<pubDate>Fri, 15 Jul 2005 00:00:00 -0500</pubDate>
</item>
</channel>
</rss>`
feed, err := Parse("https://example.org/", bytes.NewReader([]byte(data)))
if err != nil {
t.Fatal(err)
}
if feed.Entries[0].Title != "Test: bold" {
t.Errorf("Incorrect entry title, got: %s", feed.Entries[0].Title)
}
}
func TestParseEntryWithEnclosures(t *testing.T) {
data := `<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0">

View file

@ -111,7 +111,7 @@ type RSSImage struct {
type RSSItem struct {
// Title is the title of the item.
Title string `xml:"rss title"`
Title RSSTitle `xml:"rss title"`
// Link is the URL of the item.
Link string `xml:"rss link"`
@ -169,6 +169,11 @@ type RSSItem struct {
googleplay.GooglePlayItemElement
}
type RSSTitle struct {
Data string `xml:",chardata"`
Inner string `xml:",innerxml"`
}
type RSSAuthor struct {
XMLName xml.Name
Data string `xml:",chardata"`