fix(podcast): address CodeRabbit review feedback

- Remove total HTTP client timeout from episode downloads (use transport
  phase timeouts instead) and delete partial files on failure
- Keep SSE progress listener subscribed while the show view is mounted
- Return podcastEpisode (with streamId) from getNewestPodcasts
- Clean up dependent rows and MediaFiles in DeleteChannel within a tx
- Limit RSS feed body size to 32 MiB
- Block additional reserved IP ranges in the SSRF guard
- Fall back to enclosure URL when an item has no <guid>
- Handle clipboard write failure in the podcast list
This commit is contained in:
Jiho Andrew Lee 2026-09-30 15:42:14 +09:00
commit 4bebfad0ed
10 changed files with 126 additions and 52 deletions

View file

@ -316,8 +316,15 @@ func (s *podcastService) doDownload(ctx context.Context, ep *model.PodcastEpisod
s.setEpisodeError(ctx, ep, fmt.Errorf("invalid enclosure URL: %w", err))
return
}
httpClient := &http.Client{Timeout: 30 * time.Second, Transport: safeHTTPTransport}
resp, err := httpClient.Get(ep.EnclosureURL) //nolint:gosec
// No total Client.Timeout here: it would also cover reading the body and
// abort large episodes. Connection-phase limits live on safeHTTPTransport.
httpClient := &http.Client{Transport: safeHTTPTransport}
req, err := http.NewRequestWithContext(ctx, http.MethodGet, ep.EnclosureURL, nil)
if err != nil {
s.setEpisodeError(ctx, ep, err)
return
}
resp, err := httpClient.Do(req) //nolint:gosec
if err != nil {
s.setEpisodeError(ctx, ep, err)
return
@ -337,6 +344,8 @@ func (s *podcastService) doDownload(ctx context.Context, ep *model.PodcastEpisod
size, err := io.Copy(&progressWriter{ep: ep, ds: s.ds, broker: s.broker, ctx: ctx, w: f}, resp.Body)
if err != nil {
f.Close()
_ = os.Remove(dest) // don't leave a partial file behind
s.setEpisodeError(ctx, ep, err)
return
}
@ -499,7 +508,39 @@ func (s *podcastService) DeleteChannel(ctx context.Context, id string) error {
_ = os.Remove(ep.Path)
}
}
return s.ds.PodcastChannel().Delete(ctx, id)
// The channel-/episode-scoped detail tables have no FK cascade, and the
// registered MediaFiles are not FK-linked either, so clean them up here.
return s.ds.WithTx(func(tx model.DataStore) error {
for _, ep := range episodes {
if ep.StreamID != "" {
if err := tx.MediaFile().Delete(ctx, ep.StreamID); err != nil {
return err
}
}
if err := tx.PodcastPerson().SaveForEpisode(ctx, ep.ID, nil); err != nil {
return err
}
if err := tx.PodcastImage().SaveForEpisode(ctx, ep.ID, nil); err != nil {
return err
}
}
if err := tx.PodcastPerson().SaveForChannel(ctx, id, nil); err != nil {
return err
}
if err := tx.PodcastFunding().SaveForChannel(ctx, id, nil); err != nil {
return err
}
if err := tx.PodcastImage().SaveForChannel(ctx, id, nil); err != nil {
return err
}
if err := tx.PodcastPodroll().SaveForChannel(ctx, id, nil); err != nil {
return err
}
if err := tx.PodcastLiveItem().DeleteByChannel(ctx, id); err != nil {
return err
}
return tx.PodcastChannel().Delete(ctx, id)
})
}
// progressWriter wraps an io.Writer and periodically saves download progress to DB.
@ -571,7 +612,36 @@ var isReservedIP = func(ip net.IP) bool {
ip.IsLinkLocalMulticast() ||
ip.IsInterfaceLocalMulticast() ||
ip.IsMulticast() ||
ip.IsUnspecified()
ip.IsUnspecified() ||
inReservedNets(ip)
}
// reservedNets are special-purpose ranges that net.IP's helpers don't cover.
var reservedNets = func() []*net.IPNet {
var nets []*net.IPNet
for _, cidr := range []string{
"0.0.0.0/8", // "this" network
"100.64.0.0/10", // carrier-grade NAT
"192.0.0.0/24", // IETF protocol assignments
"198.18.0.0/15", // benchmarking
"64:ff9b::/96", // NAT64
} {
_, n, err := net.ParseCIDR(cidr)
if err != nil {
panic(err)
}
nets = append(nets, n)
}
return nets
}()
func inReservedNets(ip net.IP) bool {
for _, n := range reservedNets {
if n.Contains(ip) {
return true
}
}
return false
}
// AllowLoopbackHTTPForTests relaxes safeHTTPTransport's SSRF guard to permit
@ -597,6 +667,8 @@ func AllowLoopbackHTTPForTests() {
// check and the actual TCP connect (DNS rebinding) can't be used to reach
// a reserved address that validateURL alone would have caught.
var safeHTTPTransport = &http.Transport{
TLSHandshakeTimeout: 15 * time.Second,
ResponseHeaderTimeout: 30 * time.Second,
DialContext: func(ctx context.Context, network, addr string) (net.Conn, error) {
host, port, err := net.SplitHostPort(addr)
if err != nil {
@ -609,7 +681,7 @@ var safeHTTPTransport = &http.Transport{
if len(ips) == 0 {
return nil, fmt.Errorf("host %q did not resolve to any address", host)
}
var dialer net.Dialer
dialer := net.Dialer{Timeout: 15 * time.Second}
var lastErr error
for _, ip := range ips {
if isReservedIP(ip.IP) {
@ -626,6 +698,8 @@ var safeHTTPTransport = &http.Transport{
},
}
const maxFeedSize = 32 << 20 // 32 MiB
func fetchAndParse(rssURL string) (*rssFeed, error) {
if err := validateURL(rssURL); err != nil {
return nil, fmt.Errorf("invalid RSS feed URL: %w", err)
@ -637,10 +711,15 @@ func fetchAndParse(rssURL string) (*rssFeed, error) {
}
defer resp.Body.Close()
data, err := io.ReadAll(resp.Body)
// Read at most one byte past the limit so an oversized feed is detected
// without buffering it entirely in memory.
data, err := io.ReadAll(io.LimitReader(resp.Body, maxFeedSize+1))
if err != nil {
return nil, fmt.Errorf("reading RSS feed: %w", err)
}
if len(data) > maxFeedSize {
return nil, fmt.Errorf("RSS feed exceeds %d bytes", maxFeedSize)
}
return ParseRSSFeed(data)
}

View file

@ -236,15 +236,15 @@ type rssItem struct {
ItunesDur string `xml:"http://www.itunes.com/dtds/podcast-1.0.dtd duration"`
// Podcasting 2.0 episode tags
PodcastChapters podcastChapters `xml:"https://podcastindex.org/namespace/1.0 chapters"`
PodcastChapters podcastChapters `xml:"https://podcastindex.org/namespace/1.0 chapters"`
PodcastTranscripts []podcastTranscript `xml:"https://podcastindex.org/namespace/1.0 transcript"`
PodcastSeason podcastSeason `xml:"https://podcastindex.org/namespace/1.0 season"`
PodcastEpisodeNum podcastEpisodeNum `xml:"https://podcastindex.org/namespace/1.0 episode"`
PodcastSoundbite podcastSoundbite `xml:"https://podcastindex.org/namespace/1.0 soundbite"`
PodcastPersons []podcastPerson `xml:"https://podcastindex.org/namespace/1.0 person"`
PodcastLocation podcastLocation `xml:"https://podcastindex.org/namespace/1.0 location"`
PodcastLicense podcastLicense `xml:"https://podcastindex.org/namespace/1.0 license"`
PodcastImages []podcastImageTag `xml:"https://podcastindex.org/namespace/1.0 image"`
PodcastSeason podcastSeason `xml:"https://podcastindex.org/namespace/1.0 season"`
PodcastEpisodeNum podcastEpisodeNum `xml:"https://podcastindex.org/namespace/1.0 episode"`
PodcastSoundbite podcastSoundbite `xml:"https://podcastindex.org/namespace/1.0 soundbite"`
PodcastPersons []podcastPerson `xml:"https://podcastindex.org/namespace/1.0 person"`
PodcastLocation podcastLocation `xml:"https://podcastindex.org/namespace/1.0 location"`
PodcastLicense podcastLicense `xml:"https://podcastindex.org/namespace/1.0 license"`
PodcastImages []podcastImageTag `xml:"https://podcastindex.org/namespace/1.0 image"`
}
type enclosure struct {
@ -413,8 +413,14 @@ func ParseRSSFeed(data []byte) (*rssFeed, error) {
pubDate, _ := parseRSSDate(item.PubDate)
suffix := suffixFromMIME(item.Enclosure.Type, item.Enclosure.URL)
// <guid> is optional in RSS 2.0; fall back to the enclosure URL so
// episodes keep a unique identity for de-duplication on refresh.
guid := strings.TrimSpace(item.GUID)
if guid == "" {
guid = item.Enclosure.URL
}
ep := model.PodcastEpisode{
GUID: item.GUID,
GUID: guid,
Title: item.Title,
Description: desc,
PublishDate: pubDate,

View file

@ -77,6 +77,15 @@ var _ = Describe("ParseRSSFeed", func() {
Expect(feed.Episodes).To(HaveLen(2))
})
It("falls back to the enclosure URL when <guid> is missing", func() {
feed, err := podcasts.ParseRSSFeed([]byte(`<?xml version="1.0"?><rss version="2.0"><channel><title>T</title>
<item><title>No guid</title><enclosure url="https://example.com/noguid.mp3" length="1" type="audio/mpeg"/></item>
</channel></rss>`))
Expect(err).ToNot(HaveOccurred())
Expect(feed.Episodes).To(HaveLen(1))
Expect(feed.Episodes[0].GUID).To(Equal("https://example.com/noguid.mp3"))
})
It("parses episode fields correctly", func() {
feed, _ := podcasts.ParseRSSFeed([]byte(testRSSFeed))
ep := feed.Episodes[0]