Seek 5s into videos for thumbnail extraction

The 0.5s seek often landed on the initial keyframe or a fade-in,
producing blocky thumbnails. Aim ~5s in instead, with an ffprobe
midpoint fallback for clips shorter than that and a 2s → 0
retry chain if the seek still overshoots.

internal/config/config.go

@@ -70,6 +70,10 @@ type SourceConfig struct {
 	PollIntervalMinutes int    `toml:"poll_interval_minutes"`
 	DirectRoute         string `toml:"direct_route"`
 	Enabled             bool   `toml:"enabled"`
+	// Language, when set, drops feed items whose per-item language tag is
+	// present and does not match (prefix). Useful for multilingual feeds
+	// like Politico Europe that publish English + French side-by-side.
+	Language            string `toml:"language"`
 }
 
 func Load(path string) (*Config, error) {

internal/ingestion/parser.go

@@ -28,11 +28,13 @@ type FeedItem struct {
 	ArticleURL  string
 	Source      string
 	DirectRoute string
+	Language    string // per-item <language> (or dc:language) when present; "" otherwise
 	PublishedAt int64 // unix seconds from RSS pubDate / Atom published (or updated as fallback); 0 if absent or unparseable
 }
 
 var (
 	htmlTagRe = regexp.MustCompile(`<[^>]*>`)
+	wsRe      = regexp.MustCompile(`\s+`)
 	imgSrcRe  = regexp.MustCompile(`(?i)<img[^>]+src=["']([^"']+)["']`)
 )
 
@@ -59,6 +61,7 @@ func FetchFeed(feedURL string) ([]FeedItem, error) {
 			Lede:        extractLede(item.Description),
 			ImageURL:    NormalizeImageURL(extractImageURL(item)),
 			ArticleURL:  strings.TrimSpace(item.Link),
+			Language:    itemLanguage(item),
 			PublishedAt: itemPublished(item),
 		}
 		if fi.GUID == "" || fi.ArticleURL == "" {
@@ -71,6 +74,22 @@ func FetchFeed(feedURL string) ([]FeedItem, error) {
 	return items, nil
 }
 
+// itemLanguage returns the per-item language tag if the feed provides one.
+// Politico Europe puts a raw <language> inside each <item>; gofeed parks
+// unknown item-level elements in item.Custom. We also check the standard
+// dc:language extension for feeds that use Dublin Core.
+func itemLanguage(item *gofeed.Item) string {
+	if item.Custom != nil {
+		if v, ok := item.Custom["language"]; ok {
+			return strings.ToLower(strings.TrimSpace(v))
+		}
+	}
+	if item.DublinCoreExt != nil && len(item.DublinCoreExt.Language) > 0 {
+		return strings.ToLower(strings.TrimSpace(item.DublinCoreExt.Language[0]))
+	}
+	return ""
+}
+
 func itemGUID(item *gofeed.Item) string {
 	if item.GUID != "" {
 		return item.GUID
@@ -96,8 +115,10 @@ func extractLede(desc string) string {
 	if desc == "" {
 		return ""
 	}
-	text := htmlTagRe.ReplaceAllString(desc, "")
+	// Replace tags with a space so adjacent blocks like </p><p> don't fuse words.
+	text := htmlTagRe.ReplaceAllString(desc, " ")
 	text = html.UnescapeString(text)
+	text = wsRe.ReplaceAllString(text, " ")
 	return strings.TrimSpace(text)
 }
 

internal/ingestion/parser_test.go

@@ -48,6 +48,11 @@ func TestExtractLede(t *testing.T) {
 			"It's a new era—one of change.",
 			"It's a new era\u2014one of change.",
 		},
+		{
+			"adjacent block tags inject a space",
+			"<p>...end of moments</p><p>Clarence B Jones, a former...</p>",
+			"...end of moments Clarence B Jones, a former...",
+		},
 		{
 			"tags and entities combined",
 			"<p>Oil prices rose &amp; gas fell by &gt;5%.</p>",

internal/ingestion/poller.go

@@ -3,6 +3,7 @@ package ingestion
 import (
 	"context"
 	"log/slog"
+	"strings"
 	"sync"
 	"time"
 
@@ -109,6 +110,17 @@ func (p *Poller) pollOnceWithErr(ctx context.Context, src config.SourceConfig) e
 			continue
 		}
 
+		// Drop items in languages this source isn't configured to surface.
+		// Items without a language tag pass through — only filter when the
+		// feed explicitly marks an item with a non-matching language.
+		if src.Language != "" && items[i].Language != "" &&
+			!strings.HasPrefix(items[i].Language, strings.ToLower(src.Language)) {
+			slog.Info("dropping story (language filter)",
+				"guid", items[i].GUID, "source", src.Name,
+				"item_lang", items[i].Language, "want", src.Language)
+			continue
+		}
+
 		// Dedup checks first — canonical and headline use the ORIGINAL URL so
 		// they stay stable whether we end up swapping to a Wayback snapshot.
 		originalURL := items[i].ArticleURL

internal/web/static/js/settings.js

@@ -69,7 +69,7 @@
       list.innerHTML = '<p class="px-2 py-4 text-sm text-center text-[color:var(--ink)]/50">No feeds configured.</p>';
     } else {
       var lastChannel = null;
-      sources.forEach(function (src) {
+      sources.forEach(function (src, i) {
         if (src.channel !== lastChannel) {
           lastChannel = src.channel;
           var hdr = document.createElement("div");
@@ -77,7 +77,7 @@
           hdr.textContent = src.channel || "other";
           list.appendChild(hdr);
         }
-        var id = "pete-feed-" + btoa(src.name).replace(/=/g, "");
+        var id = "pete-feed-" + i;
         var row = document.createElement("label");
         row.setAttribute("for", id);
         row.className = "flex items-center justify-between gap-3 rounded-2xl px-3 py-2 hover:bg-[color:var(--ink)]/5 cursor-pointer";

internal/web/thumbs.go

@@ -17,6 +17,7 @@ import (
 	"os"
 	"os/exec"
 	"path/filepath"
+	"strconv"
 	"strings"
 	"sync"
 	"time"
@@ -305,15 +306,49 @@ func buildAvifFromFFmpeg(body []byte, outPath string) error {
 	outPNG.Close()
 	defer os.Remove(pngPath)
 
-	if err := extractVideoFrame(inPath, pngPath, "0.5"); err != nil {
-		// Some clips are shorter than the seek offset; retry from frame 0.
-		if err := extractVideoFrame(inPath, pngPath, "0"); err != nil {
-			return err
+	// Aim ~5s in — past the typical fade-in / first-GOP blockiness while
+	// still likely within the clip. For shorter clips we fall back to the
+	// midpoint (when ffprobe is available) so we don't seek off the end.
+	seek := "5"
+	if d := probeVideoDuration(inPath); d > 0 && d < 5 {
+		seek = strconv.FormatFloat(d/2, 'f', 2, 64)
+	}
+	if err := extractVideoFrame(inPath, pngPath, seek); err != nil {
+		// Seek overshot or container's funky — try 2s, then frame 0.
+		if err := extractVideoFrame(inPath, pngPath, "2"); err != nil {
+			if err := extractVideoFrame(inPath, pngPath, "0"); err != nil {
+				return err
+			}
 		}
 	}
 	return avifEncodeFromPNG(pngPath, outPath)
 }
 
+// probeVideoDuration returns the clip length in seconds, or 0 if ffprobe is
+// missing or can't read it (image-as-video fallback, malformed containers).
+func probeVideoDuration(inPath string) float64 {
+	if _, err := exec.LookPath("ffprobe"); err != nil {
+		return 0
+	}
+	ctx, cancel := context.WithTimeout(context.Background(), ffmpegTimeout)
+	defer cancel()
+	cmd := exec.CommandContext(ctx, "ffprobe",
+		"-v", "error",
+		"-show_entries", "format=duration",
+		"-of", "default=noprint_wrappers=1:nokey=1",
+		inPath,
+	)
+	out, err := cmd.Output()
+	if err != nil {
+		return 0
+	}
+	d, err := strconv.ParseFloat(strings.TrimSpace(string(out)), 64)
+	if err != nil || d <= 0 {
+		return 0
+	}
+	return d
+}
+
 // extractVideoFrame asks ffmpeg to seek to seekSec, grab one frame, and
 // downscale it to at most thumbWidth wide before writing pngPath.
 func extractVideoFrame(inPath, pngPath, seekSec string) error {