Add feed/reader mode with full-article capture at ingest

Reader mode presents the stories on a page one at a time in a focused
overlay, marking each read as it's shown. Left/right arrows (or the header
book button / `f`) page through them; read stories dim on the grid. Read
state is device-local in localStorage.

Backing this required actually capturing article bodies, which Pete wasn't
doing — it kept only the RSS <description> lede and discarded content:encoded:

- stories.content column (idempotent migration; old rows fall back to lede)
- parser keeps content:encoded as paragraph-preserving text
- article fetch already done for paywall detection now also returns its body,
  so ingest stores the richer of feed-content vs scraped body with no extra
  request (prefers the archive snapshot body for paywalled stories)
- GET /api/article?id= serves the stored text; card queries now select id and
  expose it as data-id for the reader

Tests cover content extraction, the storage round-trip, and the article
endpoint + card rendering end to end.

internal/ingestion/parser.go

@@ -31,6 +31,7 @@ type FeedItem struct {
 	GUID        string
 	Headline    string
 	Lede        string
+	Content     string // full article text from content:encoded, block structure preserved; "" when the feed ships only a snippet
 	ImageURL    string
 	ArticleURL  string
 	Source      string
@@ -43,8 +44,19 @@ var (
 	htmlTagRe = regexp.MustCompile(`<[^>]*>`)
 	wsRe      = regexp.MustCompile(`\s+`)
 	imgSrcRe  = regexp.MustCompile(`(?i)<img[^>]+src=["']([^"']+)["']`)
+
+	// Used by extractContentText to preserve paragraph structure when turning
+	// content:encoded HTML into plain text.
+	blockCloseRe  = regexp.MustCompile(`(?i)</(p|div|li|h[1-6]|blockquote|article|section|figure|figcaption|ul|ol|table|tr|pre)>`)
+	brRe          = regexp.MustCompile(`(?i)<br\s*/?>`)
+	hspaceRe      = regexp.MustCompile(`[ \t\f\r]+`)
+	manyNewlineRe = regexp.MustCompile(`\n{3,}`)
 )
 
+// maxContentChars bounds the stored article text. Generous enough for any real
+// article while keeping row size and the reader payload sane.
+const maxContentChars = 20000
+
 // FetchFeed fetches and parses an RSS/Atom feed, returning items. ua overrides
 // the User-Agent header; pass "" to use Pete's honest default bot UA.
 func FetchFeed(feedURL, ua string) ([]FeedItem, error) {
@@ -79,6 +91,7 @@ func FetchFeed(feedURL, ua string) ([]FeedItem, error) {
 			GUID:        itemGUID(item),
 			Headline:    strings.TrimSpace(item.Title),
 			Lede:        extractLede(item.Description),
+			Content:     extractContentText(item.Content),
 			ImageURL:    NormalizeImageURL(extractImageURL(item)),
 			ArticleURL:  strings.TrimSpace(item.Link),
 			Language:    itemLanguage(item),
@@ -142,6 +155,31 @@ func extractLede(desc string) string {
 	return strings.TrimSpace(text)
 }
 
+// extractContentText converts a feed's content:encoded HTML into plain text
+// with paragraph breaks preserved, for reader mode. Block-level boundaries
+// (</p>, </div>, <br>, list items, headings…) become newlines before the
+// remaining tags are stripped, so the reader can re-wrap the text into
+// paragraphs. Returns "" when the feed carried no content:encoded.
+func extractContentText(raw string) string {
+	if strings.TrimSpace(raw) == "" {
+		return ""
+	}
+	s := brRe.ReplaceAllString(raw, "\n")
+	s = blockCloseRe.ReplaceAllString(s, "\n\n")
+	s = htmlTagRe.ReplaceAllString(s, "")
+	s = html.UnescapeString(s)
+	lines := strings.Split(s, "\n")
+	for i, ln := range lines {
+		lines[i] = strings.TrimSpace(hspaceRe.ReplaceAllString(ln, " "))
+	}
+	s = manyNewlineRe.ReplaceAllString(strings.Join(lines, "\n"), "\n\n")
+	s = strings.TrimSpace(s)
+	if len(s) > maxContentChars {
+		s = s[:maxContentChars]
+	}
+	return s
+}
+
 // extractImageURL tries to find an image URL from feed item metadata.
 // Order: media:content/thumbnail, enclosures, item.Image, then <img> tags
 // scraped from content:encoded / description (catches feeds like Ars that