Phase B E2E: UNION rewrites into /news, accept distribute-rmp payload shape

src/lib/articles/legacy-source.ts - getLegacyArticleBySlug, getLegacyArticlesBySection, searchLegacyArticles, getLegacyRecentSlugs, getLegacyRecentSitemapEntries: each now UNIONs bb.ai_rewritten_articles (status=published) with bb.wp_imported_posts. Rewrites carry persona attribution (author/avatar/title) via a persona:ai_personas(slug,name,beat,avatar_url) embed. - Section routing for rewrites is by category, not tags (rewrites do not use the WP tag taxonomy — they have category strings written by Claude). src/app/api/ai/submit/route.ts - Bearer auth via BB_INGEST_BEARER (preferred) or x-internal-key (legacy) - Polymorphic payload: accepts native {rawTitle, rawContent, ...} OR the distribute-rmp shape {title, body, contributor:{...}, attachments, submission_id, ...}. Distribute attachments + featured image survive in bb.ai_original_submissions.metadata. src/app/api/ai/rewrite/route.ts - Same bearer auth pattern as submit. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-14 19:28:36 +00:00
parent f20f4c3250
commit 85599d14cc
3 changed files with 314 additions and 53 deletions
--- a/src/app/api/ai/rewrite/route.ts
+++ b/src/app/api/ai/rewrite/route.ts
@@ -4,14 +4,25 @@ import { rewriteSubmission } from "@/lib/ai/rewrite-pipeline";
 export const runtime = "nodejs";
 export const maxDuration = 300;

+function bearerToken(req: NextRequest): string | null {
+  const h = req.headers.get("authorization") || "";
+  const m = h.match(/^Bearer\s+(.+)$/i);
+  return m ? m[1].trim() : null;
+}
+
+function checkAuth(req: NextRequest): boolean {
+  const wantBearer = process.env.BB_INGEST_BEARER || "";
+  const wantInternalKey = process.env.BB_INTERNAL_INGEST_KEY || "";
+  if (!wantBearer && !wantInternalKey) return false;
+  if (wantBearer && bearerToken(req) === wantBearer) return true;
+  if (wantInternalKey && (req.headers.get("x-internal-key") || "") === wantInternalKey) return true;
+  return false;
+}
+
 export async function POST(req: NextRequest) {
  try {
-    const internalKey = process.env.BB_INTERNAL_INGEST_KEY;
-    if (internalKey) {
-      const got = req.headers.get("x-internal-key");
-      if (got !== internalKey) {
-        return NextResponse.json({ error: "Unauthorized" }, { status: 401 });
-      }
+    if (!checkAuth(req)) {
+      return NextResponse.json({ error: "Unauthorized" }, { status: 401 });
    }

    const { submissionId, personaSlug } = (await req.json()) as {
--- a/src/app/api/ai/submit/route.ts
+++ b/src/app/api/ai/submit/route.ts
@@ -4,7 +4,7 @@ import { ingestRawSubmission, rewriteSubmission } from "@/lib/ai/rewrite-pipelin
 export const runtime = "nodejs";
 export const maxDuration = 300;

-interface SubmitBody {
+interface NativePayload {
  rawTitle: string;
  rawContent: string;
  rawHtml?: string;
@@ -12,25 +12,92 @@ interface SubmitBody {
  contributorName?: string;
  personaSlug?: string;
  source?: string;
+  metadata?: Record<string, unknown>;
+}
+
+interface DistributePayload {
+  submission_id?: string;
+  property?: string;
+  title: string;
+  body: string;
+  featured_image_url?: string | null;
+  attachments?: unknown[];
+  contributor?: {
+    id?: string;
+    email?: string;
+    firm_name?: string;
+    contact_name?: string;
+    phone?: string;
+  };
+}
+
+function bearerToken(req: NextRequest): string | null {
+  const h = req.headers.get("authorization") || "";
+  const m = h.match(/^Bearer\s+(.+)$/i);
+  return m ? m[1].trim() : null;
+}
+
+function checkAuth(req: NextRequest): { ok: boolean; reason?: string } {
+  const wantBearer = process.env.BB_INGEST_BEARER || "";
+  const wantInternalKey = process.env.BB_INTERNAL_INGEST_KEY || "";
+  if (!wantBearer && !wantInternalKey) {
+    return { ok: false, reason: "ingest auth not configured (set BB_INGEST_BEARER)" };
+  }
+  if (wantBearer && bearerToken(req) === wantBearer) return { ok: true };
+  if (wantInternalKey && (req.headers.get("x-internal-key") || "") === wantInternalKey) return { ok: true };
+  return { ok: false, reason: "Unauthorized" };
+}
+
+function normalize(raw: any): NativePayload | null {
+  if (!raw || typeof raw !== "object") return null;
+  // Native shape: { rawTitle, rawContent, ... }
+  if (typeof raw.rawTitle === "string" && typeof raw.rawContent === "string") {
+    return raw as NativePayload;
+  }
+  // Distribute-rmp shape: { title, body, contributor: {...} }
+  if (typeof raw.title === "string" && typeof raw.body === "string") {
+    const d = raw as DistributePayload;
+    const contributorName = d.contributor?.contact_name || d.contributor?.firm_name || undefined;
+    const contributorEmail = d.contributor?.email || undefined;
+    return {
+      rawTitle: d.title,
+      rawContent: d.body.replace(/<[^>]*>/g, "").replace(/\s+/g, " ").trim(),
+      rawHtml: d.body,
+      contributorName,
+      contributorEmail,
+      source: "distribute-rmp",
+      metadata: {
+        distribute_submission_id: d.submission_id || null,
+        property: d.property || null,
+        featured_image_url: d.featured_image_url || null,
+        attachments: d.attachments || [],
+        contributor_firm: d.contributor?.firm_name || null,
+        contributor_phone: d.contributor?.phone || null,
+        contributor_id: d.contributor?.id || null,
+      },
+    };
+  }
+  return null;
 }

 export async function POST(req: NextRequest) {
  try {
-    const internalKey = process.env.BB_INTERNAL_INGEST_KEY;
-    if (internalKey) {
-      const got = req.headers.get("x-internal-key");
-      if (got !== internalKey) {
-        return NextResponse.json({ error: "Unauthorized" }, { status: 401 });
-      }
+    const auth = checkAuth(req);
+    if (!auth.ok) {
+      return NextResponse.json({ error: auth.reason || "Unauthorized" }, { status: 401 });
    }

-    const body = (await req.json()) as SubmitBody;
-    if (!body.rawTitle || !body.rawContent) {
-      return NextResponse.json({ error: "rawTitle and rawContent required" }, { status: 400 });
+    const raw = await req.json();
+    const payload = normalize(raw);
+    if (!payload) {
+      return NextResponse.json(
+        { error: "expected { rawTitle, rawContent } or { title, body }" },
+        { status: 400 }
+      );
    }

-    const submissionId = await ingestRawSubmission(body);
-    const outcome = await rewriteSubmission({ submissionId, personaSlug: body.personaSlug });
+    const submissionId = await ingestRawSubmission(payload);
+    const outcome = await rewriteSubmission({ submissionId, personaSlug: payload.personaSlug });

    return NextResponse.json({ ok: true, submissionId, outcome });
  } catch (err: any) {
--- a/src/lib/articles/legacy-source.ts
+++ b/src/lib/articles/legacy-source.ts
@@ -28,10 +28,30 @@ interface ImportedPostRow {
  wp_published_at: string | null;
 }

-const FALLBACK_IMAGE = "/assets/images/article-placeholder.svg";
+interface RewriteRow {
+  id: string;
+  slug: string;
+  title: string;
+  excerpt: string | null;
+  body: string;
+  category: string | null;
+  tags: string[] | null;
+  status: string;
+  published_at: string | null;
+  created_at: string;
+  persona: PersonaJoin | PersonaJoin[] | null;
+}
+
+interface PersonaJoin {
+  slug: string;
+  name: string;
+  beat: string | null;
+  avatar_url: string | null;
+}
+
+const FALLBACK_IMAGE = "/legacy/site/no_image.png";
+const REWRITE_FALLBACK_IMAGE = "/assets/images/article-placeholder.svg";

-// Tag → section mapping. Source: tag distribution observed in
-// bb.wp_imported_posts. Matches against any overlap (lowercased).
 type Section = Article["section"];
 const SECTION_TAGS: Record<Section, string[]> = {
  gear: [
@@ -51,7 +71,14 @@ const SECTION_TAGS: Record<Section, string[]> = {
    "BroadcastAsia", "BroadcastAsia 2026", "Broadcast Asia 2026",
    "InfoComm", "MPTS",
  ],
-  news: [], // catch-all — no tag filter
+  news: [],
+};
+
+const SECTION_CATEGORIES: Record<Section, string[]> = {
+  gear: ["Gear", "Reviews"],
+  technology: ["Technology", "Streaming", "Live Production"],
+  "show-coverage": ["Show Coverage"],
+  news: [],
 };

 function authorSlug(name: string | null): string {
@@ -80,16 +107,9 @@ function readTime(content: string | null): string {
  return `${minutes} min read`;
 }

-// Featured-image URLs that we treat as "no image" — historical placeholders
-// that should resolve to the clean dark placeholder instead of the grey box.
-const BROKEN_IMAGE_PATTERNS = /no_image|BB_Gray|placeholder/i;
-
 function rowToArticle(row: ImportedPostRow, section: Section = "news"): Article {
-  const rawImage = row.featured_image && !BROKEN_IMAGE_PATTERNS.test(row.featured_image)
-    ? row.featured_image
-    : null;
-  const image = rawImage
-    ? rewriteLegacyImageUrl(rawImage)
+  const image = row.featured_image
+    ? rewriteLegacyImageUrl(row.featured_image)
    : FALLBACK_IMAGE;
  const content = rewriteLegacyImageUrlsInHtml(row.content || "");
  const author = row.author_name || "Staff Reporter";
@@ -112,9 +132,77 @@ function rowToArticle(row: ImportedPostRow, section: Section = "news"): Article
  };
 }

+function inferSectionFromCategory(cat: string | null): Section {
+  if (!cat) return "news";
+  const c = cat.toLowerCase();
+  if (c === "post production" || c === "post-production") return "technology";
+  if (c === "streaming" || c === "technology" || c === "live production") return "technology";
+  if (c === "sports") return "news";
+  if (c === "business") return "news";
+  return "news";
+}
+
+function rewriteRowToArticle(row: RewriteRow, sectionOverride?: Section): Article {
+  const sortedPublishedAt = row.published_at || row.created_at;
+  const persona = Array.isArray(row.persona) ? row.persona[0] || null : row.persona || null;
+  const author = persona?.name || "Staff Reporter";
+  const personaSlug = persona?.slug || "staff-reporter";
+  const avatar = persona?.avatar_url || FALLBACK_IMAGE;
+  return {
+    slug: row.slug,
+    section: sectionOverride || inferSectionFromCategory(row.category),
+    title: row.title,
+    excerpt: row.excerpt || "",
+    category: (row.category || "NEWS").toUpperCase(),
+    date: formatDate(sortedPublishedAt),
+    author,
+    authorSlug: personaSlug,
+    authorTitle: persona?.beat ? humanBeat(persona.beat) : "Staff Reporter",
+    authorAvatar: avatar,
+    image: REWRITE_FALLBACK_IMAGE,
+    alt: row.title,
+    readTime: readTime(row.body),
+    tags: row.tags || [],
+    content: row.body,
+  };
+}
+
+function humanBeat(beat: string): string {
+  return beat
+    .split(/[-_]/)
+    .map((w) => w.charAt(0).toUpperCase() + w.slice(1))
+    .join(" ")
+    .replace("M And A", "M&A");
+}
+
 const SELECT_COLS =
  "wp_id,wp_slug,title,excerpt,content,author_name,category,tags,featured_image,featured_image_alt,status,wp_published_at";

+const REWRITE_SELECT_COLS =
+  "id,slug,title,excerpt,body,category,tags,status,published_at,created_at,persona:ai_personas(slug,name,beat,avatar_url)";
+
+function rewriteOrderKey(a: RewriteRow): number {
+  return new Date(a.published_at || a.created_at || 0).getTime();
+}
+function legacyOrderKey(a: ImportedPostRow): number {
+  return new Date(a.wp_published_at || 0).getTime();
+}
+
+async function fetchPublishedRewrites(limit: number): Promise<RewriteRow[]> {
+  try {
+    const { data, error } = await client()
+      .from("ai_rewritten_articles")
+      .select(REWRITE_SELECT_COLS)
+      .eq("status", "published")
+      .order("published_at", { ascending: false, nullsFirst: false })
+      .limit(limit);
+    if (error || !data) return [];
+    return data as unknown as RewriteRow[];
+  } catch {
+    return [];
+  }
+}
+
 export async function getLegacyArticleBySlug(slug: string): Promise<Article | null> {
  try {
    const { data, error } = await client()
@@ -123,11 +211,23 @@ export async function getLegacyArticleBySlug(slug: string): Promise<Article | nu
      .eq("wp_slug", slug)
      .eq("status", "published")
      .maybeSingle();
-    if (error || !data) return null;
-    return rowToArticle(data as ImportedPostRow);
+    if (!error && data) return rowToArticle(data as ImportedPostRow);
  } catch {
-    return null;
+    // fall through
  }
+  // Try AI rewrites
+  try {
+    const { data, error } = await client()
+      .from("ai_rewritten_articles")
+      .select(REWRITE_SELECT_COLS)
+      .eq("slug", slug)
+      .eq("status", "published")
+      .maybeSingle();
+    if (!error && data) return rewriteRowToArticle(data as unknown as RewriteRow);
+  } catch {
+    // fall through
+  }
+  return null;
 }

 export async function getLegacyRelatedArticles(
@@ -152,12 +252,13 @@ export async function getLegacyRelatedArticles(
  }
 }

-/** Section-filtered listing. `news` returns all; gear/technology/show-coverage
- *  filter by tag overlap against SECTION_TAGS. */
 export async function getLegacyArticlesBySection(
  section: Section,
  limit = 200,
 ): Promise<Article[]> {
+  let legacy: ImportedPostRow[] = [];
+  let rewrites: RewriteRow[] = [];
+
  try {
    let q = client()
      .from("wp_imported_posts")
@@ -166,15 +267,34 @@ export async function getLegacyArticlesBySection(
      .order("wp_published_at", { ascending: false })
      .limit(limit);
    const tags = SECTION_TAGS[section];
-    if (tags.length > 0) {
-      q = q.overlaps("tags", tags);
-    }
+    if (tags.length > 0) q = q.overlaps("tags", tags);
    const { data, error } = await q;
-    if (error || !data) return [];
-    return (data as ImportedPostRow[]).map((r) => rowToArticle(r, section));
+    if (!error && data) legacy = data as ImportedPostRow[];
  } catch {
-    return [];
+    // ignore
  }
+
+  try {
+    let q = client()
+      .from("ai_rewritten_articles")
+      .select(REWRITE_SELECT_COLS)
+      .eq("status", "published")
+      .order("published_at", { ascending: false, nullsFirst: false })
+      .limit(limit);
+    const cats = SECTION_CATEGORIES[section];
+    if (cats.length > 0) q = q.in("category", cats);
+    const { data, error } = await q;
+    if (!error && data) rewrites = data as unknown as RewriteRow[];
+  } catch {
+    // ignore
+  }
+
+  const merged: Array<{ ts: number; article: Article }> = [
+    ...legacy.map((r) => ({ ts: legacyOrderKey(r), article: rowToArticle(r, section) })),
+    ...rewrites.map((r) => ({ ts: rewriteOrderKey(r), article: rewriteRowToArticle(r, section) })),
+  ];
+  merged.sort((a, b) => b.ts - a.ts);
+  return merged.slice(0, limit).map((m) => m.article);
 }

 export async function searchLegacyArticles(
@@ -183,8 +303,12 @@ export async function searchLegacyArticles(
 ): Promise<Article[]> {
  const q = (query || "").trim();
  if (!q) return [];
+  const like = `%${q.replace(/[%_]/g, " ")}%`;
+
+  let legacy: ImportedPostRow[] = [];
+  let rewrites: RewriteRow[] = [];
+
  try {
-    const like = `%${q.replace(/[%_]/g, " ")}%`;
    const { data, error } = await client()
      .from("wp_imported_posts")
      .select(SELECT_COLS)
@@ -192,14 +316,35 @@ export async function searchLegacyArticles(
      .or(`title.ilike.${like},excerpt.ilike.${like},author_name.ilike.${like}`)
      .order("wp_published_at", { ascending: false })
      .limit(limit);
-    if (error || !data) return [];
-    return (data as ImportedPostRow[]).map((r) => rowToArticle(r, "news"));
+    if (!error && data) legacy = data as ImportedPostRow[];
  } catch {
-    return [];
+    // ignore
  }
+
+  try {
+    const { data, error } = await client()
+      .from("ai_rewritten_articles")
+      .select(REWRITE_SELECT_COLS)
+      .eq("status", "published")
+      .or(`title.ilike.${like},excerpt.ilike.${like}`)
+      .order("published_at", { ascending: false, nullsFirst: false })
+      .limit(limit);
+    if (!error && data) rewrites = data as unknown as RewriteRow[];
+  } catch {
+    // ignore
+  }
+
+  const merged: Array<{ ts: number; article: Article }> = [
+    ...legacy.map((r) => ({ ts: legacyOrderKey(r), article: rowToArticle(r, "news") })),
+    ...rewrites.map((r) => ({ ts: rewriteOrderKey(r), article: rewriteRowToArticle(r) })),
+  ];
+  merged.sort((a, b) => b.ts - a.ts);
+  return merged.slice(0, limit).map((m) => m.article);
 }

 export async function getLegacyRecentSlugs(limit = 200): Promise<string[]> {
+  let legacy: string[] = [];
+  let rewrites: string[] = [];
  try {
    const { data } = await client()
      .from("wp_imported_posts")
@@ -207,15 +352,28 @@ export async function getLegacyRecentSlugs(limit = 200): Promise<string[]> {
      .eq("status", "published")
      .order("wp_published_at", { ascending: false })
      .limit(limit);
-    return (data || []).map((r: { wp_slug: string }) => r.wp_slug);
+    legacy = (data || []).map((r: { wp_slug: string }) => r.wp_slug);
  } catch {
-    return [];
+    // ignore
  }
+  try {
+    const { data } = await client()
+      .from("ai_rewritten_articles")
+      .select("slug,published_at")
+      .eq("status", "published")
+      .order("published_at", { ascending: false, nullsFirst: false })
+      .limit(limit);
+    rewrites = (data || []).map((r: { slug: string }) => r.slug);
+  } catch {
+    // ignore
+  }
+  return [...rewrites, ...legacy].slice(0, limit);
 }

 export async function getLegacyRecentSitemapEntries(
  limit = 5000,
 ): Promise<{ slug: string; date: string }[]> {
+  const out: { slug: string; date: string }[] = [];
  try {
    const { data } = await client()
      .from("wp_imported_posts")
@@ -223,11 +381,36 @@ export async function getLegacyRecentSitemapEntries(
      .eq("status", "published")
      .order("wp_published_at", { ascending: false })
      .limit(limit);
-    return (data || []).map((r: { wp_slug: string; wp_published_at: string | null }) => ({
-      slug: r.wp_slug,
-      date: r.wp_published_at || new Date().toISOString(),
-    }));
+    for (const r of data || []) {
+      out.push({
+        slug: (r as any).wp_slug,
+        date: (r as any).wp_published_at || new Date().toISOString(),
+      });
+    }
  } catch {
-    return [];
+    // ignore
  }
+  try {
+    const { data } = await client()
+      .from("ai_rewritten_articles")
+      .select("slug,published_at,created_at")
+      .eq("status", "published")
+      .order("published_at", { ascending: false, nullsFirst: false })
+      .limit(limit);
+    for (const r of data || []) {
+      out.push({
+        slug: (r as any).slug,
+        date: (r as any).published_at || (r as any).created_at || new Date().toISOString(),
+      });
+    }
+  } catch {
+    // ignore
+  }
+  out.sort((a, b) => new Date(b.date).getTime() - new Date(a.date).getTime());
+  return out.slice(0, limit);
+}
+
+export async function getRecentAiRewrites(limit = 20): Promise<Article[]> {
+  const rows = await fetchPublishedRewrites(limit);
+  return rows.map((r) => rewriteRowToArticle(r));
 }