From c4d591b96254abb0c6deadde8f1b00851a4455d8 Mon Sep 17 00:00:00 2001
From: Julien Herr <julien@herr.fr>
Date: Mon, 25 May 2026 09:04:36 +0200
Subject: [PATCH] feat(ingest): detect and mark confirmation emails

---
 src/application/email-processor.test.ts | 54 +++++++++++++++++++++++++
 src/application/email-processor.ts      | 14 ++++++-
 2 files changed, 67 insertions(+), 1 deletion(-)

diff --git a/src/application/email-processor.test.ts b/src/application/email-processor.test.ts
index 1462953..66dffc1 100644
--- a/src/application/email-processor.test.ts
+++ b/src/application/email-processor.test.ts
@@ -825,6 +825,60 @@ describe("processEmail — feed icon", () => {
   });
 });
 
+describe("processEmail — confirmation detection", () => {
+  let env: ReturnType<typeof createMockEnv>;
+
+  beforeEach(async () => {
+    env = createMockEnv();
+    await env.EMAIL_STORAGE.put(
+      `feed:${VALID_FEED_ID}:config`,
+      JSON.stringify({}),
+    );
+    await seedInboundIndex(env, VALID_FEED_ID);
+  });
+
+  it("marks a confirmation email and raises pendingConfirmation", async () => {
+    const result = await processEmail(
+      makeInput({
+        subject: "Please confirm your subscription",
+        content:
+          '<p>Click <a href="https://example.com/confirm?token=abc">Confirm</a></p>',
+      }),
+      env as any,
+    );
+
+    expect(result.ok).toBe(true);
+
+    const metadata = await env.EMAIL_STORAGE.get(
+      `feed:${VALID_FEED_ID}:metadata`,
+      "json",
+    );
+    expect(metadata.pendingConfirmation).toBe(true);
+    expect(metadata.emails[0].confirmation?.links[0]).toBe(
+      "https://example.com/confirm?token=abc",
+    );
+  });
+
+  it("does not mark a regular newsletter as a confirmation", async () => {
+    const result = await processEmail(
+      makeInput({
+        subject: "Weekly Newsletter",
+        content: "<p>Here is your weekly digest of news.</p>",
+      }),
+      env as any,
+    );
+
+    expect(result.ok).toBe(true);
+
+    const metadata = await env.EMAIL_STORAGE.get(
+      `feed:${VALID_FEED_ID}:metadata`,
+      "json",
+    );
+    expect(metadata.pendingConfirmation).toBeFalsy();
+    expect(metadata.emails[0].confirmation).toBeUndefined();
+  });
+});
+
 describe("processEmail — unsubscribe capture", () => {
   let env: ReturnType<typeof createMockEnv>;
 
diff --git a/src/application/email-processor.ts b/src/application/email-processor.ts
index 2afee5a..78fa867 100644
--- a/src/application/email-processor.ts
+++ b/src/application/email-processor.ts
@@ -5,7 +5,12 @@ import { dispatchFeedEvents } from "../application/feed-events";
 import { extractEmailDomain } from "../infrastructure/favicon-fetcher";
 import { parseOneClickUnsubscribe } from "../infrastructure/unsubscribe";
 import { getAttachmentBucket } from "../infrastructure/attachments";
-import { extractInlineCids } from "../infrastructure/html-processor";
+import {
+  extractInlineCids,
+  extractLinks,
+  htmlToText,
+} from "../infrastructure/html-processor";
+import { detectConfirmation } from "../domain/confirmation";
 import { attachmentIdsForCleanup } from "./feed-cleanup";
 import { FeedRepository } from "../infrastructure/feed-repository";
 import { BackgroundScheduler } from "../infrastructure/worker";
@@ -182,6 +187,12 @@ async function storeEmail(
     return false; // signal: skipped (not stored)
   }
 
+  const confirmation = detectConfirmation({
+    subject: input.subject,
+    text: htmlToText(input.content),
+    links: extractLinks(input.content),
+  });
+
   const attachmentBucket = getAttachmentBucket(env);
   const inlineCids = extractInlineCids(input.content);
   const storedAttachments: AttachmentData[] =
@@ -218,6 +229,7 @@ async function storeEmail(
     ...(inlineIds.length > 0 ? { inlineAttachmentIds: inlineIds } : {}),
     ...(messageId ? { messageId } : {}),
     dedupHash,
+    ...(confirmation ? { confirmation: { links: confirmation.links } } : {}),
   };
 
   // Track the latest sender's domain (feed icon) and capture the RFC 8058