test(14-03): add failing spec for email-alert generic link/subject extraction

RED phase (TDD) for Task 1: pure-function tests for extractCandidateLinks, titleFromEmail, sourceNoticeIdFor — covers HTML + plaintext bodies, footer-noise removal, link cap, and D-04 no-portal-specific-parser restraint. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-23 13:37:37 +02:00
parent 6e3789a0b0
commit 4d6fbb136e
1 changed files with 134 additions and 0 deletions
@@ -0,0 +1,134 @@
+import { readFileSync } from 'fs';
+import { join } from 'path';
+import { describe, expect, it } from 'vitest';
+import {
+  EmailAlertAdapter,
+  extractCandidateLinks,
+  sourceNoticeIdFor,
+  titleFromEmail,
+} from './email-alert.adapter';
+
+/**
+ * email-alert.adapter.spec — Task 1 (test-first, TDD) proof for the generic
+ * link/subject extraction (D-04): pure functions only, no I/O, mirroring
+ * cosinex.adapter.spec.ts's pure-function spec style.
+ */
+describe('EmailAlertAdapter', () => {
+  it('declares sourceType email-alert and the single email-alert portal', () => {
+    const adapter = new EmailAlertAdapter();
+    expect(adapter.sourceType).toBe('email-alert');
+    expect(adapter.portals).toEqual(['email-alert']);
+  });
+
+  describe('extractCandidateLinks (pure)', () => {
+    it('extracts absolute http(s) hrefs from an HTML body via cheerio', () => {
+      const html = `
+        <html><body>
+          <p>Neue Ausschreibung verfügbar:</p>
+          <a href="https://portal.example/detail/123">Zur Ausschreibung</a>
+          <a href="https://portal.example/detail/456">Weiterer Link</a>
+        </body></html>
+      `;
+      const links = extractCandidateLinks(html, '');
+      expect(links).toEqual([
+        'https://portal.example/detail/123',
+        'https://portal.example/detail/456',
+      ]);
+    });
+
+    it('falls back to a plaintext URL regex when bodyHtml is null', () => {
+      const text =
+        'Neue Ausschreibung: https://portal.example/detail/789 — bitte prüfen.';
+      const links = extractCandidateLinks(null, text);
+      expect(links).toEqual(['https://portal.example/detail/789']);
+    });
+
+    it('drops footer-noise links (unsubscribe/abmelden/einstellungen/preferences)', () => {
+      const html = `
+        <a href="https://portal.example/detail/1">Detail</a>
+        <a href="https://portal.example/unsubscribe?id=1">Abbestellen</a>
+        <a href="https://portal.example/abmelden">Abmelden</a>
+        <a href="https://portal.example/einstellungen">Einstellungen</a>
+        <a href="https://mail.example/preferences">Preferences</a>
+      `;
+      const links = extractCandidateLinks(html, '');
+      expect(links).toEqual(['https://portal.example/detail/1']);
+    });
+
+    it('dedupes identical hrefs', () => {
+      const html = `
+        <a href="https://portal.example/detail/1">Detail</a>
+        <a href="https://portal.example/detail/1">Nochmal</a>
+      `;
+      const links = extractCandidateLinks(html, '');
+      expect(links).toEqual(['https://portal.example/detail/1']);
+    });
+
+    it('caps extracted links at MAX_LINKS_PER_EMAIL (link-spam safety valve)', () => {
+      const html = Array.from(
+        { length: 10 },
+        (_, i) => `<a href="https://portal.example/detail/${i}">Link ${i}</a>`,
+      ).join('\n');
+      const links = extractCandidateLinks(html, '');
+      expect(links.length).toBe(5);
+    });
+
+    it('ignores non-http(s) hrefs (mailto:, relative paths, javascript:)', () => {
+      const html = `
+        <a href="mailto:info@portal.example">Kontakt</a>
+        <a href="/relative/path">Relativ</a>
+        <a href="javascript:void(0)">JS</a>
+        <a href="https://portal.example/detail/1">Detail</a>
+      `;
+      const links = extractCandidateLinks(html, '');
+      expect(links).toEqual(['https://portal.example/detail/1']);
+    });
+
+    it('returns [] for an empty HTML body and empty plaintext', () => {
+      expect(extractCandidateLinks('', '')).toEqual([]);
+      expect(extractCandidateLinks(null, '')).toEqual([]);
+    });
+  });
+
+  describe('titleFromEmail (pure)', () => {
+    it('returns the trimmed subject when present', () => {
+      expect(titleFromEmail('  Neue Ausschreibung Nr. 123  ', 'Body text')).toBe(
+        'Neue Ausschreibung Nr. 123',
+      );
+    });
+
+    it('falls back to the first non-empty body line when subject is blank', () => {
+      const text = '\n\n  Erste inhaltliche Zeile  \nZweite Zeile';
+      expect(titleFromEmail('', text)).toBe('Erste inhaltliche Zeile');
+      expect(titleFromEmail('   ', text)).toBe('Erste inhaltliche Zeile');
+    });
+
+    it('falls back to "Unbenannte Ausschreibung" when subject and body are both empty', () => {
+      expect(titleFromEmail('', '')).toBe('Unbenannte Ausschreibung');
+      expect(titleFromEmail('', '\n\n   \n')).toBe('Unbenannte Ausschreibung');
+    });
+  });
+
+  describe('sourceNoticeIdFor (pure)', () => {
+    it('returns a stable 40-char hex id for the same link', () => {
+      const link = 'https://portal.example/detail/123';
+      const id1 = sourceNoticeIdFor(link);
+      const id2 = sourceNoticeIdFor(link);
+      expect(id1).toBe(id2);
+      expect(id1).toMatch(/^[a-f0-9]{40}$/);
+    });
+
+    it('returns different ids for different links', () => {
+      const id1 = sourceNoticeIdFor('https://portal.example/detail/1');
+      const id2 = sourceNoticeIdFor('https://portal.example/detail/2');
+      expect(id1).not.toBe(id2);
+    });
+  });
+
+  it('never branches on a specific sender email/domain (D-04 restraint — no portal-specific parser)', () => {
+    const source = readFileSync(join(__dirname, 'email-alert.adapter.ts'), 'utf8');
+    // No hardcoded sender-domain/portal-name conditional logic — generic
+    // extraction only, mirroring the RESEARCH.md Anti-Pattern guard.
+    expect(source).not.toMatch(/senderDomain|from\.includes\(|sender\.includes\(/);
+  });
+});