Run playwright UI tests

lib/server/taxonomy/asset-discovery.test.ts

@@ -0,0 +1,73 @@
+import { describe, expect, it } from 'bun:test';
+import { discoverFilingAssets } from '@/lib/server/taxonomy/asset-discovery';
+
+describe('taxonomy asset discovery', () => {
+  it('classifies assets and selects ranked instance/pdf candidates', async () => {
+    const fetchImpl = (async () => {
+      return new Response(JSON.stringify({
+        directory: {
+          item: [
+            { name: 'abc_htm.xml', size: '900000' },
+            { name: 'abc_pre.xml', size: '250000' },
+            { name: 'abc_lab.xml', size: '120000' },
+            { name: '10k_financial_statements.pdf', size: '400000' },
+            { name: 'annual_report.pdf', size: '300000' },
+            { name: 'quarter_statement.pdf', size: '200000' },
+            { name: 'exhibit99.pdf', size: '500000' }
+          ]
+        }
+      }), {
+        status: 200,
+        headers: {
+          'content-type': 'application/json'
+        }
+      });
+    }) as unknown as typeof fetch;
+
+    const result = await discoverFilingAssets({
+      cik: '0000123456',
+      accessionNumber: '0000123456-26-000001',
+      filingUrl: 'https://www.sec.gov/Archives/edgar/data/123456/000012345626000001/abc.htm',
+      primaryDocument: 'abc.htm',
+      fetchImpl
+    });
+
+    expect(result.directoryUrl).toBe('https://www.sec.gov/Archives/edgar/data/123456/000012345626000001/');
+
+    const selectedInstance = result.assets.find((asset) => asset.asset_type === 'instance' && asset.is_selected);
+    expect(selectedInstance?.name).toBe('abc_htm.xml');
+
+    const selectedPdfs = result.assets
+      .filter((asset) => asset.asset_type === 'pdf' && asset.is_selected)
+      .map((asset) => asset.name);
+    expect(selectedPdfs.length).toBe(3);
+    expect(selectedPdfs).toContain('10k_financial_statements.pdf');
+    expect(selectedPdfs).toContain('annual_report.pdf');
+    expect(selectedPdfs).toContain('quarter_statement.pdf');
+    expect(selectedPdfs).not.toContain('exhibit99.pdf');
+  });
+
+  it('falls back to filing url when SEC directory assets are unavailable', async () => {
+    const fetchImpl = (async () => {
+      return new Response('not found', { status: 404 });
+    }) as unknown as typeof fetch;
+
+    const result = await discoverFilingAssets({
+      cik: '0000123456',
+      accessionNumber: '0000123456-26-000001',
+      filingUrl: 'https://www.sec.gov/Archives/edgar/data/123456/000012345626000001/abc.xml',
+      primaryDocument: 'abc.xml',
+      fetchImpl
+    });
+
+    expect(result.assets.length).toBe(1);
+    expect(result.assets[0]).toEqual({
+      asset_type: 'instance',
+      name: 'abc.xml',
+      url: 'https://www.sec.gov/Archives/edgar/data/123456/000012345626000001/abc.xml',
+      size_bytes: null,
+      score: 6,
+      is_selected: true
+    });
+  });
+});