Don't cache embedded GEDCOM in sessionStorage (#308)

src/datasource/embedded.ts

@@ -61,7 +61,15 @@ export class EmbeddedDataSource implements DataSource<EmbeddedSourceSpec> {
       try {
         const embeddedHash = 'embedded';
         storeGedcom(embeddedHash, gedcom, new Map());
-        const data = await loadGedcom(embeddedHash, onProgress);
+        // Embedded data is volatile: the parent re-posts a fresh GEDCOM on every
+        // load (and a page reload re-runs the ready/parent_ready handshake, so
+        // the parent re-sends). That GEDCOM may inline object URLs (blob:/data:)
+        // that die with the previous document. Persisting it to sessionStorage
+        // under the constant 'embedded' key would return dead URLs on the next
+        // load, so opt out of the session cache for embedded data.
+        const data = await loadGedcom(embeddedHash, onProgress, {
+          useSessionCache: false,
+        });
         const software = getSoftware(data.gedcom.head);
         analyticsEvent('embedded_file_loaded', {
           event_label: software,

src/datasource/load_data.ts

@@ -35,9 +35,10 @@ async function prepareData(
   cacheId: string,
   images?: Map<string, string>,
   onProgress?: (status: string) => void,
+  useSessionCache = true,
 ): Promise<TopolaData> {
   const data = await convertGedcom(gedcom, images || new Map(), onProgress);
-  if (!images || images.size === 0) {
+  if (useSessionCache && (!images || images.size === 0)) {
     if (gedcom.length <= SESSION_CACHE_GEDCOM_LIMIT) {
       const dataToSerialize = {...data};
       delete dataToSerialize.images;
@@ -184,14 +185,18 @@ export async function loadFromUrl(
 export async function loadGedcom(
   hash: string,
   onProgress?: (status: string) => void,
+  options?: {useSessionCache?: boolean},
 ): Promise<TopolaData> {
-  try {
-    const cachedData = sessionStorage.getItem(hash);
-    if (cachedData) {
-      return JSON.parse(cachedData);
+  const useSessionCache = options?.useSessionCache ?? true;
+  if (useSessionCache) {
+    try {
+      const cachedData = sessionStorage.getItem(hash);
+      if (cachedData) {
+        return JSON.parse(cachedData);
+      }
+    } catch (e) {
+      console.warn('Failed to load data from session storage: ' + e);
     }
-  } catch (e) {
-    console.warn('Failed to load data from session storage: ' + e);
   }
   // Retrieve from the in-memory store (survives within-tab navigation even
   // when the GEDCOM is too large for history.pushState or sessionStorage).
@@ -203,7 +208,13 @@ export async function loadGedcom(
     );
   }
   try {
-    return await prepareData(stored.gedcom, hash, stored.images, onProgress);
+    return await prepareData(
+      stored.gedcom,
+      hash,
+      stored.images,
+      onProgress,
+      useSessionCache,
+    );
   } catch (error) {
     revokeObjectUrls(stored.images);
     throw error;

src/datasource/load_data_store.spec.ts

@@ -1,6 +1,7 @@
-import {describe, expect, it} from '@jest/globals';
+import {beforeEach, describe, expect, it} from '@jest/globals';
 import {storeGedcom} from './gedcom_store';
 import {loadGedcom} from './load_data';
+import {mockSessionStorage} from './test_helpers';
 
 // Minimal GEDCOM with 1 individual and 1 family so convertGedcom succeeds.
 const MINIMAL_GEDCOM = [
@@ -43,3 +44,31 @@ describe('loadGedcom()', () => {
     ]);
   });
 });
+
+describe('loadGedcom() session cache', () => {
+  let sessionStorageMock: {[key: string]: string};
+
+  beforeEach(() => {
+    sessionStorageMock = mockSessionStorage();
+  });
+
+  it('caches prepared data in sessionStorage by default', async () => {
+    const hash = 'cache-default-hash';
+    storeGedcom(hash, MINIMAL_GEDCOM, new Map());
+    await loadGedcom(hash);
+    expect(sessionStorageMock[hash]).toBeDefined();
+  });
+
+  it('ignores and preserves the cache when useSessionCache is false', async () => {
+    const hash = 'cache-disabled-hash';
+    sessionStorageMock[hash] = JSON.stringify({stale: true});
+    storeGedcom(hash, MINIMAL_GEDCOM, new Map());
+
+    const data = await loadGedcom(hash, undefined, {useSessionCache: false});
+
+    // Returned from the in-memory store, not the stale cache.
+    expect(data.chartData.indis).toHaveLength(1);
+    // The stale entry is neither served nor overwritten with fresh data.
+    expect(sessionStorageMock[hash]).toBe(JSON.stringify({stale: true}));
+  });
+});