[4.x] Access to Local Storage can be blocked in Chrome

[olafcm]

We updated to v4 of the client.
Since then we started to receive the following error for Chrome users:

Failed to read the 'localStorage' property from 'Window': Access is denied for this document.

The issue happens here:

function u(e) {
  var t = e.localStorage || window.localStorage,
    r = "algoliasearch-client-js-".concat(e.key),
    n = function() {
      return JSON.parse(t.getItem(r) || "{}")
    };

More info: https://www.chromium.org/for-testers/bug-reporting-guidelines/uncaught-securityerror-failed-to-read-the-localstorage-property-from-window-access-is-denied-for-this-document

Is it possible for the client to detect if the LocalStorage is available? I.e. use: https://github.com/Modernizr/Modernizr/blob/master/feature-detects/storage/localstorage.js

Thank you

[nunomaduro]

Thanks for reporting this issue, I will investigate this tomorrow morning. 👍

[nunomaduro]

@olafcm Just to be clear, your search is working, but you got the warning on the chrome console right?

[olafcm]

Yes, I tested it and search works. It probably only affects the search results caching.

[nunomaduro]

I just don't preserve the state of the hosts after the refresh, and it uses the fallback in memory cache:

algoliasearch-client-javascript/packages/algoliasearch/src/builds/browserLite.ts

Line 49 in d862a91

createInMemoryCache(),

.

I will see if we can avoid the warning on the console. Otherwise everything seems to be working as expected.

[Haroenv]

I wonder if adding a .catch which still throws, but without the default error message will avoid it showing on the console.

algoliasearch-client-javascript/packages/cache-browser-local-storage/src/createBrowserLocalStorageCache.ts

Line 31 in d862a91

.then(([value]) => value);

[nunomaduro]

We are using catch on the FallbackableCache:

algoliasearch-client-javascript/packages/cache-common/src/createFallbackableCache.ts

Line 21 in d862a91

return current.get(key, defaultValue, events).catch(() => {

That performs a catch on the driver to fallback to the next one. Leet me investigate this.

[nunomaduro]

@olafcm can you also review the pull request that I just made?

[nunomaduro]

The bug has been fixed in version 4.0.2. ⚡️

[olafcm]

Works perfect, thank you very much for the perfect and quick fix @nunomaduro 🎈