fix: Page.setContent working with unicode strings (#4433)

Fix `page.setContent` with unicode strings that exceeds the range of a 8-bit byte.

This patch implements decoding part of the [MDN's solution #4 for the "unicode problem"](https://developer.mozilla.org/en-US/docs/Web/API/WindowBase64/Base64_encoding_and_decoding#Solution_4_%E2%80%93_escaping_the_string_before_encoding_it).

Since we rely on node.js buffer to encode into base64, we don't have troubles with base64 encoding, so it is left as-is.

Fixes #4424
This commit is contained in:
Sergio M
2019-05-18 16:15:16 +02:00
committed by Andrey Lushnikov
parent 3f23bb022e
commit 60249e0bc2
3 changed files with 14 additions and 2 deletions

View File

@@ -205,7 +205,7 @@ class DOMWorld {
// lifecycle event. @see https://crrev.com/608658
await this.evaluate(base64html => {
document.open();
document.write(atob(base64html));
document.write(decodeURIComponent(atob(base64html).split('').map(c => '%' + ('00' + c.charCodeAt(0).toString(16)).slice(-2)).join('')));
document.close();
}, Buffer.from(html).toString('base64'));
const watcher = new LifecycleWatcher(this._frameManager, this._frame, waitUntil, timeout);