Skip to content

Crawlee—A web scraping and browser automation library for Node.js to build reliable crawlers. In JavaScript and TypeScript. Extract data for AI, LLMs, RAG, or GPTs. Download HTML, PDF, JPG, PNG, and other files from websites. Works with Puppeteer, Playwright, Cheerio, JSDOM, and raw HTTP. Both headful and headless mode. With proxy rotation.

Fix flaky test

Ondra Urban35df7ca9eb27
diff --git a/test/live_view/live_view_server.js b/test/live_view/live_view_server.js
index 74c1a807d..055c293d0 100644
--- a/test/live_view/live_view_server.js
+++ b/test/live_view/live_view_server.js
@@ -102,18 +102,21 @@ describe('LiveViewServer', () => {
it('should not store more than maxScreenshotFiles screenshots', async () => {
const snapshots = [];
socket.on('snapshot', s => snapshots.push(s));
- await Promise.all(Array(3).fill(null).map(() => lvs.serve(fakePage)));
- const screenshots = await Promise.all(snapshots
- .sort((a, b) => a.screenshotIndex - b.screenshotIndex)
- .map((snap, idx) => {
- expect(snap.screenshotIndex).to.be.eql(idx);
- return snap;
- })
- .map(({ screenshotIndex }) => rqst(`${BASE_URL}/screenshot/${screenshotIndex}`)));
- screenshots.forEach((s, idx) => expect(s).to.be.eql(`screenshot${idx}`));
- const files = await readdir(LOCAL_STORAGE_SUBDIR);
- expect(files).to.have.lengthOf(2);
- files.forEach((f, idx) => expect(f).to.be.eql(`${idx + 1}`));
+ await Promise.all(Array(5).fill(null).map(() => lvs.serve(fakePage)));
+ const files = await new Promise((resolve, reject) => {
+ const interval = setInterval(async () => {
+ const files = await readdir(LOCAL_STORAGE_SUBDIR); // eslint-disable-line no-shadow
+ if (files.length === 2) {
+ clearInterval(interval);
+ resolve(files);
+ }
+ }, 10);
+ setTimeout(() => {
+ clearInterval(interval);
+ reject(new Error('Files were not deleted in 2000ms.'));
+ }, 2000);
+ });
+ files.forEach((f, idx) => expect(f).to.be.eql(`${idx + 3}`));
});
});
});