AI/firecrawl
1
0
Fork 0
mirror of https://git.mirrors.martin98.com/https://github.com/mendableai/firecrawl synced 2025-06-04 11:24:40 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update removeUnwantedElements.test.ts

Browse Source
This commit is contained in:
Nicolas 2024-06-26 21:20:11 -03:00
parent 388ce3cbce
commit 042f81ddf2
1 changed files with 28 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
apps/api/src/scraper/WebScraper/utils/__tests__/removeUnwantedElements.test.ts
View File

@ -144,4 +144,32 @@ describe('removeUnwantedElements', () => {
expect(result).not.toContain('<aside>');
expect(result).not.toContain('<footer>');
});
it('should include specified tags with class', () => {
const html = `<div><main class="main-content">Main Content</main><aside class="remove">Remove</aside><footer class="footer-content">Footer Content</footer></div>`;
const options: PageOptions = { onlyIncludeTags: ['.main-content', '.footer-content'] };
const result = removeUnwantedElements(html, options);
expect(result).toContain('<main class="main-content">Main Content</main>');
expect(result).toContain('<footer class="footer-content">Footer Content</footer>');
expect(result).not.toContain('<aside class="remove">');
});
it('should include specified tags with id', () => {
const html = `<div><main id="main-content">Main Content</main><aside id="remove">Remove</aside><footer id="footer-content">Footer Content</footer></div>`;
const options: PageOptions = { onlyIncludeTags: ['#main-content', '#footer-content'] };
const result = removeUnwantedElements(html, options);
expect(result).toContain('<main id="main-content">Main Content</main>');
expect(result).toContain('<footer id="footer-content">Footer Content</footer>');
expect(result).not.toContain('<aside id="remove">');
});
it('should include specified tags with mixed class and id', () => {
const html = `<div><main class="main-content">Main Content</main><aside id="remove">Remove</aside><footer id="footer-content">Footer Content</footer></div>`;
const options: PageOptions = { onlyIncludeTags: ['.main-content', '#footer-content'] };
const result = removeUnwantedElements(html, options);
expect(result).toContain('<main class="main-content">Main Content</main>');
expect(result).toContain('<footer id="footer-content">Footer Content</footer>');
expect(result).not.toContain('<aside id="remove">');
});
});