I find that VS Code does really good URL recognition / extraction- better than any methods found in the internet. Since VS Code is opensource, where can I find the code? Probably regex rules.
I searched the source code with keyword url or extract, but can not find any clues.
Not 100% percent sure that this is it, but from my search through the source code, I think you might be looking for src/vs/editor/common/languages/linkComputer.ts, which uses a state-machine-based approach.
You can see in src/vs/editor/browser/services/editorWorkerService.ts, in EditorWorkerService's constructor, there's a comment saying "register default link-provider and default completions-provider", and a call to client.computeLinks, which is defined in src/vs/editor/common/services/editorSimpleWorker.ts in the EditorSimpleWorker class' computeLinks function, which in turn calls computeLinks from linkComputer.ts. That comment lines up nicely with the description of DocumentLinkProvider<T>'s provideDocumentLinks function, which states:
Note that the editor ships with a default provider that detects
http(s)andfilelinks.
Fun fact: you can find where it's used in https://github.com/search?q=repo%3Amicrosoft%2Fvscode%20computeLinks&type=code, including src/vs/workbench/contrib/output/common/outputLinkComputer.ts and src/vs/workbench/contrib/terminalContrib/links/browser/terminalUriLinkDetector.ts.
The integrated terminal in VS Code has various other capabilities for links (including URLs, file links, folder links, and "word links"). The integrated uses xtermjs, which supports OSC 8 hyperlinks (see also How to create links in xterm.js). VS Code also has several builtin link detector patterns specifically designed to recognize things like file links (that don't use file://) in command outputs of popular commandline tools. See src/vs/workbench/contrib/terminalContrib/links/browser/terminalLinkParsing.ts. Also, the word links feature has some extra bits to search following text in a line for info about line numbers and such (see 1.88 release notes).
The Output Panel also has some special link detection patterns, which you can find in src/vs/workbench/contrib/output/common/outputLinkComputer.ts.
As mentioned above, there's the DocumentLinkProvider extension API that extension can use to contribute links to documents. If you want to find out how a specific extension is doing that, you'll need to post a question specifically about that extension.
vscode uses ripgrep under the hood. You can find more information about the implementation here https://github.com/BurntSushi/ripgrep
httporhttps, these strings should be part of the regex used, maybehttps?://