René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"urlopen() hangs and eats 100% CPU on some https urls (git bisect found problematic commit)","articleBody":"## Summary\n\nRustPython 0.5.0 (and also current `main`) hangs and starts eating 100% CPU (one core) on first or second iteration of following script:\n\n```python\nimport urllib.request\nfor i in range(3):\n    print(i)\n    req=urllib.request.Request('https://crates.io/api/v1/crates/tokio')\n    with urllib.request.urlopen(req) as r:\n        r.read()\n```\n\nInteresting thing is, reproducibility depends on what exactly HTTPs server sends to the client. Just HTTPs plus relatively large file is not enough. For example, bug completely disappears when I replace `https://crates.io/api/v1/crates/tokio` with `https://gist.githubusercontent.com/khaykov/a6105154becce4c0530da38e723c2330/raw/41ab415ac41c93a198f7da5b47d604956157c5c3/gistfile1.txt`.\n\nI was able to `git bisect` this to a single problematic commit - 073adbd8da244b6203e8f32009b905af390c1a3d (https://github.com/RustPython/RustPython/pull/7418). Reverting it on top of 0.5.0 fixes this issue for me. And I have no idea about how to fix this properly.\n\n## Expected\n\nRustPython should terminate after three iterations.\n\n## Actual\n\nRustPython hangs and starts eating 100% CPU\n\n## Python Documentation\n\nN/A\n","author":{"url":"https://github.com/im-0","@type":"Person","name":"im-0"},"datePublished":"2026-05-17T07:44:51.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/7891/RustPython/issues/7891"}