René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Optimize `pathlib.PurePath.__fspath__()`","articleBody":"# Feature or enhancement\r\n\r\nReturn an *unnormalized* path from `pathlib.PurePath.__fspath__()`\r\n\r\n# Pitch\r\n\r\nCode like `open(Path('./README.txt'))` or `Path('/home//barney').iterdir()` shouldn't require us to normalize the path in pathlib (e.g. remove `.` segments, doubled slashes, etc), as OS APIs are perfectly happy with unnormalized paths.\r\n\r\nWe can improve the performance of most `Path` methods, and the effective performance of passing a `Path` object to any API that accepts `os.PathLike`, by skipping normalization in `__fspath__()`.\r\n\r\n# Prerequisites\r\n\r\nPathlib must not normalize paths on construction:\r\n\r\n- https://github.com/python/cpython/issues/101362\r\n\r\nPathlib's normalization must not change the meaning of paths:\r\n\r\n- https://github.com/python/cpython/issues/65238\r\n- https://github.com/python/cpython/issues/78079\r\n\r\n# Previous discussion\r\n\r\n- https://github.com/python/cpython/pull/101560\n\n\u003c!-- gh-linked-prs --\u003e\n### Linked PRs\n* gh-112397\n\u003c!-- /gh-linked-prs --\u003e\n","author":{"url":"https://github.com/barneygale","@type":"Person","name":"barneygale"},"datePublished":"2023-03-17T16:24:14.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":5},"url":"https://github.com/102783/cpython/issues/102783"}