This project provides a file-like object that fetches parts of a file using HTTP range requests as needed.

>>> from seekablehttpfile import SeekableHttpFile
>>> from zipfile import ZipFile

>>> f = SeekableHttpFile("https://files.pythonhosted.org/packages/cb/90/599c79a248dcae6935331113649de5d75427e320efde21b583648b498584/tensorflow_intel-2.14.0-cp310-cp310-win_amd64.whl")  # 284MB
>>> # use as normal, for example with ZipFile
>>> z = ZipFile(f)
>>> len(z.namelist())
9414
>>> # find out how much we actually read
>>> f.stats
{'num_requests': 4, 'optimistic_bytes_read': 256000, 'lazy_bytes_read': 1078669, 'satisfied_from_cache': 2}

Users of this library should be able to use Python 3.7 or above. This is validated by tests, and is the version encoded in Requires-Python.

Development requires Python 3.12 (or 3.13), realistically. CI validates that it works on 3.10 and above, but you will get less checking in flake8. Expect older versions to be dropped as soon as there is a version of mypy, black, etc that is not simultaneously compatible with all development versions.

Notably, this means that while this project might function (and pass tests) on Python 3.7, it's likely that you can't typecheck against it.

seekablehttpfile is copyright Tim Hatch, and licensed under the MIT license. I am providing code in this repository to you under an open source license. This is my personal repository; the license you receive to my code is from me and not from my employer. See the LICENSE file for details.