Fix #512 - Isolate directories for generating hashes #556
Conversation
Some distributions have incompatible type of files with same name on the root directory. For example, matplotlib-2.0.2.tar.gz has a directory named "LICENSE" on the root, which may conflict many other distributions, causing errors like: IOError: [Errno 20] Not a directory: u'/tmp/tmpz/LICENSE/LICENSE_STIX' So we need to isolate unpacking directory of each distributions.
|
So this makes it possible to unpack distribution packages with conflicting files inside them. I'm just thinking that why the code unpacks the downloaded packages in the first place, since it only needs to hash the byte contents of the package. I think the code should be simplified to not do any unpacking, but rather just hash the file contents. This should fix the bug and also make it faster and smarter since unnecessary unpacking is avoided. I already tested this with Prequ. I'll create a PR for this to pip-tools too. |
|
@suutari yeah, you are right and I considered that too, but just wanted a quick patch to have minimal impact on the status quo, i.e. depending on pip's internal API. For pip >= 10.0.0 (pypa/pip@95bcf8c), the whole implementation of |
|
This should probably be closed now that #557 is merged. Right? |
|
Indeed, thanks for the reminder. |
Some distributions have incompatible type of files with same name on the root directory. For example,
matplotlib-2.0.2.tar.gzhas a directory named "LICENSE" on the root, which may conflict with many other distributions, causing errors like:IOError: [Errno 20] Not a directory: u'/tmp/tmpz/LICENSE/LICENSE_STIX'So we need to isolate unpacking directory of each distributions.
Contributor checklist