Add ignore argument to copytree that lets one filter stuff out

[pjbull]

shutil.copytree has an ignore argument that lets one filter stuff out. Would be nice to support that too, but maybe it can be a future enhnacement.

https://docs.python.org/3/library/shutil.html#shutil.copytree

Originally posted by @jayqi in #142 (comment)

---

Seems like shutil just supports callables, and we may even just be able to point users to the same pattern constructor: https://docs.python.org/3/library/shutil.html#shutil.ignore_patterns

[aaossa]

Is this issue still available? If that's the case I could work on this.

Seems like applying the filter in this section should do the trick right?

cloudpathlib/cloudpathlib/cloudpath.py

Lines 778 to 786 in 203f6aa

	for subpath in self.iterdir():
	if subpath.is_file():
	subpath.copy(
	destination / subpath.name, force_overwrite_to_cloud=force_overwrite_to_cloud
	)
	elif subpath.is_dir():
	subpath.copytree(
	destination / subpath.name, force_overwrite_to_cloud=force_overwrite_to_cloud
	)

Python does something similar I think, by iterating the paths and ignoring those that must be ignored. Relevant code:

• Using the ignore callable to retrieve set of ignored names: https://github.com/python/cpython/blob/a4ac14faa5c2be433738dfbbed14b0eaa1a2399e/Lib/shutil.py#L456
• USing the output of the callable to skip ignored names: https://github.com/python/cpython/blob/a4ac14faa5c2be433738dfbbed14b0eaa1a2399e/Lib/shutil.py#L464-L466

About the callable that defines which objects to ignore, seems like Python's implementation is already good enough, maybe just use the same object (i.e., use the shutil version)? Relevant Python source code: https://github.com/python/cpython/blob/a4ac14faa5c2be433738dfbbed14b0eaa1a2399e/Lib/shutil.py#L441-L451

[pjbull]

@aaossa Thanks for digging in here. First pass read of what you outlined seems right to me, thanks. Happy to take a PR with an implementation like that. Thanks!

[aaossa]

Sure, I'll be working on this in the upcoming days 👌

[aaossa]

Created the PR with the proposed implementation in #272 . Every test is passed, including two tests added to confirm that the argument works as expected.

[pjbull]

Merged into the main branch, thanks!

[noamsgl]

Could not find the ignore argument in the Documentation (https://cloudpathlib.drivendata.org/v0.6/api-reference/cloudpath/#cloudpathlib.cloudpath.CloudPath.copytree)
Is there a reason this was left out?

[jayqi]

Hi @noamsgl, this is only available in v0.11.0 and higher. You linked to the v0.6 version of the docs. There's a menu that lets you change the version at the top left of the documentation website.