Issue with path in windows

[lguerard]

The parse_path functions sometimes creates issues on windows when parsing a direct path copied.

• Using the script parameter for file "U:\VAMP\guerard\LS7\new copy\SmalltileandZ - Copy.czi" returns 'path':'U:\\VAMP\\guerard\\LS7\\new copy\\'
• Directly using parse_path("U:\VAMP\guerard\LS7\new copy\SmalltileandZ - Copy.czi") gives 'path': 'U:\\VAMP\\guerard\\LS7\new copy\\' which then later can cause issues

[lguerard]

Adding path = path.lower().replace("\n",r"\\n") at the beggining of the script fixes it, but it doesn't feel like the proper way to do it... Let me know if you have any ideas @ehrenfeu

[lguerard]

Same thing is happening with \t

[ehrenfeu]

Sorry, I need to see the two results side by side:

           "U:\VAMP\guerard\LS7\new copy\SmalltileandZ - Copy.czi"
parse_path("U:\VAMP\guerard\LS7\new copy\SmalltileandZ - Copy.czi"

'path':'U:\\VAMP\\guerard\\LS7\\new copy\\'
'path': 'U:\\VAMP\\guerard\\LS7\new copy\\'

Now I'm confused. Could you point out or explain differently what the issue is exactly? The only thing I'm seeing is the space after the colon in the second section, but I guess that's just a copy-paste thing, no?

[imagejan]

• Directly using parse_path("U:\VAMP\guerard\LS7\new copy\SmalltileandZ - Copy.czi") gives 'path': 'U:\\VAMP\\guerard\\LS7\new copy\\' which then later can cause issues

Does it help calling parse_path with a raw string (r in front of the quotes)?

parse_path(r"U:\VAMP\guerard\LS7\new copy\SmalltileandZ - Copy.czi")

Otherwise Python interprets \t and \n as tab and newline, respectively...

[ehrenfeu]

Ah, I see the \n in the second path now...

@imagejan is probably right, calling it like this should work - the question is then rather if we can enforce this from within the function (as in, avoiding the string to be interpreted wrongly), or if that's already too late.

➡️ I'll need to look into this!

[rohangirishrao]

I guess replacing the \n back to "proper" alphabet at the top of the method should work, right? Converting to a raw string inside the function would be too late.
path = path.replace("\n", "\\n").replace("\t", "\\t") ?

[ehrenfeu]

@lguerard could you create an additional test for this (preferred), or modify the existing one here:

python-imcflibs/tests/test_pathtools.py

Lines 37 to 44 in 2affd76

	def test_parse_path_windows():
	path = r'C:\foo\bar'
	parsed = parse_path(path)
	assert parsed['orig'] == path
	assert parsed['full'] == r'C:/foo/bar'
	assert parsed['fname'] == 'bar'
	assert parsed['dname'] == 'foo'

[ehrenfeu]

@lguerard sorry if that's a stupid question, but is this issue supposed to be fixed by the changes from the PR's mentioned above (#24 & #26)?