Does os.path relpath produce an incorrect relative path?

[BlindAnagram]

I am wondering whether I have misunderstood the semantics of os.path
relpath or whether it has a bug.

Here is a short test program:

-----------------------------------------------------------
from os.path import relpath, split

src_path = 'C:\\lib\\src\\'

vcx_path = 'C:\\build.vs22\\lib\\lib.vcxproj'
rel_path = relpath(src_path, vcx_path)
print(f"{vcx_path = }\n{src_path = }\n{rel_path = }\n")

vcx_path = 'C:\\build.vs22\\lib\\'
rel_path = relpath(src_path, vcx_path)
print(f"{vcx_path = }\n{src_path = }\n{rel_path = }\n")

vcx_path = 'C:\\build.vs22\\lib\\lib.vcxproj'
rel_path = relpath(src_path, split(vcx_path)[0])
print(f"{vcx_path = }\n{src_path = }\n{rel_path = }\n")
-----------------------------------------------------------

and its output with Python 3.11:

-----------------------------------------------------------
vcx_path = 'C:\\build.vs22\\lib\\lib.vcxproj'
src_path = 'C:\\lib\\src\\'
rel_path = '..\\..\\..\\lib\\src'

vcx_path = 'C:\\build.vs22\\lib\\'
src_path = 'C:\\lib\\src\\'
rel_path = '..\\..\\lib\\src'

vcx_path = 'C:\\build.vs22\\lib\\lib.vcxproj'
src_path = 'C:\\lib\\src\\'
rel_path = '..\\..\\lib\\src'
-----------------------------------------------------------

The first of these three results produces an incorrect relative path
because relpath does not strip off any non-directory tails before
comparing paths.

Is this a bug or my misunderstanding of relpath semantics?

Comments would be appreciated as I don't waste developer time if this is
not a bug.

regards

Brian

[Eryk Sun]

On 5/25/23, BlindAnagram <BlindA...@nowhere.org> wrote:
>
> vcx_path = 'C:\\build.vs22\\lib\\lib.vcxproj'
> src_path = 'C:\\lib\\src\\'
> rel_path = '..\\..\\..\\lib\\src'
>

> [snip]

>
> The first of these three results produces an incorrect relative path
> because relpath does not strip off any non-directory tails before
> comparing paths.

The start path is assumed to be a directory, which defaults to the
current working directory, and the input paths are first resolved as
absolute paths. In order to reach src_path from vcx_path, one has to
traverse 3 levels up to the root directory.

https://docs.python.org/3/library/os.path.html#os.path.relpath

[Greg Ewing]

On 25/05/23 7:49 pm, BlindAnagram wrote:
> The first of these three results produces an incorrect relative path
> because relpath does not strip off any non-directory tails before
> comparing paths.

It has no way of knowing whether a pathname component is a directory
or not. It's purely an operation on strings, it doesn't look in the
file system.

--
Greg

[MRAB]

On 2023-05-25 16:53, Eryk Sun wrote:
> On 5/25/23, BlindAnagram <BlindA...@nowhere.org> wrote:
>>

>> vcx_path = 'C:\\build.vs22\\lib\\lib.vcxproj'
>> src_path = 'C:\\lib\\src\\'
>> rel_path = '..\\..\\..\\lib\\src'
>>

>> [snip]

>>
>> The first of these three results produces an incorrect relative path
>> because relpath does not strip off any non-directory tails before
>> comparing paths.
>

> The start path is assumed to be a directory, which defaults to the
> current working directory, and the input paths are first resolved as
> absolute paths. In order to reach src_path from vcx_path, one has to
> traverse 3 levels up to the root directory.
>
> https://docs.python.org/3/library/os.path.html#os.path.relpath

Well, it's not necessarily the optimal relative path, because it's not
always necessary to go all the way up to the root, as in the first example.

[cactus]

Thanks to all for their comments.

I was hoping that there would be an direct way in Python to find a relative path between two paths in which the convention is that all directories end with '\\' so that it is possible to distinguish the 'tail' references to files as not part of the path comparison.

But I will just have to remember to strip the file tails myself.

Surprisingly (for me at least) the alternative provided by the pathlib module 'relative_to' method doesn't provide for full relative path computation. I was expecting this would offer everything that os.path offers but it doesn't in this case.

Brian

[cactus]

On Thursday, 25 May 2023 at 17:57:21 UTC+1, MRAB wrote:

[Eryk Sun]

On 5/26/23, cactus <riem...@gmail.com> wrote:
>
> Surprisingly (for me at least) the alternative provided by the pathlib
> module 'relative_to' method doesn't provide for full relative path
> computation. I was expecting this would offer everything that os.path
> offers but it doesn't in this case.

Starting with Python 3.12, the relative_to() method has a walk_up
parameter. It defaults to False for the sake of backward
compatibility.

https://docs.python.org/3.12/library/pathlib.html#pathlib.PurePath.relative_to

[BlindAnagram]

Thanks, it is good to know that this is coming.

Brian