bmcfee
commented
1 year ago Confirmed that this is due to this line:
https://github.com/fatiando/pooch/blob/a965902d26015453ac00269597a23b83d85db644/pooch/core.py#L660
For example, the 1.6.0 behavior would have been:
In [2]: line = "Karissa_Hobbs_-_Let's_Go_Fishin'.ogg"
In [3]: line.split()
Out[3]: ["Karissa_Hobbs_-_Let's_Go_Fishin'.ogg"]In 1.7, it becomes:
In [11]: shlex.split(line)
Out[11]: ['Karissa_Hobbs_-_Lets_Go_Fishin.ogg'](note the missing single-quotes)
A quick fix here is to use shlex.quote:
In [12]: shlex.split(shlex.quote(line))
Out[12]: ["Karissa_Hobbs_-_Let's_Go_Fishin'.ogg"]I don't know if this will be safe across platforms, but it would at least fix the bug I'm seeing.
dokempf
santisoler
leouieda
Description of the problem:
This just popped up in the librosa test suite after the environments upgraded from pooch 1.6 to 1.7.
We have a registry of data files that are fetched by pooch, as listed here: https://github.com/librosa/librosa/blob/c800e74f6a6ec5c27e0fa978d7355943cce04359/librosa/util/example_data/registry.txt Note that three of the files (lines 22-24) contain an apostrophe:
In pooch 1.6, this works as expected. However, after upgrading to 1.7, reading these files fails (see below). I believe this is a regression caused by #315 .
Full code that generated the error Minimal example via librosa:
This will call out to our example loader: https://github.com/librosa/librosa/blob/c800e74f6a6ec5c27e0fa978d7355943cce04359/librosa/util/files.py#L43-L98
Full error message
System information
conda listbelow:Output of conda list (it's bloated, but you asked for it :))