Open Taki-B opened 5 years ago
It's in the 3MF system. I suspect that if you loaded a .3MF file from an umlaut path it wouldn't load.
@Samir55
I will take a look, LoH .
If that's the case, make sure you use boost::nowide
like we do for the other IO::
classes
boost::nowide is used in reading the extracted files from 3mf but the problem is that miniz doesn't extract the file at the first place. I am looking for an easy fix.
@lordofhyphens @Samir55
Problem isn't in miniz as I've tested it by itself and it can work with a string like this "Geräte" in it's path.
I think it's in the 23_3mf.t perl file I've fixed it by decoding the input path and output path to utf8 using Encode::decode_utf8()
from Encode
, and it's working properly.
But I'm not sure where else should I add this fix beside this test case as it will cause a problem every time read_tmf
is called with special characters like 'ä'.
Considering this is a perl side issue it should be fixed before calling Slic3r::IO::TMF::read(input_file, THIS)
which is in c++.
When the cloned code resides in a directory path including umlauts (/home/taki/Geräte/3D-Drucker/Slic3r_GIT/Slic3r/ in my case) parts of the code won't compile.
Steps to reproduce the behavior:
Desktop (please complete the following information):
git describe --tag
1.3.0-1243-g21eb603c)Additional context The bug is easy to circumvent: just rename the directory and everything runs as expected