Closed floppym closed 2 years ago
Thank you, the changes look good.
Just on Fedora the encoding test shows 'charset=utf-8' for all files except usb.ids, where it produces 'charset=unknown-8bit', I've allowed it too.
I'll merge the request if you're fine with my modification.
Just on Fedora the encoding test shows 'charset=utf-8' for all files except usb.ids, where it produces 'charset=unknown-8bit', I've allowed it too.
That sounds like a bug on Fedora's file command, or some difference in the data we are comparing. I think adding "unknown-8bit" as a valid option sort of invalidates the purpose of the test.
That said, if you want to add that workaround for now and fix the issue later, that's ok by me.
Hmm it seems the bug might be in Gentoo's file command. Attempting to run iconv
on the usb.ids file fails.
% iconv -f utf-8 -t utf-16 /usr/share/hwdata/usb.ids > foo.txt
iconv: illegal input sequence at position 692405
Given this discrepancy between distros, I think your change is reasonable.
I pushed a different approach: let's use iconv to validate the data instead of file.
Thank you, merged!
I am updating Gentoo Linux to use the hwdata package instead of our homegrown hwids package, and ran into a few issues when running
make check
.