When I ran the demo codes, it caused the error.

code

import unidic2ud
nlp=unidic2ud.load("kindai")
s=nlp("其國を治めんと欲する者は先づ其家を齊ふ")
print(s)

cmd output

------------------- ERROR DETAILS ------------------------
arguments: -r site-packages\unidic2ud\mecabrc -d site-packages\unidic2ud\download\gendai
[ifs] no such file or directory: site-packagesunidic2udmecabrc
----------------------------------------------------------

After I changed the file "unidic2ud/unidic2ud.py", line 224 and 231, the code can run.

224 self.mecab=Tagger("-r "+r+" -d "+d).parse
# self.mecab=Tagger("-r '"+r+"' -d '"+d+"'").parse
231 self.mecab=Tagger("-r "+r+" -d "+unidic_lite.DICDIR).parse
# self.mecab=Tagger("-r '"+r+"' -d '"+unidic_lite.DICDIR+"'").parse

A call to shlex.split(args) within mecab strips \\ characters from paths unless the paths are enclosed by " characters.

line 224 in unidic2ud.py does not do this:
self.mecab=Tagger("-r "+r+" -d "+d).parse

thus the library does not work with downloaded dictionaries on windows.
this modified line fixes the problem.

self.mecab=Tagger(f"""-r "{r}" -d "{d}" """).parse

mecab itself does enclose paths in this way within it's own code, so I find it highly unlikely that this modification will break anything on other platforms.

All the best :)