Using the Spotify plugin, it is unable to correctly search for albums & artists with Chinese/Japanese characters as its being converted to Ascii before the search. Resulting in 0 results from spotify.

Searches instead should keep the characters as the original (like how MusicBrainz works) so Spotify can return the correct results as spotify doesn't support searching with Pinyin/Romaji

Problem

With a folder containing songs from 盗作 - ヨルシカ (or any album/artist using Chinese characters)

Running this command in verbose (-vv) mode:

$ beet -vv import . -t

Led to this problem:

spotify: Searching Spotify for 'album:Dao Zuo  artist:yorushika'
spotify: Found 0 result(s) from Spotify for 'album:Dao Zuo  artist:yorushika'

The album name is getting converted to ascii when searched, resulting in 0 results.
In this case, 盗作 -> Dao Zuo

The album it should be finding:
https://open.spotify.com/album/6pZ0SrZCP8Bm28L6JhMtBy?si=X2sNSW5oTAGuJBNaVCAD0w

Setup

• OS: Arch Linux
• Python version: 3.13.2
• beets version: 2.2.0

My configuration (output of beet config) is:

directory: ~/Music

plugins:
    - embedart
    - fetchart
    - lyrics
    - lastgenre
    - spotify
    - scrub

import:
    move: yes


spotify:
    source_weight: 0.0
    mode: open
    show_failures: on

This is most likely due to unidecode.unidecode(query) which converts all characters into ascii prior to the search.