Closed sjtuross closed 2 years ago
@bakerboy I would have thought that URL encoding the search term is spec-compliant?
@Qstick?
seems to be double encoded on cross-seeds end
ok i see as well:
> const str = "%25E8%259C%2598%25E8%259B%259B%25E4%25BE%25A0%25EF%25BC%259A%25E5%25B9%25B3%25E8%25A1%258C%25E5%25AE%2587%25E5%25AE%2599+Spider+Man+Into+The+Spider+Verse+2018"
> decodeURIComponent(str)
'%E8%9C%98%E8%9B%9B%E4%BE%A0%EF%BC%9A%E5%B9%B3%E8%A1%8C%E5%AE%87%E5%AE%99+Spider+Man+Into+The+Spider+Verse+2018'
> decodeURIComponent(decodeURIComponent(str))
'蜘蛛侠:平行宇宙+Spider+Man+Into+The+Spider+Verse+2018'
v4.3.3
Currently non-Latin characters in torrent name is got encoded and then sent to indexers as search term. So no torrents can be matched.
For example, these Chinese characters
蜘蛛侠:平行宇宙
are encoded to%E8%9C%98%E8%9B%9B%E4%BE%A0%EF%BC%9A%E5%B9%B3%E8%A1%8C%E5%AE%87%E5%AE%99
as part of the search term as shown in the cross-seed log and prowlarr log.Is there a way to send the original term or the decoded version?
cross-seed log
prowlarr log