Solution for Japanese title in "EN" Calibre

[kovidgoyal]

calibre does create romaji *provided the interface language is set to Japanese*. And does so using pykakasi which incidentally was spun out of calibre code to do that by Hiroshi Miura who contributed said code to calibre in the first place.

East asian scripts all re-use the same unicode codepoints so it is *not possible* to transliterate them without using a base language. That base language is the user interface language in calibre.

[chaley]

We added a "tweak" to calibre permitting you to specify the base language for transliterations. The changes are in calibre source and will be in the next preview/release.

The tweak documentation is:

#: East Asian language to use for transliteration
# Setting this tweak will make calibre use the specified language as the "base"
# language when transliterating East Asian languages to ASCII. This might be
# useful if you run calibre in English but want text transliterated to
# Japanese. The valid values are 'ja', 'kr', 'vn', 'zh', and '' (empty string).
# The empty string means use the user interface language as the base language.
# Any value not in the above list will be treated as the empty string.
# Example: east_asian_base_language = 'ja'
east_asian_base_language = ''

I tested it with the Gutenberg book "苦悶の欄" by Earl Derr Biggers (https://www.gutenberg.org/ebooks/39287). With the tweak empty the title was transliterated to "Ku Men noLan". With the tweak set to 'ja' it was transliterated to "Kumon no Ran". Is this transliteration correct? I have no idea.

The tweak applies to newly added books.