When Calibre convert, the input language is zh-tw, but the output language become zh

[wealthtsay]

When Calibre convert a book, the input book language is "zh-tw"(traditional Chinese), but the output book language will become "zh"(Simplified Chinese).

I want the output book language (zh-tw) is the same as the input book language (zh-tw).

Please help to correct this error in later software updates.

For example: epub (content.opf)

Input book: <dc:language>zh-tw</dc:language>
Output book: <dc:language>zh</dc:language>

[kovidgoyal]

calibre does not support country specific language variants. So you will need to either correct it manually or simply leave the langauge field in calibre blank, in which case it wont set the language.

[wealthtsay]

My input format is epub, output format is mobi. How either correct it manually on mobi files ?

[wealthtsay]

If I simply leave the langauge field in calibre blank on epub file. But the converted mobi files language is "zh", not "zh-tw". How to resolve it ? Thank you

[kovidgoyal]

Convert to azw3 to MOBI. ANd you can edit the azw3 in the calibre editor.

[wealthtsay]

When I use the calibre editor and open the metadata.opf to edit the AZW3 language from "zho" to "zh-tw" , and save it. But when I edit this azw3 again, the language changes back to "zho"

How to resolve it ?

[kovidgoyal]

Then I'm afraid you are out of luck, sorry. Looking a the AZW3 output plugin code, it looks like it also does not support language + country metadata.

[wealthtsay]

Can Calibre solve this problem in future updates?

Thank you very much

[kovidgoyal]

It's certainly possible, but it is a fairly large amount of work (one has to write code to parse and validate http://www.rfc-editor.org/rfc/rfc5646.txt codes and then check that they actually work on a kindle).

[wealthtsay]

Ok. Please resolve this issue at next system update as soon as possible, thank you very much.

[yuleshow]

Yes, please resolve this. It's very important for Chinese readers. Chinese has two systems, Simplified Chinese (zh) and Traditional Chinese (zh-TW), both are a conflict with each other.

[pdurrant]

Calibre is open source, and Kovid is open to pull requests. If it's an issue that severely affects Chinese readers, perhaps one that also has technical skills will look at implementing it.

[jhowell]

Amazon only officially added support for books in Traditional Chinese in May of this year. Books sold by Amazon in Traditional Chinese are only delivered in KFX format, not KF8/AZW3. I do not know if this is due to a limitation in the KF8 renderer, which is based on an outdated version of webkit, or whether it is just intended to promote lock in to their ecosystem.

The KFX Output plugin for calibre can convert Traditional Chinese into KFX format when used in conjunction with the latest version of Amazon's Kindle Previewer. A KFX supporting Traditional Chinese is output by the Previewer for language codes of zh-CHT, zh-Hant, zh-HK, zh-MO, and zh-TW.

The KFX Output plugin works around the problem discussed in this thread by examining the source EPUB to see if the actual HTML content specifies a country code and if so moving it into the OPF before conversion.

Added: Currently KFX only supports horizontal text for Traditional Chinese, but it appears that Amazon is working on handling vertical text in the future.

[DNSB]

And perhaps we should ask which language designation system should be used. Should Kovid attempt to support such forms as zh-Hant and zh-Hans for traditional and simplified? Or older deprecated forms? See IANA language subtag registry for more information on RFC5646 supported tags.

[DNSB]

Originally Posted by wealthtsay

When Calibre convert a book, the input book language is "zh-tw"(traditional Chinese), but the output book language will become "zh"(Simplified Chinese).

Minor correction. zh is not simplified Chinese, zh is generic Chinese language support. zh-TW is an obsolete form. See https://ISO Language Codes (639-1 and 693-2) and IETF Language Types for more information.