Solution for Japanese title in "EN" Calibre

[Maverynthia]

So Calibre has the BUG (Yes it's a BUG Mr. Goyal not a FEATURE) if you are in the EN version to convert Japanese titles to Chinese Pinyin which is unacceptable because the books state they are in language field they are Japanese and Chinese isn't Japanese, so why aren't they converted to romaji.

Now the solution that put forth by Mr. Goyal is to "Just put your Calibre into Japanese!" Well some of us have translated books that use the Japanese title or some of us WANT to use the Japanese titles OR some of us are learning Japanese and having the whole program in Japanese is well, silly as we don't understand it (yet). THIS IS NOT A SOLUTION AT ALL.

SOLUTION:
You want to go into the C:\Program Files\Calibre2\resources\localization or the (x86) if on 32bit Calibre under 64bit windows and extract the locales.zip. (Might want to make a backup) Keep the locales.zip open in the archiver and open the ja folder. In the extracted folder open your flavor of English (or other language), AU, CA, or GB and grab the message.do plop it into the ja folder in the archive.
(Theoretically you can just move it from the EN_** folder into the ja folder and rezip too...)
Set Calibre to "Japanese" and it should be in English, however converting all the titles to romaji and not pinyin.
You might need to do this at every update unfortunately as this BUG isn't going to get fixed and I don't know about setting custom locales to where you can make this permanent.

Note this doesn't auto censor book titles you don't want people to see ;3

Disclaimer: I take no responsibility if you accidentally delete all your books and cause the universe to implode while doing this. Also not responsible if this causes any other bugs to manifest.

[BetterRed]

@Maverynthia - Have you tried putting a copy of the ..\Calibre2\resources\localization\locales.zip file into the configuration resources folder and making the changes there - i.e. %appdata%\calibre\resources\localization\locales.z ip

See ==>> Overriding icons, templates, et cetera

Assuming that works as it should, your modified locales.zip will be used and installing a new version of calibre will leave it as is.

I have successfully overridden a number of items in Calibre2/resources.

BR

[Maverynthia]

Originally Posted by BetterRed

@Maverynthia - Have you tried putting a copy of the ..\Calibre2\resources\localization\locales.zip file into the configuration resources folder and making the changes there - i.e. %appdata%\calibre\resources\localization\locales.z ip

See ==>> Overriding icons, templates, et cetera

Assuming that works as it should, your modified locales.zip will be used and installing a new version of calibre will leave it as is.

I have successfully overridden a number of items in Calibre2/resources.

BR

Thanks, I wasn't sure where the folder exactly was


I tried putting it in the folder and it appears that Calibre ignores the locales.zip. So.. :/

[Kumabjorn]

Originally Posted by BetterRed

@Maverynthia - Have you tried putting a copy of the ..\Calibre2\resources\localization\locales.zip file into the configuration resources folder and making the changes there - i.e. %appdata%\calibre\resources\localization\locales.z ip

See ==>> Overriding icons, templates, et cetera

Assuming that works as it should, your modified locales.zip will be used and installing a new version of calibre will leave it as is.

I have successfully overridden a number of items in Calibre2/resources.

BR

If you do this, Japanese books will show up in Japanese or in romaji? Will that work with books in tategaki as well?

Twitter is to communications as haiku is to literature.

[C08]

i'm sorry to necro this thread, but has there been any updates on this issue? i couldn't get the proposed solution to work on the newest version of calibre. currently i'm adding japanese books to calibre and eeeverything is in romaji when converted. it shouldn't be necessary to manually fix the file names back to the original japanese - just keep everything, even the file paths and folders, in the original language! how do i do that?

[DNSB]

Originally Posted by C08

i'm sorry to necro this thread, but has there been any updates on this issue? i couldn't get the proposed solution to work on the newest version of calibre. currently i'm adding japanese books to calibre and eeeverything is in romaji when converted. it shouldn't be necessary to manually fix the file names back to the original japanese - just keep everything, even the file paths and folders, in the original language! how do i do that?

If you mean the names of the files in the library directories? That is by design. calibre works on multiple OS and file systems and some of them do not support Unicode. In the calibre GUI, you should be able to see the original title, author, etc. which are stored in the database. Manually modifying the filenames in the calibre library is an easy route to a corrupted library.

Simply put, treat the calibre library as a black box and use calibre to manage it. Stop trying to peek under the skirts!

To quote Microsoft's documentation:

NTFS stores file names in Unicode. In contrast, the older FAT12, FAT16, and FAT32 file systems use the OEM character set.

[BetterRed]

Stop pretending it's not an issue. Calibre provides a single key shortcut to open library book folders where the transliterated names are exposed.

If you edit a format file then the relevant editor (including calibre's EPUB editor) will almost invariably show the file name on the title bar. If you make use of Sigil's integration with GitHub you'll see it there too.

I don't have a solution, but pretending it's not a issue for some isn't a solution. And can we drop 'peeking under skirts' - it's a crime, currently a serious issue in South Korea. Even this libertine four score cis-gender male finds it offensive.

BR

[DNSB]

So you would have Kovid et alia drop compatibility to allow use of Unicode filenames?

What Sigil integration with GitHub? Do you mean where Sigil stores checkpoints?

Explaining the reasoning behind the filenames calibre uses is pretending that it is not an issue for some? An interesting approach to logic.

However it was nice to see that you agree with me that looking directly at the contents of a calibre library is a criminal act.

[C08]

Originally Posted by DNSB

If you mean the names of the files in the library directories? That is by design. calibre works on multiple OS and file systems and some of them do not support Unicode. In the calibre GUI, you should be able to see the original title, author, etc. which are stored in the database. Manually modifying the filenames in the calibre library is an easy route to a corrupted library.

i am on windows 11, and it supports japanese characters in the file name. yes, i see the orignal IN calibre, but that doesn't really help when the output after converting is in romaji. i don't use calibre for media organization, only format conversion. i have no interest in using it as a library tool, so i'm not worried about corruption at all.
it's difficult and chaotic to locate anything with file names and paths suddenly in romaji. and extremely tedious having to individually rename all the files back to japanese.
is there really no option to consistently use unicode/the original characters for the output file? i mean i get that not all os's support it, but are there no options for those that do?

[Sirtel]

Originally Posted by C08

i am on windows 11, and it supports japanese characters in the file name. yes, i see the orignal IN calibre, but that doesn't really help when the output after converting is in romaji. i don't use calibre for media organization, only format conversion. i have no interest in using it as a library tool, so i'm not worried about corruption at all.
it's difficult and chaotic to locate anything with file names and paths suddenly in romaji. and extremely tedious having to individually rename all the files back to japanese.
is there really no option to consistently use unicode/the original characters for the output file? i mean i get that not all os's support it, but are there no options for those that do?

What happens if you use the Save to Disk feature to save the books outside calibre? Are the filenames still wrong?

[BetterRed]

Originally Posted by C08

i am on windows 11, and it supports japanese characters in the file name. yes, i see the orignal IN calibre, but that doesn't really help when the output after converting is in romaji. i don't use calibre for media organization, only format conversion. i have no interest in using it as a library tool, so i'm not worried about corruption at all.
it's difficult and chaotic to locate anything with file names and paths suddenly in romaji. and extremely tedious having to individually rename all the files back to japanese.
is there really no option to consistently use unicode/the original characters for the output file? i mean i get that not all os's support it, but are there no options for those that do?

If all you want to do is convert, you could forego use of the library manager (aka the GUI) and use the ebook-convert command line tool ==>> Command Line Interface. This gives the user direct control over the output file name.

If you want to do the conversion within the library manager, then you need to use Save to Disk (or Send to Device) with a suitable template to create an 'output' file with a 'name' based on the metadata. See Preferences->Save to Disk and Send to Device.

You could probably automate this (i.e. do a Save to Disk immediately after a Conversion) using the Action Chains plugin

BR

[C08]

Originally Posted by Sirtel

What happens if you use the Save to Disk feature to save the books outside calibre? Are the filenames still wrong?

To my surprise, the files come out.. ok-ish! i mean there are some titles that are shortened or omitted, but i think this has to do with the metadata more than the inner workings of calibre. i never even considered that this would work, why would saving a file again when it's already saved in the library make any difference? it seemed like a redundant extra step - i've never understood the point of the save function to begin with. a conversion implies that a new file is created and therefore saved to disk at the end. having it work this way is very strange (to me at least).

Originally Posted by BetterRed

If all you want to do is convert, you could forego use of the library manager (aka the GUI) and use the ebook-convert command line tool ==>> Command Line Interface. This gives the user direct control over the output file name.

If you want to do the conversion within the library manager, then you need to use Save to Disk (or Send to Device) with a suitable template to create an 'output' file with a 'name' based on the metadata. See Preferences->Save to Disk and Send to Device.

You could probably automate this (i.e. do a Save to Disk immediately after a Conversion) using the Action Chains plugin

BR

this seems like a more streamlined solution, will have a look into the automation possibilities! thanks!



i still think it should be possible to have the library file(s) and path in the original languages characters. maybe not by default, but at least as an option. i mean imagine peoples reactions if path and file name was automatically translated to, say, cyrillic characters with no other alternatives. "what are you even doing there, you're not supposed to see this part anyway" would not be an acceptable answer. pfff.

it's also surprising that (as the first poster points out) calibre doesn't even go straight for proper romaji for japanese characters, it's via pinyin...! a very strange choice imo.


but seeing as this solved my problem, i'll leave these issues be and get to converting !

[Sirtel]

It seems strange to you because you see Calibre as a converter, but Calibre is first and foremost a library manager - that is its main purpose. And you're supposed to work with your files via the GUI, not directly.

Save to Disk is meant for working with/manipulating the files themselves - it saves them outside Calibre, so you can do whatever you want with them without messing up the library structure.

There have been many posters with similar problems and questions, who only used Calibre for converting and therefore didn't understand why Calibre has such a rigid way with the files in the library. They thought Calibre was first and foremost a converter. But actually Calibre was created as a library managing tool and many decisions about its basic functioning reflect that. Converting is just one of its features, not its raison d'être.

[BetterRed]

@Sirtel - You are avoiding the substantive issue: the Romanisation of Japanese script would normally transliterate to a form known as romaji , but calibre transliterates Japanese to pinyin which is the Chinese equivalent. Look them up in WikiP. sushi is romaji, shou si is pinyin.

Example: calibre will transliterate Война и миръ to Voina i mir, which most Russian speakers would find acceptable, but if calibre transliterated it to Guerra e pace, i.e. to Italian, I guarantee most Russian speakers would find it unacceptable.

See post #7 for why it matters.

I don't know why calibre doesn't create romaji for Japanese scripts. I would imagine transliterations are done by a third party library - perhaps the ICU library - International Components for Unicode.

And if it was simple to remedy I am certain KG would have done it years ago.

BR

[Comfy.n]

Originally Posted by BetterRed

I don't know why calibre doesn't create romaji for Japanese scripts. I would imagine transliterations are done by a third party library - perhaps the ICU library - International Components for Unicode.

I believe the 3rd party library is pykakasi.

And if it was simple to remedy I am certain KG would have done it years ago.

BR

KG once mentioned that he would welcome patches from someone who understands japanese for issues like this:
https://bugs.launchpad.net/calibre/+...850/comments/6

I figure @kindlefere might help somehow...(their github has main info in chinese, though)