Special characters in author names

[LadyKate]

I am wondering if I have missed a setting that will change accented characters to the unaccented version for author names.

I usually remove the accented character as when I manually type an author for a search I'm not going to be putting in the accent.

eg: author Zoë Sharp
If I type in Zoe Sharp it will show no matches.

Any ideas?

[kovidgoyal]

If you set your calibre interface language to english, searching unaccented characters will match accented ones.

[LadyKate]

Thank you. I'll have to double check. As I'm in quebec I may have a system setting for french on here somewhere

[rkillings]

I want to do the reverse but don't know how. I want the calibre interface to be English, and I take care to preserve accented characters in author names and book titles (e.g. "Čapek, Karel", "Pääbo, Svante",…). calibre displays the metadata correctly -- but then it strips the accents from the file and folder names when it saves to disk. My OS (Windows 7) has no problem displaying correctly accented file and folder names. How do I tell calibre NOT to "asciiize" when saving to disk?

[kovidgoyal]

Preferences->saving to disk

[rkillings]

Sorry, I was using the wrong terminology. Preferences->Saving books to disk->Convert non-English characters to English equivalents is and has been unchecked.
I meant to say, Saving books to the calibre library itself (which is itself a save-to-disk operation, no?;-).
I'm keeping synced copies of my library on several different cloud storage services, now that those companies are offering so much free space. It's (relatively) easy to access my full library from an iPad or other tablet and download any title, but if the accented characters have been stripped from the filenames by calibre, the ebook reader apps I've tried won't show them in the author and title fields, even though I've put them in the calibre metadata entries.
Now, I grant that none of the reader apps, iBooks in particular, is as feature-rich as I'd like. Few of them let me do any ex-post editing of the author or title field. It may just be too early in the development of this technology… by others, not by the calibre developers!
Thanks for the good work, Kovid.

[theducks]

Originally Posted by rkillings

Sorry, I was using the wrong terminology. Preferences->Saving books to disk->Convert non-English characters to English equivalents is and has been unchecked.
I meant to say, Saving books to the calibre library itself (which is itself a save-to-disk operation, no?;-).
I'm keeping synced copies of my library on several different cloud storage services, now that those companies are offering so much free space. It's (relatively) easy to access my full library from an iPad or other tablet and download any title, but if the accented characters have been stripped from the filenames by calibre, the ebook reader apps I've tried won't show them in the author and title fields, even though I've put them in the calibre metadata entries.
Now, I grant that none of the reader apps, iBooks in particular, is as feature-rich as I'd like. Few of them let me do any ex-post editing of the author or title field. It may just be too early in the development of this technology… by others, not by the calibre developers!
Thanks for the good work, Kovid.

NO

You may NOT change the way calibre stores a library.
http://manual.calibre-ebook.com/faq....tory-structure

Calibre MUST work across OS platforms so it uses characters and paths that work on all of them.

[rkillings]

Actually, save-to-disk without asciiizing is adequate to one of my purposes -- just not the library backup purpose.

What I did not realize is that selecting every title in the library and saving it all to disk will duplicate the folder structure of the calibre library, but *with* any accented characters in names and titles as they appear in calibre metadata. The output won't *be* a calibre library with metadata json file, but it will serve as vehicle for getting books onto a tablet via cloud data storage alone. (Until more apps are able to use calibre metadata, as Calibre Cloud on Android does.)

Furthermore, for the save-to-disk operation I can select only the ebook format I'm going to use on a given tablet -- useful, because I usually have both epub and mobi.

[user743]

Originally Posted by theducks

NO

You may NOT change the way calibre stores a library.
http://manual.calibre-ebook.com/faq....tory-structure

Calibre MUST work across OS platforms so it uses characters and paths that work on all of them.

no one said anything about changing the folder structure.

there is no reason that if anyone sais anything similar . everyone starts SCREAMING!

[pdurrant]

Originally Posted by rkillings

It's (relatively) easy to access my full library from an iPad or other tablet and download any title, but if the accented characters have been stripped from the filenames by calibre, the ebook reader apps I've tried won't show them in the author and title fields, even though I've put them in the calibre metadata entries.

Your problem here isn't the folder names, but the metadata inside the ebook files. Calibre does NOT update the files stored in the calibre library when you change the book's metadata.

I beleive it will update the ebook's internal metadata when you do a Save To Disk.

[BoyBlue]

Yes, that's a very old post, indeed.
But the rather simple problem it raises has still not been solved.
When i raised the issue recently (may 2025), Calibre's developer, seemingly quite annoyed, replied that this was a very old issue that has been discussed at length and that I should simply google the subject.
Very well. But why then is Calibre still replacing French accented characters in folder and file names, while correctly storing the author and title data with their proper accented spelling in the opf metadata file ?
The conversion of accented characters to their English equivalent does not happen by magic. Somewhere in Calibre's code, a function processes those names and outputs modified unaccented names. That's not right !
So, why not simply put an end to that stripping/converting ? Because Calibre's search and classify functions would get more complicated? Because the representation of accented characters is not the same across all OS ?
Come on...

[DNSB]

Some people are still using systems that do not support UTF-8 or even a small chunk of it. Using the limited ASCII character set for the filenames in the library where the user should not be digging around in the files simply ensures that cross-platform compatibility works. The calibre GUI will happily show the accented characters or basically any UTF-8 character as part of the book title or the author name or other fields.

Save to disk or send to device can use the full UTF-8 character set. Though I've seen multiple posts where the complaint is made that the book title, author name or even the contents of the book are filled with the unknown character glyph since the font used does not support that glyph.

Out of curiosity, since this is your first post on MobileRead, where did you raise the question?

[JSWolf]

Originally Posted by BoyBlue

Yes, that's a very old post, indeed.
But the rather simple problem it raises has still not been solved.
When i raised the issue recently (may 2025), Calibre's developer, seemingly quite annoyed, replied that this was a very old issue that has been discussed at length and that I should simply google the subject.
Very well. But why then is Calibre still replacing French accented characters in folder and file names, while correctly storing the author and title data with their proper accented spelling in the opf metadata file ?
The conversion of accented characters to their English equivalent does not happen by magic. Somewhere in Calibre's code, a function processes those names and outputs modified unaccented names. That's not right !
So, why not simply put an end to that stripping/converting ? Because Calibre's search and classify functions would get more complicated? Because the representation of accented characters is not the same across all OS ?
Come on...

You don't understand how calibre works. The library is under the hood and should be left alone. How it's setup is of no concern. All the filenames could be random numbers and it still would not matter. So just ignore the library. You are not to touch the library. You are to leave it alone.

When you do a search, it searches for the data you see in the GUI. The GUI will have your special characters as long as you made sure of it. When you do a save to disk, you can get a filename that contains these special characters as long as the OS you are using supports them. When you send to your Kobo you can have the metadata embedded so it will have your special characters so you can see them on your Kobo as part of the title and/or author name(s).

So what calibre does with the filename for the library is of no consequence and you just let it be.