Save Epub to disk.Creates extra folder.

[Johansen68]

I have been using Calibre for 15 years and have never encountered this problem before.

I save 5 ebooks by one author using "Save to disk", "Save only Epub format to disk". The files end up in two folders.
One file has the author's name/folder name wrong.

I have, as far as I know, fixed the metadata and all that. The file is in the correct folher in the library.

(I know I can probably use "save to disk in a single folder" as a workaround, but I fear something is wrong with the file somwhere.)

Any ideas?

J

The highlighted file is the one...
https://imgur.com/a/FuKnm5z

[Sirtel]

I think the filename is too long, so there's some automatic shortening going on. Try to make the title shorter.

[Johansen68]

Thank you.
That was the solution. Shorten book title.

[DNSB]

For me, when editing the metadata, if I see a colon in the book title, the portion after the title goes into the subtitle if it is worth keeping (all too many Amazon ebooks tend to stuff the series and other useless information into the title).

I.e. The Horrific Sufferings of the Mind-Reading Monster Hercules Barefoot: His Wonderful Love and His Terrible Hatred would be broken into a title of The Horrific Sufferings of the Mind-Reading Monster Hercules Barefoot and a subtitle of His Wonderful Love and His Terrible Hatred.

[Moonbase59]

It’s unfortunate that Calibre tries to stuff a correctly tagged subtitle into the book title (separated by a ': ').

I never liked that, would have preferred a separate field. But since the "subtitle" is often indeed filled with useless information (series info, or just "Novel"), I mostly also remove it, and never bothered to make an extra column for it.

[JSWolf]

Originally Posted by Moonbase59

It’s unfortunate that Calibre tries to stuff a correctly tagged subtitle into the book title (separated by a ': ').

I never liked that, would have preferred a separate field. But since the "subtitle" is often indeed filled with useless information (series info, or just "Novel"), I mostly also remove it, and never bothered to make an extra column for it.

On a Kobo Reader, you can have the subtitle show separately from the title.

[DNSB]

Originally Posted by Moonbase59

It’s unfortunate that Calibre tries to stuff a correctly tagged subtitle into the book title (separated by a ': ').

I never liked that, would have preferred a separate field. But since the "subtitle" is often indeed filled with useless information (series info, or just "Novel"), I mostly also remove it, and never bothered to make an extra column for it.

And when you have multiple colons in the title field? A quick check shows that I have several hundred books where the original title had more than one colon.

Some Aspects of remote healthcare: Managing the gap in rural and remote health care: A review of the current literature

That one at lease correctly formats the title/subtitles.

Then we have Amazon's approach where they love stuffing the series and other useless bits and bobs into the ebook title.

Children of Time: Winner of the Arthur C. Clarke Award for Best Science Fiction Novel (The Children of Time Novels): A Novel

Since my Kobo ereaders will display the subtitle—if it is worth saving—I do have a subtitle field to save it into for transfer.

[Moonbase59]

Originally Posted by DNSB

And when you have multiple colons in the title field? A quick check shows that I have several hundred books where the original title had more than one colon.

As I say. Probably a bit short-sighted by Calibre to build one opinionated title string from two separate metadata items. Would be better if Calibre had a native, separate column for subtitle.

[DNSB]

Sadly while using a colon to separate the title and subtitle is common usage, I can't find any standard that specifies that usage.

Opinionated? I have seldom thought of a computer program as being able to be opinionated.

I just went through my library. Of 17,142 books, I have 113 with subtitles so ~.66%. Oddly, I found exactly 11 where the ePub3 standard for a subtitle was used:

    <dc:title id="t2">2019 Edition Fully Revised and Updated</dc:title>
    <meta refines="#t2" property="title-type">subtitle</meta>

    <dc:title id="subtitle">A History of Computer Entertainment</dc:title>
    <meta refines="#subtitle" property="title-type">subtitle</meta>

Oddly, for 9 of those, I added that metadata element when editing in Sigil.

[Moonbase59]

I guess it has historical reasons. As far as I remember EPUB2 has subtitle not defined, while EPUB3 has the "refines" like "main" and "subtitle". Since Calibre has been around a long time, it might be doing it this way for compatibility reasons.

As the OP noticed, this may lead to file naming problems with long titles. Calibre shortens long parts of the path to something that is deemed Windoze-compatible, apparently.

I mostly care about EPUB3 nowadays (in spite of my Tolino Vision 5 always complaining "Opening this book can lead to errors. Do you still want to open it?").

[Quoth]

Originally Posted by Moonbase59

As I say. Probably a bit short-sighted by Calibre to build one opinionated title string from two separate metadata items. Would be better if Calibre had a native, separate column for subtitle.

My Calibre library does have a separate subtitle column. Also a series column. This works well with any Kobo ereader. Irrelevant if epub2 or epub3 or kepub format.

If I was fussed on reading on my Kindles I'd do custom template for title combining subtitle and series, as Amazon or publishers now often do. Often when I import azw3 / mobi/ kfx /epub from any source of download I have to cut up and paste the title to title, subtitle and series. Too often the author has to be edited, even if only title case instead of all caps.

[Quoth]

I only use ereaders and apps that work with the metadata. I don't care what the filenames are as long as they are compatible with the target (and many devices use FAT32, so it's that not "windoze" that is the compatibility issue. Also filenames need to work with SMB share copies and backups.
I've had Amazon offer me downloads with characters illegal on FAT32 or Windows NTFS (I'm using Ext4 on Linux) and those work on the PC locally and with Add to Calibre. They won't work with most ereaders or copying to an SMB share (even if it's a Linux rather than Windows server).

For example having a ":" in file or directory name works locally on Linux (and probably MacOS?) but a recipe for disaster. The Calibre programmer sensibly takes a lowest common denominator approach to "Save to disk/device" by default.

[DNSB]

I have unfond memories of one little idiot who insisted on using /\/\ as the first characters of his filenames on MacOS 9 and then whinged when he couldn't save his files to the server share. Evidently, MikeL301 was too fugly.