[FileType Plugin] GetFileName

[BetterRed]

Originally Posted by kovidgoyal

Because they can run in places where there is no db, such as the command line tools on a system with no library.

- thanks - it's always good to know why things are so

BR

[Rellwood]

@all who care...

I like the idea of a set library. I often have too many different versions of the same column throughout old and new libraries...(they multiply worse than rabbits), having one column dedicated throughout the system would be great. Especially when I do a copy to from one library to another and the data is on different columns.

[AnselmD]

Only for information:

After switching to a new library, and configuring the GetFileName (0.0.12) plugin, I must restart calibre, otherwise it does not work.

[Terisa de morgan]

Thank you, @AnselmD, I'm aware of it and I'm trying to solve it.

[wjhill]

Hi folks -- first off, thanks for this plugin, and I wish I'd discovered it before yesterday; I have a large library and it would have been lovely to keep the original filenames for all those books!

After reviewing the plugin and starting to configure it, I realized I might be able to help folks who, like me, had a few questions around details of the process.

As I worked through config, and ensuring the plugin worked, I wrote up a doc in Word and uploaded it to my Google Drive for review:

https://drive.google.com/open?id=0B6...203aHpEZ3dkbnM

(Let me know if you have issues seeing it; it should be set to View for Anyone, and I did test for such on a non-logged in browser.)

This is written against the 0.0.12 version of the plugin, for future reference.

I'm more than open to edits in the next little while, as well as putting it somewhere else more appropriate. I'm also open to putting a FOSS license on it so people can take over updating it, if so moved -- since I can't own/update this doc long-term.

In the meantime, I'm happy to take corrections/edits/improvements to content, although I'd like to avoid major changes to formatting until I have a more permanent home.

[AnselmD]

Originally Posted by kovidgoyal

Optimizing one-shot additions of very large numbers of books has never really been something I have looked at. AS adding themis typically a one-time operation and should be done in smaller batches anyway, that way if something goes wrong/crashes you dont have to redo them all.

I prefer to do it in one batch without making any interaction. if something goes wrong, I am prepared to do it again. I am now able to do it with calibre command line tool, see:

calibredb add --help

With closed GUI, setting environment variable cal_lib to the new and empty library, using Ubuntu (with ext4) I am importing the epub books:

calibredb add -r -d --dont-notify-gui --library-path=$cal_lib .

Using the command line tool is important to finish in a reasonable time!
Additional I used some performance improvements during the import:

Setting the database metadata.db to RAM Disk. Setting the Journalling Mode of Sqlite for the time of the import to Write-Ahead Logging (with ext4).

Importing 106000 epub books (with and without images) of:

Mirroring How-To - Gutenberg
http://www.gutenberg.org/wiki/Gutenb...rroring_How-To

in less than 16 hours. This includes generating al opf backup files. So there is no need to call calibredb backup_metadata after the import.

[Giuseppe Chillem]

Have been any new developments on this plugin ?

[Terisa de morgan]

Aside of some small changes for switching libraries, that I'll try to publish on weekend, no.