elibsrv - light EPUB aggregation server

[mateuszv]

Hi there,

Not sure if this is the right place for such announces.. If not, please redirect me where appropriate

I have a collection of ebooks in EPUB format, and I was looking for a software that I could point at the directory where my EPUB files are, and it would scan them, providing a OPDS / HTML interface to access them with some sorting capabilities.

The only tools I found were either too complex for what I need (eg. Calibre, which requires plenty of heavy dependencies), or not standalone (like COPS, that require a Calibre install).

So I created my own tool, named elibsrv. elibsrv is a little tool for Linux (but should work on BSD just as well), that scans a collection of EPUB files from time to time, store metada inside a database, and provides an OPDS and HTML interface to access ebooks.

I published it under a free, open-source license, hoping it might come handy to other people, too.

The project is hosted here:
http://elibsrv.sourceforge.net

I also provide a public 'Demo' with a few public domain ebooks, here (you can access it both with a web browser and an OPDS device):
http://elibsrv-demo.viste.fr

[canpolat]

Nice work. This is very similar to what I had in mind. I am also planing to start hacking something to catalog my books and represent them in a grid view with some filtering options. I don't like Calibre importing my files in it's own folder structure.

In the demo, when I click on a book, it downloads a file named index.php which I couldn't open with my text editor.

[mateuszv]

Originally Posted by canpolat

In the demo, when I click on a book, it downloads a file named index.php which I couldn't open with my text editor.

Hello,

This is normal - what you think is an 'index.php' file is in fact the EPUB book itself. To use it via a web browser, you will have to install an EPUB reader first. I can only recommend using Firefox + the very good "EPUBReader" plugin.

[eschwartz]

Shouldn't the ebook get renamed then?

I don't think EPUBReader will open a book if the file extension is bad. What is the mimetype on the download?

[tuskin]

Windows 8.1 desktop:
Chrome works nicely and opens (or downloads) index.epub
Firefox works but wants to open (or download) index.php
IE thinks it's an RSS feed

MetroIE doesn't work - wants to download or open an xml-file.

tuskin

[mateuszv]

Hi all,

Thanks for all the tests!

BTW, if you wonder why the demo server feels a bit 'sluggish', it's because it runs on an prehistoric 500MHz VIA CPU in my basement

About the MIME type: elibsrv returns files using the mimetype "application/epub+zip". I think (?) this is the right one to use, right?

About browsers that do not want to open files: I don't really know what exactly every browser base its decision on.. I tested only using Firefox + EPUBReader on my Linux desktop, and it works fine.

I will see what I can do to make it easier on other browsers, too.

[mateuszv]

Hello,

I implemented a new way of building URLs inside elibsrv, and released a new version of the software a few minutes ago.

Now, every possible browser should be able to understand what the files are, and how they should be named.

I hope it will work for all of you now.

[eschwartz]

Example download file url is: http://elibsrv-demo.viste-family.net/srv/www/elibsrv-demo.viste-family.net/getfile/3022351156/Bram Stoker - Dracula (en).epub

Seems like your url-building is taking the whole filesystem path over there. Probably not terribly secure to hand people any kind of data about your computer, also it seems wasteful.

But nitpicky details aside, this project looks good!

[mateuszv]

Originally Posted by eschwartz

Seems like your url-building is taking the whole filesystem path over there. Probably not terribly secure to hand people any kind of data about your computer, also it seems wasteful.

Yes, you're obviously right - dunno how I could miss this
That's fixed now. Thanks for pointing this out. Don't hesitate to let me know if you see any other of such 'details', I'll be happy to fix them.

[ersplus]

First Sorry for my bad english...

I love the concept of this software. I can only encourage you to continue in that voice. This is a great alternative to Calibre. Is there a way to install properly on a Synology NAS? Has anyone been successful? Any idea?

[mateuszv]

Originally Posted by ersplus

Is there a way to install properly on a Synology NAS? Has anyone been successful? Any idea?

I have no experience at all with Synology. I read right now on google that it's some kind of Linux-based device, so.. elibsrv might work, but I really doubt that you will be able to install required dependencies on such NAS-oriented thing (a PostgreSQL database, PHP with GD, a gcc compiler, etc...).

An alternative way could be to use the NAS just as a storage, and mount it eg. via NFS on some micro machine with a "normal" Linux. This could be a really nice job for a Raspberry Pi device (I actually have one of these and it works flawlessly under Raspbian!).

[tuskin]

Windows 8.1: Chrome and Firefox open the epubs in the associated program (Calibre viewer). No change in IEs - xml download/RSS feed.

On Android 4.0.4 (Samsung Note 10.1) the stock browser also thinks it's an RSS feed and informs me that Google Reader has been discontinued. Firefox works - opens the books in Aldiko.

tuskin

[mateuszv]

Originally Posted by tuskin

Windows 8.1: Chrome and Firefox open the epubs in the associated program (Calibre viewer). No change in IEs - xml download/RSS feed.

On Android 4.0.4 (Samsung Note 10.1) the stock browser also thinks it's an RSS feed and informs me that Google Reader has been discontinued. Firefox works - opens the books in Aldiko.

I see - you are talking about the fact that the browsers do not even access to the html interface - at first I though it was about interpreting epubs as XML

Now I understand, and it's in fact logic: elibsrv decides what to answer with (OPDS or HTML) looking at how the browser presents itself. I detect Firefox, Opera, Chrome, and probably not much more. When elibsrv don't recognize the client, it assumes it is an OPDS reader (since elibsrv was designed primarily with OPDS devices in mind, the HTML frontend is only a 'bonus' thing I coded).

To avoid this "smart" autodetection, and force HTML output, you could use such URL:
http://elibsrv-demo.viste-family.net/?f=html

I'd be curious to know how the browsers that aren't working for you presents themselves... Would you have any way of checking this? Or maybe could your provide me the three last digits of your (public) IP address, so I'd know what I have to look for in my server logs?

[ersplus]

Originally Posted by mateuszv

I have no experience at all with Synology. I read right now on google that it's some kind of Linux-based device, so.. elibsrv might work, but I really doubt that you will be able to install required dependencies on such NAS-oriented thing (a PostgreSQL database, PHP with GD, a gcc compiler, etc...).

I have post your work in French Nas Forum, I hope that someone will find a way to adapt

Thanks for your work again

[Adminer]

Hi!

Thank you for your code. Did you think to do that on flat file database? I'm thinking to do the same in PHP - but on my windows localhost.
That localhost and ebook portal may be for me and for my family.
I was trying to install your code on UniServerZ from:
http://sourceforge.net/projects/mini...form%20Server/

but my phpadmin when I import sql file into elibsrvdb databese, said an error:

Error
SQL query:
CREATE TABLE books (file VARCHAR NOT NULL, crc32 BIGINT NOT NULL PRIMARY KEY, author VARCHAR NOT NULL, title VARCHAR NOT NULL, language VARCHAR NOT NULL, description VARCHAR NOT NULL);
MySQL said: Documentation

#1064 - You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'NOT NULL, crc32 BIGINT NOT NULL PRIMARY KEY, author VARCHAR NOT NULL, title VARC' at line 1

What do you think about to do this code without any MySQL database - but only on flat file database and txt files?

I'm in razorCMS 2.xx support team and this CMS is perfect to be a frontdoor for your plugin code. I was created many www portals with more then 100 pages - and they works perfect on flat database.

CMS on flat file database - works even on your pendrive - using localhost software. It's the power of easy nice looking skins. Bloging or comments may be there too - you can write a simple article about books, eReaders, etc...
And we may have password protected pages with ebooks. And member area for your family who may wants to download recomended ebooks too. Sorting tools. Pagination. Tree menu. Importing files one by one from folders. Voting script... Everything with the power of PHP and jQuery.

It's may be full private ebook www portal for you and your family with many possibilities - this plugins may have password secured access to the ebook pages, tokens for ebook filenames, members area for downloading/uploading ebooks or to upload new free ebook proposals for me by my friends.

But everything should works on flat file database, and txt files - and should be very easy to run - by one click. MySQL is good, you are buying a simple car - not a bus station ;-)
Flat file database is perfect to use for all home users.
What do you think about? And to create your code portable?

Regards