Calibre2Web

[prky]

Originally Posted by itimpi

When adding it to the Online Catalogs section of Stanza on the iPhone you did select "Add Stanza Catalog" rather than"Add Web Page"?

Ahhhhhh.



I edited an existing entry I had (when I'd just thrown some e-books up on my web server) which of course had been added as a web page, rather than a Stanza Catalog.

All working now - thanks heaps for your really quick response.

prk.

[prky]

I've been working on entering my library and I've noticed an issue if the title has brackets in it (like these ones).

In this case, the XML info generated by calibre2web doesn't contain the link to the epub file (but does have the summary, cover, etc), and so I don't get a download link while browsing on my iphone.

If I edit the XML data on the web server manually and put the epub link in, I can then download from the iphone (so the epub file is there and working).

Additionally, if I edit the metadata in calibre, remove the brackets from the title, re-run calibre2web and copy the catalogues and books back over to the web server, it then works fine.

Should brackets in a book's title work, or are they considered invalid characters?

Cheers,

prk.

[itimpi]

Anthing that calibre accepts should work.

I will look at it later today to see if I can spot why they are not working - sure to be a minor bug of some sort.

[itimpi]

As expected the failure to accept round brackets in the title was a program bug. A bit of code trying to remove the (#) part from the name scanned for the first bracket and if there was one in the name did the wrong thing. Changing it to scan backwards from the end fixed the problem.

An updated version (0.97) of calibre2Web is now attached to the first post in this thread which fixes this issue.

[prky]

Thanks for the fix - works a treat!

Along the way, I noticed that the new version of calibre2web is creating _CATALOG (without the S), although the readme still says to add the URL as _CATALOGS/catalog.xml:

---
Calibre2Web v0.97

DSN for metadata.db is 'CALIBRE'
Catalog Path is 'C:\Users\prk\Calibre Library\'
Creating Catlog files in 'C:\Users\prk\Calibre Library\_CATALOG'
---

Looks like it doesn't matter which it is, as long as the URL added in the iPhone and the folder path on the server match.

For the moment I've just symlinked _CATALOGS to _CATALOG and all works fine.

Not sure whether it's meant to be _CATALOGS or _CATALOG, but might be worth updating either the program or readme to be consistent.

Again - thanks heaps for your work on this, and the really fast turnaround on the bug fix - I really appreciate it.

prk.

[itimpi]

It is meant to be _CATALOGS. I had changed it to _CATALOG for test purposes aned forgot to change it back. I will upload a fixed 0.97 (I do not think it is worth upgrading the version number for this).

[artbatista]

I have come to one annoyance that I hope you can fix. It has to do with case sensitivity.

Say you import a book in calibre:

Fake Author - Fake title.epub

the you change title in calibre to:

Fake Author - Fake Title.epub

Calibre ignores the changed case of the title (in Windows anyway) and it changes the title in the metadata, but it does not change the case in the actual filename.

Now when I upload the stanza directory to my Apache server under linux, the download is broken because of the disparity in case between the catalog filename (I assume you get the filenames from the metadata) and the actual filename.

It may be only a problem if you run calibre in windows and then upload the files to an apache /linux server, but it sure annoys the hell out of me.

I bugged this to Kovid, but he said he won't fix/it's not a problem. Perhaps you could adjust the script to check the actual filename?

Thank you for your time.

Art

[itimpi]

I am afraid that I have not found a way to handle this in the script. As you say in Windows you cannot easily detect a filename case mismatch.

The way I personally handle this is to temporarily add a character to the title in Calibre. I then remove it when I have the case correct. This sequence causes renames that correct the case issue.

[dpierron]

I made a Java tool heavily inspired by Calibre2Web... I hope it'll help those without Windows - me included

[dpierron]

I released a new version of my calibredbxtract tool, which adds support for generating authors, series and recent additions catalogs.

[Yorkiesman]

Calibre2web is working on my website and created a layout which runs perfect from Mozilla but Stanza comes up with the following error when I had the catalog:-

Catalog Error
Error reading catalog data. Please try again later.

I have a removed everything from the library and added 2 simple epub to find the fault the books have standard titles and authors no punctuation.

I'm stumped if anybody's got an idea.

[itimpi]

Not sure - that error means one of two thing:
- There is an error in the XML. To catch this I normally load the page into IE as it flags up any XML errors.
- The URL is wrong for Stanza. The URL is likely to be case significant (depending on your web server) so that is worth checking. Also make sure it is set up as a "atalog" rather than as a "web page".

[Yorkiesman]

Thanks for your reply itimpi. I tried running from another webserver and it worked. It's a problem with access from the webserver but I cannot find what is wrong.

[fender]

Thank you very much!
I created a catalog of books in portuguese
http://blogoio.com/stanza/list/catalog.xml


from Brasil

[itimpi]

Glad you liked it!

Note that there is also the Calibre2opds program (sticky topic at the top of the forum) which is superseding calibre2web. Calibre2opds is now probably richer in function (it generates both XML and HTML catalogs) and being written in java is cross-platform.