[Metadata Source Plugin] ANOBII (& inmondadori.it)

[nephtys59]

Thanks for the update, I tested it on a couple of books and it worked fine.
I'll test it further in the next few days and let you know if I find bugs.
Thanks again so much for this plugin!

[dave9000]

There's a small problem with character encoding (windows-1252).
E.g. this sentence:
"... uno dei più famosi scrittori ..."
should be
"... uno dei più famosi scrittori ..."

It looks like the original text is encoded TWICE with UTF-8.
Thanks for the useful plugin anyway!

P.S. the following patch in worker.py seems to fix the issue:
OLD: raw = raw.decode('windows-1252', errors='replace')
NEW: raw = raw.decode('utf-8', errors='replace')

[Pr.BarnArt]

Originally Posted by dave9000

There's a small problem with character encoding (windows-1252).
E.g. this sentence:
"... uno dei più famosi scrittori ..."
should be
"... uno dei più famosi scrittori ..."

It looks like the original text is encoded TWICE with UTF-8.
Thanks for the useful plugin anyway!

P.S. the following patch in worker.py seems to fix the issue:
OLD: raw = raw.decode('windows-1252', errors='replace')
NEW: raw = raw.decode('utf-8', errors='replace')

Thx. I will do a check .
Can you give me title and writer please.
Do you use anobii or bol-it as source?

[dave9000]

Try e.g.:
"Le Case Di Iszm" by "Jack Vance"

I find 4 encoding errors in the description:
nell'impossibilità (nell'impossibilità)
perché (perchè)
C'è (c'è)
però (però)

I used anobii.

[Pr.BarnArt]

There is a new update available: 0.3.0BETA!!
(see the startpage of this topic)

Originally Posted by dave9000

There's a small problem with character encoding (windows-1252).
the following patch in worker.py seems to fix the issue:
OLD: raw = raw.decode('windows-1252', errors='replace')
NEW: raw = raw.decode('utf-8', errors='replace')

Hey Dave, thanks for your reply.
I changed the the decoding for anobii(now utf-8), and didnot change the decoding for bol.it ( windows-1252).
Just because of the diacritics I changed them both to windows-1252 in the previous version 0.2.0.

Diactritics gives a lot of problems and they are not easy to handle. Also because I don't have many references to check, I am dependent of the booktitles and author names that are given to me. So please check and stay replying . References are always welcome.

[dave9000]

Thanks for the update!
I tested it with "Iain M. Banks - Pensa a fleba".
This title is listed in both anobii and bol: the comments are retrieved correctly from both sites.

By the way, it would be nice, in a future update, to skip the downloading of the default cover from bol.
At the moment I just run a script that removes the cover.jpg files with a size of 15941 bytes: I guess there should be a clever way to do that ;-)

[Pr.BarnArt]

Originally Posted by dave9000

By the way, it would be nice, in a future update, to skip the downloading of the default cover from bol.
At the moment I just run a script that removes the cover.jpg files with a size of 15941 bytes: I guess there should be a clever way to do that ;-)

At bol.it there is no default name for the default cover. We have to look at its size. Please give me a title/.athor with the default cover, so I can take a look at it.

[dave9000]

Try:
"Marion Zimmer Bradley - La Signora Delle Tempeste"
"Arthur C. Clarke - Culla"

[Pr.BarnArt]

Originally Posted by dave9000

By the way, it would be nice, in a future update, to skip the downloading of the default cover from bol.
At the moment I just run a script that removes the cover.jpg files with a size of 15941 bytes: I guess there should be a clever way to do that ;-)

I did some testing, but I dont get the size of the cover by headers.
THe next step would be the same as you do, looking at the size of the downloaded no-cover.jpg, i.c. 9361 (using culla)
The properties of the no-cover-image says: 9.024 bytes and size on disk:12.288 bytes.

I can filter out images of that size (including other images of the same size).
But I still doubt if that will work, because you are talking about a size of 15941. That could imply that there are different sizes of the no-cover available image.

BTW culla has a no-cover, but the other title you gave returned a real cover .
Do you know some more titles?

[dave9000]

Hi,
have a look in attachment at the cover.jpg that I get from bol for these titles:

Bernard Blanc - Bad Chili
Brigitte Aubert - I Diari Di Casa March
Carlo Bini - Manoscritto Di Un Prigioniero
Clive Barker - Abarat
David Baldacci - The Camel Club
Emily Bronte - Cime Tempestose
J. G. Ballard - Millennium People
Marion Zimmer Bradley - Dark Satanic
Neal Barrett - Aldair in Albion
Ray Bradbury - Morte a Venice
Robert Bloch - La Sciarpa
Stableford Brian - Il Risveglio Dei Creatori
Umberto Boccioni - Taccuini Futuristi
William P. Blatty - Gemini Killer

Maybe the bol server tries to guess the language of the client browser and returns a localized default cover ...

[Pr.BarnArt]

Hey Dave,

Sorry it took that long, but it was holidaytime for me.
I made some changes to the plugin for you to test.
It skips covers with a size of 9361.
in the configuration of the plugin you can set an extra size, that also will be skipped.

Just install this one (0.3.1.a) from file, then start calibre again in a doscommand window(cmd) with: calibre-debug -g

When you use the plugin this way the dos-window shows you the size that is set in the configuration file and the actual size of the current image.
If your 'no-image-available'-image size is different from 9361 then put that size in the configuration field.


I dont know if there are different sizes, so lets find out.

[dave9000]

Working perfectly, thanks!

[Pr.BarnArt]

Originally Posted by dave9000

Working perfectly, thanks!

Hey, what is the size you are using? Or sizes.

[dave9000]

version 0.3.1.a works correctly with the default size you put into it, I tought that it was 15941 ...

[Pr.BarnArt]

@Dave9000
Default size is 9361 and the extra size option is set to 0 by default.
So if you did not set the extra size, the plugin doesnot use the size 15941.
So, in that case the extra option is not needed.
Thanks for your reply