Web Services [DONE/DENIED]

old granted and denied feature requests

Moderator: AniDB

Locked
chuhwi
Posts: 2
Joined: Mon Oct 04, 2004 8:07 pm

Web Services [DONE/DENIED]

Post by chuhwi »

My request is that you create some sort of Web Services interface to anidb.net (not necessarily based on something fancy like SOAP; simple GET url's that return XML would suffice). In particular, I would like to be able to search the database by name, retrieve animedb's AID for an anime and obtain basic information on that anime (features requiring user authentication such as anime list editing would probably be overkill).

Because a given anime will often be known by several titles, and may have various TV, OVA, and movie versions, it is currently extremely difficult to uniquely identify it (consider the usefulness of books' ISBN numbers). While within any given database (anidb.net, animenfo, etc.) this problem is solved, providing a machine-accessible, global method of identifying anime (based on animedb's ID's for anime) would prove extremely useful.

Such a system would have numerous applications. For example, it would enable the creation of a system to automatically catalog and normalize a collection of anime video-files. Additionally, it would enable bittorrent sites to provide anidb AID's to accompany files, allowing clients to automatically recognize and download desired series (this is currently possible with animesuki to an extremely limited extent via their xml feed).

It would probably take less than an hour to enable searching in this manner based on an arbitrary xml format, and it would be really, really useful.

For an example of a roughly similar site that provides (much more complicated) web services, see the book website allconsuming.net
Skywalka
Posts: 889
Joined: Tue Sep 16, 2003 7:57 pm

Post by Skywalka »

Errm... normalize Anime files? Isn't that was AoM is for? AoM's goal is to have a somewhat offline version of AniDB available. Files are currently identified via their MD5, CRC32 and SHA1 hash, I think it is not necessary to build another client which uses an identification method that is not as exact as the hashes.

Additionally usually a file is not in AniDB for a couple of hours/days, untill it is added manually by a user. Currently having a BitTorrent .torrent file accompanied by a link to AniDB is easy, simply add the file yourself, share the file and link to the AniDB entry via a weblink. Once you have the file downloaded you take it out of the BT client and then AoM automatically renames it. That's what I do every day.

Maybe you could be a little more specific. What is the advantage of services like allconsuming.net compared to what the webinterface of anidb.net provides at the moment? "Downloading entire series" via an AnimeID - I don't understand that criterium. You want a client where you start downloading ONE file via BT and then the client should find all the other available BTs that are available for that Anime? Something like an automated download client which you simply tell that you like this and that series and it automatically gets more files of the same kind?

Sounds funny. Somewhat like TiVo.

But I honestly don't think this is really possible. First of all I AniDB does not catalogue .torrents, just ed2k links. This way you would only be able to mass-add ed2k links/files to your ed2k client. Currently AniDB does not support mass-add of a series by a certain group, so I guess EXP would not support such a client.

Currently there is a new client in development for notification (UDP API), I guess that is the thing that EXP is interested in the most at the moment, a way for automatically informing the users that something new arrived without them hammering the webserver all the time. Getting the files should still be the job of the user I guess.
chuhwi
Posts: 2
Joined: Mon Oct 04, 2004 8:07 pm

Post by chuhwi »

AoM and the closed API do not fill meet my desire since they don't allow anyone the creation of new applications using animedb. The UDP api is impractical for potential uses I see since it focuses on file hashes (and therefore specific, known files) to the exclusion of more useful, general information, and uses its own protocol rather than simple XML.

I am more interested in simple access to the more general, comprehensive information anidb provides about anime series rather than specific released files (which is obviously somewhat brittle), and for this reason the scope of my proposed API is quite different than that of the current UDP API.

With regards to automatic downloading of series, I was proposing a system by which trackers and other servers, with minimal effort, could provide a system that would be far more flexible than attempting to have all the intelligence on the end of either the client or anidb.

I don't know all the uses such an interface could be put to, but I feel that if it existed it would be used in creative ways. At any rate, I feel it would serve a different purpose from the current API since it would be more general and much simpler to program for.
exp
Site Admin
Posts: 2438
Joined: Tue Oct 01, 2002 9:42 pm
Location: Nowhere

Post by exp »

I will think about this.

BYe!
EXP
exp
Site Admin
Posts: 2438
Joined: Tue Oct 01, 2002 9:42 pm
Location: Nowhere

Post by exp »

BTW,

do you know any pages and tools which would actually use such a feature? url?
or are you mainly speculating?

BYe!
EXP
Forse
Posts: 1
Joined: Thu Oct 07, 2004 8:05 pm
Contact:

Post by Forse »

exp wrote:BTW,

do you know any pages and tools which would actually use such a feature? url?
or are you mainly speculating?

BYe!
EXP
If you created xml for say series pages then any program could read data from it. Okey let's say you made series info available via xml then I could for example fetch that info and display it...well anywhere. Other websites could benefit from this as they could display anime info on their website.

If you were to create xml I know my website would benefit from it at least. You could make available only basic info (series name,kanji,year,genre) and for more info users would have to go to anidb.net which would bring more traffic to ya. Another benefit is that websites/programs would display "info privided by anidb.net" or something like this which would raise awareness about anidb.

You would only need to make one file which fetches row from database and formats it according to XML standarts. This can be very easily done with perl/php.
Guest

Post by Guest »

I was just speculating, but I know that I would definitely use it at least.
exp
Site Admin
Posts: 2438
Joined: Tue Oct 01, 2002 9:42 pm
Location: Nowhere

Post by exp »

Well,

I'd like to see some more concrete usages, i.e. a list of pages and programms which would make use of such a feature and in what way.

BYe!
EXP
fahrenheit
AniDB Staff
Posts: 438
Joined: Thu Apr 08, 2004 1:43 am
Location: Portugal

Post by fahrenheit »

i didn't quite understood the thread, but what i think, is that one of the uses of this that is being talked about in the thread is to implement some away to retrieve up-to-date information on a anime, being that information:

title, jap title, english title, synonyms, genres, type, episodes, year, producer, website, rating, reviews and a link for anidb for complete info.

this i supose would acess a xml of the anime pages instead of the:

http://anidb.ath.cx/perl-bin/animedb.pl ... e&aid=1554

by

http://anidb.ath.cx/perl-bin/animedb.pl ... l&aid=1554

so someone with a parser could display that information on their website for example to give a description of the animes they like, a bit like the export templates but with allways up-to-date information.

Or I could have missed complety the thread and its not anything of this..
Elberet
Posts: 778
Joined: Sat Jul 19, 2003 8:14 pm

Post by Elberet »

Wouldn't this be rather counter productive? IIRC, it's unwanted to access the AniDB web interface by means other then a webbrowser in order to e.g. spider the anime list and reuse it for other purposes.

Creating an XML interface or extending the current webinterface so that it creates XML pages containing the usual information upon a special request flag would not only make it much easier to fetch stuff from AniDB and reuse it somewhere else, it would also sanctionize it.

However... Such a feauture would be quite useful to me even. I was thinking about developing a script for my homepage that takes an AniDB XML mylist export, reads it into a local mysql database and displays this information within my own website. This would allow me to add my own comments, links or other material to anime entries without having to write full-fledged reviews for AniDB - but so far, having to update the local database via manually downloaded export archives was a major turn off.
SonicFX
Posts: 1
Joined: Mon Dec 04, 2006 10:37 pm

Post by SonicFX »

I'm a little bit confused... Title Says [Done/Denied]... is it Done or is it Denied? I was searching this Forum to find some ways to request Informations whitout Socket-Connection (which are hard to implement for me)... I Realy would like such a XML-Relpy... But can't find information if its implemented or not.
When I change the url to http://anidb.info/perl-bin/animedb.pl?show=xml&aid=... the reply is a XML document but only a error message in it...

Is that a implemented feature?
If yes... Is there a tread about it or a documentation?

Thx Anyway...
SonicFX
Locked