bambi73 Wrote:Yes, I welcome any reasonable improvements ... which doesn't conflict with my ideas .
Please add some note what did you changed because comparing it with my changes will be PITA.
And i don't think scraper parser environment is so bad, it does good job for regular usage, only maybe AniDB scraper is bit too big monster for it
Back when i actively developed this scraper i thought about some Python support and already did some preliminary work on it, but it always ended like "Too much work with unsure results" .
Right now AniDB scraper works fine and fullfil all my needs so it's enough for me.
well I mean, parsing xml/html with a regex itself is badass. There were changes I wanted to do, that were quite difficult. I'll list an example of a change I want, I don't know how to do.
How did you do the python work? I didn't know the scraper system in xbmc exposed had any hooks for python. (I don't have xbmc building on my machine, so I havent really seen how the plugin system works). If it's really possible, I dont mind working on the port. I'm sure it will be way better than struggling with regex and these continuous transformations.
so to merge the 2 xmls I used this tool
http://tools.decisionsoft.com/xmldiff.html. It seems pretty good.
In the anidb.xml you listed to me yesterday, I saw this regexp
(?i)<a href="
http://anidb\.net/perl-bin/animedb\.pl\?...\d+)"[^>]*>(.*?)</a>
after show=anime you have &\; instead of &. Any idea why?
I have 2 changes I have made.
1) for the fanart lookups in addition to the main title, official english, synonym x-jat and first synonym en, I have added the short en. (I needed this for madoka. There is a better solution, so let me talk about that)
2) the real change I made was to be able to import the other special episodes into xbmc. Trailers, Parodies, Other, and Ops and Eds that are all recognized by anidb. But the solution I have isnt very nice.
xbmc stores season and episode number as integers. So I have mapped episodes like T01 and P01 to very high season numbers. This works for me. Sadly the episodes get displayed by xbmc as being season 115 and so on. Tell me if you are interested in these changes.
There are 5 changes I want to make, but couldnt quite figure out.
1) For some reason the anidb image that shows up doesnt scale properly in the "tv show properties" window. I'm wondering how this logic works, especially because I'd also like to use this with the mediaInfo 2 view of xbmc.
2) I'd like to use the anidb image scaled as if it was fanart, when there is no fanart on tvdb. Is this possible?
3) Some searches on tvdb have the right result but in the wrong order. Again here is where a general purpose language like python would be better. I have a show called a.li.ce. The tvdb result returns alice as the first and a.li.ce further down. but it gets the fanart of the first show which is wrong. Maybe ranking the options by the best match would help.
4) the search for tvdb url runs through some options, but not all options. Specifically, I think it should run through all the english synonym titles, not just the first one. For example the current show Mahou Shoujo Madoka Magika, the api returns the first english synonym as pmagi, which is not on tvdb. But some of the other synonyms are. Is it possible to select all synonyms? I had to add the option of using the short english title because I could not figure out how to use all synonyms.
5) The last piece is about anime movies. Sadly tvdb doesn't have any fanart for them. Do you have any ideas of how we might find some fanart for them?