Due to MAL html update.

Specifically:

def self.top(options = {})

      page = options[:page] || 1
      limit = (page.to_i - 1) * 30
...

Referenced ticket in MyAniList: https://github.com/croberts22/MyAniList/issues/650

cURL requests throughout the API are unnecessarily bulky. This could be rectified by having a function generate preset cURL requests without having to always specify the environment or interface.

Could we just remove the auth field? Not sure what the point of it is.

Related to #18.

Unfortunately there is no separation of titles within a section (such as "synonyms") by DOM element, so it's not possible to differentiate between commas as part of the title and commas used as a separator.

However, for the English title there is only expected to be one title. Therefore, this issue could be fixed for just the English title by not splitting the text by comma and just returning a single-element array.

As of a day or two ago, no synopses are being retrieved by the API.

This is probably related to #27; triage and see what URLs are showing up. We may be getting a 403.

Scraping the comma separated list of producers would be nice to have.

Contact Crave to get IP address(es) whitelisted.

#12 originally solved this issue, but it looks like the HTML formatting is causing a regression.

MAL has recently added a "Premiered" field in the Information section of an anime page. For example: Premiered: Fall 2014.

Related anime/manga are failing to parse, again due to the new structure of MAL's HTTP. Revisit the regex.

It looks like the rank is pulling in additional information, such as the superscript of '2'. Here are two examples that have bad data:
http://umal-api.coreyjustinroberts.com/1.1/anime/23281
http://umal-api.coreyjustinroberts.com/1.1/anime/4382

Newlines aren't represented correctly when sending JSON back.

Due to #41.

It seems that some pages on MAL might still be using an older layout and as such the members_scores for older shows aren't populating correctly.
http://myanimelist.net/anime/11049 returns a members_score of null.
http://myanimelist.net/anime/28999/Charlotte returns the members_score correctly.

I dug around a bit but afraid my skills just aren't good enough to fix this, would you please look into this?

Get the following error when trying to retrieve any anime:

Have tried rolling back to an earlier version but the problem remains. Suspect that something on one of the pages has changed. Error can be reproduced by going to any anime link.

Unfortunately my Ruby skills isn't very good and I don't have time to play with it right now.

Also due to an HTML change. Should be a fairly simple fix.

It looks like there were some redesigns on anime pages.
http://umal-api.coreyjustinroberts.com/1.1/anime/6547 returns:
{"error":"unknown-error","details":"Exception message: Error scraping anime with ID=6547. Original exception: undefined method text' for nil:NilClass."}`

This is a big project. I would like to move it to Django via Heroku for easier maintenance.

Currently only the total member count is scraped, but it would be good to also scrape the number of users that have voted. This is in parentheses after the score.

I think it would be good to scrape the user recommendations for the anime page. That is, just the title, URL, and anime id. This will allow links to the recommended anime to be added to the iPhone app, similar to prequels, sequels, etc.

For example:
http://myanimelist.net/anime/1689/Byousoku_5_Centimeter
http://umal-api.coreyjustinroberts.com/1.1/anime/1689

This is due to MAL changing up the HTML a bit, making Nokogiri scanning the wrong item. Instead of looking for <h2>Related [Anime|Manga]</h2>, it should be </div>Related [Anime|Manga]</h2>.