Issues pull requests apiv3

[janinko]

Overview
I cleaned up GHIssue and GHPullRequest and make them relevant to api v3

Added SmallUser representing reference to user. In api this user is represented by:

    "user": {
        "login": "janinko",
        "avatar_url": "https://secure.gravatar.com/avatar/bf8425ee432f7e9748f9e1c540a125ed?d=https://a248.e.akamai.net/assets.github.com%2Fimages%2Fgravatars%2Fgravatar-user-420.png",
        "url": "https://api.github.com/users/janinko",
        "gravatar_id": "bf8425ee432f7e9748f9e1c540a125ed",
        "id": 644267
    },

Added DetailedPullRequest - when retrieving pull request by id, it has more attributes then pull requests obtained by GHRepository.getPullRequests(). The attributes are:

merged_by - SmallUser
review_comments - int
additions - int
merged - boolean
mergeable - Boolean
deletions - int
mergeable_state - String ("unknown", "dirty", "clean")
changed_files - int

Possible issues
I'm not sure if change in b40677a is OK. I changed return value of GHRepository.getPullRequest(int) from GHPullRequest to its subclass GHDetailedPullRequest. IMO it shouldn't be causing any problems.

[buildhive]

Kohsuke Kawaguchi » github-api #26 SUCCESS
This pull request looks good
(what's this?)

[kohsuke]

Thank you for adding this.

As you noted, in various places, GitHub V3 API distinguishes shallow data representation vs full data representation. I've been meaning to deal with this by adding lazy population support --- the object will populate itself if a field is requested that it doesn't have yet.

I think that is easier to use for the client, and we can avoid having multiple classes for conceptually the same thing.

I've used your change as the opportunity to implement it, and I've updated your change accordingly. Specifically, I got rid of GHSmallUser in favor of GHUser, and ditto for GHDetailedPullRequest.

I've released 1.32 with this change.