NAME

Pithub::Base - Github v3 base class for all Pithub modules

VERSION

version 0.01017

DESCRIPTION

All Pithub modules inherit from Pithub::Base, even Pithub itself. So all attributes listed here can either be set in the constructor or via the setter on the objects.

If any attribute is set on a Pithub object, it gets automatically set on objects, that get created by a method call on the Pithub object. This is very convenient for attributes like the "token" or the "user" and "repo" attributes.

The "user" and "repo" attributes are special: They get even set on method calls that require both of them. This is to reduce verbosity, especially if you want to do a lot of things on the same repo. This also works for other objects: If you create an object of Pithub::Repos where you set the "user" and "repo" attribute in the constructor, this will also be set once you get to the Pithub::Repos::Keys object via the keys method.

Examples:

    # just to demonstrate the "magic"
    print Pithub->new( user => 'plu' )->repos->user;          # plu
    print Pithub::Repos->new( user => 'plu' )->keys->user;    # plu

    # and now some real use cases
    my $p = Pithub->new( user => 'plu', repo => 'Pithub' );
    my $r = $p->repos;

    print $r->user;    # plu
    print $r->repo;    # pithub

    # usually you would do
    print $r->get( user => 'plu', repo => 'Pithub' )->content->{html_url};

    # but since user + repo has been set already
    print $r->get->content->{html_url};

    # of course parameters to the method take precedence
    print $r->get( user => 'miyagawa', repo => 'Plack' )->content->{html_url};

    # it even works on other objects
    my $repo = Pithub::Repos->new( user => 'plu', repo => 'Pithub' );
    print $repo->watching->list->first->{login};

ATTRIBUTES

auto_pagination

api_uri

Defaults to https://api.github.com.

Examples:

    my $users = Pithub::Users->new( api_uri => 'https://api-foo.github.com' );

    # ... is the same as ...

    my $users = Pithub::Users->new;
    $users->api_uri('https://api-foo.github.com');

jsonp_callback

If you want to use the response directly in JavaScript for example, Github supports setting a JSONP callback parameter.

Examples:

    my $p = Pithub->new( jsonp_callback => 'loadGithubData' );
    my $result = $p->users->get( user => 'plu' );
    print $result->raw_content;

The result will look like this:

    loadGithubData({
        "meta": {
            "status": 200,
            "X-RateLimit-Limit": "5000",
            "X-RateLimit-Remaining": "4661"
        },
        "data": {
            "type": "User",
            "location": "Dubai",
            "url": "https://api.github.com/users/plu",
            "login": "plu",
            "name": "Johannes Plunien",
            ...
        }
    })

Be careful: The content method will try to decode the JSON into a Perl data structure. This is not possible if the jsonp_callback is set:

    # calling this ...
    print $result->content;

    # ... will throw an exception like this ...
    Runtime error: malformed JSON string, neither array, object, number, string or atom,
    at character offset 0 (before "loadGithubData( ...

There are two helper methods:

clear_jsonp_callback: reset the jsonp_callback attribute
has_jsonp_callback: check if the jsonp_callback attribute is set

per_page

By default undef, so it defaults to Github's default. See also: http://developer.github.com/v3/#pagination.

Examples:

    my $users = Pithub::Users->new( per_page => 100 );

    # ... is the same as ...

    my $users = Pithub::Users->new;
    $users->per_page(100);

There are two helper methods:

clear_per_page: reset the per_page attribute
has_per_page: check if the per_page attribute is set

prepare_request

This is a CodeRef and can be used to modify the HTTP::Request object on a global basis, before it's being sent to the Github API. It's useful for setting MIME types for example. See also: http://developer.github.com/v3/mimes/. This is the right way to go if you want to modify the HTTP request of all API calls. If you just want to change a few, consider sending the prepare_request parameter on any method call.

Let's use this example from the Github docs:

Html

application/vnd.github-issue.html+json

Return html rendered from the body's markdown. Response will include body_html.

Examples:

    my $p = Pithub::Issues->new(
        prepare_request => sub {
            my ($request) = @_;
            $request->header( Accept => 'application/vnd.github-issue.html+json' );
        }
    );

    my $result = $p->get(
        user     => 'miyagawa',
        repo     => 'Plack',
        issue_id => 209,
    );

    print $result->content->{body_html};

Please compare to the solution where you set the custom HTTP header on the method call, instead globally on the object:

    my $p = Pithub::Issues->new;

    my $result = $p->get(
        user     => 'miyagawa',
        repo     => 'Plack',
        issue_id => 209,
        options  => {
            prepare_request => sub {
                my ($request) = @_;
                $request->header( Accept => 'application/vnd.github-issue.html+json' );
            },
        }
    );

    print $result->content->{body_html};

repo

This can be set as a default repo to use for API calls that require the repo parameter to be set. There are many of them and it can get kind of verbose to include the repo and the user for all of the calls, especially if you want to do many operations on the same user/repo.

Examples:

    my $c = Pithub::Repos::Collaborators->new( repo => 'Pithub' );
    my $result = $c->list( user => 'plu' );

There are two helper methods:

clear_repo: reset the repo attribute
has_repo: check if the repo attribute is set

token

If the OAuth token is set, Pithub will sent it via an HTTP header on each API request. Currently the basic authentication method is not supported.

ua

By default a LWP::UserAgent object, but it can be anything that implements the same interface.

user

This can be set as a default user to use for API calls that require the user parameter to be set.

Examples:

    my $c = Pithub::Repos::Collaborators->new( user => 'plu' );
    my $result = $c->list( repo => 'Pithub' );

There are two helper methods:

clear_user: reset the user attribute
has_user: check if the user attribute is set

It might make sense to use this together with the repo attribute:

    my $c = Pithub::Repos::Commits->new( user => 'plu', repo => 'Pithub' );
    my $result = $c->list;
    my $result = $c->list_comments;
    my $result = $c->get('6b6127383666e8ecb41ec20a669e4f0552772363');

METHODS

request

This method is the central point: All Pithub are using this method for making requests to the Github. If Github adds a new API call that is not yet supported, this method can be used directly. It accepts an hash with following keys:

method: mandatory string, one of the following:
- DELETE
- GET
- PATCH
- POST
- PUT
path: mandatory string of the relative path used for making the API call.
data: optional data reference, usually a reference to an array or hash. It must be possible to serialize this using JSON::Any. This will be the HTTP request body.
options: optional hash reference to set additional options on the request. So far prepare_request is supported. See more about that in the examples below. So this can be used on every method which maps directly to an API call.
params: optional hash reference to set additional GET parameters. This could be achieved using the prepare_request in the options hashref as well, but this is shorter. It's being used in list method of Pithub::Issues for example.

Usually you should not end up using this method at all. It's only available if Pithub is missing anything from the Github v3 API. Though here are some examples how to use it:

Same as "list" in Pithub::Issues:

    my $p      = Pithub->new;
    my $result = $p->request(
        method => 'GET',
        path   => '/repos/plu/Pithub/issues',
        params => {
            state     => 'closed',
            direction => 'asc',
        }
    );

Same as "get" in Pithub::Users:

    my $p = Pithub->new;
    my $result = $p->request(
        method => 'GET',
        path   => '/users/plu',
    );

Same as "create" in Pithub::Gists:

    my $p      = Pithub->new;
    my $method = 'POST';
    my $path   = '/gists';
    my $data   = {
        description => 'the description for this gist',
        public      => 1,
        files       => { 'file1.txt' => { content => 'String file content' } }
    };
    my $result = $p->request(
        method => $method,
        path   => $path,
        data   => $data,
    );

Same as "get" in Pithub::GitData::Trees:

    my $p       = Pithub->new;
    my $method  = 'GET';
    my $path    = '/repos/miyagawa/Plack/issues/209';
    my $data    = undef;
    my $options = {
        prepare_request => sub {
            my ($request) = @_;
            $request->header( Accept => 'application/vnd.github-issue.html+json' );
        },
    };
    my $result = $p->request(
        method  => $method,
        path    => $path,
        data    => $data,
        options => $options,
    );

This method always returns a Pithub::Result object.

AUTHOR

Johannes Plunien <plu@cpan.org>

COPYRIGHT AND LICENSE

This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5 programming language system itself.

CPAN RT