Multiple calls to the GitHub API

Question

I have the following class:

public class IssueWithComments : Issue
{
    public IReadOnlyList<IssueComment> FullComments { get; set; }

    public IssueWithComments()
        : this(new List<IssueComment>())
    {
    }

    public IssueWithComments(IEnumerable<IssueComment> comments)
    {
        FullComments = comments.ToList();

    }

    public static IssueWithComments FromIssue(Issue issue, IEnumerable<IssueComment> comments)
    {
        return new IssueWithComments(comments) 
        {
            Assignee = issue.Assignee,
            Body = issue.Body,
            ClosedAt = issue.ClosedAt,
            Comments = issue.Comments,
            CreatedAt = issue.CreatedAt,
            HtmlUrl = issue.HtmlUrl,
            Labels = issue.Labels,
            Milestone = issue.Milestone,
            Number = issue.Number,
            PullRequest = issue.PullRequest,
            State = issue.State,
            Title = issue.Title,
            UpdatedAt = issue.UpdatedAt,
            Url = issue.Url,
            User = issue.User
        };
    }
}

where Issue is a type defined in the Octokit.Net namespace.

I populate my objects by calling the following code:

public static async Task<IReadOnlyList<IssueWithComments>> GetAllForRepositoryWithComments(this IIssuesClient client, string owner, string name)
{
     var issues = await client.GetAllForRepository(owner, name);
     List<IssueWithComments> all = new List<IssueWithComments>();

     foreach (var i in issues)
     {
         all.Add(IssueWithComments.FromIssue(i, await client.Comment.GetAllForIssue(owner, name, i.Number)));
     }

     return all;
 }

This works, but just takes 7.640 seconds on average according to System.Diagnostics.Stopwatch (plus that I can see it taking a long time to load).

Task.WhenAll could work, but the way to find the Issue it is referring to is as follows:

x.FullComments.First().HtmlUrl.Segments[4]

Is there a better way to either run the loop or should I go with Task.WhenAll and drill into the URL?

almaz · Accepted Answer · 2015-06-08 11:12:03Z

2

Since you're loading all comments for all issues, I think

await client.Comment.GetAllForRepository(owner, name)

is what you're looking for. Otherwise, if you do need to use GetAllForIssue, try using Task.WhenAll to run comments retrieval in parallel:

public static async Task<IReadOnlyList<IssueWithComments>> GetAllForRepositoryWithComments(IIssuesClient client, string owner, string name)
{
    var issues = await client.GetAllForRepository(owner, name);

    return await Task.WhenAll(issues.Select(async i => IssueWithComments.FromIssue(i, await client.Comment.GetAllForIssue(owner, name, i.Number)));
}

answered Jun 8, 2015 at 11:12

almaz

7,27219 silver badges30 bronze badges

\$\begingroup\$ The first option would not work, as there is still no way to comfortably map the results to the appropriate issue. Your code - while very marginally slower then a fix I was working on - is much cleaner so you get the accept. \$\endgroup\$

Mord Zuber
– Mord Zuber

2015-06-08 13:13:45 +00:00
Commented Jun 8, 2015 at 13:13
\$\begingroup\$ By the way this cuts it down from the 7.640 seconds mentioned in the question to an average if 1.5 seconds \$\endgroup\$

Mord Zuber
– Mord Zuber

2015-06-08 13:15:38 +00:00
Commented Jun 8, 2015 at 13:15

Add a comment |

Stack Exchange Network

Multiple calls to the GitHub API

1 Answer 1

You must log in to answer this question.

Hot Network Questions

Multiple calls to the GitHub API

1 Answer 1

You must log in to answer this question.

Related

Hot Network Questions