MichaelCurrin/README.md

## README.md

      
    Raw
  

              README.md
            
          
    Get Pull Requests using GH's GraphQL API

How to get Pull Requests data using Github in the browser, or using the API to allow for automating reporting or building in values into a website.
Resources


GitHub GQL API: https://developer.github.com/v4/
Explorer: https://developer.github.com/v4/explorer/
Github community topics - these were used to understand the syntax needed.

GraphQL Get all pull request for an org
GraphQL Open Pull Requets only


Queries

There some approaches followed in this gist.
Get Pull Requests for the user

Use the pullRequests object on the viewer (current user). And showing the repo associated. This is not so clean.
Or get PRs withing repos of the view.
Search Pull Requests against other repos

The approaches above are not so flexible.
This approach uses a search that one can do in the browser on Github and use that query in a search function in the GQL API.

Click Pull Requests on Github or go to https://github.com/pulls .
Change the search parameters or enter your own search.
Use that search for the search GQL file in this gist.

Note that type should be ISSUE to get Pull Requests. Alternatively, get USER or REPOSITORY objects for other searches.
The default order is newest first. You can use oldest first with sort:created-asc.


Here are sample searches:

Default:

Get all own PRs for my user, whether for my own repos or not. Includes private repos by default and excludes archived repos.
is:open is:pr author:MichaelCurrin archived:false


Get all my contributions to public repos not owned by me:

is:merged is:pr is:public archived:false author:MichaelCurrin -user:MichaelCurrin
Search results on Github: github.com/pulls?q=is%3Amerged+is%3Apr+...


PRs in atom/atom and atom/github after a given data:

is:pr created:>2019-04-01 repo:atom/atom repo:atom/github


Running queries

Explorer

Use the explorer linked in Resources, copy and paste a query and run.
Command-line

Request

How to run a query using curl and an auth token.
curl -H "Authorization: bearer token" -X POST -d " \
 { \
   \"query\": \"query { viewer { login }}\" \
 } \
" https://api.github.com/graphql
See Forming calls in the docs for details.
Results

See below for simplified sample data for search query described earlier.
{
  "issueCount": 1,
  "edges": [
    {
      "node": {
        "number": 105,
        "title": "fix: Allow config flag to take a parameter",
        "repository": {
          "nameWithOwner": "docsifyjs/docsify-cli"
        },
        "createdAt": "2020-05-17T18:59:33Z",
        "mergedAt": "2020-05-18T08:39:27Z",
        "url": "https://github.com/docsifyjs/docsify-cli/pull/105",
        "changedFiles": 1,
        "additions": 1,
        "deletions": 1
      }
    }
  ]
}
Future development

I like the search approach so want to use that more.

The query can be parametized - send JSON data as the search query. Unfortunately the query has to be prepated in JSON - you can't send just username as an argument (unless GQL has a way to concatenate strings).
Adding paging for higher volume searches.
Extend to search for other types not just PRs.
Add a parser to turn the data in a CSV, or embed on a website.


## my_prs.gql
{
  viewer {
    pullRequests(first: 100, states: CLOSED, orderBy: {field: UPDATED_AT, direction: DESC}) {
      nodes {
        repository {
          name
          isPrivate
        }
      }
    }
  }
}

## my_prs_by_repo.gql
{
  viewer {
    repositories(first: 5) {
      nodes {
        name
        pullRequests(last: 5) {
          nodes {
            number
            url
            state
            reviews(last: 100) {
              nodes {
                author {
                  login
                }
                state
              }
            }
          }
        }
      }
    }
  }
}

## search_my_merged_prs_on_public_repos.gql
{
  search(query: "is:merged is:pr is:public archived:false author:MichaelCurrin -user:MichaelCurrin", type: ISSUE, first: 100) {
    issueCount
    edges {
      node {
        ... on PullRequest {
          number
          title
          repository {
            nameWithOwner
          }
          createdAt
          mergedAt
          url
          changedFiles
          additions
          deletions
        }
      }
    }
  }
}
	{
	viewer {
	pullRequests(first: 100, states: CLOSED, orderBy: {field: UPDATED_AT, direction: DESC}) {
	nodes {
	repository {
	name
	isPrivate
	}
	}
	}
	}
	}
	{
	viewer {
	repositories(first: 5) {
	nodes {
	name
	pullRequests(last: 5) {
	nodes {
	number
	url
	state
	reviews(last: 100) {
	nodes {
	author {
	login
	}
	state
	}
	}
	}
	}
	}
	}
	}
	}
	{
	search(query: "is:merged is:pr is:public archived:false author:MichaelCurrin -user:MichaelCurrin", type: ISSUE, first: 100) {
	issueCount
	edges {
	node {
	... on PullRequest {
	number
	title
	repository {
	nameWithOwner
	}
	createdAt
	mergedAt
	url
	changedFiles
	additions
	deletions
	}
	}
	}
	}
	}