Optimize github anno-prs and github checkout-prs for the case of 500+ PRs in a repo

[PawelLipski]

As for now, all PRs are first fetched, even for git machete github checkout-prs <single-PR-number>, for the sake of traversing the PR chain upwards. This takes time due to the pagination limits (100 PRs at most in a single page, as of Sep 2023).

This is rather hard to optimize reasonably... we can consider fetching PRs for the matching set of head branches and/or author (only first match the PRs that have the given author).
/search/issues (v3) REST API cannot be used for that purpose as it does not include PR-specific fields that we can't do without (like head and base branches), only the stuff that's common to issues and PRs. Also, there's no /search/pulls.

Sample relevant v3 query:

curl -sL -H"Authorization: Bearer $(gh auth token)" https://api.github.com/search/issues?q=is:pr+state:open+repo:VirtusLab/git-machete+author:PawelLipski | jq -C | less -r

Thus, our only chance is GraphQL (v4) API. A sample query:

curl -sL -H"Authorization: Bearer $(gh auth token)" https://api.github.com/graphql -d "{\"query\": $(jq -R -s . < query.graphql)}" | jq -C | less -r

where query.graphql has:

{
  search(query: "is:pr author:PawelLipski repo:VirtusLab/git-machete", type: ISSUE, first: 100) {
  ### search(query: "is:pr author:PawelLipski repo:VirtusLab/git-machete", type: ISSUE, first: 100, after: "<endCursor>") {
    issueCount
    edges {
      node {
        ... on PullRequest {
          title
          url
          headRefName
          baseRefName
        }
      }
    }
    pageInfo {
      endCursor
      hasNextPage
    }
  }
}