Examples
These calls are read-only. Each one asks for more than the repository or
organisation list itself. The library snippets assume token is already set.
The commands use GITHUB_TOKEN or gh auth token when --token is omitted.
Workflows and Topics
Fetch both for every repository in an organisation.
from lupaxa.github_extractor import GitHubExtractor
client = GitHubExtractor(token)
repos = client.repositories_by_org("acme", resources=("topics", "workflows"))
for repo in repos:
names = [] if repo.topics is None else repo.topics.names
workflows = [] if repo.workflows is None else [item.name for item in repo.workflows]
print(repo.full_name)
print(" topics:", ", ".join(names) or "(none)")
print(" workflows:", ", ".join(workflows) or "(none)")
for failure in repo.errors:
print(" missing:", failure.addon, failure.status, failure.message)
The JSON from the command includes topics.names and a workflows list on
each repository. total_count is not in that list.
One Repository, Topics and Workflow Runs
repo = client.repository(
"acme/widget",
resources=("topics", "workflow_runs"),
)
topic_names = [] if repo.topics is None else repo.topics.names
run_names = [] if repo.workflow_runs is None else [run.name for run in repo.workflow_runs]
print(repo.full_name, topic_names, run_names)
Topics, Workflows, Languages, Commits, and Issues
Every extra is a name in resources.
repos = client.repositories_by_org(
"acme",
resources=("topics", "workflows", "languages", "commits", "issues"),
)
for repo in repos:
print(repo.full_name, repo.languages)
print(" commits:", 0 if repo.commits is None else len(repo.commits))
print(" issues:", 0 if repo.issues is None else len(repo.issues))
A missing extra is an empty list and is recorded on errors with status
404, 409, or 410. The repository is still returned.
One User, Several Extras
Releases, tags, topics, workflows, pull requests, and milestones are names in the same list.
repos = client.user_repositories(
username="ada",
resources=("releases", "tags", "topics", "workflows", "pulls", "milestones"),
)
github-extractor repos user --username ada --releases --tags --topics --workflows --pulls --milestones
Omit --username to use the authenticated user. That is
user_repositories() with no username.
Visible Repositories, Private Ones Dropped
repos = client.visible_repositories(
skip_private=True,
include=("acme/*",),
resources=("topics", "languages"),
)
--include matches full_name. --include-name matches the repository name
only.
Every Organisation, Selected Repositories
repos = client.organisation_repositories(
include_orgs=("acme*", "labs*"),
ignore=("*/*-archive",),
resources=("workflows", "topics", "commits"),
)
github-extractor repos organisations --include-org 'acme*' --include-org 'labs*' --ignore '*/*-archive' --workflows --topics --commits
repos organizations is not a command. The organisation list itself accepts
both spellings: organisations and organizations.
Names Only
slugs=True returns strings and does not request topics, workflows, or any
other extra, even if those flags are set.
names is a list of strings such as acme/widget.
Organisation Members, Teams, and Hooks
orgs = client.organisations(resources=("members", "teams", "hooks"))
for org in orgs:
member_logins = [] if org.members is None else [member.login for member in org.members]
team_names = [] if org.teams is None else [team.name for team in org.teams]
print(org.login, member_logins, team_names)
Authenticated User
user = client.authenticated_user(resources=("emails", "repos", "starred"))
print(user.login)
print([] if user.repos is None else [repo.full_name for repo in user.repos])
Authenticated User, Repository Extras on repos
repository_resources attaches extras to each repository in the user's
repos list.
user = client.authenticated_user(
resources=("repos",),
repository_resources=("topics", "commits"),
)
for repo in user.repos or []:
names = [] if repo.topics is None else repo.topics.names
commit_count = 0 if repo.commits is None else len(repo.commits)
print(repo.full_name, names, commit_count)
A Single Endpoint
get is the escape hatch for a path the named methods do not cover. The
result is JSON, not a model. --max-pages still applies.
Cap a Large Walk
client = GitHubExtractor(token, max_pages=2, max_workers=4)
repos = client.repositories_by_org("acme", resources=("workflows", "topics", "issues", "commits"))
github-extractor repos org acme --workflows --topics --issues --commits --max-pages 2 --max-workers 4
--timeout is the per-request timeout in seconds. The default is 10.