GitHarbor

Read the documentation!

GitHarbor User Guide

GitHarbor is a unified interface for interacting with Git hosting platforms. It provides a consistent API to work with repositories hosted on GitHub, GitLab, Azure DevOps, Gitea, CodeBerg and Bitbucket.

Getting Started

The main entry point is the create_repository() function which accepts a repository URL and platform-specific credentials:

from githarbor import create_repository

# GitHub repository
repo = create_repository("https://github.com/owner/repo", token="github_pat_...")

# GitLab repository
repo = create_repository("https://gitlab.com/owner/repo", token="glpat-...")

# Azure DevOps repository
repo = create_repository(
    "https://dev.azure.com/org/project/_git/repo",
    token="azure_pat"
)

[!TIP] Always use personal access tokens (PATs) for authentication. Never hardcode tokens in your source code.

Working with Repositories

Basic Repository Information

# Get repository name and default branch
print(repo.name)
print(repo.default_branch)

# Get language statistics
languages = repo.get_languages()
# Returns: {"Python": 10000, "JavaScript": 5000}

# Get recent activity statistics
activity = repo.get_recent_activity(
    days=30,
    include_commits=True,
    include_prs=True,
    include_issues=True
)

Branches and Tags

# List all branches
branches = repo.list_branches()

# Get specific branch
main_branch = repo.get_branch("main")

# List all tags
tags = repo.list_tags()

# Get specific tag
tag = repo.get_tag("v1.0.0")

# Compare branches
diff = repo.compare_branches(
    base="main",
    head="feature",
    include_commits=True,
    include_files=True,
    include_stats=True
)

Commits

# Get specific commit
commit = repo.get_commit("abcd1234")

# List commits with filters
commits = repo.list_commits(
    branch="main",
    since=datetime(2024, 1, 1),
    until=datetime(2024, 2, 1),
    author="username",
    path="src/",
    max_results=100
)

# Search commits
results = repo.search_commits(
    query="fix bug",
    branch="main",
    path="src/",
    max_results=10
)

Issues and Pull Requests

# List open issues
open_issues = repo.list_issues(state="open")

# Get specific issue
issue = repo.get_issue(123)

# List pull requests
prs = repo.list_pull_requests(state="open")  # or "closed" or "all"

# Get specific pull request
pr = repo.get_pull_request(456)

Releases

# Get latest release
latest = repo.get_latest_release(
    include_drafts=False,
    include_prereleases=False
)

# List all releases
releases = repo.list_releases(
    include_drafts=False,
    include_prereleases=True,
    limit=10
)

# Get specific release
release = repo.get_release("v1.0.0")

Repository Content

# Download single file
repo.download(
    path="README.md",
    destination="local/README.md"
)

# Download directory recursively
repo.download(
    path="src/",
    destination="local/src",
    recursive=True
)

# Iterate through files
for file_path in repo.iter_files(
    path="src/",
    ref="main",
    pattern="*.py"
):
    print(file_path)

CI/CD Workflows

# List all workflows
workflows = repo.list_workflows()

# Get specific workflow
workflow = repo.get_workflow("workflow_id")

# Get specific workflow run
run = repo.get_workflow_run("run_id")

Contributors

# Get repository contributors
contributors = repo.get_contributors(
    sort_by="commits",  # or "name" or "date"
    limit=10
)

Error Handling

GitHarbor provides specific exceptions for different error cases:

from githarbor.exceptions import (
    RepositoryNotFoundError,
    ResourceNotFoundError,
    FeatureNotSupportedError
)

try:
    repo = create_repository("https://github.com/nonexistent/repo")
except RepositoryNotFoundError:
    print("Repository does not exist")

try:
    issue = repo.get_issue(999999)
except ResourceNotFoundError:
    print("Issue does not exist")

try:
    repo.some_unsupported_method()
except FeatureNotSupportedError:
    print("This feature is not supported by this repository provider")

[!NOTE] Not all features are supported by all platforms. Operations that aren't supported will raise FeatureNotSupportedError.

[!IMPORTANT] Be mindful of API rate limits when making many requests. Consider implementing retries and delays in your code.