Skip to main content

Pierre Git Storage SDK for Python

Project description

pierre-storage

Pierre Git Storage SDK for Python applications.

Installation

pip install pierre-storage

Usage

Basic Setup

from pierre_storage import GitStorage

# Initialize the client with your name and key
storage = GitStorage({
    "name": "your-name",  # e.g., 'v0'
    "key": "your-key",  # Your API key in PEM format
})

Creating a Repository

# Create a new repository with auto-generated ID
repo = await storage.create_repo()
print(repo.id)  # e.g., '123e4567-e89b-12d3-a456-426614174000'

# Create a repository with custom ID
custom_repo = await storage.create_repo({"id": "my-custom-repo"})
print(custom_repo.id)  # 'my-custom-repo'

Finding a Repository

found_repo = await storage.find_one({"id": "repo-id"})
if found_repo:
    url = await found_repo.get_remote_url()
    print(f"Repository URL: {url}")

Getting Remote URLs

The SDK generates secure URLs with JWT authentication for Git operations:

# Get URL with default permissions (git:write, git:read) and 1-year TTL
url = await repo.get_remote_url()
# Returns: https://t:JWT@your-name.code.storage/repo-id.git

# Configure the Git remote
print(f"Run: git remote add origin {url}")

# Get URL with custom permissions and TTL
read_only_url = await repo.get_remote_url({
    "permissions": ["git:read"],  # Read-only access
    "ttl": 3600,  # 1 hour in seconds
})

# Available permissions:
# - 'git:read'   - Read access to Git repository
# - 'git:write'  - Write access to Git repository
# - 'repo:write' - Create a repository

Working with Repository Content

Once you have a repository instance, you can perform various Git operations:

repo = await storage.create_repo()
# or
repo = await storage.find_one({"id": "existing-repo-id"})

# Get file content (streaming)
response = await repo.get_file_stream({
    "path": "README.md",
    "ref": "main",  # optional, defaults to default branch
})
text = await response.aread()
print(text.decode())

# List all files in the repository
files = await repo.list_files({
    "ref": "main",  # optional, defaults to default branch
})
print(files["paths"])  # List of file paths

# List branches
branches = await repo.list_branches({
    "limit": 10,
    "cursor": None,  # for pagination
})
print(branches["branches"])

# List commits
commits = await repo.list_commits({
    "branch": "main",  # optional
    "limit": 20,
    "cursor": None,  # for pagination
})
print(commits["commits"])

# Get branch diff
branch_diff = await repo.get_branch_diff({
    "branch": "feature-branch",
    "base": "main",  # optional, defaults to main
})
print(branch_diff["stats"])
print(branch_diff["files"])

# Get commit diff
commit_diff = await repo.get_commit_diff({
    "sha": "abc123...",
})
print(commit_diff["stats"])
print(commit_diff["files"])

Creating Commits

The SDK provides a fluent builder API for creating commits with streaming support:

# Create a commit
result = await (
    repo.create_commit({
        "target_branch": "main",
        "commit_message": "Update docs",
        "author": {"name": "Docs Bot", "email": "docs@example.com"},
    })
    .add_file_from_string("docs/changelog.md", "# v2.0.1\n- add streaming SDK\n")
    .add_file("docs/readme.md", b"Binary content here")
    .delete_path("docs/legacy.txt")
    .send()
)

print(result["commit_sha"])
print(result["ref_update"]["new_sha"])
print(result["ref_update"]["old_sha"])  # All zeroes when ref is created

The builder exposes:

  • add_file(path, source, options) - Attach bytes from various sources
  • add_file_from_string(path, contents, encoding, options) - Add text files (defaults to UTF-8)
  • delete_path(path) - Remove files or folders
  • send() - Finalize the commit and receive metadata

send() returns a result with:

{
    "commit_sha": str,
    "tree_sha": str,
    "target_branch": str,
    "pack_bytes": int,
    "blob_count": int,
    "ref_update": {
        "branch": str,
        "old_sha": str,  # All zeroes when the ref is created
        "new_sha": str,
    }
}

If the backend reports a failure, the builder raises a RefUpdateError containing the status, reason, and ref details.

Options:

  • target_branch (required): Branch name (without refs/heads/ prefix)
  • expected_head_sha (optional): Branch or commit that must match the remote tip
  • commit_message (required): The commit message
  • author (required): Dictionary with name and email
  • committer (optional): Dictionary with name and email (defaults to author)

Files are chunked into 4 MiB segments, allowing streaming of large assets without buffering in memory.

The target_branch must already exist on the remote repository. To seed an empty repository, omit expected_head_sha; the service will create the first commit only when no refs are present.

Streaming Large Files

The commit builder accepts async iterables, allowing streaming of large files:

async def file_chunks():
    """Generate file chunks asynchronously."""
    with open("/tmp/large-file.zip", "rb") as f:
        while chunk := f.read(1024 * 1024):  # Read 1MB at a time
            yield chunk

result = await (
    repo.create_commit({
        "target_branch": "assets",
        "expected_head_sha": "abc123...",
        "commit_message": "Upload latest design bundle",
        "author": {"name": "Assets Uploader", "email": "assets@example.com"},
    })
    .add_file("assets/design-kit.zip", file_chunks())
    .send()
)

Restoring Commits

You can restore a repository to a previous commit:

result = await repo.restore_commit({
    "target_branch": "main",
    "target_commit_sha": "abc123...",  # Commit to restore to
    "expected_head_sha": "def456...",  # Optional: current HEAD for safety
    "commit_message": "Restore to stable version",
    "author": {"name": "DevOps", "email": "devops@example.com"},
})

print(result["commit_sha"])
print(result["ref_update"])

API Reference

GitStorage

class GitStorage:
    def __init__(self, options: GitStorageOptions) -> None: ...
    async def create_repo(self, options: CreateRepoOptions = None) -> Repo: ...
    async def find_one(self, options: FindOneOptions) -> Optional[Repo]: ...
    def get_config(self) -> GitStorageOptions: ...

Repo

class Repo:
    @property
    def id(self) -> str: ...

    async def get_remote_url(self, options: GetRemoteURLOptions = None) -> str: ...
    async def get_file_stream(self, options: GetFileOptions) -> Response: ...
    async def list_files(self, options: ListFilesOptions = None) -> ListFilesResult: ...
    async def list_branches(self, options: ListBranchesOptions = None) -> ListBranchesResult: ...
    async def list_commits(self, options: ListCommitsOptions = None) -> ListCommitsResult: ...
    async def get_branch_diff(self, options: GetBranchDiffOptions) -> GetBranchDiffResult: ...
    async def get_commit_diff(self, options: GetCommitDiffOptions) -> GetCommitDiffResult: ...
    async def pull_upstream(self, options: PullUpstreamOptions = None) -> None: ...
    async def restore_commit(self, options: RestoreCommitOptions) -> RestoreCommitResult: ...
    def create_commit(self, options: CreateCommitOptions) -> CommitBuilder: ...

Type Definitions

All types are provided via TypedDict for better IDE support:

from pierre_storage.types import (
    GitStorageOptions,
    CreateRepoOptions,
    GetRemoteURLOptions,
    CommitSignature,
    # ... and more
)

Webhook Validation

The SDK includes utilities for validating webhook signatures:

from pierre_storage import validate_webhook, parse_push_event

# Validate webhook signature
result = validate_webhook(
    payload=request.body.decode(),  # Raw payload string
    signature=request.headers["X-Pierre-Signature"],
    secret="your-webhook-secret",
    options={"max_age_seconds": 300},  # 5 minutes
)

if result["valid"]:
    # Parse the event
    data = json.loads(request.body)

    if result["event_type"] == "push":
        event = parse_push_event(data)
        print(f"Push to {event['ref']}")
        print(f"Commit: {event['before']} -> {event['after']}")
else:
    print(f"Invalid webhook: {result['error']}")

Authentication

The SDK uses JWT (JSON Web Tokens) for authentication. When you call get_remote_url(), it:

  1. Creates a JWT with your name, repository ID, and requested permissions
  2. Signs it with your private key (ES256, RS256, or EdDSA)
  3. Embeds it in the Git remote URL as the password

The generated URLs are compatible with standard Git clients and include all necessary authentication.

Error Handling

The SDK provides specific error classes:

from pierre_storage import ApiError, RefUpdateError

try:
    repo = await storage.create_repo({"id": "existing"})
except ApiError as e:
    print(f"API error: {e.message}")
    print(f"Status code: {e.status_code}")

try:
    result = await builder.send()
except RefUpdateError as e:
    print(f"Ref update failed: {e.message}")
    print(f"Status: {e.status}")
    print(f"Reason: {e.reason}")
    print(f"Ref update: {e.ref_update}")

Development

Setup

# Create virtual environment and install dependencies
python3 -m venv venv
source venv/bin/activate
pip install -e ".[dev]"

# Or use Moon
moon run git-storage-sdk-python:setup

# Run tests
pytest

# Run tests with coverage
pytest --cov=pierre_storage --cov-report=html

# Type checking
mypy pierre_storage

# Linting
ruff check pierre_storage

Building

python -m build

License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pierre_storage-0.1.3.tar.gz (24.1 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

pierre_storage-0.1.3-py3-none-any.whl (22.1 kB view details)

Uploaded Python 3

File details

Details for the file pierre_storage-0.1.3.tar.gz.

File metadata

  • Download URL: pierre_storage-0.1.3.tar.gz
  • Upload date:
  • Size: 24.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.0

File hashes

Hashes for pierre_storage-0.1.3.tar.gz
Algorithm Hash digest
SHA256 3cd0c78daa189986a69004416e4d3e31913251645ed774af4b945c340f08db77
MD5 75bb508bc3c3ff83acf22c0098321174
BLAKE2b-256 c32c855bfc1f52c04d7bc45f90b3da5d93f937968896f578ee15190b86ea32c9

See more details on using hashes here.

File details

Details for the file pierre_storage-0.1.3-py3-none-any.whl.

File metadata

  • Download URL: pierre_storage-0.1.3-py3-none-any.whl
  • Upload date:
  • Size: 22.1 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.0

File hashes

Hashes for pierre_storage-0.1.3-py3-none-any.whl
Algorithm Hash digest
SHA256 d56e5af713c9b2779e29e42e2f3c93ebca20150f7e526f25c92b4596fa1d02bd
MD5 9b2a9b6342d2c60f72e71edfea2ded46
BLAKE2b-256 867e8d9645c7ac9a4f72d063a7bd172127aef1431bf7d3147f977162d32e9aca

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page