epregistry

Read the documentation!

Overview

The Entry Point Registry system provides a convenient way to manage and access Python entry points. It offers two different approaches to work with entry points:

• Group-based access: Work with all entry points in a specific group
• Module-based access: Work with all entry points provided by a specific module

This flexibility makes it particularly useful for plugin systems, extensions, or any modular components in Python applications.

Basic Usage

Group-based Registry

When you want to work with entry points organized by their group:

from epregistry import EntryPointRegistry

# Create a registry for console scripts
registry = EntryPointRegistry[Callable]("console_scripts")

# Get and load an entry point
script = registry.load("my-script")

# Get all entry points in the group
all_scripts = registry.get_all()

Module-based Registry

When you want to work with all entry points provided by a specific module:

from epregistry import ModuleEntryPointRegistry

# Create a registry for a specific module
registry = ModuleEntryPointRegistry[Any]("your_module_name")

# Get all groups that have entry points from this module
groups = registry.groups()

# Get entry points for a specific group
group_eps = registry.get_group("console_scripts")

# Load all entry points for a group
loaded_eps = registry.load_group("console_scripts")

💡 Tip: Type Hints Use the generic type parameter to specify the expected type of your entry points. For example, EntryPointRegistry[Callable] indicates that the entry points are callable objects.

Working with Group-based Registry

Get Entry Points

# Get an entry point (returns None if not found)
entry_point = registry.get("script_name")

# Get and load an entry point (returns None if not found)
loaded_entry_point = registry.load("script_name")

# Get an entry point with exception handling
try:
    entry_point = registry["script_name"]
except KeyError:
    print("Entry point not found")

Working with Multiple Entry Points

# Get all entry point names
names = registry.names()

# Get all entry points as a dictionary
all_entry_points = registry.get_all()  # dict[str, EntryPoint]

# Load all entry points
loaded_points = registry.load_all()  # dict[str, T]

Working with Module-based Registry

Get Entry Points by Group

# Get all entry points for a specific group
eps = registry.get_group("console_scripts")

# Load all entry points for a group
loaded_eps = registry.load_group("console_scripts")

Access All Entry Points

# Get all groups that contain entry points from this module
groups = registry.groups()

# Get all entry points organized by group
all_eps = registry.get_all()  # dict[str, list[EntryPoint]]

# Load all entry points from all groups
loaded_eps = registry.load_all()  # dict[str, list[T]]

Common Operations

# Check if an entry point exists
if "script_name" in registry:
    print("Entry point exists")

# Get the total number of entry points
count = len(registry)

# Iterate over entry points
for entry_point in registry:
    print(entry_point.name)

Advanced Features

Metadata Access

# For group-based registry
metadata = registry.get_metadata("script_name")
print(f"Module: {metadata['module']}")
print(f"Attribute: {metadata['attr']}")
print(f"Distribution: {metadata['dist']}")
print(f"Version: {metadata['version']}")

Extension Point Directory

# For group-based registry
directory = registry.get_extension_point_dir("script_name")
print(f"Extension is installed at: {directory}")

Discovery and Search

from epregistry import (
    available_groups,
    filter_entry_points,
    search_entry_points,
    list_distributions,
)

# Get all available groups
groups = available_groups()

# Filter entry points
flask_eps = filter_entry_points(group="flask.*")
pytest_eps = filter_entry_points(distribution="pytest")
test_eps = filter_entry_points(name_pattern="test_*")

# Search across all entry points
results = search_entry_points(
    "test",
    include_groups=True,
    include_names=True,
    include_distributions=True
)

# List all distributions with entry points
distributions = list_distributions()

💡 Tip: Filtering Patterns The filtering system supports wildcards:

• * matches any number of characters
• ? matches exactly one character
• Patterns are case-insensitive

Package and Distribution Name Conversion

The package also contain some helpers to convert between package and distribution names. The mapping in this case is also cached, only the first conversion may take long to build the index.

from epregistry import package_to_distribution, distribution_to_package

# Convert package name to distribution
dist_name = package_to_distribution("PIL")  # Returns 'Pillow'
dist_name = package_to_distribution("requests")  # Returns 'requests'

# Convert distribution to primary package
pkg_name = distribution_to_package("Pillow")  # Returns 'PIL'
pkg_name = distribution_to_package("requests")  # Returns 'requests'

Integration with Package Management

The Entry Point Registry integrates with Python's importlib.metadata system, making it compatible with:

• 📦 setuptools
• 📦 poetry
• Other packaging tools that follow the entry points specification

📝 Note: Automatic Caching Both registry types implement automatic caching of entry points for better performance. The cache is initialized on first use and shared across all registry instances.