Happen to you that as a researcher, machine learning practitioner, you run multiple versions of your code in different places? like colab, kaggle, your computer. I had the same problem, so I created this package to help me, and maybe us, to overcome this problem by downloading the repository from GitHub (for private you need to provide a token), and work with it. Then if you want to change something, do it on your computer and push it, then update it in the code.
raise an ISSUE.
import agri
agri.authenticate("__token__")
my_repo = agri.import_repo("mmaleki92/test_repo")
print(agri.get_repo_structure("test_repo"))Happy researching.
A simple guide to using the repository browser to work with GitHub repositories.
Here's a simple example to get started:
# Import the library
import agri
# Authenticate with your GitHub token
agri.authenticate("your_github_token")
# Import a repository
my_repo = agri.import_repo("username/repo_name")
# Access a Python file or module
my_module = my_repo.folder.module_name # Automatically looks for module_name.py
# Use functions from the module
result = my_module.some_function()# Import a repository
my_repo = agri.import_repo("username/repo_name")
# Import a specific branch
dev_branch = agri.import_repo("username/repo_name", branch="development")# Show repository structure during import
my_repo = agri.import_repo("username/repo_name", show_structure=True)
# View structure of an already imported repository
structure = agri.get_repo_structure("username/repo_name")
print(structure)
# List all imported repositories
repos = agri.list_imported_repos()
print(repos)# List top-level folders and files
print(dir(my_repo))
# Navigate through directories
print(dir(my_repo.src))
print(dir(my_repo.src.utils))# Update to get latest changes
updated_repo = agri.update_repo("username/repo_name")Files are lazily loaded - they're only executed when you access them:
# Access a module (this will execute the Python file)
helper = my_repo.src.utils.helper # Automatically finds helper.py
# Use functions or classes from the module
result = helper.process_data(my_data)import os
# Get file path without executing code
file_path = os.path.join(my_repo.__path__, "README.md")
# Read content
with open(file_path, 'r') as f:
content = f.read()
print(content)# Create and commit a new file
agri.create_file_and_commit(
"username/repo_name",
file_content="print('Hello, world!')",
repo_file_path="examples/hello.py",
message="Add hello world example"
)# Commit a single file
agri.commit_files(
"username/repo_name",
local_source="/path/to/local/file.py",
repo_target="src/utils/file.py",
message="Add utility file"
)
# Commit multiple files
agri.commit_files(
"username/repo_name",
local_source={
"/path/to/file1.py": "src/file1.py",
"/path/to/file2.py": "src/file2.py"
},
message="Add multiple files"
)# Delete a file or directory
agri.delete_files_and_commit(
"username/repo_name",
repo_file_paths="old_file.py",
message="Remove old file"
)
# Delete multiple files/directories
agri.delete_files_and_commit(
"username/repo_name",
repo_file_paths=["file1.py", "old_folder/"],
message="Remove unused files"
)Here's a complete example workflow:
import agri
# Step 1: Authenticate
agri.authenticate("your_github_token")
# Step 2: Import a repository
my_repo = agri.import_repo("username/my-project")
# Step 3: Navigate and use repository contents
config = my_repo.src.config # Access config.py file
settings = config.DEFAULT_SETTINGS # Get a variable from the file
# Step 4: Make changes and commit them
agri.create_file_and_commit(
"username/my-project",
file_content=f"""
# Updated configuration
DEFAULT_SETTINGS = {settings}
DEBUG = True
""",
repo_file_path="src/config.py",
message="Update configuration settings"
)That's it! This quick guide covers the basics of working with repositories, accessing files, and committing changes.