Many programming languages provide support for environment isolation, such as Node.js with node_modules, Golang with go mod, and Python with virtualenv and pyvenv. To create a dependency snapshot, the pip freeze > requirements.txt command is often used to generate a requirements.txt file. In some scenarios, this method suffices, but in more complex situations, requirements.txt falls short.
requirements.txt
appdirs==1.4.3
astroid==2.3.3
attrs==19.3.0
black==19.3b0
certifi==2019.11.28
chardet==3.0.4
click==7.1.1
et-xmlfile==1.0.1
Flask==1.1.1
gevent==1.4.0
greenlet==0.4.15
idna==2.9
isort==4.3.21
itsdangerous==1.1.0
jdcal==1.4.1
Jinja2==2.11.1
lazy-object-proxy==1.4.3
MarkupSafe==1.1.1
mccabe==0.6.1
numpy==1.18.2
openpyxl==3.0.3
pandas==1.0.3
pylint==2.4.4
python-dateutil==2.8.1
pytz==2019.3
requests==2.23.0
six==1.14.0
tinydb==3.15.2
toml==0.10.0
typed-ast==1.4.1
urllib3==1.25.8
Werkzeug==1.0.0
wrapt==1.11.2
The requirements.txt file only records the versions of dependencies. If the official PyPI source is slow, and you need to use a faster domestic mirror, you typically have to use pip install -i or modify the global pip.conf file.
When a project uses a specific Python version, this version cannot be reflected in requirements.txt; it must be documented in the README or other documentation, and the correct Python version must be manually invoked when creating a virtual environment.
When projects need to use code optimization tools like flake8, pylint, or black, these dependencies are also written into requirements.txt by the pip freeze command, even though they are not needed in the production environment.
Pipfile
The emergence of Pipenv has solved these issues. Pipenv is a Python dependency management tool released by Kenneth Reitz in January 2017, and it uses Pipfile to replace requirements.txt.
[[source]]
name = "pypi"
url = "https://pypi.doubanio.com/simple"
verify_ssl = false
[dev-packages]
isort = "*"
black = "==19.3b0"
pylint = "*"
[packages]
flask = "*"
tinydb = "*"
pandas = "*"
requests = "*"
gevent = "*"
openpyxl = "*"
[requires]
python_version = "3.6"
Benefit 1: More Detailed Records
Compared to requirements.txt, Pipfile includes the pip source settings, allowing different environments for different projects. Dependencies are divided into dev and default environments; for example, pylint, flake8, and black can be placed in the dev dependencies.
Benefit 2: Fewer Manual Activations of Virtual Environments
Pipenv integrates the use of virtualenv, pyvenv, and pip commands, reducing the number of times you need to manually activate a virtual environment. To run main.py using the pyvenv module, you first need to execute source venv/bin/activate to activate the virtual environment, then execute python main.py. With pipenv, you only need to execute pipenv run main.py in the project’s root directory to automatically activate the current virtual environment and run main.py. If you need to install dependencies, simply execute pipenv install xxx without first activating the virtual environment and using pip install xxx.
Benefit 3: Lock Mechanism
Adding or removing installed packages from the Pipfile generates a Pipfile.lock to lock the versions and dependency information of the installed packages. The pipfile.lock file allows you to precisely restore dependency versions.
Common Commands
# Initialize a virtual environment (you can specify the Python version)
$ pipenv --python 3.6.9
# Activate the virtual environment for the current project
$ pipenv shell
# Install development dependencies
$ pipenv install black --dev
# Generate a lock file
$ pipenv lock