06. pip¶
Overview¶
pip is the standard package installer for Python, used to install, upgrade, remove, and manage Python packages from package indexes such as the Python Package Index (PyPI). It is the default tool for dependency management in most Python environments and is commonly used in local development, virtual environments, containers, and CI/CD pipelines.
Primary use cases include:
(1) Installing third-party libraries from PyPI
(2) Managing project dependencies via requirement files
(3) Upgrading or removing installed packages
(4) Inspecting installed distributions and their metadata
pip is designed to work across platforms and integrates with Python’s packaging ecosystem, including wheels, virtual environments, and package indexes.
Installation and Scope¶
Default Availability¶
In most modern Python distributions, pip is included by default.
(1) Python 3.4+ includes pip through the ensurepip module
(2) Python installers for Windows and macOS typically install pip automatically
(3) Many Linux distributions provide pip through system package managers
Environment Scope¶
pip can operate in different installation scopes.
| Scope | Description | Typical Use Case |
|---|---|---|
| System-wide | Installs packages into the global Python environment | OS-managed environments, base containers |
| User site | Installs packages into the user’s home directory | Restricted environments without root access |
| Virtual environment | Installs packages isolated from the system Python | Project-specific dependencies |
Recommended installation practice
When working on application or library development:
(1) Create a virtual environment using python -m venv <env>
(2) Activate the environment
(3) Install dependencies using pip inside the virtual environment
This prevents dependency conflicts and avoids modifying the system Python.
Why pip Became the Standard Tool¶
pip became the dominant package installer in the Python ecosystem due to technical, ecosystem, and governance factors.
(1) Official endorsement
(1.1) pip is recommended in the official Python documentation as the standard installation tool.
(1.2) It is bundled with modern Python releases, increasing adoption by default.
(2) Integration with PyPI
(2.1) pip integrates directly with the Python Package Index (PyPI).
(2.2) It supports both public and private package indexes.
(3) Wheel support
(3.1) Native support for wheel distributions enables fast, reliable installations.
(3.2) Binary wheels reduce the need for local compilation.
(4) Dependency resolution (4.1) Modern versions include a robust dependency resolver. (4.2) It enforces compatibility constraints declared by packages.
(5) Ecosystem compatibility (5.1) Supported by most Python tools, frameworks, and cloud platforms. (5.2) Widely integrated into CI/CD systems and container workflows.
(6) Extensibility and configuration (6.1) Supports configuration files and environment variables. (6.2) Allows integration with internal artifact repositories.
Adoption implication
For most Python-based systems:
(1) Assume pip is available in the runtime environment
(2) Standardize dependency management around pip and requirements.txt
(3) Validate compatibility when using alternative package managers
Core Concepts and Components¶
Key Concepts¶
| Component | Description |
|---|---|
| Package | A distributable Python project, typically published on PyPI |
| Distribution | A packaged release of a project (wheel or source archive) |
| Wheel | A built distribution format (.whl) designed for fast installation |
| Requirement | A specification of a package and optional version constraints |
| Package index | A repository of packages, such as PyPI |
Supported Distribution Types¶
| Type | File Extension | Description |
|---|---|---|
| Wheel | .whl | Prebuilt binary distribution; preferred by pip |
| Source distribution | .tar.gz, .zip | Requires local build during installation |
pip Commands Reference¶
pip provides 14 primary commands for package management and environment inspection.
| Command | Description |
|---|---|
install | Install package |
uninstall | Uninstall package |
inspect | Inspect package |
list | List installed packages |
show | Show information about package |
freeze | Output installed packages in requirements format |
check | Verify installed packages have compatible dependencies |
download | Download package |
wheel | Build wheel archives for a package and all of its dependencies |
hash | Compute hashes of package archives |
search | Search PyPI for packages whose name or summary matches the given terms |
cache | Inspect and manage pip's wheel cache |
config | Manage pip's configuration |
debug | Show information useful for debugging pip |
Search command behavior
The search command depends on PyPI’s XML-RPC API and may not be supported by all package indexes.
Before relying on pip search:
(1) Verify that the target index supports search functionality
(2) Consider searching directly on the package index website
Package Discovery and Installation Flow¶
This flow describes how pip locates and installs a package. It is triggered when a user executes a command such as pip install <package>, and it results in the package being downloaded and installed into the target environment.
High-Level Flow (pip and Package Index)¶
flowchart LR
A[User: pip install package] --> B[pip client]
B --> C[Package Index<br>PyPI or custom index]
C --> D[Package metadata]
D --> E[Select compatible distribution]
E --> F[Download wheel or source]
F --> G[Install into environment]Package Search and Selection Steps¶
(1) The user runs pip install <package> in a Python environment. (2) pip reads configuration, including index URLs and authentication settings. (3) pip sends a request to the configured package index (default: PyPI). (4) The index returns package metadata and available distributions. (5) pip evaluates version constraints and environment compatibility. (6) pip selects the most suitable distribution, preferring wheels when available. (7) The selected distribution is downloaded to the local cache. (8) If the distribution is a wheel, it is installed directly. (9) If the distribution is a source archive, it is built into a wheel before installation. (10) The package and its dependencies are installed into the target environment.
Index and trust configuration
When installing from external package indexes:
(1) Verify the package index URL is trusted
(2) Use HTTPS connections for all package sources
(3) Consider enabling hash-checking mode for reproducible and secure installs
Limitations and Constraints¶
(1) pip does not provide full environment management; it relies on tools such as venv or virtualenv.
(2) Dependency resolution may fail if incompatible version constraints are specified.
(3) System-wide installations may conflict with OS package managers.
(4) Source distributions require build tools and compilers on the host system.
(5) Some packages may provide wheels only for specific platforms or Python versions.
System Python considerations
On Linux or managed systems:
(1) Avoid installing packages into the system Python using sudo pip install
(2) Use virtual environments for application dependencies
(3) Use the system package manager for OS-level Python packages
Reference¶
(1) https://docs.python.org/3/installing/index.html