Pylint

4.0.5 · active · verified Sat Mar 28

Pylint is a widely used static code analyzer for Python. It inspects source code without execution to detect programming errors, enforce coding standards (like PEP 8), identify code smells, and suggest refactorings. It provides a detailed report and a code quality score. Pylint maintains an active development status with regular releases, including minor versions for new checks and bug fixes, and major versions that may introduce breaking changes to options or output formats.

Warnings

breaking Pylint 4.x removes direct access to package metadata via `pylint.__pkginfo__`. Users should migrate to `importlib.metadata` for retrieving package information.
Fix: Replace `from pylint import __pkginfo__` with `from importlib import metadata; metadata.metadata('pylint')`.
breaking Pylint 4.x officially drops support for *launching* Pylint with Python 3.9. While it can still *lint* code written for Python 3.9 (or older) using `--py-version=3.9`, the environment running Pylint itself must be Python 3.10.0 or newer.
Fix: Ensure that the Python interpreter used to install and run Pylint is 3.10.0 or newer. If linting older Python code, use the `--py-version` flag.
breaking In Pylint 3.x, the behavior of enabling or disabling individual messages changed. The order of `--enable` or `--disable` flags (or their counterparts in configuration files) now strictly dictates their application, meaning later options can override earlier ones.
Fix: Review configuration files and command-line arguments to ensure the intended message enabling/disabling order is maintained. Avoid redundant or conflicting `enable=all`/`disable=all` with specific message controls.
gotcha Pylint's default configuration can be very verbose, especially on legacy or less-PEP 8 compliant projects. This often leads to an overwhelming number of messages upon first run.
Fix: During initial adoption, start with `--errors-only` or disable convention/refactor messages (e.g., `--disable=C,R`) and progressively re-enable checks as code quality improves. Generate a `.pylintrc` file (`pylint --generate-rcfile > .pylintrc`) for granular control over messages and checkers.
gotcha The `invalid-name` message often triggers due to Pylint's default expectation of `UPPERCASE` for module-level constants, which might conflict with project-specific naming conventions.
Fix: Customize the regular expressions for naming conventions in your `.pylintrc` file. For instance, modify `--const-rgx` or `--variable-rgx` to match your project's style.
gotcha Pylint relies on Python's module resolution mechanism. Incorrect `PYTHONPATH` settings or running Pylint from an unexpected working directory can lead to it analyzing the wrong module version or failing to import modules, resulting in `import-error` messages.
Fix: Always run Pylint from the project root or a directory where Python can correctly resolve your modules. Verify your `PYTHONPATH` environment variable if encountering import issues.

Install

pip install pylint Standard Installation
pip install pylint[spelling] With Spelling Checks (requires Enchant C library)

Imports

Run
```
from pylint.lint import Run
```
Used for programmatic invocation of Pylint.
TextReporter
```
from pylint.reporters.text import TextReporter
```
Useful for capturing Pylint's output to a custom stream when running programmatically.

Quickstart

The most common way to use Pylint is via the command line, simply by pointing it to a Python file, module, or package. For integration into tools or CI/CD pipelines, Pylint can also be invoked programmatically using `pylint.lint.Run`, allowing for custom reporters and output handling. The quickstart demonstrates both methods.

import os
import io
from pylint.lint import Run
from pylint.reporters.text import TextReporter

# Create a dummy Python file for linting
with open('my_module.py', 'w') as f:
    f.write('"""A simple module."""
MY_CONSTANT = 10
def my_function(arg1, arg2):  # arg2 is unused
    return arg1 + MY_CONSTANT
')

# --- Command Line Usage ---
print("\n--- Command Line Pylint ---")
os.system('pylint my_module.py')

# --- Programmatic Usage ---
print("\n--- Programmatic Pylint ---")
# Capture output to a string buffer
pylint_output = io.StringIO()
reporter = TextReporter(pylint_output)

# Run Pylint on the module, disabling reports for cleaner output
# do_exit=False prevents sys.exit() on linting issues
Run(['my_module.py', '--disable=C0114,C0116,R0903', '--reports=no'], reporter=reporter, do_exit=False)

print(pylint_output.getvalue())

# Clean up the dummy file
os.remove('my_module.py')

view raw JSON →