Library API

Lintro exposes a small, stable Python API for embedding checks, formatting, and tests in your own programs.

from lintro.api import check, fmt, test

result = check(paths=["src"], tools="ruff")
if not result.success:
    raise SystemExit(result.exit_code)

Each entry point returns a LintroResult:

from dataclasses import dataclass


@dataclass(frozen=True)
class LintroResult:
    action: str      # "check", "fmt", or "test"
    exit_code: int   # 0 on success, non-zero when issues/errors occur

    @property
    def success(self) -> bool: ...

Available functions (all keyword-only):

• check(...) — run linting/quality tools against paths.
• format(...) — auto-fix formatting issues. Also exported as fmt.
• test(...) — run pytest through Lintro’s output formatting.

Exceptions raised during execution propagate normally to the caller, and tool output is written directly to the console (no buffering).

Migration note

Previous releases exposed thin programmatic wrappers in the command modules (lintro.cli_utils.commands.check.check, lintro.cli_utils.commands.format.format_code, and lintro.cli_utils.commands.test.test). Internally these invoked the Click commands through click.testing.CliRunner, a test helper that swallowed exceptions into a result object and buffered stdout/stderr.

Those wrappers still exist for backward compatibility, but now delegate to the real library API in lintro.api. External callers should migrate to lintro.api.check / lintro.api.format (fmt) / lintro.api.test, which:

• return a structured LintroResult instead of a Click Result;
• let exceptions propagate instead of capturing them; and
• do not redirect or buffer console output.

CliRunner is no longer used anywhere in the lintro package; it remains only in the test suite, where it is the appropriate tool for exercising the CLI.