rune/scripts/polyglot_lint.py

#!/usr/bin/env python3
"""polyglot_lint.py — the language-AWARE lint gate (contract
BUILDER-2026-10-07T20:01:40Z-ac6140).

Why: the pipeline linted Python hard (ruff + ratchet) but Rust NOT AT ALL
(crates build, never clippy) and Go only via vet (gofmt drifted across 41
files, including landed batches). This gate applies the RIGHT linter to
whatever changed:

    Python → ruff check
    Go     → go vet (per module) + gofmt -l (formatting is a gate)
    Rust   → cargo clippy -- -D warnings (per changed crate) + cargo fmt --check
    other  → noted and skipped

Modes:
    (default)        changed files vs the push range (git diff against
                     origin/master / the last push ref; falls back to the
                     last commit when no upstream exists)
    --all            full tree
    --files a,b,c    explicit list

Exit non-zero on any finding. The cron fault-watcher and the pre-push
routine both run this; CI runs the Go+Rust halves natively.

Usage:
    python3 scripts/polyglot_lint.py            # changed-only
    python3 scripts/polyglot_lint.py --all      # everything
"""

import argparse
import os
import subprocess
import sys
from pathlib import Path

REPO = Path(__file__).resolve().parent.parent

LANG_BY_EXT = {
    ".py": "python",
    ".go": "go",
    ".rs": "rust",
}
LANG_BY_PATH = {
    "src/go/": "go",
    "src/rust/": "rust",
}


def changed_files(explicit=None, all_files=False):
    if explicit:
        return list(explicit)
    if all_files:
        return [
            str(p.relative_to(REPO))
            for p in REPO.rglob("*")
            if p.is_file() and not str(p).startswith((".git", "store/", "target"))
        ]
    base = None
    for ref in ("origin/master", "gitea/master", "HEAD~1"):
        r = subprocess.run(["git", "rev-parse", "--verify", ref],
                           capture_output=True, text=True, cwd=REPO)
        if r.returncode == 0:
            base = ref
            break
    if base is None:
        return []
    r = subprocess.run(["git", "diff", "--name-only", f"{base}...HEAD"],
                       capture_output=True, text=True, cwd=REPO)
    return [f for f in r.stdout.splitlines() if f.strip()]


def group_by_language(files):
    groups = {"python": [], "go": [], "rust": []}
    for f in files:
        lang = None
        for prefix, l in LANG_BY_PATH.items():
            if f.startswith(prefix):
                lang = l
                break
        if lang is None:
            lang = LANG_BY_EXT.get(Path(f).suffix)
        if lang:
            groups[lang].append(f)
    return groups


def run(cmd, cwd=None):
    r = subprocess.run(cmd, cwd=cwd or REPO, capture_output=True, text=True)
    return r.returncode, r.stdout, r.stderr


def lint_python(files):
    findings = []
    if not files:
        return findings
    rc, out, err = run(["ruff", "check", *files])
    if rc != 0:
        findings.append(f"ruff: {out}{err}")
    return findings


def lint_go(files):
    findings = []
    modules = sorted({str(Path(f).parent) for f in files})
    # collapse to module roots (dirs containing go.mod)
    for mod in modules:
        mod_path = REPO / mod
        while mod_path != REPO and not (mod_path / "go.mod").exists():
            mod_path = mod_path.parent
        if mod_path == REPO:
            continue
        rel = mod_path.relative_to(REPO)
        rc, out, err = run(["go", "vet", "./..."], cwd=mod_path)
        if rc != 0:
            findings.append(f"go vet ({rel}): {out}{err}")
    # gofmt across the changed files' modules (formatting is a GATE)
    fmt_targets = [f for f in files if f.endswith(".go")]
    if fmt_targets:
        rc, out, _ = run(["gofmt", "-l", *fmt_targets])
        if rc != 0:
            findings.append(f"gofmt: failed to run")
        elif out.strip():
            findings.append(
                "gofmt unformatted (run `gofmt -w` on these):\n  "
                + "\n  ".join(out.splitlines())
            )
    return findings


def lint_rust(files):
    findings = []
    # files may arrive relative (changed-only mode) or absolute (--all);
    # normalize before relative_to (the ValueError from mixed forms broke
    # the gate on every all-rust changed set).
    crates = sorted({str((f if os.path.isabs(f) else REPO / f)).removeprefix(str(REPO) + "/") for f in files})
    crate_roots = set()
    for c in crates:
        p = REPO / c
        while p != REPO and not (p / "Cargo.toml").exists():
            p = p.parent
        if p != REPO:
            crate_roots.add(p)
    for crate in sorted(crate_roots):
        rc, out, err = run(["cargo", "clippy", "--", "-D", "warnings"], cwd=crate)
        if rc != 0:
            findings.append(f"clippy ({crate.name}):\n{out}{err}")
        rc, out, err = run(["cargo", "fmt", "--check"], cwd=crate)
        if rc != 0:
            findings.append(f"rustfmt ({crate.name}):\n{out}{err}")
    return findings


def main():
    ap = argparse.ArgumentParser()
    ap.add_argument("--all", action="store_true", help="lint the full tree")
    ap.add_argument("--files", help="comma-separated explicit file list")
    args = ap.parse_args()

    explicit = args.files.split(",") if args.files else None
    files = changed_files(explicit, args.all)
    if not files:
        print("polyglot lint: no changed files detected — nothing to do")
        return 0

    groups = group_by_language(files)
    print(f"polyglot lint: {len(files)} changed → "
          f"python={len(groups['python'])} go={len(groups['go'])} "
          f"rust={len(groups['rust'])}")

    all_findings = []
    all_findings += lint_python(groups["python"])
    all_findings += lint_go(groups["go"])
    all_findings += lint_rust(groups["rust"])

    if all_findings:
        print(f"POLYGLOT LINT: {len(all_findings)} finding(s):")
        for f in all_findings:
            print(f"  - {f}")
        return 1
    print("polyglot lint: clean")
    return 0


if __name__ == "__main__":
    sys.exit(main())