Initial check-mermaid.py — Mermaid block validator via mmdc in podman

2026-05-26 14:07:49 +00:00
parent ab5871b3b1
commit faabb91e62
1 changed files with 146 additions and 0 deletions
@@ -0,0 +1,146 @@
+#!/usr/bin/env python3
+"""Syntax-check every ```mermaid block in repo markdown files.
+
+Runs `mmdc` (Mermaid CLI) in a transient podman container; treats any
+non-zero exit as a syntax failure and reports the offending file:line.
+
+Usage:
+    check-mermaid.py                  # check everything under cwd
+    check-mermaid.py docs/            # check one path
+    check-mermaid.py --image foo      # override container image
+    check-mermaid.py --no-pull docs/  # skip upfront podman pull
+
+Exit 0 if all blocks parse, 1 otherwise. Honors the standard CI
+convention.
+"""
+
+from __future__ import annotations
+
+import argparse
+import os
+import shutil
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+
+DEFAULT_IMAGE = "docker.io/minlag/mermaid-cli:latest"
+SKIP_DIRS = {".git", "node_modules", ".venv", "venv", "vendor"}
+
+
+def find_markdown(root: Path) -> list[Path]:
+    out: list[Path] = []
+    for dirpath, dirnames, filenames in os.walk(root):
+        dirnames[:] = [d for d in dirnames if d not in SKIP_DIRS]
+        for fn in filenames:
+            if fn.endswith(".md"):
+                out.append(Path(dirpath) / fn)
+    return sorted(out)
+
+
+def extract_blocks(md: Path) -> list[tuple[int, str]]:
+    """Return [(start_line_1indexed, body), ...] for every ```mermaid block."""
+    lines = md.read_text().splitlines()
+    out: list[tuple[int, str]] = []
+    i = 0
+    while i < len(lines):
+        if lines[i].strip() == "```mermaid":
+            start = i + 1  # the line after the fence is line start+1 (1-indexed)
+            body_lines: list[str] = []
+            i += 1
+            while i < len(lines) and not lines[i].startswith("```"):
+                body_lines.append(lines[i])
+                i += 1
+            out.append((start + 1, "\n".join(body_lines) + "\n"))
+        i += 1
+    return out
+
+
+def check_block(image: str, body: str) -> tuple[bool, str]:
+    """Run mmdc on the body. Returns (ok, stderr_or_empty)."""
+    # mmdc requires real files on a bind mount — using stdin/stdout
+    # confuses puppeteer's file watching. Use a temp dir mounted into
+    # the container. World-writable so the container's UID can write
+    # the SVG output.
+    with tempfile.TemporaryDirectory() as td:
+        td_path = Path(td)
+        td_path.chmod(0o777)
+        (td_path / "in.mmd").write_text(body)
+        r = subprocess.run(
+            [
+                "podman", "run", "--rm",
+                "-v", f"{td}:/data:Z",
+                image,
+                "-i", "/data/in.mmd",
+                "-o", "/data/out.svg",
+                "--quiet",
+            ],
+            capture_output=True, text=True, timeout=60,
+        )
+        if r.returncode == 0:
+            return True, ""
+        return False, (r.stderr or r.stdout).strip()
+
+
+def main() -> int:
+    ap = argparse.ArgumentParser(description=__doc__.split("\n", 1)[0])
+    ap.add_argument("paths", nargs="*", default=["."],
+                    help="files or dirs to scan (default: current dir)")
+    ap.add_argument("--image", default=DEFAULT_IMAGE,
+                    help=f"container image to use (default {DEFAULT_IMAGE})")
+    ap.add_argument("--no-pull", action="store_true",
+                    help="skip the upfront `podman pull` step")
+    args = ap.parse_args()
+
+    if not shutil.which("podman"):
+        print("podman not in PATH", file=sys.stderr)
+        return 2
+
+    if not args.no_pull:
+        subprocess.run(["podman", "pull", "--quiet", args.image],
+                       capture_output=True, check=False)
+
+    # Collect targets
+    targets: list[Path] = []
+    for p_arg in args.paths:
+        p = Path(p_arg)
+        if p.is_dir():
+            targets.extend(find_markdown(p))
+        elif p.is_file():
+            targets.append(p)
+    targets = sorted(set(targets))
+
+    total_blocks = 0
+    failures: list[tuple[Path, int, str]] = []
+
+    for md in targets:
+        blocks = extract_blocks(md)
+        if not blocks:
+            continue
+        for line, body in blocks:
+            total_blocks += 1
+            ok, err = check_block(args.image, body)
+            marker = "\033[32m✓\033[0m" if ok else "\033[31m✗\033[0m"
+            print(f"  {marker} {md}:{line}")
+            if not ok:
+                failures.append((md, line, err))
+
+    print()
+    if failures:
+        print(f"\033[31m{len(failures)} of {total_blocks} block(s) FAILED:\033[0m\n")
+        for md, line, err in failures:
+            print(f"=== {md}:{line} ===")
+            # Trim noisy puppeteer stack; keep the parse error
+            for line_ in err.splitlines():
+                if not line_.strip():
+                    continue
+                print(f"  {line_}")
+            print()
+        return 1
+    print(f"\033[32m{total_blocks} block(s) OK\033[0m")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())