#!/usr/bin/env python3
"""Phase 5 audit for cross-request/static state risks in site code files."""
from __future__ import annotations

import argparse
import json
from dataclasses import asdict, dataclass
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]
SITE = ROOT / "site"

PATTERNS = [
    ("static local/global", "static "),
    ("once hook", "ONCE("),
    ("init hook", "INIT("),
    ("background task", "task_"),
]

SKIP_PARTS = {".git", "tmp", "work", "bin", "pkg", "__pycache__"}
CODE_SUFFIXES = {".uce", ".h"}
DOC_SUFFIXES = {".txt"}


@dataclass
class Finding:
    path: str
    line: int
    kind: str
    severity: str
    text: str
    note: str


def iter_files(include_doc_text: bool) -> list[Path]:
    suffixes = CODE_SUFFIXES | (DOC_SUFFIXES if include_doc_text else set())
    result: list[Path] = []
    for path in SITE.rglob("*"):
        if not path.is_file():
            continue
        if any(part in SKIP_PARTS for part in path.parts):
            continue
        if path.suffix not in suffixes:
            continue
        result.append(path)
    return sorted(result)


def note_for(kind: str, severity: str) -> str:
    if severity == "documentation":
        return "Documentation prose mention; useful for terminology review, not a direct static-state migration finding."
    if kind in {"once hook", "init hook"}:
        return "Audit behavior under per-request wasm workspaces; ONCE/INIT may need host-side cache semantics if used for cross-request state."
    if kind == "background task":
        return "Task APIs cross request lifetimes; verify they are host handles, not guest statics."
    return "Check whether state is request-local, immutable, or intentionally persistent; unit statics reset per wasm workspace."


def severity_for(path: Path) -> str:
    return "documentation" if path.suffix in DOC_SUFFIXES else "code"


def scan(include_doc_text: bool) -> list[Finding]:
    findings: list[Finding] = []
    for path in iter_files(include_doc_text):
        try:
            lines = path.read_text(encoding="utf-8").splitlines()
        except UnicodeDecodeError:
            continue
        severity = severity_for(path)
        for lineno, line in enumerate(lines, 1):
            stripped = line.strip()
            if severity == "code" and (stripped.startswith("//") or stripped.startswith("# ")):
                continue
            for kind, needle in PATTERNS:
                if needle in line:
                    findings.append(Finding(
                        path=str(path.relative_to(ROOT)),
                        line=lineno,
                        kind=kind,
                        severity=severity,
                        text=stripped[:180],
                        note=note_for(kind, severity),
                    ))
    return findings


def write_reports(findings: list[Finding], out_dir: Path) -> None:
    out_dir.mkdir(parents=True, exist_ok=True)
    (out_dir / "site-static-audit.json").write_text(json.dumps([asdict(f) for f in findings], indent=2) + "\n")
    lines = ["# Phase 5 site static-state audit", ""]
    if not findings:
        lines.append("No candidate cross-request/static-state patterns found.")
    else:
        code_count = sum(1 for f in findings if f.severity == "code")
        doc_count = sum(1 for f in findings if f.severity == "documentation")
        lines.append(f"Findings: {code_count} code, {doc_count} documentation prose.")
        lines.append("")
        lines.extend(["| file | line | severity | kind | code | note |", "|---|---:|---|---|---|---|"])
        for f in findings:
            code = f.text.replace("|", "\\|")
            note = f.note.replace("|", "\\|")
            lines.append(f"| {f.path} | {f.line} | {f.severity} | {f.kind} | `{code}` | {note} |")
    (out_dir / "site-static-audit.md").write_text("\n".join(lines) + "\n")


def main() -> int:
    parser = argparse.ArgumentParser(description="Phase 5 site static/cross-request-state audit")
    parser.add_argument("--include-doc-text", action="store_true", help="include site/doc .txt prose as documentation-severity findings")
    parser.add_argument("--out-dir", default="/tmp/uce/wasm-phase5")
    args = parser.parse_args()
    out_dir = Path(args.out_dir)
    findings = scan(args.include_doc_text)
    write_reports(findings, out_dir)
    print(f"Found {len(findings)} candidate static/cross-request patterns")
    print(f"wrote {out_dir / 'site-static-audit.json'} and {out_dir / 'site-static-audit.md'}")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())