#!/usr/bin/env node
/**
 * Secret-scanning CI gate.
 *
 * Scans every Git-tracked text file looking for hardcoded secrets: source-host
 * tokens, cloud/API credentials, private keys, JWT-shaped tokens, embedded URL
 * credentials, and high-entropy assignment literals on secret-like names.
 *
 * Findings on `.env.example` variants or test/fixture/mock/stub paths are
 * recorded as "reviewed" rather than failures. Lines that contain obvious
 * placeholder markers (`placeholder`, `example`, `your-`, `<...>`, `${...}`,
 * `xxxxx`, `dummy`, `REDACTED`, ...) are skipped entirely.
 *
 * Exit 1 on any real (non-reviewed) finding, 0 otherwise.
 */
import fs from "node:fs";
import path from "node:path";
import { spawnSync } from "node:child_process";
import url from "node:url";

export const SKIP_DIR_NAMES = new Set([
  "node_modules",
  "dist",
  "build",
  "coverage",
  ".react-router",
  ".git",
  ".cache",
  "generated",
  ".next",
  ".turbo",
]);

const SKIP_FILE_SUFFIXES = [
  ".lock",
  "-lock.json",
  ".png",
  ".jpg",
  ".jpeg",
  ".svg",
  ".ico",
  ".webp",
  ".pdf",
  ".woff",
  ".woff2",
  ".ttf",
  ".eot",
  ".zip",
  ".gz",
  ".br",
  ".wasm",
  ".mp4",
];

const SKIP_FILE_NAMES = new Set(["pnpm-lock.yaml"]);

// Allowlisted env-example filenames whose findings are logged as reviewed.
const ALLOWLISTED_FILENAMES = new Set([
  ".env.example",
  ".env.local.example",
  ".env.production.example",
  ".env.test",
]);

// Exact path components that mark an intentionally synthetic test surface.
const ALLOWLISTED_PATH_COMPONENTS = new Set([
  "test",
  "tests",
  "__tests__",
  "fixture",
  "fixtures",
  "mock",
  "mocks",
  "stub",
  "stubs",
]);

// Self-reference: this scanner file describes the patterns it scans for, so
// matches against its own source must be treated as reviewed.
const SELF_FILENAMES = new Set([
  "verify-secrets.mjs",
  "verify-secrets.test.mjs",
]);

// Placeholder markers that disqualify a line from being a real secret.
const PLACEHOLDER_MARKERS = [
  "placeholder",
  "example",
  "your-",
  "your_",
  "replace-with",
  "replace_with",
  "xxxxx",
  "dummy",
  "redacted",
  "changeme",
  "change-me",
  "change_me",
  "todo",
  "fixme",
];

export function lineHasPlaceholder(line) {
  const lower = line.toLowerCase();
  for (const marker of PLACEHOLDER_MARKERS) {
    if (lower.includes(marker)) return true;
  }
  if (/<[^>]+>/.test(line)) return true;
  if (/\{[^}]+\}/.test(line)) return true;
  if (line.includes("${")) return true;
  return false;
}

export const PATTERNS = [
  { id: "github-pat", regex: /\bgh[opurs]_[A-Za-z0-9_]{20,}\b/g },
  { id: "github-fine-grained-pat", regex: /\bgithub_pat_[A-Za-z0-9_]{20,}\b/g },
  { id: "gitlab-pat", regex: /\bglpat-[A-Za-z0-9_-]{20,}\b/g },
  { id: "npm-access-token", regex: /\bnpm_[A-Za-z0-9]{30,}\b/g },
  { id: "pypi-access-token", regex: /\bpypi-[A-Za-z0-9_-]{40,}\b/g },
  { id: "huggingface-token", regex: /\bhf_[A-Za-z0-9]{30,}\b/g },
  { id: "google-api-key", regex: /\bAIza[0-9A-Za-z_-]{35}\b/g },
  { id: "aws-access-key-id", regex: /\bAKIA[0-9A-Z]{16}\b/g },
  { id: "payment-secret-live", regex: /\bsk_live_[A-Za-z0-9]{20,}\b/g },
  {
    id: "payment-webhook-live",
    regex: /\bwhsec_[A-Za-z0-9]{20,}\b/g,
    requiresLineContains: "live",
  },
  { id: "slack-bot-token", regex: /\bxoxb-\d+-\d+-\d+-[A-Za-z0-9]+\b/g },
  {
    id: "private-key-header",
    regex: /-----BEGIN (?:RSA |EC |DSA |OPENSSH |PGP )?PRIVATE KEY-----/g,
  },
  { id: "age-secret-key", regex: /\bAGE-SECRET-KEY-1[0-9A-Z]{20,}\b/g },
  {
    id: "jwt-like",
    regex:
      /\beyJ[A-Za-z0-9_-]{10,}\.eyJ[A-Za-z0-9_-]{10,}\.[A-Za-z0-9_-]{10,}\b/g,
  },
  {
    id: "generic-high-entropy-assignment",
    regex:
      /\b(?:password|passphrase|secret|token|auth_?token|api_?key|access_?key|client_?secret|private_?key|signing_?key|webhook_?secret)\s*[:=]\s*(?:['"][A-Za-z0-9+/=_-]{24,}['"]|(?=[A-Za-z0-9+/=_-]{32,}\b)(?=[A-Za-z0-9+/=_-]*\d)[A-Za-z0-9+/=_-]{32,})/gi,
  },
  {
    id: "embedded-url-credentials",
    regex: /\b(?:https?|postgres(?:ql)?|redis):\/\/[^\s/:@]+:[^\s/@]{8,}@[^\s]+/gi,
  },
];

export function isAllowlistedPath(absPath) {
  const basename = path.basename(absPath);
  if (SELF_FILENAMES.has(basename)) return true;
  if (ALLOWLISTED_FILENAMES.has(basename)) return true;
  const components = path.normalize(absPath).toLowerCase().split(path.sep);
  if (components.some((component) => ALLOWLISTED_PATH_COMPONENTS.has(component))) return true;
  return /\.(?:test|spec)\.[^.]+$/i.test(basename);
}

export function scanText(text, { filename = "" } = {}) {
  const findings = [];
  const lines = text.split(/\r?\n/);
  for (let i = 0; i < lines.length; i += 1) {
    const line = lines[i];
    if (lineHasPlaceholder(line)) continue;
    for (const pattern of PATTERNS) {
      pattern.regex.lastIndex = 0;
      for (const match of line.matchAll(pattern.regex)) {
        if (
          pattern.requiresLineContains &&
          !line.toLowerCase().includes(pattern.requiresLineContains)
        ) {
          continue;
        }
        const trailingText = line.slice((match.index ?? 0) + match[0].length);
        if (
          pattern.id === "generic-high-entropy-assignment" &&
          /:\s*[A-Z][A-Z0-9_]*$/.test(match[0]) &&
          /^\s*[,;)]/.test(trailingText)
        ) {
          continue;
        }
        findings.push({
          file: filename,
          line: i + 1,
          patternId: pattern.id,
          match: match[0],
        });
      }
    }
  }
  return findings;
}

function shouldSkipFile(absPath) {
  const basename = path.basename(absPath);
  if (SKIP_FILE_NAMES.has(basename)) return true;
  const lower = basename.toLowerCase();
  for (const suffix of SKIP_FILE_SUFFIXES) {
    if (lower.endsWith(suffix)) return true;
  }
  return false;
}

function listFilesRecursive(dir) {
  if (!fs.existsSync(dir)) return [];
  const out = [];
  const stack = [dir];
  while (stack.length > 0) {
    const current = stack.pop();
    let entries;
    try {
      entries = fs.readdirSync(current, { withFileTypes: true });
    } catch {
      continue;
    }
    for (const entry of entries) {
      if (entry.isDirectory()) {
        if (SKIP_DIR_NAMES.has(entry.name)) continue;
        stack.push(path.join(current, entry.name));
      } else if (entry.isFile()) {
        const abs = path.join(current, entry.name);
        if (shouldSkipFile(abs)) continue;
        out.push(abs);
      }
    }
  }
  return out;
}

export function collectFiles(root) {
  const tracked = spawnSync("git", ["-C", root, "ls-files", "-z"], {
    encoding: "utf8",
    stdio: ["ignore", "pipe", "ignore"],
  });
  if (tracked.status === 0) {
    return tracked.stdout
      .split("\0")
      .filter(Boolean)
      .map((relativePath) => path.join(root, relativePath))
      .filter((absolutePath) => {
        if (shouldSkipFile(absolutePath)) return false;
        try {
          return fs.lstatSync(absolutePath).isFile();
        } catch {
          return false;
        }
      });
  }
  return listFilesRecursive(root);
}

export function isKnownLocalCredential(finding) {
  if (finding.patternId !== "embedded-url-credentials") return false;
  const rawUrl = finding.match
    .split('"')[0]
    .split("'")[0]
    .replace(/^postgres:\/\//, "postgresql://");
  try {
    const parsed = new URL(rawUrl);
    return (
      parsed.protocol === "postgresql:" &&
      ["localhost", "127.0.0.1", "::1"].includes(parsed.hostname) &&
      [["postgres", "postgres"], ["hivegrid", "hivegrid"]].some(
        ([username, password]) =>
          parsed.username === username && parsed.password === password,
      )
    );
  } catch {
    return false;
  }
}

export function evaluate(root = process.cwd()) {
  const files = collectFiles(root);
  const real = [];
  const reviewed = [];
  for (const file of files) {
    let content;
    try {
      content = fs.readFileSync(file, "utf8");
    } catch {
      continue;
    }
    const findings = scanText(content, { filename: file });
    if (findings.length === 0) continue;
    const allowlisted = isAllowlistedPath(file);
    for (const finding of findings) {
      if (allowlisted || isKnownLocalCredential(finding)) {
        reviewed.push(finding);
      } else {
        real.push(finding);
      }
    }
  }
  return {
    scanned: files.length,
    patterns: PATTERNS.length,
    real,
    reviewed,
  };
}

export function formatFinding(finding, rootRel) {
  const rel = path.relative(rootRel, finding.file) || finding.file;
  return `${rel}:${finding.line}: ${finding.patternId} matched [redacted length=${finding.match.length}]`;
}

const isMain =
  import.meta.url === url.pathToFileURL(process.argv[1] ?? "").href;

/* eslint-disable no-console -- CLI reporter writes to stdout/stderr. */
if (isMain) {
  const root = process.cwd();
  const result = evaluate(root);
  if (result.real.length === 0) {
    console.log("PASS no secret-shaped strings found");
  } else {
    for (const finding of result.real) {
      console.error(`FAIL ${formatFinding(finding, root)}`);
    }
  }
  if (result.reviewed.length > 0) {
    for (const finding of result.reviewed) {
      console.log(`reviewed ${formatFinding(finding, root)}`);
    }
  }
  console.log(
    `\nSummary: scanned files: ${result.scanned}, patterns: ${result.patterns}, real: ${result.real.length}, reviewed: ${result.reviewed.length}`,
  );
  if (result.real.length > 0) {
    process.exit(1);
  }
}
/* eslint-enable no-console */