/**
 * earth-now.co
 * Author:  Simon-Pierre Boucher
 * Contact: contact@spboucher.ai
 * File:    apps/ingest/src/archive.ts
 * Purpose: Raw payload archive — every ingested byte is kept (with sha256) for future re-fitting; S3-shaped interface, local FS for now
 */

import { createHash } from "node:crypto";
import { mkdir, writeFile } from "node:fs/promises";
import { join } from "node:path";

export interface ArchiveOptions {
  /** Archive root. Defaults to $RAW_ARCHIVE_DIR, then "data/raw". */
  dir?: string;
  /** Timestamp override for deterministic tests. */
  now?: Date;
}

export interface ArchiveResult {
  rawPath: string;
  checksumPath: string;
  sha256: string;
}

/**
 * Write `${dir}/${sourceId}/${timestamp}.raw` plus a sibling `.sha256` checksum file.
 * We keep ALL raw data for re-fitting; when this moves to S3 only this module changes.
 */
export async function archiveRaw(
  sourceId: string,
  payload: string | Uint8Array,
  options: ArchiveOptions = {},
): Promise<ArchiveResult> {
  if (!/^[a-z0-9_-]+$/i.test(sourceId))
    throw new Error(`archiveRaw: invalid sourceId '${sourceId}' (path-safe [a-z0-9_-] only)`);
  const dir = options.dir ?? process.env.RAW_ARCHIVE_DIR ?? "data/raw";
  const now = options.now ?? new Date();
  // ISO timestamp made filename-safe: 2026-08-09T12:00:00.000Z → 2026-08-09T12-00-00-000Z
  const stamp = now.toISOString().replace(/[:.]/g, "-");
  const targetDir = join(dir, sourceId);
  await mkdir(targetDir, { recursive: true });

  const data = typeof payload === "string" ? Buffer.from(payload, "utf8") : Buffer.from(payload);
  const sha256 = createHash("sha256").update(data).digest("hex");
  const fileName = `${stamp}.raw`;
  const rawPath = join(targetDir, fileName);
  const checksumPath = `${rawPath}.sha256`;

  await writeFile(rawPath, data);
  await writeFile(checksumPath, `${sha256}  ${fileName}\n`, "utf8");
  return { rawPath, checksumPath, sha256 };
}