prop-ai-hr/scripts/stage-legacy-knowledge.mjs

#!/usr/bin/env node

import fs from 'node:fs/promises';
import path from 'node:path';
import { fileURLToPath } from 'node:url';

export function parseArgs(argv) {
  const options = {
    baseUrl: process.env.AIHR_LEGACY_BASE_URL || 'https://wygj-api.localhost',
    token: process.env.AIHR_LEGACY_TOKEN,
    clientid: process.env.AIHR_LEGACY_CLIENT_ID,
    batchSize: 50,
    afterAttachmentId: 0,
    maxBatches: 1,
    manifest: 'tmp/legacy-knowledge-stage-manifest.json',
    execute: false
  };
  for (let index = 0; index < argv.length; index += 1) {
    const value = argv[index];
    if (value === '--base-url') options.baseUrl = argv[++index];
    else if (value === '--token') options.token = argv[++index];
    else if (value === '--clientid') options.clientid = argv[++index];
    else if (value === '--batch-size') options.batchSize = Number(argv[++index]);
    else if (value === '--after-attachment-id') options.afterAttachmentId = Number(argv[++index]);
    else if (value === '--max-batches') options.maxBatches = Number(argv[++index]);
    else if (value === '--manifest') options.manifest = argv[++index];
    else if (value === '--execute') options.execute = true;
    else if (value === '--dry-run') options.execute = false;
    else throw new Error(`Unknown option: ${value}`);
  }
  if (!Number.isInteger(options.batchSize) || options.batchSize < 20 || options.batchSize > 100) {
    throw new Error('--batch-size must be an integer between 20 and 100');
  }
  if (!Number.isSafeInteger(options.afterAttachmentId) || options.afterAttachmentId < 0) {
    throw new Error('--after-attachment-id must be a non-negative integer');
  }
  if (!Number.isInteger(options.maxBatches) || options.maxBatches < 1) {
    throw new Error('--max-batches must be a positive integer');
  }
  if (options.execute && !options.token) {
    throw new Error('--token or AIHR_LEGACY_TOKEN is required with --execute');
  }
  return options;
}

async function readManifest(file) {
  try {
    const parsed = JSON.parse(await fs.readFile(file, 'utf8'));
    if (parsed.schemaVersion !== 1 || !Array.isArray(parsed.batches)) {
      throw new Error('unsupported manifest schema');
    }
    return parsed;
  } catch (error) {
    if (error.code === 'ENOENT') return null;
    throw new Error(`Cannot read manifest ${file}: ${error.message}`);
  }
}

async function writeManifest(file, manifest) {
  const absolute = path.resolve(file);
  await fs.mkdir(path.dirname(absolute), { recursive: true });
  const temporary = `${absolute}.${process.pid}.tmp`;
  await fs.writeFile(temporary, `${JSON.stringify(manifest, null, 2)}\n`, { encoding: 'utf8', mode: 0o600 });
  await fs.rename(temporary, absolute);
}

function headers(options) {
  const result = { accept: 'application/json' };
  if (options.token) {
    result.Authorization = options.token.startsWith('Bearer ') ? options.token : `Bearer ${options.token}`;
  }
  if (options.clientid) result.clientid = options.clientid;
  return result;
}

async function stageBatch(options, cursor, fetchImpl) {
  const url = new URL('/api/knowledge/quality/legacy/stage', `${options.baseUrl.replace(/\/$/, '')}/`);
  url.searchParams.set('afterAttachmentId', String(cursor));
  url.searchParams.set('limit', String(options.batchSize));
  const response = await fetchImpl(url, { method: 'POST', headers: headers(options) });
  const body = await response.json();
  if (!response.ok || body.code !== 200 || !body.data) {
    throw new Error(`legacy stage failed with HTTP ${response.status} code ${body.code ?? '<missing>'}`);
  }
  return body.data;
}

async function previewBatch(options, cursor, fetchImpl) {
  const url = new URL('/api/knowledge/quality/legacy/preview', `${options.baseUrl.replace(/\/$/, '')}/`);
  url.searchParams.set('afterAttachmentId', String(cursor));
  url.searchParams.set('limit', String(options.batchSize));
  const response = await fetchImpl(url, { method: 'GET', headers: headers(options) });
  const body = await response.json();
  if (!response.ok || body.code !== 200 || !body.data) {
    throw new Error(`legacy preview failed with HTTP ${response.status} code ${body.code ?? '<missing>'}`);
  }
  return body.data;
}

export async function run(options, dependencies = {}) {
  const fetchImpl = dependencies.fetchImpl || globalThis.fetch;
  const now = dependencies.now || (() => new Date());
  const previous = await readManifest(options.manifest);
  const resumeCursor = previous?.status === 'IN_PROGRESS' || previous?.status === 'FAILED'
    ? Number(previous.cursor || 0)
    : options.afterAttachmentId;
  if (!options.execute) {
    const preview = options.token ? await previewBatch(options, resumeCursor, fetchImpl) : null;
    return {
      mode: 'DRY_RUN',
      baseUrl: options.baseUrl,
      batchSize: options.batchSize,
      maxBatches: options.maxBatches,
      cursor: resumeCursor,
      manifest: path.resolve(options.manifest),
      preview,
      note: preview
        ? 'Authenticated preview completed without mutation. Add --execute to stage data into review or quarantine.'
        : 'Offline validation only. Pass --token for a read-only server preview; add --execute to stage data.'
    };
  }

  const manifest = previous && previous.status !== 'COMPLETE'
    ? previous
    : { schemaVersion: 1, status: 'IN_PROGRESS', cursor: resumeCursor, batches: [] };
  manifest.status = 'IN_PROGRESS';
  manifest.updatedAt = now().toISOString();
  await writeManifest(options.manifest, manifest);

  let cursor = Number(manifest.cursor || resumeCursor);
  for (let batchNumber = 0; batchNumber < options.maxBatches; batchNumber += 1) {
    try {
      const result = await stageBatch(options, cursor, fetchImpl);
      const entry = {
        sequence: manifest.batches.length + 1,
        startedAfterAttachmentId: cursor,
        discovered: result.discovered,
        staged: result.staged,
        reparsed: result.reparsed,
        quarantined: result.quarantined,
        stagedAssetIds: result.stagedAssetIds || [],
        failedAttachmentIds: result.failedAttachmentIds || [],
        nextCursor: result.nextCursor,
        moreAvailable: Boolean(result.moreAvailable),
        completedAt: now().toISOString()
      };
      manifest.batches.push(entry);
      if (entry.failedAttachmentIds.length > 0) {
        manifest.status = 'FAILED';
        manifest.lastError = `Batch contains ${entry.failedAttachmentIds.length} failed attachment(s)`;
        manifest.cursor = cursor;
        await writeManifest(options.manifest, manifest);
        throw new Error(`${manifest.lastError}; cursor was not advanced`);
      }
      cursor = Number(result.nextCursor || cursor);
      manifest.cursor = cursor;
      manifest.status = result.moreAvailable ? 'IN_PROGRESS' : 'COMPLETE';
      delete manifest.lastError;
      await writeManifest(options.manifest, manifest);
      if (!result.moreAvailable || result.discovered === 0) break;
    } catch (error) {
      if (manifest.status !== 'FAILED') {
        manifest.status = 'FAILED';
        manifest.lastError = error.message;
        manifest.cursor = cursor;
        manifest.updatedAt = now().toISOString();
        await writeManifest(options.manifest, manifest);
      }
      throw error;
    }
  }
  return {
    mode: 'EXECUTE',
    status: manifest.status,
    cursor: manifest.cursor,
    batches: manifest.batches.length,
    manifest: path.resolve(options.manifest)
  };
}

if (fileURLToPath(import.meta.url) === process.argv[1]) {
  run(parseArgs(process.argv.slice(2)))
    .then((result) => console.log(JSON.stringify(result, null, 2)))
    .catch((error) => { console.error(error.message); process.exitCode = 1; });
}