process.env.DATABASE_URL ??=
  "postgresql://sec_cybert:sec_cybert@localhost:5432/sec_cybert";

import { readFile } from "node:fs/promises";
import { ne } from "drizzle-orm";
import { db } from "../db";
import * as schema from "../db/schema";
import { generateAssignments, printAssignmentStats } from "../lib/assignment";

const SAMPLED_IDS_PATH =
  "/home/joey/Documents/sec-cyBERT/labelapp/.sampled-ids.json";

async function main() {
  // 1. Read sampled paragraph IDs
  console.log("Reading sampled paragraph IDs...");
  const raw = await readFile(SAMPLED_IDS_PATH, "utf-8");
  const paragraphIds: string[] = JSON.parse(raw);
  console.log(`  ${paragraphIds.length} paragraph IDs loaded`);

  // 2. Read annotator IDs from DB (exclude admin)
  console.log("Loading annotators...");
  const annotators = await db
    .select({ id: schema.annotators.id })
    .from(schema.annotators)
    .where(ne(schema.annotators.id, "admin"));
  const annotatorIds = annotators.map((a) => a.id).sort();
  console.log(`  ${annotatorIds.length} annotators: ${annotatorIds.join(", ")}`);

  // 3. Generate BIBD assignments
  console.log("Generating BIBD assignments...");
  const assignments = generateAssignments(paragraphIds, annotatorIds, 3);

  // 4. Print stats before inserting
  printAssignmentStats(assignments, annotatorIds);

  // 5. Insert into DB in batches
  console.log("\nInserting assignments into DB...");
  const BATCH_SIZE = 1000;
  for (let i = 0; i < assignments.length; i += BATCH_SIZE) {
    const batch = assignments.slice(i, i + BATCH_SIZE);
    await db
      .insert(schema.assignments)
      .values(batch)
      .onConflictDoNothing();

    const progress = Math.min(i + BATCH_SIZE, assignments.length);
    console.log(`  Inserted ${progress}/${assignments.length} assignments`);
  }

  console.log("Assignment complete.");
  process.exit(0);
}

main().catch((err) => {
  console.error("Assignment failed:", err);
  process.exit(1);
});