/**
 * Keencrest: find business emails for a list of websites (Google Sheets).
 * Uses the Website Contact Details Scraper by Data Gleaner on Apify.
 * https://apify.com/datagleaner/website-contact-details-scraper
 *
 * SETUP
 * 1. Sheet layout: row 1 is a header. Websites in column A from row 2 down.
 *    Columns B (email), C (status) and D (reason) receive the results.
 * 2. Extensions > Apps Script > paste this file > Project Settings > Script properties:
 *    add APIFY_TOKEN = your own Apify API token (Apify Console > Settings > Integrations).
 * 3. Reload the sheet. A "Keencrest" menu appears. Run "Find emails (selected rows or all)".
 *
 * COST: $0.002 per site where an email is found, nothing for sites without one.
 * MAX_USD below is a hard cap per batch. Batches of BATCH sites keep each request short.
 */
const BATCH = 100;      // sites per request
const MAX_USD = 1;      // maxTotalChargeUsd per request
const ACTOR = "datagleaner~website-contact-details-scraper";

function onOpen() {
  SpreadsheetApp.getUi().createMenu("Keencrest")
    .addItem("Find emails (selected rows or all)", "enrichEmails")
    .addToUi();
}

function toDomain_(s) {
  return String(s).trim().replace(/^https?:\/\//i, "").replace(/^www\./i, "").split("/")[0].toLowerCase();
}

function enrichEmails() {
  const token = PropertiesService.getScriptProperties().getProperty("APIFY_TOKEN");
  if (!token) throw new Error("Add the script property APIFY_TOKEN first (see the setup notes at the top).");
  const sheet = SpreadsheetApp.getActiveSheet();
  const last = sheet.getLastRow();
  if (last < 2) return;
  const sel = sheet.getActiveRange();
  // Selected rows if the selection is inside the data, otherwise every row.
  let first = 2, count = last - 1;
  if (sel && sel.getRow() >= 2 && sel.getRow() <= last && sel.getNumRows() > 0 && sel.getNumRows() < last - 1) {
    first = sel.getRow(); count = Math.min(sel.getNumRows(), last - first + 1);
  }
  const raw = sheet.getRange(first, 1, count, 1).getValues().map(r => String(r[0]).trim());
  const rows = []; // {idx, domain}
  raw.forEach((v, i) => { if (v) rows.push({ idx: i, domain: toDomain_(v) }); });
  const out = raw.map(() => ["", "", ""]);
  for (let s = 0; s < rows.length; s += BATCH) {
    const chunk = rows.slice(s, s + BATCH);
    const domains = Array.from(new Set(chunk.map(r => r.domain)));
    const url = "https://api.apify.com/v2/acts/" + ACTOR + "/run-sync-get-dataset-items" +
      "?maxTotalChargeUsd=" + MAX_USD + "&fields=domain,bestEmail,status,failureReason";
    const res = UrlFetchApp.fetch(url, {
      method: "post", contentType: "application/json",
      headers: { Authorization: "Bearer " + token },
      payload: JSON.stringify({ websites: domains, maxPagesPerSite: 4 }),
      muteHttpExceptions: true
    });
    if (res.getResponseCode() !== 201) throw new Error(res.getResponseCode() + " " + res.getContentText().slice(0, 300));
    // Output order is not the input order: join on domain.
    const byDomain = {};
    JSON.parse(res.getContentText()).forEach(i => { byDomain[i.domain] = i; });
    chunk.forEach(r => {
      const i = byDomain[r.domain] || {};
      out[r.idx] = [i.bestEmail || "", i.status || "missing", i.failureReason || ""];
    });
  }
  sheet.getRange(first, 2, count, 3).setValues(out);
}