Shifted/scripts/build-vehicle-catalog.js

const fs = require("fs");
const path = require("path");
const https = require("https");
const { parse } = require("csv-parse/sync");

const START_TOP_YEAR = 1970;
const END_TOP_YEAR = 1991;
const START_FULL_YEAR = 1992;
const END_YEAR = 2026;

const OUTPUT_PATH = path.join(__dirname, "..", "components", "data", "vehicleCatalog.json");

const EXTRA_MAKES = [
  "AMC",
  "Alfa Romeo",
  "Audi",
  "Buick",
  "Cadillac",
  "Chevrolet",
  "Chrysler",
  "Datsun",
  "Dodge",
  "Fiat",
  "Ford",
  "GMC",
  "Honda",
  "Jeep",
  "Land Rover",
  "Lincoln",
  "Mazda",
  "Mercedes-Benz",
  "Mercury",
  "Nissan",
  "Oldsmobile",
  "Peugeot",
  "Plymouth",
  "Pontiac",
  "Renault",
  "Saab",
  "Subaru",
  "Toyota",
  "Volkswagen",
  "Volvo",
];

function fetchUrl(url) {
  return new Promise((resolve, reject) => {
    https
      .get(url, (res) => {
        if (res.statusCode && res.statusCode >= 400) {
          reject(new Error(`HTTP ${res.statusCode} for ${url}`));
          res.resume();
          return;
        }
        let data = "";
        res.on("data", (chunk) => {
          data += chunk;
        });
        res.on("end", () => resolve(data));
      })
      .on("error", reject);
  });
}

function normalizeSpaces(value) {
  return value.replace(/\s+/g, " ").trim();
}

function tokenize(value) {
  return normalizeSpaces(value)
    .replace(/[.,;:]/g, " ")
    .replace(/\.+/g, " ")
    .replace(/&/gi, "&")
    .replace(/[()]/g, " ")
    .split(/\s+/)
    .filter(Boolean);
}

function buildMakeMatchers(makes) {
  const uniqueMakes = Array.from(new Set(makes));
  uniqueMakes.sort((a, b) => b.length - a.length);
  return uniqueMakes.map((make) => ({
    make,
    tokens: make.toLowerCase().split(/\s+/),
  }));
}

function parseTopList(description, makeMatchers) {
  if (!description) return [];
  const lower = description.toLowerCase();
  const marker = "sold:";
  const idx = lower.indexOf(marker);
  if (idx === -1) return [];

  let list = description.slice(idx + marker.length).trim();
  list = list.split("...")[0];
  list = list.replace(/\.$/, "");

  const rawTokens = tokenize(list);
  const lowerTokens = rawTokens.map((t) => t.toLowerCase());

  const results = [];
  let i = 0;

  const matchMakeAt = (start) => {
    for (const entry of makeMatchers) {
      const { tokens } = entry;
      if (start + tokens.length > lowerTokens.length) continue;
      let matched = true;
      for (let j = 0; j < tokens.length; j += 1) {
        if (lowerTokens[start + j] !== tokens[j]) {
          matched = false;
          break;
        }
      }
      if (matched) return entry;
    }
    return null;
  };

  while (i < lowerTokens.length) {
    const match = matchMakeAt(i);
    if (!match) {
      i += 1;
      continue;
    }

    const make = match.make;
    const startModel = i + match.tokens.length;
    let endModel = startModel;

    while (endModel < lowerTokens.length) {
      if (matchMakeAt(endModel)) break;
      endModel += 1;
    }

    const modelTokens = rawTokens.slice(startModel, endModel);
    const model = normalizeSpaces(modelTokens.join(" "));
    if (model) {
      results.push({ make, model });
    }

    i = endModel;
  }

  return results;
}

async function loadFullList() {
  const rows = [];

  for (let year = START_FULL_YEAR; year <= END_YEAR; year += 1) {
    const url = `https://raw.githubusercontent.com/abhionlyone/us-car-models-data/master/${year}.csv`;
    // eslint-disable-next-line no-console
    console.log(`Downloading ${url}`);
    const csv = await fetchUrl(url);
    const records = parse(csv, { columns: true, skip_empty_lines: true });
    for (const record of records) {
      const make = String(record.make || "").trim();
      const model = String(record.model || "").trim();
      if (!make || !model) continue;
      rows.push({ year, make, model, trim: "Base" });
    }
  }

  return rows;
}

async function loadTopList(makeMatchers) {
  const rows = [];

  for (let year = START_TOP_YEAR; year <= END_TOP_YEAR; year += 1) {
    const url = `https://carorigins.com/top-12-cars-of-${year}/`;
    // eslint-disable-next-line no-console
    console.log(`Downloading ${url}`);
    let html = "";
    try {
      html = await fetchUrl(url);
    } catch (err) {
      // eslint-disable-next-line no-console
      console.warn(`Skipping ${year}: ${err.message}`);
      continue;
    }

    const match = html.match(/<meta[^>]+property=\"og:description\"[^>]+content=\"([^\"]+)\"/i);
    if (!match) {
      // eslint-disable-next-line no-console
      console.warn(`Skipping ${year}: og:description not found`);
      continue;
    }

    const description = match[1];
    const vehicles = parseTopList(description, makeMatchers);

    for (const vehicle of vehicles) {
      rows.push({ year, make: vehicle.make, model: vehicle.model, trim: "Base" });
    }
  }

  return rows;
}

async function main() {
  const fullRows = await loadFullList();
  const makesFromFull = fullRows.map((row) => row.make);
  const makeMatchers = buildMakeMatchers([...makesFromFull, ...EXTRA_MAKES]);
  const topRows = await loadTopList(makeMatchers);

  const allRows = [...topRows, ...fullRows];
  const seen = new Set();
  const deduped = [];

  for (const row of allRows) {
    const key = `${row.year}|${row.make}|${row.model}|${row.trim}`;
    if (seen.has(key)) continue;
    seen.add(key);
    deduped.push(row);
  }

  deduped.sort((a, b) => {
    if (a.year !== b.year) return a.year - b.year;
    if (a.make !== b.make) return a.make.localeCompare(b.make);
    if (a.model !== b.model) return a.model.localeCompare(b.model);
    return a.trim.localeCompare(b.trim);
  });

  fs.writeFileSync(OUTPUT_PATH, JSON.stringify(deduped, null, 2));

  // eslint-disable-next-line no-console
  console.log(`Wrote ${deduped.length} rows to ${OUTPUT_PATH}`);
}

main().catch((err) => {
  // eslint-disable-next-line no-console
  console.error(err);
  process.exit(1);
});