Function humanizeSlug

humanizeSlug(slug: string): string

Turns a product slug back into something close to the store's product title.

Slugs flatten punctuation to hyphens, so this reverses the common cases:

v-v / w-w / w-v → (v/v), and marks the number before it as a percent
concentration context → percent: 10-aqueous-solution → 10% aqueous solution
leading digit runs → locants: 1-2-dichloroethane → 1,2-dichloroethane
three-digit groups → thousands: 1-000ppm → 1,000ppm
ascending pairs → ranges: 20-30-mesh → 20-30 mesh
other digit pairs → decimals: 0-100-normal → 0.100 normal
1g-l / 10-g-l → 1g/L / 10 g/L
p-xylene → keeps the stereo/position prefix hyphen

Measured at ~74% exact match (case- and comma-insensitive) against real product titles; the remainder is unrecoverable from a slug (commas, editorial parentheticals, degree signs, #, =, +). The real title is recovered from the product page later, so this only needs to be good enough to rank matches.

Parameters

slug: string
A product slug, e.g. sodium-metasilicate-anhydrous.

Returns string

The humanized product name.

Example

humanizeSlug('isopropyl-alcohol-70-v-v-aqueous-solution');
// "isopropyl alcohol 70% (v/v) aqueous solution"

Source

export function humanizeSlug(slug: string): string {
  const parts = slug.split('-').filter(Boolean);
  const out: string[] = [];

  for (let i = 0; i < parts.length; i++) {
    const cur = parts[i];
    const next = parts[i + 1];
    const prev = out[out.length - 1];

    // "1-2-dichloroethane" -> "1,2-dichloroethane": leading digits are locants.
    if (
      i === 0 &&
      DIGITS.test(cur) &&
      next &&
      DIGITS.test(next) &&
      parts[i + 2] &&
      !DIGITS.test(parts[i + 2])
    ) {
      out.push(`${cur},${next}-${parts[i + 2]}`);
      i += 2;
      continue;
    }

    // "p-xylene", "tert-butyl": stereo/position prefixes keep their hyphen.
    if (i === 0 && next && CHEM_PREFIX.test(cur)) {
      out.push(`${cur}-${next}`);
      i++;
      continue;
    }

    // "v-v" -> "(v/v)"; a bare number in front of it is a percentage.
    if (next && RATIO.test(cur) && RATIO.test(next)) {
      if (prev && /\d$/.test(prev)) out[out.length - 1] = `${prev}%`;
      out.push(`(${cur}/${next})`);
      i++;
      continue;
    }

    // "1g-l" / "10-g-l" -> "1g/L" / "10 g/L"
    if (next === 'l' && VOLUME_UNIT.test(cur.replace(/^\d+/, ''))) {
      out.push(`${cur}/L`);
      i++;
      continue;
    }

    if (next && DIGITS.test(cur)) {
      // Leading zero is always a decimal: "0-0855" -> "0.0855"
      if (cur.startsWith('0') && DIGITS.test(next)) {
        out.push(`${cur}.${next}`);
        i++;
        continue;
      }
      // Three digits following is a thousands group: "1-000ppm" -> "1,000ppm"
      if (LEADS_3_DIGITS.test(next)) {
        out.push(`${cur},${next}`);
        i++;
        continue;
      }
      // Ascending multi-digit pair is a range: "20-30-mesh" -> "20-30 mesh"
      if (DIGITS.test(next) && cur.length >= 2 && next.length >= 2 && Number(next) > Number(cur)) {
        out.push(`${cur}-${next}`);
        i++;
        continue;
      }
      // Otherwise a decimal: "6-0-normal" -> "6.0 normal"
      if (DIGITS.test(next)) {
        out.push(`${cur}.${next}`);
        i++;
        continue;
      }
    }

    out.push(cur);
  }

  for (let i = 0; i < out.length - 1; i++) {
    // "10 aqueous solution" -> "10% aqueous solution"; also catches decimals.
    if (/^\d+([.,]\d+)?$/.test(out[i]) && CONCENTRATION_CONTEXT.has(out[i + 1])) {
      out[i] = `${out[i]}%`;
    }
    // Conductivity standards: "1,500 s solution" -> "1,500 µS solution"
    if (out[i + 1] === 's' && /^\d+,\d{3}$/.test(out[i])) out[i + 1] = 'µS';
  }

  return out.join(' ');
}