Opening hours validator

Before you import a few thousand rows of scraped store hours, find out which ones are junk. This checks a hours string, reports a confidence score, lists any segment it could not read, and names the days left undefined.

…

What counts as a problem

• Unparsed segments — text the parser refused to guess at. Usually a note like "hours vary in summer" that a human has to resolve.
• Undefined days — a day never mentioned as open or closed. Most importers treat that as closed, which is often wrong.
• Zero-length ranges — 9-9, typically a typo for a full day or for 9-9pm.
• Low confidence — below about 0.6, review by hand before importing.

Validating in bulk

The /v1/bulk endpoint takes up to 500 strings per call and returns the same confidence and unparsed_segments fields for each, so you can sort a whole catalogue worst-first and only look at the tail.

Do this at scale

This page runs on the public OpenHours API. The same call handles bulk input — up to 500 hours strings per request — and returns all three formats at once. Free tier is 100 requests/day with no signup.

curl -X POST https://smeltworks.com/openhours/v1/parse \
  -H 'content-type: application/json' \
  -d '{"text":"Mon-Fri 9am-5pm, Sat 10-2, closed Sun"}'

Starter — $9/mo, 10k req Growth — $29/mo, 100k req