# Changelog of IndiaMART Scraper - Suppliers, Exporters, Prices & GST (`haketa/indiamart-scraper`) Actor

- **URL**: https://apify.com/haketa/indiamart-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/haketa/indiamart-scraper.md

## Changelog

### 1.1.0 — 2026-11-01

- **Quality verification:** Confirmed a fresh run of **540 unique, de-duplicated records** across 6 broad keywords (`industrial valves`, `basmati rice`, `led lights`, `submersible pump`, `cotton fabric`, `plywood`) in **94 seconds** on a US datacenter proxy. Core/identity fields (`productName`, `productUrl`, `displayId`, `companyName`, `companyProfileUrl`) came back at **100%**, with `gstNumber` 100%, `supplierWebsite` 100%, `callNumber` 100%, `supplierRating` 100%, `specifications` 98%, `iecNumber` 96%, `price` 93% and `exportCountries` 83% — at a **0% duplicate rate** on product ID.
- **Bug fix — clean descriptions:** Rebuilt the HTML-entity decoder in `cleanText`. Product `description` (and other text) now resolves the full range of entities IndiaMART emits — `–` → `–`, `—` → `—`, `“`/`”` → `“ ”`, `…` → `…`, `°` → `°`, `×` → `×`, `±` → `±`, `é` → `é`, non-breaking spaces, plus **all numeric entities** (` `, `–`, hex ` `). Previously these were left as raw codes in ~18% of records; they now render as proper, human-readable text. Unknown entities are still left untouched (no crashes).
- **README refresh:** Replaced the output sample with a real record from the verification run, refreshed the "Proven results" table with the fresh per-keyword coverage numbers, and updated the maintenance note.
- **Verified:** Pagination (export storefront, 2 pages per term) and `autoExpand` category fan-out reliably hit the per-keyword target; all output field names and proxy behaviour confirmed against the live `export.indiamart.com` search API.