# Changelog of 2GIS Business Details and Contacts Scraper (`arjun_code/2gis-business-details-scraper`) Actor

- **URL**: https://apify.com/arjun\_code/2gis-business-details-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/arjun\_code/2gis-business-details-scraper.md

## Changelog

### 0.2.18

- Clarified the direct-URL lookup positioning in the Store and SEO descriptions.
- Added ready-to-run public task links and use cases to the English and Russian README sections.

### 0.2.17

- Stop requesting more business URLs when the run's charge limit is reached; report skipped URLs in `OUTPUT`.
- Document the event pricing in English and Russian.

### 0.2.16

- Charge the `business-detail` event for successful business records, without charging the event for lookup errors.

### 0.2.15

- Added English and Russian Actor description and README guidance. The scraping workflow and output fields are unchanged.

### 0.2.14

- Parse the business detail embedded in HTTP 410 responses before treating the page as unavailable.

### 0.2.13

- Set the public input limit to 20 business URLs per run.

### 0.2.12

- Switched detail requests to `curl_cffi` Chrome impersonation to improve compatibility with 2GIS pages.

### 0.2.11

- Changed the display title to identify the URL-based business details and contacts workflow.
- On a detail request timeout, use one residential request instead of another datacenter request.

### 0.2.10

- Corrected city and canonical URL parsing for language-prefixed 2GIS links.
- Limited request waiting and added one new-connection retry for timeouts.
- Reworked the Dataset table into business and contact views.
- Replaced the illustrative README response with an excerpt from a real run.

### 0.2.0

- Repositioned the Actor for direct 2GIS business URL lookups.
- Replaced browser-based keyword discovery with lightweight HTTP detail requests.
- Switched the HTTP request layer to `curl_cffi`.
- Added five-request `curl_cffi` concurrency using a lightweight HTTP/1.1 request path.
- Added automatic datacenter proxy rotation for reliable cloud requests.
- Removed residential proxy, Chromium, DrissionPage, and external website enrichment.
- Reduced the default Actor memory from 1024 MB to 256 MB.

### 0.1.0

- Added the original browser-based keyword search prototype.