Bing Searcher

The BingSearcher is a specialized search engine scraper for Bing, built on top of the WebSearcher framework. It uses browser-based scraping to navigate through search results effectively while handling Bing's specific features.

Features

• Browser-based Execution: Uses Playwright (via @isdk/web-fetcher) to simulate real user behavior, helping to bypass basic anti-bot measures.
• Pagination: Automatically handles multi-page results using Bing's first parameter.
• Data Extraction: Extracts titles, URLs, and snippets.
• Date Extraction: Supports optional deep extraction of publication dates from result pages.
• International Edition Support: Specifically designed to handle regional redirects.

Usage

import { BingSearcher, WebSearcher } from '@isdk/ai-tools'

// Register if not already done
WebSearcher.register(BingSearcher)

const results = await WebSearcher.search('bing', 'typescript design patterns', {
  limit: 20,
  international: true, // Force Bing International Edition
  region: 'US',
  timeRange: 'month',
})

Options

In addition to standard SearchOptions, BingSearcher supports:

| Option | Type | Description | | :-------------- | :----------------- | :--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | international | boolean | If true, appends ensearch=1 to the URL. This forces Bing to use the International Edition, bypassing the redirect to cn.bing.com in China. Defaults to true if region is not 'CN' or language is not 'zh'. | | region | string | ISO 3166-1 alpha-2 country code (e.g., 'US', 'GB'). Maps to cc parameter. | | language | string | ISO 639-1 language code (e.g., 'en', 'fr'). Maps to setlang parameter. | | timeRange | string \| object | Filter by time: 'hour' (fallbacks to 'day'), 'day', 'week', 'month', 'year'. Also supports custom range object: { from: string, to?: string }. | | safeSearch | string | 'off', 'moderate', 'strict'. | | needDate | boolean | If true, will attempt to fetch and extract publication dates for each result. |

Implementation Details

• Engine: browser (Headless mode by default, can be configured).
• Pagination Type: url-param (parameter: first).
• Selector: Targets #b_results for main results and li.b_algo for individual items.