clean-web-scraper 4.0.6 → 4.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/example-usage.js CHANGED
@@ -231,7 +231,7 @@ async function mondoweiss ( enable )
231
231
  includeMetadata: true,
232
232
  metadataFields: ["author", "articleTitle", "pageTitle", "description", "dateScrapedDate"],
233
233
  crawlingDelay: 0,
234
- batchSize: 10
234
+ batchSize: 20
235
235
  };
236
236
  return await runScraper( config, enable );
237
237
  }
@@ -303,11 +303,11 @@ async function palestineremembered ( enable )
303
303
 
304
304
  void async function main ()
305
305
  {
306
- const palianswersScraper = await palianswers( true );
307
- const decolonizepalestineScraper = await decolonizepalestine( true );
308
- const khameneiIrFreePalestineTagScraper = await khameneiIrFreePalestineTag( true );
309
- const electronicintifadaScraper = await electronicintifada( true );
310
- const standWithPalestineScraper = await standWithPalestine( true );
306
+ const palianswersScraper = await palianswers( false );
307
+ const decolonizepalestineScraper = await decolonizepalestine( false );
308
+ const khameneiIrFreePalestineTagScraper = await khameneiIrFreePalestineTag( false );
309
+ const electronicintifadaScraper = await electronicintifada( false );
310
+ const standWithPalestineScraper = await standWithPalestine( false );
311
311
  const mondoweisScraper = await mondoweiss( true );
312
312
  const bdsmovementScraper = await bdsmovement( false );
313
313
  const palestinerememberedScraper = await palestineremembered( false );
package/main.js CHANGED
@@ -110,7 +110,7 @@ class WebScraper
110
110
 
111
111
  async processUrl ( url, depth, queue )
112
112
  {
113
- console.log( `Processing URL: ${url}` );
113
+ console.log( `Processing URL: ${url}. Total visited: ${this.visited.size}. Total processed: ${this.allProcessedContent.length}.` );
114
114
  if ( this.hasReachedMax( depth ) )
115
115
  {
116
116
  return;
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "clean-web-scraper",
3
- "version": "4.0.6",
3
+ "version": "4.1.0",
4
4
  "main": "main.js",
5
5
  "scripts": {
6
6
  "start": "node main.js",