fix(crawler/rust): dedupe

mendableai · Jan 30, 2025 · a7eb2f7 · a7eb2f7
1 parent c88176a
commit a7eb2f7
Showing 1 changed file with 2 additions and 2 deletions.
diff --git a/apps/api/src/scraper/WebScraper/crawler.ts b/apps/api/src/scraper/WebScraper/crawler.ts
@@ -401,13 +401,13 @@ export class WebCrawler {
 
   public async extractLinksFromHTML(html: string, url: string) {
     try {
-      return (await this.extractLinksFromHTMLRust(html, url)).map(x => {
+      return [...new Set((await this.extractLinksFromHTMLRust(html, url)).map(x => {
         try {
           return new URL(x, url).href
         } catch (e) {
           return null;
         }
-      }).filter(x => x !== null) as string[];
+      }).filter(x => x !== null) as string[])];
     } catch (error) {
       this.logger.error("Failed to call html-transformer! Falling back to cheerio...", {
         error,