diff --git a/apps/api/src/__tests__/e2e_withAuth/index.test.ts b/apps/api/src/__tests__/e2e_withAuth/index.test.ts index 331283e..39825c4 100644 --- a/apps/api/src/__tests__/e2e_withAuth/index.test.ts +++ b/apps/api/src/__tests__/e2e_withAuth/index.test.ts @@ -266,7 +266,7 @@ describe("E2E Tests for API Routes", () => { urls.forEach((url: string) => { expect(url.startsWith("https://wwww.mendable.ai/blog/")).toBeFalsy(); }); - }, 60000); // 60 seconds + }, 90000); // 90 seconds it("should return a successful response with a valid API key and limit to 3", async () => { const crawlResponse = await request(TEST_URL) diff --git a/apps/api/src/scraper/WebScraper/crawler.ts b/apps/api/src/scraper/WebScraper/crawler.ts index adfd435..9340aa8 100644 --- a/apps/api/src/scraper/WebScraper/crawler.ts +++ b/apps/api/src/scraper/WebScraper/crawler.ts @@ -25,7 +25,7 @@ export class WebCrawler { initialUrl, includes, excludes, - maxCrawledLinks, + maxCrawledLinks = 10000, limit = 10000, generateImgAltText = false, maxCrawledDepth = 10, @@ -152,7 +152,7 @@ export class WebCrawler { inProgress?: (progress: Progress) => void, ): Promise<{ url: string, html: string }[]> { const queue = async.queue(async (task: string, callback) => { - if (this.crawledUrls.size >= this.maxCrawledLinks) { + if (this.crawledUrls.size >= Math.min(this.maxCrawledLinks, this.limit)) { if (callback && typeof callback === "function") { callback(); } @@ -176,14 +176,14 @@ export class WebCrawler { if (inProgress && newUrls.length > 0) { inProgress({ current: this.crawledUrls.size, - total: this.maxCrawledLinks, + total: Math.min(this.maxCrawledLinks, this.limit), status: "SCRAPING", currentDocumentUrl: newUrls[newUrls.length - 1].url, }); } else if (inProgress) { inProgress({ current: this.crawledUrls.size, - total: this.maxCrawledLinks, + total: Math.min(this.maxCrawledLinks, this.limit), status: "SCRAPING", currentDocumentUrl: task, });