From dba96998e35a813a447d020e52114be63f922b38 Mon Sep 17 00:00:00 2001 From: Nicolas Date: Thu, 3 Oct 2024 18:56:51 -0300 Subject: [PATCH] Update fetch.ts --- .../src/scraper/WebScraper/scrapers/fetch.ts | 21 +++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/apps/api/src/scraper/WebScraper/scrapers/fetch.ts b/apps/api/src/scraper/WebScraper/scrapers/fetch.ts index 88659797..0df3be72 100644 --- a/apps/api/src/scraper/WebScraper/scrapers/fetch.ts +++ b/apps/api/src/scraper/WebScraper/scrapers/fetch.ts @@ -35,7 +35,9 @@ export async function scrapWithFetch( }); if (response.status !== 200) { - Logger.debug(`⛏️ Axios: Failed to fetch url: ${url} with status: ${response.status}`); + Logger.debug( + `⛏️ Axios: Failed to fetch url: ${url} with status: ${response.status}` + ); logParams.error_message = response.statusText; logParams.response_code = response.status; return { @@ -48,7 +50,10 @@ export async function scrapWithFetch( const contentType = response.headers["content-type"]; if (contentType && contentType.includes("application/pdf")) { logParams.success = true; - const { content, pageStatusCode, pageError } = await fetchAndProcessPdf(url, pageOptions?.parsePDF); + const { content, pageStatusCode, pageError } = await fetchAndProcessPdf( + url, + pageOptions?.parsePDF + ); logParams.response_code = pageStatusCode; logParams.error_message = pageError; return { content, pageStatusCode: response.status, pageError }; @@ -57,7 +62,11 @@ export async function scrapWithFetch( logParams.success = true; logParams.html = text; logParams.response_code = response.status; - return { content: text, pageStatusCode: response.status, pageError: null }; + return { + content: text, + pageStatusCode: response.status, + pageError: null, + }; } } catch (error) { if (error.code === "ECONNABORTED") { @@ -67,7 +76,11 @@ export async function scrapWithFetch( logParams.error_message = error.message || error; Logger.debug(`⛏️ Axios: Failed to fetch url: ${url} | Error: ${error}`); } - return { content: "", pageStatusCode: error.response?.status ?? null, pageError: logParams.error_message }; + return { + content: "", + pageStatusCode: error.response?.status ?? null, + pageError: logParams.error_message, + }; } finally { const endTime = Date.now(); logParams.time_taken_seconds = (endTime - logParams.startTime) / 1000;