update console outputs

Author: timothycarambat

collector/processLink/convert/generic.js

@@ -106,16 +106,13 @@ function validatedHeaders(headers = {}) {
  */
 async function getPageContent({ link, captureAs = "text", headers = {} }) {
   try {
-    const runtimeSettings = new RuntimeSettings();
-    const launchArgs = runtimeSettings.get("browserLaunchArgs");
-    console.log("launchArgs", launchArgs);
-
     let pageContents = [];
+    const runtimeSettings = new RuntimeSettings();
     const loader = new PuppeteerWebBaseLoader(link, {
       launchOptions: {
         headless: "new",
         ignoreHTTPSErrors: true,
-        args: launchArgs,
+        args: runtimeSettings.get("browserLaunchArgs"),
       },
       gotoOptions: {
         waitUntil: "networkidle2",

docker/.env.example

@@ -363,4 +363,9 @@ GID='1000'
 # Specify the target languages for when using OCR to parse images and PDFs.
 # This is a comma separated list of language codes as a string. Unsupported languages will be ignored.
 # Default is English. See https://tesseract-ocr.github.io/tessdoc/Data-Files-in-different-versions.html for a list of valid language codes.
-# TARGET_OCR_LANG=eng,deu,ita,spa,fra,por,rus,nld,tur,hun,pol,ita,spa,fra,por,rus,nld,tur,hun,pol
+# TARGET_OCR_LANG=eng,deu,ita,spa,fra,por,rus,nld,tur,hun,pol,ita,spa,fra,por,rus,nld,tur,hun,pol
+
+# Runtime flags for built-in pupeeteer Chromium instance
+# This is only required on Linux machines running AnythingLLM via Docker
+# and do not want to use the --cap-add=SYS_ADMIN docker argument
+# ANYTHINGLLM_CHROMIUM_ARGS="--no-sandbox,--disable-setuid-sandbox"

server/.env.example

@@ -362,3 +362,8 @@ TTS_PROVIDER="native"
 # This is a comma separated list of language codes as a string. Unsupported languages will be ignored.
 # Default is English. See https://tesseract-ocr.github.io/tessdoc/Data-Files-in-different-versions.html for a list of valid language codes.
 # TARGET_OCR_LANG=eng,deu,ita,spa,fra,por,rus,nld,tur,hun,pol,ita,spa,fra,por,rus,nld,tur,hun,pol
+
+# Runtime flags for built-in pupeeteer Chromium instance
+# This is only required on Linux machines running AnythingLLM via Docker
+# and do not want to use the --cap-add=SYS_ADMIN docker argument
+# ANYTHINGLLM_CHROMIUM_ARGS="--no-sandbox,--disable-setuid-sandbox"