Allow user to specify args for chromium process so they dont need SYS_ADMIN on container. (#4397)

* allow user to specify args for chromium process so they dont need SYS_ADMIN perms

* use arg flag content

* update console outputs

.github/workflows/dev-build.yaml

@@ -6,7 +6,7 @@ concurrency:
 
 on:
   push:
-    branches: ['upload-ui-ux'] # put your current branch to create a build. Core team only.
+    branches: ['3999-chromium-flags'] # put your current branch to create a build. Core team only.
     paths-ignore:
       - '**.md'
       - 'cloud-deployments/*'

collector/processLink/convert/generic.js

@@ -5,6 +5,7 @@ const {
 const { writeToServerDocuments } = require("../../utils/files");
 const { tokenizeString } = require("../../utils/tokenizer");
 const { default: slugify } = require("slugify");
+const RuntimeSettings = require("../../utils/runtimeSettings");
 
 /**
  * Scrape a generic URL and return the content in the specified format
@@ -106,10 +107,12 @@ function validatedHeaders(headers = {}) {
 async function getPageContent({ link, captureAs = "text", headers = {} }) {
   try {
     let pageContents = [];
+    const runtimeSettings = new RuntimeSettings();
     const loader = new PuppeteerWebBaseLoader(link, {
       launchOptions: {
         headless: "new",
         ignoreHTTPSErrors: true,
+        args: runtimeSettings.get("browserLaunchArgs"),
       },
       gotoOptions: {
         waitUntil: "networkidle2",

collector/utils/runtimeSettings/index.js

@@ -27,6 +27,16 @@ class RuntimeSettings {
       // Value must be explicitly "true" or "false" as a string
       validate: (value) => String(value) === "true",
     },
+    browserLaunchArgs: {
+      default: [],
+      validate: (value) => {
+        let args = [];
+        if (Array.isArray(value)) args = value.map((arg) => String(arg.trim()));
+        if (typeof value === "string")
+          args = value.split(",").map((arg) => arg.trim());
+        return args;
+      },
+    },
   };
 
   constructor() {

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"

server/utils/collectorApi/index.js

@@ -38,6 +38,7 @@ class CollectorApi {
       },
       runtimeSettings: {
         allowAnyIp: process.env.COLLECTOR_ALLOW_ANY_IP ?? "false",
+        browserLaunchArgs: process.env.ANYTHINGLLM_CHROMIUM_ARGS ?? [],
       },
     };
   }

server/utils/helpers/updateENV.js

@@ -1167,6 +1167,9 @@ function dumpENV() {
 
     // Allow disabling of streaming for generic openai
     "GENERIC_OPENAI_STREAMING_DISABLED",
+
+    // Specify Chromium args for collector
+    "ANYTHINGLLM_CHROMIUM_ARGS",
   ];
 
   // Simple sanitization of each value to prevent ENV injection via newline or quote escaping.