Attributes within <html> tag are stripped in extension, but not in cloud version

Help and Support
1

Hello,

I’m seeing divergent behavior between the Chrome version and distill.io version of the monitor.

Specifically, in the Chrome version, attributes of the <html> tag itself appear stripped, while in the cloud version, they are retained (the desired behavior for me).

Using the config json below when monitoring https://chatgpt.com/ works correctly in the cloud monitor and captures the “data-build” <html> attribute, but shows up as “undefined” in the local extension version.

Moreover, when l view the diff in source view in the extension, I can see that it has performed some source sanitization, like stripping the <html> attributes (which I happen to need), and stripping the <meta> tag entirely (which doesn’t concern my use case). For example:

<html><head><link rel="modulepreload" href="https://cdn.oaistatic.com/assets/manifest-316c9336.js">

Whereas when viewing the same diff in the cloud monitor version, no such sanitization is performed, and the “data-build” attribute within the <html> tag which I need to monitor for changes is still present:

<html lang="en-US" data-build="prod-eefe4fe3afec2efde62cdc0669965885fde88495" dir="ltr" class=""><head><meta charset="UTF-8"><meta name="viewport" content="width=device-width, initial-scale=1"><link rel="modulepreload" href="https://cdn.oaistatic.com/assets/manifest-316c9336.js">

And lastly, the config json I’m using:

{
  "selections": [
    {
      "frames": [
        {
          "index": 0,
          "excludes": [],
          "includes": [
            {
              "expr": "/html",
              "type": "xpath",
              "fields": [
                {
                  "type": "attribute",
                  "name": "data-build"
                }
              ]
            }
          ]
        }
      ],
      "dynamic": false,
      "delay": 0
    }
  ],
  "regexp": {
    "expr": "",
    "flags": "gim"
  },
  "ignoreEmptyText": true,
  "includeStyle": false,
  "viewport": {},
  "dataAttr": "text",
  "params": {},
  "blockAdsAndCookies": false
}

Any thoughts as to whether I can do anything locally to disable the sanitization behavior?

Thanks in advance!

Tom

1 Like
2

thanks for flagging this @hey_tommy. the extension’s module to extract html out is slightly different from cloud monitor’s. we will get this normalized and let you know. cheers!

1 Like
3

Thanks - I appreciate the reply.
Cheers!