Conversation
ahmedawwad
commented
Feb 11, 2025
- Page location, remove the www., so that all hostnames will be unified (www.nasa.gov will be nasa.gov)
- All other data points for .gov and .mil we'll remove the www.
- All other data points for non (gov and mil), do not remove www.
- Added additional dimension "hostname_dimension" to fix the sub-property leaky filters.
- Added whitelist parameters (product1 and wwa) to NOAA/DOC agency
- Added whitelist parameters (Topic and TitleID) for NASA agency
- Moved the lon and lat parameters from DOC agency only to the general parameters as they might exist in different agencies
Page location, we’ll remove the www., so that all hostnames will be unified (www.nasa.gov will be nasa.gov) All other data points for .gov and .mil we'll remove the www. All other data points for non (gov and mil), do not remove www. Added additional dimension "hostname_dimension" to fix sub-property leaky filters. Added whitelist parameter (product1 and wwa) for NOAA/DOC agency Added whitelist parameter (Topic and TitleID) for NASA agency Moved the lon and lat parameters from DOC agency only to the general parameters as they might exist in different agencies
| function _setAllowedQS() { | ||
| var queries = { | ||
| "default": ["utm_id", "utm_source", "utm_medium", "utm_campaign", "utm_term", "utm_content", "utm_source_platform", "utm_creative_format", "utm_marketing_tactic", "gbraid", "wbraid", "_gl", "gclid", "dclid", "gclsrc", "affiliate", "dap-dev-env", "v"], | ||
| "default": ["utm_id", "utm_source", "utm_medium", "utm_campaign", "utm_term", "utm_content", "utm_source_platform", "utm_creative_format", "utm_marketing_tactic", "gbraid", "wbraid", "_gl", "gclid", "dclid", "gclsrc", "affiliate", "dap-dev-env", "v", "lat", "lon"], |
Check failure
Code scanning / CodeQL
Polynomial regular expression used on uncontrolled data
|
@ahmedawwad Could you explain how "hostname_dimension" will solve the leaky filter problem? It's not clear to me. |
Currently we are using the full URL to filter data into subproperties. The data leaks happen when a hostname exists in a query string parameter, not just the hostname. it's included where it shouldn't. |
| [oCONFIG.MAIN_HOSTNAME_DIMENSION]: oCONFIG.HOSTNAME.toLowerCase() | ||
| }; | ||
| ((document.referrer && -1 !== document.referrer.search(location.hostname)) ? (r.page_referrer = _scrubbedURL(document.referrer)) : document.referrer); | ||
| ((document.referrer && -1 !== document.referrer.search(location.hostname)) ? (r.page_referrer = _scrubbedURL(document.referrer)) : (r.page_referrer = document.referrer)); |
There was a problem hiding this comment.
@ahmedawwad Could you explain this page_referrer change?