Skip to content


Subversion checkout URL

You can clone with
Download ZIP


umlauts in sitesearch keywords #3450

ryrun opened this Issue · 12 comments

5 participants


I love the new features of 1.9, you are great guys. Good work.
I've only a small problem with tracked keywords from sitesearch. Umlauts wont be displayed correctly. I saw that all urls with umlaut keywords are also not displayed correctly in the page urls view.
how does piwik detect the charset of a page?
Keywords: piwik charset sitesearch


Can't confirm this on my installation of Piwik 1.9. Please provide more information about your site (charset, php, etc.). Maybe provide a screenshot.




Site is using charset "iso-8859-15".


Url-Parameter of the piwik request:

Should be tracked as "Tte".


Thanks for the report. We should urldecode the keyword twice


Iam not sure if i have the same problem as megadroid. its also with the "umlauts" but i dont see the squared question marks instead my search words just end (they get cut off) when an umlaut comes up...



(In [7289]) Fixes #3450, handle non-utf-8 character encoding in query parameters of tracked URLs.


In 1.9.2 the problem occurs again. Please see changes in reencodeParameters-function.


Agreed, problem show up again with 1.9.2.



is there maybe a fix for this or can we revert to a previous version of function that works in 1.9.2?


(In [7521]) Site search improvements

  • Support &kwd[]= notation fixes #3454
  • Support random text case fixes #3539
  • Fixes umlauts regression with non utf8 encoding: Fixes #3450
  • Adding setPageCharset() method to Tracking API FIxes #3565

(In [7523]) Refs #3450 Fix regression in the Javascript tracker

@ryrun ryrun added this to the 1.10 - Piwik 1.10 milestone
@mattab mattab was assigned by ryrun
This issue was closed.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.