3 Replies Latest reply: Aug 12, 2014 4:54 AM by Erik Korme RSS

    Bug? Nordic (æøå) characters garbled using $INCLUDE and UTF-8 encoded .qvs

    Jon Martin Gulliksen

      [Qlivkiew 11.20 SR3, using -prj folders]

       

      My source files contain field names using Norwegian characters.

       

      I am moving scripts out of the QVWs and into QVS for improved version control (Git), and encountered the following issue:

       

      When using Norwegian/Nordic characters in a QVS that is stored with UTF-8 encoding, I get garbage letters back when executing the script. It runs fine when the same text sits in the QVW.

       

      I have a partially successful workaround by saving the QVS as Windows-1252 instead, so this seems to be encoding-related. But, this is not a solution as it confuses my version control, and, UTF is after all the new standard and this should frankly load without issues.

       

      Screenshot from attatched example QVW, where the same data is loaded twice, with different results:

      utf8vs1252.png

       

      The two .qvs files contain inline tables with identical data (different table/fieldname to separate on load), but have been saved with the respective encodings in an external text editor (Sublime Text 2):

       

      Table_UTF8:

      LOAD * INLINE [

          Encoded_with_UTF8

          æøå

      ];

       

      Table_1252:

      LOAD * INLINE [

          Encoded_with_1252:

          æøå

      ];

       

      The QVW load script has these lines:

       

      $(include=utf-8.qvs)

      $(include=1252.qvs)

       

      My questions are:

       

      - Is there a way to specify which encoding should be used when $INCLUDE-ing (similar to when you LOAD/STORE)?

       

      - Anyone else is able to reproduce this?

       

      - Could it be a bug in Qlikview? And if so, how can I report it to Qliktech? I can not find any "submit case" section in my Customer Portal.