Flat-File MA accented character conversion RRS feed

  • Question

  • Hello, we are using FIM 2010 R2 SP1 (4.1.3599) with a flat file MA to import/sync a csv flat-file generated by a HR system. We are using UTF-8 code page. The HR system has many international people in it, so there are many accented or diacritic characters.  We need to convert those to regular English non-accented equivalents.  I know we could write sync rule or rules extension code to do this. However, there is an option on the select attributes properties page, under advanced a checkbox that says "Replace accented characters with non-accented variants", which looked like an easy way to fix this issue.  However, enabling that option does not seem to change anything during the import/sync process.  Has anyone had any success using that option, and is there anything we are missing?
    Tuesday, March 24, 2015 1:42 AM

All replies

  • Can you use the ISO-8859-1 encoding, instead?

    Nosh Mernacaj, Identity Management Specialist

    Tuesday, March 24, 2015 1:15 PM
  • Possibly, I can follow-up with the vendor of the HR application producing the output file we are reading.  If we do that, will the option to replace accented characters work?
    Tuesday, March 24, 2015 5:44 PM
  • I am thinking for you to change the encoding in the File MA to read input as ISO-8859-1. 

    Nosh Mernacaj, Identity Management Specialist

    Tuesday, March 24, 2015 5:47 PM
  • Here is another option.

    1. Take the HR feed and parse it, removing the special chars with some script.

    2. Use the parsed file as input in your File MA.

    Nosh Mernacaj, Identity Management Specialist

    Tuesday, March 24, 2015 5:50 PM
  • We tried multiple code pages in the HR ma, including multiple Latin variations (don't see anything specifically called ISO-8859-1).  However, none of them showed the various characters correctly except for UTF-8.  We were getting a lot of bad conversion characters (like tm and copywright) using the other code pages.  The names are very global with multiple character sets involved (Middle Eastern, Asian and European).   
    Tuesday, March 24, 2015 6:51 PM
  • Yes, I understand we could use a script to convert the characters.  We just wanted to fully eliminate the built-in replace accented characters option before adding the additional work/complexity of a custom coded approach to the solution.
    Tuesday, March 24, 2015 6:54 PM
  • The work I proposed is outside FIM.  Clean file before consuming it in FIM. this is not an uncommon approach.

    I know it is not the best, but under circumstances,

    Nosh Mernacaj, Identity Management Specialist

    • Proposed as answer by Nosh Mernacaj Thursday, May 28, 2015 6:47 PM
    Tuesday, March 24, 2015 7:26 PM