Community
Participate
Working Groups
Created attachment 276071 [details] Description of the issue using diagrams Please see attached file for description with diagrams.
From the docx: In my previous post to the forum (https://www.eclipse.org/forums/index.php/m/1787606/), I wrote I exported the viewmodel file, and saved it as UTF-8 document with external editor, then imported back/replaced original, and it worked. But this time (with the latest release) it did not work. Since I see no issue regarding UTF-8 handling in EMF part, I suspect somewhere in the code (EMF Forms or ECP) that generates viewmodel (xxxx.view file) you may have code fragment that assumes ascii only, which I could not pinpoint and need somebody’s help. Yes there seems to be a problem with handling UTF-8 when generating the view model. We will check this.
New Gerrit change created: https://git.eclipse.org/r/132736
Gerrit change https://git.eclipse.org/r/132736 was merged to [release_1.19.0]. Commit: http://git.eclipse.org/c/emfclient/org.eclipse.emf.ecp.core.git/commit/?id=cf2e55892c331a6b9c902c78cdbd5598e4c6d3fc
In order to guarantee compatibility of view models to Ecore models which use non-ASCII characters, we enforce saving view models as UTF-8 in EMF Forms. This is done by explicitly setting the encoding to UTF-8 whenever we save a resource containing a view model. Besides supporting UTF-8 encoded Ecore models, this also allows using non-ASCII characters in view models (e.g. for labels or the name of a table column).
*** Bug 542746 has been marked as a duplicate of this bug. ***