Hi all, I see there is special code to handle the Content-Location header. It seems a special parsing mode has been added to the StructuredFieldParser (setFoldingPreserved), during commits to support content-location. The very weird thing is that MaximalBodyDescriptor calls setFoldingPreserved with a FALSE parameter, that is also the default for that class. So in mime4j code no one calls that code with "true". As part as another refactoring (fold/unfold consistency) I was going to remove that special handling. My readings about the Content-Location specs seems to suggest that we just need to remove any whitespace after unfolding.
Is there any real world message about how MUAs (supporting this header) encode it? Stefano
