> >right. i've now verified that InitText() is never being called, which > >would have initialized ct->c_termproc. so that means an initialization > >of ct->c_ctinitfnx is being missed. and now i really _am_ lost. :-) > > How did you verify that InitText() was never called? I set a breakpoint > for it and it is definitely being called for text parts, at least on the > messages I tested.
Same for me. Your profile entry, Paul, looks like it works properly for me. I even inserted "tr '[a-z]' '[A-Z]'" in the pipeline and it upper-cased the content. Is your locale something like en_US.iso88591? If so, the check_charset() at the end of InitText() would pass, so ct->c_termproc wouldn't be loaded. If InitText() is in fact called, that could explain what you're seeing. But that part of InitText() hasn't changed since 2008, and appears to work as documented in the mhshow(1) man page. parse_display_string() is new, but it mainly factors out two bunches of nearly identical code. It also has fixes to the quoting of escape sequences. It's handling of ct->c_termproc should be identical to 1.5. David _______________________________________________ Nmh-workers mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/nmh-workers
