[OS X TeX] Unwanted conversion of Unicode characters to \UTF{code}

[Luís Fernando Sequeira]

Hello

I have a friend using TeXShop that is writing documents in Portuguese. 
The documents are encoded in UTF-8, as they should be.

Every time a non-ASCII character is entered, it looks ok, like áéçÕ, etc.
But neither pdflatex with inputenc, nor xelatex with fonspec are able to typeset.

What happens is these characters are strangely converted to something like \UTF{00E3},
which cause the typeset to fail at that point. 


Even weirder, if he sends the files to me via e-mail, I receive them with these unwanted substitutions, so it seems there is something in reading these files that automatically converts them. For example:

Original file:
\author{João António-dos-Santos}

Comes out as:
\author{Jo\UTF{00E3}o Ant\UTF{00F3}nio-dos-Santos}

when sent to me via e-mail...

Any ideas of how to deal with it?

TIA

Luís Sequeira