[TxMt] Textencoding problems renders scripts useless

[Stefan]

Hi list,

generally, I use textmate to write source code for various languages
and systems. Everything works nicely.

Textmate is set to use ISO-Latin1.

 From time to time, I need to copy text fragments from Word oder PDF  
documents
to PHP scripts. As soon save the file with the copied text, the  
encoding of textmate
document suddenly changes. All diacritical characters get converted  
to funny
chars. All in all, the resulting PHP isn't any longer useable.

It took me some time to exactly figure out the reasons. But, I failed  
to solve it.

Does someone has an idea regarding this?

BTW: If I try this using Eclipse, the IDE reports a problem saving  
the file, since
the encoding changed. I have to start windows and copy the Word text  
fragments -
since Eclipse on Windows works just fine.

An ideas?

Kind regards,

Stefan