Why does the doc() function reject non-ASCII character

Hi Folks,

I was given a large number of XML files to convert to HTML using an XSLT
program I have written.

I collected all the filenames into a file.

In my XSLT program I have a loop that:

    - reads in a filename
    - uses the filename in the doc() function to read in the file
    - process the file

The problem is that some filenames contain odd characters. Here is an example
of a filename:

    PalaisAfacettes_jpg.xml

Notice the two dots over the 'y' character (diaeresis).

When my XSLT program gets to that filename it throws an error:

    Failed to read input file ...

Why?

Why do I get that error?

Why does the doc() function care if the filename contains a diaeresis?

Second question: do you have suggestions on how to locate in my file (that
contains all the filenames) the offending filenames?

/Roger