I've built a script that scans a folder for PDF files and then creates an XML file in the ofrmat indicated in the documentation, complete with the <?xml> header etc.
The file is well-formed, and uses only the tags indicated in the documentation. When I pass the file as an argument to pdfmerge.exe on teh command line, I get a dialog box:
"There is no Unicode byte order mark. Cannot switch to Unicode."
I'd use the old-style command file format, except that I have unicode filenames.
PdfMerge version 1.19.3361.21520