I think I was wrong … Jikes could very well affect the character encoding: writing class files is not subject to character encoding as I wrote in my previous mail. However, I forgot that jikes has to read its source files with some encoding: reading java files is subject to character encoding!!
I think what happens on your machine is this: you have configured Tomcat to use UTF-8, thus it generates Java files with UTF-8 encoding. Then, Tomcat calls jikes to read these files, and jikes uses the machine’s default encoding to read the java files. The machine’s default encoding is not UTF-8, and as a consequence, the string constants in your Java files get corrupted.
If you invoke jikes from the command line, you can specify “-encoding utf8” to override the default encoding. Unfortunatly, I don’t know how to specify additional options when invoking jikes from withing Tomcat.
Anybody who knows how to specify additional options for jikes?
Michael
Software AG Germany, Darmstadt
#webMethods#Tamino#API-Management