#221 Charset problem in exe file description field.

[Garret Wilson]

I'm generating an executable from Maven via the launch4j-maven-plugin. Because my Maven configuration (which gets passed to the Launch4J Maven Plugin) is in XML, I am confident that the source encoding has no problem, as XML ultimately gets parsed as Unicode. My project description is <description>FooBar™</description>, and this is used as the Launch4J file description:

<fileDescription>${project.description}</fileDescription>

Unfortunately in the resulting foobar.xml, if I right click and look at Details > "File description", it shows FooBar?, not FooBar™. Apparently the incorrect charset was used in the file. (I don't know what charset a .exe file expects for this metadata.)

Note that the copyright symbol © shows up fine in the "Copyright" field. This is not surprising as the copyright symbol is U+00A9 (within ISO-8859-1 and even ASCII), while the trademark symbol is U+2122. So the copyright symbol is more lenient, as you could get the charset incorrect and still have the copyright symbol show up as it is the same value in many charsets.

(Note that I already searched the tickets for "charset", "Unicode", and "encoding" and didn't find anything that seemed relevant.)