Movement of chapel source code causes same directory modules to be ignored

[lydia-duncan]

The Chapel compiler resolves unknown use statements first by looking for the module name in the internal and standard modules, and then by looking for the module name in the current directory where the source was located. However, pyChapel compilation of chapel source files moves a copy of the source code to a different location and then compiles that copy, so Chapel can't detect the modules because they aren't present. This means that in order to utilize code that is in the same directory as your Chapel file, one must specify this directory via the "module_dirs" argument.

@safl, what are your thoughts on this?