porst17 / surfer.js

A GWT frontend for surf (Java based renderer for algebraic surfaces)
1 stars 1 forks source link

How do I know which part of the code is client side code and which part is server side code? #2

Closed porst17 closed 10 years ago

porst17 commented 10 years ago

E.g. the parser is implemented as a server side module and hence we don't need to add the sources of the ANTLR library. How did you do this separation?

fpetrola commented 10 years ago

GWT use different folders for client and server side, if you take a look at surfer.js/src/org/test you will see "client" and "server" folders. In this particular case server folder holds the class that calls ANTLR for parsing equation. This is the entry point for server side service, but there are several classes which are dependencies of this service that are automatically solved by GWT Eclipse plugin. In case of client side classes they are also solved by the plugin, and it will compile only the classes that are used by the client side entry point in depth, this entry point is GwtSurferExperiment class.

porst17 commented 10 years ago

Ok. I think I can use the sourceSets approach in gradle as documented here to compile server and client code separately. Still, it seems necessary to find out all the dependent classes. I think that's doable for me.