Re: https://github.com/fusesource/fuseide/issues/151

I’m thinking it would be a good idea to break our wizard up into a number of pages.

Page 1:

• Provides the following inputs:
• camel context path
• transformation file name 
• transformation id
• source and target type (no path to definitions though)
• Finish enabled (next disabled) if transformation is Java to Java
• Layout tweaked to present source type and target type dropdowns in the same row
• .e.g.   [Source Type]  —>   [Target Type]
• Add “Other” as an option to the types

Page 2:

• This page configures the source type.  If source type is Java, this page is skipped.
• Two options on this page (mutually exclusive):
• If the type is XML, XSD, JSON, or JSON Schema, provide browser for file as we do today
• If the type is “Other”, the user needs to provide two pieces of data:
• select from a drop-down of existing data format definitions (this can be queried from the config builder)
• specify the source class used for the mapping
• If target type != Java, then Next is enabled and Finish is disabled.  If target type == Java, reverse that.

Page 3:

• This page configures the target type. 
• Two options on this page (mutually exclusive):
• If the type is XML, XSD, JSON, or JSON Schema, provide browser for file as we do today
• If the type is “Other”, the user needs to provide two pieces of data:
• select from a drop-down of existing data format definitions (this can be queried from the config builder)
• specify the target class used for the mapping
• Finish is enabled once the target configuration is specified.