Fix failure to generate runtime types with duplicate schema
The equality contract for TypeDefinition has changed in yangtools-6,
so that TypeDefinitions compare as equal if they represent the same
*semantic* type.
In BindingRuntimeTimes we made the assumption that there is at most
a 1:1 mapping of schema-to-type and used a BiMap to track that
relationship.
Fix that assumption by keeping the maps separate, and teach
RuntimeTypeGenerator to be careful and use an IdentityHashMap to
build that for us.
JIRA: MDSAL-600
Change-Id: Iff5dc7116bf1e5e7c78050d4161282097a3ff033
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>