[Feature] Reuse weaver.url to define process locations

[fmigneault]

Context

Until at least weaver 1.8, processes are stored with the 3 displayed fields below, directly using the weaver.url value. While this is acceptable and works for base use case, it causes problems if the server configuration that uses weaver is modified for any reason as all these locations become invalid.
Processes deployed before the server reconfiguration would all point to invalid locations, meaning they cannot be correctly executed nor report the correct references in the response bodies.

For example, starting two server configurations from scratch that specify different weaver.url setting, the results are are follows after deploying some process. This is not the case of builtin processes which are redeployed on each weaver startup.

Todo

Since weaver does not require to have the full URL in those fields, corresponding relative 'path' should be stored instead of the full URL. The Process properties should me modified to provide a getter that combines weaver.url with the corresponding 'path' stored in db.

weaver/weaver/datatype.py

Lines 673 to 686 in 5e2aa56

	@property
	def processDescriptionURL(self): # noqa: N802
	# type: () -> Optional[AnyStr]
	return self.get("processDescriptionURL")
	@property
	def processEndpointWPS1(self): # noqa: N802
	# type: () -> Optional[AnyStr]
	return self.get("processEndpointWPS1")
	@property
	def executeEndpoint(self): # noqa: N802
	# type: () -> Optional[AnyStr]
	return self.get("executeEndpoint")

Since these path can also refer to a remote provider locations, the paths must not be assumed to only start with the /processes.