Closed cfra closed 3 years ago
Right, this is the intended behavior if somewhat confusing for beginners. It is the the load pipe which persists data to the internal storage, not the whole update entrypoint so pipes after load (you can have multiple load statements) has no magic side-effect on the result of the load.
In order to achieve what you want you need to use the "load ... via" or "load ... cleanup" construction in order for the result of the filtering to get applied to the repository. For instance try this (abbreviated example):
- when onlyips:
- filter:
- "!//md:EntityDescriptor[md:IDPSSODescriptor]"
- when update:
- load:
- "http://www.aai.dfn.de/fileadmin/metadata/dfn-aai-test-metadata.xml verify dfn-aai.pem cleanup onlyidps"
The difference between via
and cleanup
is that cleanup
happens before validation and via
after.
Code Version
1.1.5
Expected Behavior
I have a filter configured in my update pipeline.
I expect pyFF to apply the filter during the loading process, so only entities matching that filter should be listed when browsing metadata or querying via MDQ.
Current Behavior
All entities loaded from the source are shown.
Possible Solution
I have a really hard time grasping how the pipelines work. Some more documentation and examples would be really great.
Steps to Reproduce
Use this pipeline:
Browse pyFF's webinterface.