JAV-479 Discovery optimize #261
Conversation
…anged URL cause vertx found wrong failureHandler solution:changed URL only saved in VertxServerRequestToHttpServletRequest, not save to HttpServerRequest
| context.setCurrentNode(parent); | ||
|
|
||
| DiscoveryTreeNode child = filter.discovery(context, parent); | ||
| if (child == null) { |
There was a problem hiding this comment.
If child is null, will generate a NullPointerException, it's the same
There was a problem hiding this comment.
yes, but from null exception, we can not know which filter throw it.
| return data == null; | ||
| } | ||
|
|
||
| private boolean isMapEmpty() { |
There was a problem hiding this comment.
If the data is not a map , we will get the class case exception.
I don't like these three method.
There was a problem hiding this comment.
yes, i don't like this too.
but data in cache is not always the same type.
eg: in discoveryTree, before endpointFilter, discoveryTreeNode data is Map<String, MicroserviceInstance>, after endpointFilter, discoverTreeNode data is Collection<Endpoint> or Collection<CseServer>
caller know what's data type in cache, and call different method, if invoke wrong method, then it's a bug, developers should fix it.
There was a problem hiding this comment.
isXXXEmpty methods have no problems, because they are private, and used by data type.
not so good methods are: mapData/collectionData/arrayData
4 participants
details is in https://servicecomb.atlassian.net/browse/JAV-479
in this issue, create new filter mechanism, old filters will rewrite in new Filter later.