matthieucham commented on pull request #7293:
URL: https://github.com/apache/beam/pull/7293#issuecomment-763758511
Hey guys, sorry to bother you, I don't know if it's the right place but I
have some issues with what as been commited in this PR: It seems to me that
multiple stages AggregationQuery pipelines are not correctly handled, because
of this method:
```java
public MongoCursor<Document> apply(@UnknownKeyFor @NonNull @Initialized
MongoCollection<Document> collection) {
if (this.bucket() != null) {
if (this.mongoDbPipeline().size() == 1) {
this.mongoDbPipeline().add(this.bucket());
} else {
this.mongoDbPipeline().set(this.mongoDbPipeline().size() -
1, this.bucket());
}
}
return collection.aggregate(this.mongoDbPipeline()).iterator();
}
```
The .set(...) looks very odd to me as it replaces the last stage of the
pipeline, I don't get how that could be correct ? (Note that I'm new to MongoDB
so maybe that explains...)
Moreover, I wanted to test and override this by providing my own QueryFn,
but it appears in MongoDBIO implementation that this function will be
force-casted to either FilterQuery or AggregationQuery. Shouldn't we be able to
provide any function as long as the contract is respected ?
Hopefully, you're still reading this thread. Please point me to the right
place if I should post this comment elsewhere
Thanks
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
