Missing documentation on filtering datasets on document metadata

The following analytics doc page: https://docs.couchbase.com/server/6.0/analytics/5_ddl.html#Creating_and_dropping

on creating and dropping datasets, provides syntax and examples on filtering documents only based on fields in the document body.

We can also filter dataset based on document metadata, although the syntax for filtering on metadata is not that straight-forward.

The following example filters a dataset on document id:

Bucket name: beer-sample

Dataset name: brewery-name

create dataset ` brewery-name` on `beer-sample` where meta(`beer-sample`).id like "21st%";

Note that in N1QL while creating a Global Secondary Index, one does not need to give the bucket name as part of meta() function. However, we need to do so in Analytics only for Couchbase Server version 6.0.0.

From version 6.5.0 and onwards, we will be fixing this syntax issue to just use meta().id instead of meta(`bucket-name`).id . This has been addressed as part of MB-32115 .