duplicate query blocks should increment counter in dataframe name (but should still work on duplicate document)

1. Create a query block (gets named as `query_1`)
2. Duplicate this block (it gets named as `query_1` too, but it should be named as `query_2`)

Note: If the name has been changed, let's say to `my_df`, and then you duplicate it, it should be `query_3`.

We should still keep the original names when duplicating the block while duplicating the entire parent document.

Start by tracing the query-block duplication entry point and the logic that assigns dataframe names. Reproduce the issue steps for an unchanged name and a renamed `my_df`, then verify that duplicating a whole parent document preserves original names while standalone duplication increments the counter.

Written by the indexing model from the issue text.