feat: group a materialized view by GROUP BY (#4222)

A view could only map source rows one to one, or one to many through a
FROM-position item, so nothing could see all the rows sharing a key.

A view query now takes `GROUP BY expr, ...` with aggregate projections,
planned and executed by DataFusion over the lance scan. A group spans
fragments, so a grouped view is recomputed in full whenever its source
changes; each row's provenance is its group's smallest source row id.
Such a query is stored as definition format 2, so a reader that predates
grouping reports the view as unrefreshable instead of failing to parse
it.


Stacked on #4190.

python/python/lancedb/materialized_view.py

@@ -29,9 +29,10 @@ SelectArg = Union[
 ]
 
 
-DEFINITION_FORMAT = 1
-"""The stored layout this version reads: ``{"format": 1, "query": "<SQL>"}``.
-A ``kind`` key beside it is for readers older than the format number."""
+DEFINITION_FORMAT = 2
+"""The newest stored layout this version reads: ``{"format": N, "query": "<SQL>"}``,
+format 2 being a query with ``GROUP BY``. A ``kind`` key beside it is for
+readers older than the format number."""
 
 
 @dataclass
@@ -40,7 +41,7 @@ class MaterializedViewDefinition:
 
         SELECT columns
         FROM [ns.]table [, function(args) AS alias | , UNNEST(column) AS alias]
-        [WHERE predicate] [LIMIT n]
+        [WHERE predicate] [GROUP BY expr, ...] [LIMIT n]
 
     A Function in ``FROM`` position yields one row per element it returns.
     """

python/python/tests/test_materialized_views.py

@@ -522,7 +522,7 @@ def test_stored_queries_and_legacy_layouts_are_read():
 
     # A newer writer's layout is reported, never guessed at.
     for newer in (
-        {"format": 2, "query": query},
+        {"format": 3, "query": query},
         {"kind": "select_v3", "source_table": "people"},
     ):
         with pytest.raises(NotImplementedError, match="cannot refresh"):