fix: table read query deduplicates identical rows (#2608)

tests/trace/test_table_query.py

@@ -359,3 +359,44 @@ def test_table_query_with_duplicate_row_digests(client: WeaveClient):
     )
 
     assert len(res.rows) == 3
+
+
+def test_duplicate_table_with_identical_rows(client: WeaveClient):
+    data = [{"val": i} for i in range(10)]
+
+    res1 = client.server.table_create(
+        tsi.TableCreateReq(
+            table=tsi.TableSchemaForInsert(
+                rows=data,
+                project_id=client._project_id(),
+            ),
+        )
+    )
+
+    assert len(res1.row_digests) == 10
+
+    # now create the same table with the same data
+    res2 = client.server.table_create(
+        tsi.TableCreateReq(
+            table=tsi.TableSchemaForInsert(
+                rows=data,
+                project_id=client._project_id(),
+            ),
+        )
+    )
+
+    assert len(res2.row_digests) == 10
+
+    # this is the same table!
+    assert res1.digest == res2.digest
+
+    res = client.server.table_query(
+        tsi.TableQueryReq(
+            project_id=client._project_id(),
+            digest=res1.digest,
+            sort_by=[tsi.SortBy(field="val", direction="asc")],
+        )
+    )
+
+    # this is the same table, so we should get the same number of rows
+    assert len(res.rows) == 10

weave/trace_server/table_query_builder.py

@@ -45,6 +45,7 @@ def make_natural_sort_table_query(
             FROM tables
             WHERE project_id = {{{project_id_name}: String}}
             AND digest = {{{digest_name}: String}}
+            LIMIT 1
         )
         ARRAY JOIN row_digests AS row_digest
     ) AS t ON tr.digest = t.row_digest
@@ -98,6 +99,7 @@ def make_standard_table_query(
                 FROM tables
                 WHERE project_id = {{{project_id_name}: String}}
                 AND digest = {{{digest_name}: String}}
+                LIMIT 1
             )
             ARRAY JOIN row_digests AS row_digest
         ) AS t ON tr.digest = t.row_digest