feat(exasol): implement approx_nunique, std, var

Author: jcrist

ibis/backends/sql/compilers/exasol.py

@@ -6,7 +6,7 @@
 import ibis.common.exceptions as com
 import ibis.expr.datatypes as dt
 import ibis.expr.operations as ops
-from ibis.backends.sql.compilers.base import NULL, SQLGlotCompiler
+from ibis.backends.sql.compilers.base import NULL, STAR, SQLGlotCompiler
 from ibis.backends.sql.datatypes import ExasolType
 from ibis.backends.sql.dialects import Exasol
 from ibis.backends.sql.rewrites import (
@@ -63,7 +63,6 @@ class ExasolCompiler(SQLGlotCompiler):
         ops.RegexSearch,
         ops.RegexSplit,
         ops.RowID,
-        ops.StandardDev,
         ops.Strftime,
         ops.StringJoin,
         ops.StringSplit,
@@ -77,7 +76,6 @@ class ExasolCompiler(SQLGlotCompiler):
         ops.TimestampSub,
         ops.TypeOf,
         ops.Unnest,
-        ops.Variance,
     )
 
     SIMPLE_OPS = {
@@ -180,6 +178,16 @@ def visit_StringConcat(self, op, *, arg):
         any_args_null = (a.is_(NULL) for a in arg)
         return self.if_(sg.or_(*any_args_null), NULL, self.f.concat(*arg))
 
+    def visit_CountDistinct(self, op, *, arg, where):
+        if where is not None:
+            arg = self.if_(where, arg, NULL)
+        return self.f.count(sge.Distinct(expressions=[arg]))
+
+    def visit_CountStar(self, op, *, arg, where):
+        if where is not None:
+            return self.f.sum(self.cast(where, op.dtype))
+        return self.f.count(STAR)
+
     def visit_CountDistinctStar(self, op, *, arg, where):
         raise com.UnsupportedOperationError(
             "COUNT(DISTINCT *) is not supported in Exasol"

ibis/backends/sql/dialects.py

@@ -76,6 +76,7 @@ class Generator(Postgres.Generator):
         TRANSFORMS = Postgres.Generator.TRANSFORMS.copy() | {
             sge.Interval: _interval,
             sge.GroupConcat: _group_concat,
+            sge.ApproxDistinct: rename_func("approximate_count_distinct"),
         }
         TYPE_MAPPING = Postgres.Generator.TYPE_MAPPING.copy() | {
             sge.DataType.Type.TIMESTAMPTZ: "TIMESTAMP WITH LOCAL TIME ZONE",

ibis/backends/tests/test_aggregation.py

@@ -13,7 +13,6 @@
 from ibis import literal as L
 from ibis.backends.tests.errors import (
     ClickHouseDatabaseError,
-    ExaQueryError,
     GoogleBadRequest,
     ImpalaHiveServer2Error,
     MySQLNotSupportedError,
@@ -298,7 +297,6 @@ def mean_and_std(v):
                     raises=AttributeError,
                     reason="'IntegerColumn' object has no attribute 'notany'",
                 ),
-                pytest.mark.notimpl(["exasol"], raises=ExaQueryError),
             ],
         ),
         param(
@@ -311,7 +309,6 @@ def mean_and_std(v):
                     raises=AttributeError,
                     reason="'IntegerColumn' object has no attribute 'any'",
                 ),
-                pytest.mark.notimpl(["exasol"], raises=ExaQueryError),
             ],
         ),
         param(
@@ -336,7 +333,6 @@ def mean_and_std(v):
                     raises=AttributeError,
                     reason="'IntegerColumn' object has no attribute 'notall'",
                 ),
-                pytest.mark.notimpl(["exasol"], raises=ExaQueryError),
             ],
         ),
         param(
@@ -349,7 +345,6 @@ def mean_and_std(v):
                     raises=AttributeError,
                     reason="'IntegerColumn' object has no attribute 'all'",
                 ),
-                pytest.mark.notimpl(["exasol"], raises=ExaQueryError),
             ],
         ),
         param(
@@ -485,7 +480,7 @@ def mean_and_std(v):
             id="bit_and",
             marks=[
                 pytest.mark.notimpl(
-                    ["polars", "mssql"],
+                    ["polars", "mssql", "exasol"],
                     raises=com.OperationNotDefinedError,
                 ),
                 pytest.mark.notimpl(["druid"], strict=False, raises=AssertionError),
@@ -500,7 +495,7 @@ def mean_and_std(v):
             id="bit_or",
             marks=[
                 pytest.mark.notimpl(
-                    ["polars", "mssql"],
+                    ["polars", "mssql", "exasol"],
                     raises=com.OperationNotDefinedError,
                 ),
                 pytest.mark.notyet(
@@ -514,7 +509,7 @@ def mean_and_std(v):
             id="bit_xor",
             marks=[
                 pytest.mark.notimpl(
-                    ["polars", "mssql"],
+                    ["polars", "mssql", "exasol"],
                     raises=com.OperationNotDefinedError,
                 ),
                 pytest.mark.notyet(
@@ -532,40 +527,12 @@ def mean_and_std(v):
 @pytest.mark.parametrize(
     ("ibis_cond", "pandas_cond"),
     [
-        param(
-            lambda _: None,
-            lambda _: slice(None),
-            marks=pytest.mark.notimpl(
-                ["exasol"],
-                raises=(com.OperationNotDefinedError, ExaQueryError),
-                strict=False,
-            ),
-            id="no_cond",
-        ),
+        param(lambda _: None, lambda _: slice(None), id="no_cond"),
         param(
             lambda t: t.string_col.isin(["1", "7"]),
             lambda t: t.string_col.isin(["1", "7"]),
-            marks=[
-                pytest.mark.notimpl(
-                    ["exasol"],
-                    raises=(com.OperationNotDefinedError, ExaQueryError),
-                    strict=False,
-                ),
-            ],
             id="is_in",
         ),
-        param(
-            lambda _: ibis._.string_col.isin(["1", "7"]),
-            lambda t: t.string_col.isin(["1", "7"]),
-            marks=[
-                pytest.mark.notimpl(
-                    ["exasol"],
-                    raises=(com.OperationNotDefinedError, ExaQueryError),
-                    strict=False,
-                ),
-            ],
-            id="is_in_deferred",
-        ),
     ],
 )
 def test_reduction_ops(