[FEA] Report all unsupported operations for a query in cudf.polars

python/cudf_polars/cudf_polars/dsl/ir.py

@@ -161,6 +161,12 @@ def evaluate(self, *, cache: MutableMapping[int, DataFrame]) -> DataFrame:
         )  # pragma: no cover
 
 
+@dataclasses.dataclass
+class ErrorNode(IR):
+    def evaluate(self, *, cache: MutableMapping[int, DataFrame]) -> DataFrame:
+        return pl.DataFrame()
+
+
 @dataclasses.dataclass
 class PythonScan(IR):
     """Representation of input from a python function."""

python/cudf_polars/cudf_polars/dsl/translate.py

@@ -20,11 +20,24 @@
 
 from cudf_polars.dsl import expr, ir
 from cudf_polars.typing import NodeTraverser
-from cudf_polars.utils import dtypes
+from cudf_polars.utils import dtypes, other
 
 __all__ = ["translate_ir", "translate_named_expr"]
 
 
+def debug(func):
+    def wrapper(*args, **kwargs):
+        try:
+            print(args, kwargs)
+            return func(*args, **kwargs)
+        except NotImplementedError:
+            if other._env_get_bool("CUDF_POLARS_DEBUG_MODE", default=False):
+                return ir.ErrorNode(args[0].get_schema())
+            raise
+
+    return wrapper
+
+
 class set_node(AbstractContextManager[None]):
     """
     Run a block with current node set in the visitor.
@@ -66,11 +79,14 @@ def __exit__(self, *args: Any) -> None:
 def _translate_ir(
     node: Any, visitor: NodeTraverser, schema: dict[str, plc.DataType]
 ) -> ir.IR:
+    if other._env_get_bool("CUDF_POLARS_DEBUG_MODE", default=False):
+        return ir.ErrorNode(schema)
     raise NotImplementedError(
         f"Translation for {type(node).__name__}"
     )  # pragma: no cover
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.PythonScan, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -83,6 +99,7 @@ def _(
     return ir.PythonScan(schema, options, predicate)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Scan, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -120,13 +137,15 @@ def _(
     )
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Cache, visitor: NodeTraverser, schema: dict[str, plc.DataType]
 ) -> ir.IR:
     return ir.Cache(schema, node.id_, translate_ir(visitor, n=node.input))
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.DataFrameScan, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -141,6 +160,7 @@ def _(
     )
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Select, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -151,6 +171,7 @@ def _(
     return ir.Select(schema, inp, exprs, node.should_broadcast)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.GroupBy, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -169,6 +190,7 @@ def _(
     )
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Join, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -185,6 +207,7 @@ def _(
     return ir.Join(schema, inp_left, inp_right, left_on, right_on, node.options)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.HStack, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -195,6 +218,7 @@ def _(
     return ir.HStack(schema, inp, exprs, node.should_broadcast)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Reduce, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -205,6 +229,7 @@ def _(
     return ir.Reduce(schema, inp, exprs)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Distinct, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -216,6 +241,7 @@ def _(
     )
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Sort, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -226,13 +252,15 @@ def _(
     return ir.Sort(schema, inp, by, node.sort_options, node.slice)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Slice, visitor: NodeTraverser, schema: dict[str, plc.DataType]
 ) -> ir.IR:
     return ir.Slice(schema, translate_ir(visitor, n=node.input), node.offset, node.len)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Filter, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -243,6 +271,7 @@ def _(
     return ir.Filter(schema, inp, mask)
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.SimpleProjection,
@@ -252,6 +281,7 @@ def _(
     return ir.Projection(schema, translate_ir(visitor, n=node.input))
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.MapFunction, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -266,6 +296,7 @@ def _(
     )
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.Union, visitor: NodeTraverser, schema: dict[str, plc.DataType]
@@ -275,13 +306,15 @@ def _(
     )
 
 
+@debug
 @_translate_ir.register
 def _(
     node: pl_ir.HConcat, visitor: NodeTraverser, schema: dict[str, plc.DataType]
 ) -> ir.IR:
     return ir.HConcat(schema, [translate_ir(visitor, n=n) for n in node.inputs])
 
 
+@debug
 def translate_ir(visitor: NodeTraverser, *, n: int | None = None) -> ir.IR:
     """
     Translate a polars-internal IR node to our representation.

python/cudf_polars/cudf_polars/utils/other.py

@@ -0,0 +1,30 @@
+# SPDX-FileCopyrightText: Copyright (c) 2024 NVIDIA CORPORATION & AFFILIATES.
+# SPDX-License-Identifier: Apache-2.0
+
+"""Other utilities."""
+
+from __future__ import annotations
+
+import os
+
+
+def _env_get_int(name, default):
+    """Get the integer value of the environment variable."""
+    try:
+        return int(os.getenv(name, default))
+    except (ValueError, TypeError):
+        return default
+
+
+def _env_get_bool(name, default):
+    """Get the the boolean value of the environment variable."""
+    env = os.getenv(name)
+    if env is None:
+        return default
+    as_a_int = _env_get_int(name, None)
+    env = env.lower().strip()
+    if env == "true" or env == "on" or as_a_int:
+        return True
+    if env == "false" or env == "off" or as_a_int == 0:
+        return False
+    return default