From 35ddb010433e73bf5fdd93de277635213ac24739 Mon Sep 17 00:00:00 2001
From: Andreas Kloeckner <inform@tiker.net>
Date: Tue, 27 Aug 2024 17:51:56 -0500
Subject: [PATCH] Numpy actx: warn (not error) on no user-provided transforms
---
arraycontext/impl/numpy/__init__.py | 19 ++++++++++++++++---
1 file changed, 16 insertions(+), 3 deletions(-)
diff --git a/arraycontext/impl/numpy/__init__.py b/arraycontext/impl/numpy/__init__.py
index 7d724b8..77b7b49 100644
--- a/arraycontext/impl/numpy/__init__.py
+++ b/arraycontext/impl/numpy/__init__.py
@@ -43,6 +43,7 @@ from arraycontext.context import (
ArrayOrContainerOrScalar,
ArrayOrContainerOrScalarT,
NumpyOrContainerOrScalar,
+ UntransformedCodeWarning,
)
@@ -116,9 +117,21 @@ class NumpyArrayContext(ArrayContext):
# }}}
def transform_loopy_program(self, t_unit):
- raise ValueError("NumpyArrayContext does not implement "
- "transform_loopy_program. Sub-classes are supposed "
- "to implement it.")
+ from warnings import warn
+ warn("Using the base "
+ f"{type(self).__name__}.transform_loopy_program "
+ "to transform a translation unit. "
+ "This is a no-op and will result in unoptimized C code for"
+ "the requested optimization, all in a single statement."
+ "This will work, but is unlikely to be performant."
+ f"Instead, subclass {type(self).__name__} and implement "
+ "the specific transform logic required to transform the program "
+ "for your package or application. Check higher-level packages "
+ "(e.g. meshmode), which may already have subclasses you may want "
+ "to build on.",
+ UntransformedCodeWarning, stacklevel=2)
+
+ return t_unit
def tag(self,
tags: ToTagSetConvertible,
--
GitLab