GCI107 UseTorchFromNumpy #AI #Python #DLG #Build

[cleophass]

Verified that the rule does not exist.
Confirmed that the rule is not listed in Rules.MD, so a new ID GCI107 was created.
Added a corresponding Python unit test.
Updated the CHANGELOG accordingly.

[github-actions]

This PR has been automatically marked as stale because it has no activity for 30 days.
Please add a comment if you want to keep the issue open. Thank you for your contributions!

[dedece35]

delete this class and use UtilsAST already available on "main"

[dedece35]

• use GCI109 rule id in all your PR
• please add an IT
• add a new PR with GCI109 in rules-spec projet

[Copilot]

Pull Request Overview

This PR introduces a new static analysis rule GCI107 that detects inefficient patterns when creating PyTorch tensors from NumPy arrays. The rule enforces using torch.from_numpy() instead of torch.tensor() for better memory efficiency and performance.

• Implements rule GCI107 to encourage memory-efficient tensor creation from NumPy arrays
• Adds comprehensive test coverage for various import patterns and usage scenarios
• Creates utility functions to support AST analysis for the new rule

Reviewed Changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
src/main/java/org/greencodeinitiative/creedengo/python/checks/UseTorchFromNumpy.java	Main rule implementation that detects torch.tensor() calls with NumPy arrays
src/main/java/org/greencodeinitiative/creedengo/python/checks/Utils.java	Utility class providing helper methods for AST analysis
src/test/java/org/greencodeinitiative/creedengo/python/checks/UseTorchFromNumpyTest.java	Unit test class for the new rule
src/test/resources/checks/useTorchFromNumpy.py	Test file with various scenarios for compliant and non-compliant code
src/main/java/org/greencodeinitiative/creedengo/python/PythonRuleRepository.java	Registers the new rule in the repository
CHANGELOG.md	Documents the addition of the new rule

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[Copilot]

The logic is inverted - this condition will return early when it should continue processing. It should use OR (||) instead of AND (&&) to check if NEITHER condition matches torch.tensor.

Suggested change

	if (!TORCH_TENSOR_FUNCTION.equals(Utils.getQualifiedName(callExpression)) && !TORCH_TENSOR_FUNCTION.equals(callExpression.callee().firstToken().value()+"."+callExpression.calleeSymbol().name())) {
	if (!TORCH_TENSOR_FUNCTION.equals(Utils.getQualifiedName(callExpression)) || !TORCH_TENSOR_FUNCTION.equals(callExpression.callee().firstToken().value()+"."+callExpression.calleeSymbol().name())) {

[Copilot]

Potential NullPointerException when callExpression.calleeSymbol() returns null. The code should check for null before calling .name().

Suggested change

	if (!TORCH_TENSOR_FUNCTION.equals(Utils.getQualifiedName(callExpression)) && !TORCH_TENSOR_FUNCTION.equals(callExpression.callee().firstToken().value()+"."+callExpression.calleeSymbol().name())) {
	var calleeSymbol = callExpression.calleeSymbol();
	if (!TORCH_TENSOR_FUNCTION.equals(Utils.getQualifiedName(callExpression)) &&
	!(calleeSymbol != null && TORCH_TENSOR_FUNCTION.equals(callExpression.callee().firstToken().value() + "." + calleeSymbol.name()))) {

[Copilot]

The changelog entry states 'the rule isn't finished yet' which appears inconsistent with a completed implementation. This should be updated to reflect the actual state.

Suggested change

	- [#86](https://github.com/green-code-initiative/creedengo-python/pull/86) Add rule GCI107 Torch from numpy, the rule isn't finished yet
	- [#86](https://github.com/green-code-initiative/creedengo-python/pull/86) Add rule GCI107 Torch from numpy. This rule checks for efficient use of Torch tensors created from NumPy arrays.

[github-actions]

This PR has been automatically marked as stale because it has no activity for 30 days.
Please add a comment if you want to keep the issue open. Thank you for your contributions!