Undefined References

Confiance : high

undefined-referencespython-errorsimport-failuresmodule-loadingcode-qualityruntime-errorsassistant-rhcritical-bugsnameerrorimporterrorattributeerrorproduction-failures

Programming errors where code references classes, functions, or variables that are never defined or properly imported, resulting in NameError, ImportError, or AttributeError exceptions at runtime. Particularly problematic in Python applications where import-time errors can cause complete module loading failures.

Common Patterns

Missing Import Statements

# Error: BaseRetriever referenced but never imported
class SparseRetriever(BaseRetriever):  # NameError at class definition
    pass

# Error: List and Document used in type hints without import
def process_documents(docs: List[Document]) -> List[Document]:  # NameError
    return docs

Incomplete Module Dependencies

# Error: date module referenced but not imported
def today_fr():
    return date.today().strftime("%Y-%m-%d")  # NameError: name 'date' is not defined

Package Structure Issues

# Error: Incorrect import path
from config import settings  # Should be from src.rag.config
from pipeline import RAGPipeline  # Module not in current path

Impact on Production Systems

In the assistant-rh system, undefined references caused:

Complete module loading failures
Immediate runtime crashes on function calls
Broken entry points preventing application startup
Cascade failures across dependent modules

Prevention Strategies

Static Analysis

Use tools like pylint, mypy, or pyflakes to catch undefined references
Implement pre-commit hooks for import validation
Regular code review focusing on import statements

Testing Practices

Import testing in CI/CD pipelines
Module loading validation tests
Smoke tests for all entry points

Development Workflow

IDE configuration for import assistance
Explicit import statements over wildcard imports
Regular refactoring to maintain clean dependencies

Common in RAG Systems

Undefined references are particularly problematic in rag-systems due to:

Complex inter-module dependencies
Dynamic loading of retrieval components
Type annotation requirements for document processing
Integration with multiple AI/ML libraries