229 lines
7.9 KiB
Python
229 lines
7.9 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
wikParse Test Framework
|
|
=======================
|
|
Comprehensive testing framework for all wikParse components.
|
|
"""
|
|
|
|
import json
|
|
import os
|
|
import sys
|
|
import tempfile
|
|
import sqlite3
|
|
import pathlib
|
|
from typing import Dict, List, Any, Optional
|
|
|
|
# Add scripts directory to path
|
|
SCRIPT_DIR = pathlib.Path(__file__).parent.parent / "scripts"
|
|
sys.path.insert(0, str(SCRIPT_DIR))
|
|
|
|
from transform_wiktionary import WiktionaryTransformer
|
|
from InflectionProcessor import InflectionProcessor, UniversalInflectionCompressor
|
|
|
|
class TestFramework:
|
|
"""Base test framework with common utilities."""
|
|
|
|
def __init__(self):
|
|
self.test_results = {
|
|
"passed": 0,
|
|
"failed": 0,
|
|
"errors": [],
|
|
"warnings": []
|
|
}
|
|
self.temp_files = []
|
|
|
|
def assert_equal(self, actual, expected, message=""):
|
|
"""Assert that two values are equal."""
|
|
if actual == expected:
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Expected: {expected}"
|
|
error_msg += f"\n Actual: {actual}"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def assert_not_equal(self, actual, expected, message=""):
|
|
"""Assert that two values are not equal."""
|
|
if actual != expected:
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Values should not be equal but both are: {actual}"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def assert_true(self, condition, message=""):
|
|
"""Assert that a condition is true."""
|
|
if condition:
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Condition is False"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def assert_false(self, condition, message=""):
|
|
"""Assert that a condition is false."""
|
|
if not condition:
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Condition is True"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def assert_is_instance(self, obj, cls, message=""):
|
|
"""Assert that an object is an instance of a class."""
|
|
if isinstance(obj, cls):
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Expected type: {cls}"
|
|
error_msg += f"\n Actual type: {type(obj)}"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def assert_in(self, member, container, message=""):
|
|
"""Assert that a member is in a container."""
|
|
if member in container:
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Member not found in container"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def assert_not_in(self, member, container, message=""):
|
|
"""Assert that a member is not in a container."""
|
|
if member not in container:
|
|
self.test_results["passed"] += 1
|
|
return True
|
|
else:
|
|
self.test_results["failed"] += 1
|
|
error_msg = f"Assertion failed: {message}"
|
|
error_msg += f"\n Member found in container but should not be"
|
|
self.test_results["errors"].append(error_msg)
|
|
return False
|
|
|
|
def create_temp_file(self, content="", suffix=".json"):
|
|
"""Create a temporary file and return its path."""
|
|
temp_file = tempfile.NamedTemporaryFile(mode='w', suffix=suffix, delete=False)
|
|
if content:
|
|
temp_file.write(content)
|
|
temp_file.close()
|
|
self.temp_files.append(temp_file.name)
|
|
return temp_file.name
|
|
|
|
def cleanup(self):
|
|
"""Clean up temporary files."""
|
|
for file_path in self.temp_files:
|
|
try:
|
|
os.unlink(file_path)
|
|
except:
|
|
pass
|
|
self.temp_files = []
|
|
|
|
def print_summary(self):
|
|
"""Print test summary."""
|
|
total = self.test_results["passed"] + self.test_results["failed"]
|
|
print("\n" + "="*60)
|
|
print("TEST SUMMARY")
|
|
print("="*60)
|
|
print(f"Total tests: {total}")
|
|
print(f"Passed: {self.test_results['passed']}")
|
|
print(f"Failed: {self.test_results['failed']}")
|
|
|
|
if total > 0:
|
|
success_rate = (self.test_results['passed'] / total) * 100
|
|
print(f"Success rate: {success_rate:.1f}%")
|
|
|
|
if self.test_results['errors']:
|
|
print(f"\nErrors: {len(self.test_results['errors'])}")
|
|
for error in self.test_results['errors']:
|
|
print(f" - {error}")
|
|
|
|
if self.test_results['warnings']:
|
|
print(f"\nWarnings: {len(self.test_results['warnings'])}")
|
|
for warning in self.test_results['warnings']:
|
|
print(f" - {warning}")
|
|
|
|
return self.test_results["failed"] == 0
|
|
|
|
class SchemaValidator:
|
|
"""Schema validation utilities."""
|
|
|
|
@staticmethod
|
|
def validate_universal_schema(entry: Dict[str, Any]) -> bool:
|
|
"""Validate an entry against the universal schema."""
|
|
required_fields = ["word", "pos", "senses"]
|
|
|
|
# Check required fields
|
|
for field in required_fields:
|
|
if field not in entry:
|
|
return False
|
|
|
|
# Check field types
|
|
if not isinstance(entry["word"], str):
|
|
return False
|
|
|
|
if not isinstance(entry["pos"], str):
|
|
return False
|
|
|
|
if not isinstance(entry["senses"], list):
|
|
return False
|
|
|
|
# Validate senses structure
|
|
for sense in entry["senses"]:
|
|
if not isinstance(sense, dict):
|
|
return False
|
|
|
|
return True
|
|
|
|
class TestDataLoader:
|
|
"""Load test data from various sources."""
|
|
|
|
@staticmethod
|
|
def load_sample_data(sample_name: str) -> Dict[str, Any]:
|
|
"""Load sample data from samples directory."""
|
|
samples_dir = pathlib.Path(__file__).parent.parent / "samples"
|
|
|
|
# Try different paths
|
|
possible_paths = [
|
|
samples_dir / "german" / f"{sample_name}.json",
|
|
samples_dir / "french" / f"{sample_name}.json",
|
|
samples_dir / f"{sample_name}.json"
|
|
]
|
|
|
|
for path in possible_paths:
|
|
if path.exists():
|
|
with open(path, 'r', encoding='utf-8') as f:
|
|
return json.load(f)
|
|
|
|
raise FileNotFoundError(f"Sample data not found: {sample_name}")
|
|
|
|
@staticmethod
|
|
def load_jsonl_data(file_path: str) -> List[Dict[str, Any]]:
|
|
"""Load JSONL data from file."""
|
|
entries = []
|
|
with open(file_path, 'r', encoding='utf-8') as f:
|
|
for line in f:
|
|
if line.strip():
|
|
entries.append(json.loads(line.strip()))
|
|
return entries
|
|
|
|
if __name__ == "__main__":
|
|
print("wikParse Test Framework")
|
|
print("Run specific test modules instead of this framework directly.") |