Spaces:
Sleeping
Sleeping
File size: 16,464 Bytes
680d86c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 | """Tests for inference pipeline."""
import pytest
import tempfile
import os
from unittest.mock import Mock, patch, MagicMock
import pandas as pd
import numpy as np
from src.phising_detection.inference.pipeline import PhishingDetectionPipeline
class TestPhishingDetectionPipeline:
"""Tests for PhishingDetectionPipeline class."""
def test_init_without_urlscan(self):
"""Test initialization without URLScan API key."""
pipeline = PhishingDetectionPipeline(
model_name="test_model",
model_version=1
)
assert pipeline.model_name == "test_model"
assert pipeline.model_version == 1
assert pipeline.model is None
assert pipeline.scaler is None
assert pipeline.feature_names is None
assert pipeline.urlscan_client is None
def test_init_with_urlscan(self):
"""Test initialization with URLScan API key."""
with patch('src.phising_detection.inference.pipeline.URLScanClient') as mock_client:
pipeline = PhishingDetectionPipeline(
model_name="test_model",
urlscan_api_key="test_key"
)
mock_client.assert_called_once_with(api_key="test_key")
assert pipeline.urlscan_client is not None
def test_is_loaded_false(self):
"""Test is_loaded returns False when model not loaded."""
pipeline = PhishingDetectionPipeline()
assert pipeline.is_loaded() is False
def test_is_loaded_true(self):
"""Test is_loaded returns True when model is loaded."""
pipeline = PhishingDetectionPipeline()
pipeline.model = Mock()
pipeline.scaler = Mock()
pipeline.feature_names = ['feature1', 'feature2']
assert pipeline.is_loaded() is True
@patch('src.phising_detection.inference.pipeline.connect_to_hopsworks')
@patch('src.phising_detection.inference.pipeline.joblib.load')
def test_load_model_from_hopsworks(self, mock_joblib_load, mock_connect):
"""Test loading model from Hopsworks."""
# Setup mocks
mock_project = Mock()
mock_mr = Mock()
mock_model_registry = Mock()
mock_model_registry.version = 1
mock_model_registry.download.return_value = "/tmp/model_dir"
mock_project.get_model_registry.return_value = mock_mr
mock_mr.get_model.return_value = mock_model_registry
mock_connect.return_value = mock_project
# Mock model and scaler
mock_model = Mock()
mock_scaler = Mock()
mock_joblib_load.side_effect = [mock_model, mock_scaler]
# Create temporary feature_names file
with tempfile.TemporaryDirectory() as tmpdir:
feature_names_path = os.path.join(tmpdir, "feature_names.txt")
with open(feature_names_path, 'w') as f:
f.write("feature1\nfeature2\nfeature3")
# Mock download to return our temp dir
mock_model_registry.download.return_value = tmpdir
# Test
pipeline = PhishingDetectionPipeline(model_name="test_model")
pipeline.load_model_from_hopsworks()
# Assertions
assert pipeline.model == mock_model
assert pipeline.scaler == mock_scaler
assert pipeline.feature_names == ['feature1', 'feature2', 'feature3']
mock_connect.assert_called_once()
mock_mr.get_model.assert_called_once_with("test_model")
@patch('src.phising_detection.inference.pipeline.connect_to_hopsworks')
@patch('src.phising_detection.inference.pipeline.joblib.load')
def test_load_model_with_version(self, mock_joblib_load, mock_connect):
"""Test loading specific model version from Hopsworks."""
# Setup mocks
mock_project = Mock()
mock_mr = Mock()
mock_model_registry = Mock()
mock_model_registry.version = 2
mock_model_registry.download.return_value = "/tmp/model_dir"
mock_project.get_model_registry.return_value = mock_mr
mock_mr.get_model.return_value = mock_model_registry
mock_connect.return_value = mock_project
mock_joblib_load.side_effect = [Mock(), Mock()]
with tempfile.TemporaryDirectory() as tmpdir:
feature_names_path = os.path.join(tmpdir, "feature_names.txt")
with open(feature_names_path, 'w') as f:
f.write("feature1")
mock_model_registry.download.return_value = tmpdir
# Test with specific version
pipeline = PhishingDetectionPipeline(model_name="test_model", model_version=2)
pipeline.load_model_from_hopsworks()
# Should request version 2
mock_mr.get_model.assert_called_once_with("test_model", version=2)
def test_preprocess_features_not_loaded(self):
"""Test preprocessing fails when model not loaded."""
pipeline = PhishingDetectionPipeline()
features = {'feature1': 10, 'feature2': 20}
with pytest.raises(ValueError) as exc_info:
pipeline.preprocess_features(features)
assert "Model not loaded" in str(exc_info.value)
def test_preprocess_features_success(self):
"""Test successful feature preprocessing."""
# Setup pipeline with mock components
pipeline = PhishingDetectionPipeline()
pipeline.model = Mock()
pipeline.feature_names = [
'domain_age_days',
'secure_percentage',
'has_umbrella_rank',
'umbrella_rank',
'has_tls',
'tls_valid_days',
'url_length',
'subdomain_count'
]
# Mock scaler
mock_scaler = Mock()
mock_scaler.transform.return_value = np.array([[1.5, 0.8, 5000, 365, 25, 1]])
pipeline.scaler = mock_scaler
# Test features
features = {
'domain_age_days': 3000,
'secure_percentage': 95.0,
'has_umbrella_rank': 1,
'umbrella_rank': 5000,
'has_tls': 1,
'tls_valid_days': 365,
'url_length': 25,
'subdomain_count': 1
}
result = pipeline.preprocess_features(features)
# Assertions
assert isinstance(result, pd.DataFrame)
assert list(result.columns) == pipeline.feature_names
assert len(result) == 1
mock_scaler.transform.assert_called_once()
def test_preprocess_features_with_missing_features(self):
"""Test preprocessing handles missing features."""
pipeline = PhishingDetectionPipeline()
pipeline.model = Mock()
pipeline.feature_names = [
'domain_age_days', 'secure_percentage', 'has_umbrella_rank',
'umbrella_rank', 'has_tls', 'tls_valid_days', 'url_length', 'subdomain_count'
]
mock_scaler = Mock()
# Mock the scaler to return the same shape as input continuous features (6 features)
mock_scaler.transform.return_value = np.array([[1.0, 2.0, 3.0, 4.0, 5.0, 6.0]])
pipeline.scaler = mock_scaler
# Only provide some features (missing subdomain_count)
features = {
'domain_age_days': 3000,
'secure_percentage': 95.0,
'has_umbrella_rank': 1,
'umbrella_rank': 5000,
'has_tls': 1,
'tls_valid_days': 365,
'url_length': 25
}
result = pipeline.preprocess_features(features)
# Should add missing subdomain_count and all features should be present
assert 'subdomain_count' in result.columns
assert len(result.columns) == 8 # All 8 features should be present
assert list(result.columns) == pipeline.feature_names
def test_predict_not_loaded(self):
"""Test prediction fails when model not loaded."""
pipeline = PhishingDetectionPipeline()
features = {'feature1': 10}
with pytest.raises(ValueError) as exc_info:
pipeline.predict(features)
assert "Model not loaded" in str(exc_info.value)
def test_predict_phishing(self):
"""Test prediction for phishing URL."""
# Setup pipeline
pipeline = PhishingDetectionPipeline()
pipeline.feature_names = [
'domain_age_days', 'secure_percentage', 'has_umbrella_rank',
'umbrella_rank', 'has_tls', 'tls_valid_days', 'url_length', 'subdomain_count'
]
# Mock model
mock_model = Mock()
mock_model.predict_proba.return_value = np.array([[0.2, 0.8]]) # 80% phishing
mock_model.predict.return_value = np.array([1]) # Phishing
pipeline.model = mock_model
# Mock scaler
mock_scaler = Mock()
mock_scaler.transform.return_value = np.array([[1.0, 2.0, 3.0, 4.0, 5.0, 6.0]])
pipeline.scaler = mock_scaler
# Test features
features = {
'domain_age_days': 10,
'secure_percentage': 50.0,
'has_umbrella_rank': 0,
'umbrella_rank': 999999,
'has_tls': 0,
'tls_valid_days': 0,
'url_length': 150,
'subdomain_count': 5
}
result = pipeline.predict(features)
# Assertions
assert result['prediction'] == "PHISHING"
assert result['is_phishing'] is True
assert result['confidence'] == 0.8
assert result['phishing_probability'] == 0.8
assert result['legitimate_probability'] == 0.2
def test_predict_legitimate(self):
"""Test prediction for legitimate URL."""
# Setup pipeline
pipeline = PhishingDetectionPipeline()
pipeline.feature_names = [
'domain_age_days', 'secure_percentage', 'has_umbrella_rank',
'umbrella_rank', 'has_tls', 'tls_valid_days', 'url_length', 'subdomain_count'
]
# Mock model
mock_model = Mock()
mock_model.predict_proba.return_value = np.array([[0.9, 0.1]]) # 90% legitimate
mock_model.predict.return_value = np.array([0]) # Legitimate
pipeline.model = mock_model
# Mock scaler
mock_scaler = Mock()
mock_scaler.transform.return_value = np.array([[1.0, 2.0, 3.0, 4.0, 5.0, 6.0]])
pipeline.scaler = mock_scaler
# Test features
features = {
'domain_age_days': 3000,
'secure_percentage': 95.0,
'has_umbrella_rank': 1,
'umbrella_rank': 5000,
'has_tls': 1,
'tls_valid_days': 365,
'url_length': 25,
'subdomain_count': 1
}
result = pipeline.predict(features)
# Assertions
assert result['prediction'] == "LEGITIMATE"
assert result['is_phishing'] is False
assert result['confidence'] == 0.9
assert result['phishing_probability'] == 0.1
assert result['legitimate_probability'] == 0.9
def test_predict_url_without_urlscan(self):
"""Test predict_url fails without URLScan client."""
pipeline = PhishingDetectionPipeline()
pipeline.model = Mock()
pipeline.scaler = Mock()
pipeline.feature_names = ['feature1']
with pytest.raises(ValueError) as exc_info:
pipeline.predict_url("https://example.com")
assert "URLScan client not initialized" in str(exc_info.value)
@patch('src.phising_detection.inference.pipeline.extract_features')
def test_predict_url_success(self, mock_extract_features):
"""Test successful end-to-end URL prediction."""
# Setup pipeline
pipeline = PhishingDetectionPipeline()
pipeline.feature_names = [
'domain_age_days', 'secure_percentage', 'has_umbrella_rank',
'umbrella_rank', 'has_tls', 'tls_valid_days', 'url_length', 'subdomain_count'
]
# Mock model
mock_model = Mock()
mock_model.predict_proba.return_value = np.array([[0.7, 0.3]])
mock_model.predict.return_value = np.array([0])
pipeline.model = mock_model
# Mock scaler
mock_scaler = Mock()
mock_scaler.transform.return_value = np.array([[1.0, 2.0, 3.0, 4.0, 5.0, 6.0]])
pipeline.scaler = mock_scaler
# Mock URLScan client
mock_urlscan_client = Mock()
mock_scan_result = {
'task': {'uuid': 'test-uuid-123'},
'page': {'domainAgeDays': 3000},
'stats': {'securePercentage': 95}
}
mock_urlscan_client.submit_and_wait.return_value = mock_scan_result
pipeline.urlscan_client = mock_urlscan_client
# Mock extracted features
extracted_features = {
'domain_age_days': 3000,
'secure_percentage': 95.0,
'has_umbrella_rank': 1,
'umbrella_rank': 5000,
'has_tls': 1,
'tls_valid_days': 365,
'url_length': 25,
'subdomain_count': 1
}
mock_extract_features.return_value = extracted_features
# Test
result = pipeline.predict_url("https://example.com")
# Assertions
assert result['prediction'] == "LEGITIMATE"
assert result['confidence'] == 0.7
assert result['features'] == extracted_features
assert result['scan_uuid'] == 'test-uuid-123'
mock_urlscan_client.submit_and_wait.assert_called_once_with("https://example.com")
def test_predict_url_scan_fails(self):
"""Test predict_url handles URLScan failure."""
# Setup pipeline
pipeline = PhishingDetectionPipeline()
pipeline.model = Mock()
pipeline.scaler = Mock()
pipeline.feature_names = ['feature1']
# Mock URLScan client that returns None
mock_urlscan_client = Mock()
mock_urlscan_client.submit_and_wait.return_value = None
pipeline.urlscan_client = mock_urlscan_client
# Test
result = pipeline.predict_url("https://example.com")
# Should return error
assert 'error' in result
assert result['prediction'] == "ERROR"
assert result['confidence'] == 0.0
@patch('src.phising_detection.inference.pipeline.extract_features')
def test_predict_url_exception_handling(self, mock_extract_features):
"""Test predict_url handles exceptions gracefully."""
# Setup pipeline
pipeline = PhishingDetectionPipeline()
pipeline.model = Mock()
pipeline.scaler = Mock()
pipeline.feature_names = ['feature1']
# Mock URLScan client
mock_urlscan_client = Mock()
mock_urlscan_client.submit_and_wait.return_value = {'task': {}}
pipeline.urlscan_client = mock_urlscan_client
# Mock extract_features to raise exception
mock_extract_features.side_effect = Exception("Test error")
# Test
result = pipeline.predict_url("https://example.com")
# Should return error
assert 'error' in result
assert result['prediction'] == "ERROR"
assert "Test error" in result['error']
def test_predict_numerical_stability(self):
"""Test prediction handles edge cases in probabilities."""
# Setup pipeline
pipeline = PhishingDetectionPipeline()
pipeline.feature_names = [
'domain_age_days', 'secure_percentage', 'has_umbrella_rank',
'umbrella_rank', 'has_tls', 'tls_valid_days', 'url_length', 'subdomain_count'
]
# Mock model with extreme probabilities
mock_model = Mock()
mock_model.predict_proba.return_value = np.array([[0.999999, 0.000001]])
mock_model.predict.return_value = np.array([0])
pipeline.model = mock_model
mock_scaler = Mock()
mock_scaler.transform.return_value = np.array([[1.0, 2.0, 3.0, 4.0, 5.0, 6.0]])
pipeline.scaler = mock_scaler
features = {
'domain_age_days': 3000,
'secure_percentage': 95.0,
'has_umbrella_rank': 1,
'umbrella_rank': 5000,
'has_tls': 1,
'tls_valid_days': 365,
'url_length': 25,
'subdomain_count': 1
}
result = pipeline.predict(features)
# Should handle extreme values correctly
assert result['confidence'] > 0.99
assert result['prediction'] == "LEGITIMATE"
assert result['phishing_probability'] < 0.01
|