Spaces:
Sleeping
Sleeping
File size: 23,751 Bytes
a3934b1 7bbd836 a3934b1 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 |
# test_integration_workflows.py
"""
Integration tests for complete verification workflows.
Tests end-to-end workflows including:
- Full verification workflow: select dataset → review message → provide feedback → view results → export CSV
- Session resumption workflow
- Error recovery workflows
"""
import pytest
from datetime import datetime
from src.core.verification_models import (
VerificationSession,
TestMessage,
)
from src.core.verification_store import JSONVerificationStore
from src.core.message_queue_manager import MessageQueueManager
from src.core.verification_feedback_handler import VerificationFeedbackHandler
from src.core.verification_metrics import VerificationMetricsCalculator
from src.core.verification_csv_exporter import VerificationCSVExporter
from src.core.test_datasets import TestDatasetManager
class TestCompleteVerificationWorkflow:
"""Tests for complete verification workflow."""
def test_full_workflow_select_dataset_to_export_csv(
self, temp_storage_dir, test_data_generator, assertion_helpers
):
"""
Test full workflow: select dataset → review message → provide feedback → view results → export CSV
This test verifies the complete end-to-end workflow of the verification mode.
"""
# Step 1: Initialize storage and create session
store = JSONVerificationStore(storage_dir=temp_storage_dir)
# Step 2: Select a dataset (using mixed scenarios for variety)
dataset = TestDatasetManager.MIXED_SCENARIOS_DATASET
assert dataset is not None
assert len(dataset.messages) > 0
# Step 3: Create a verification session
session = test_data_generator.create_verification_session(
session_id="workflow_test_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
# Step 4: Initialize message queue
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
# Step 5: Create feedback handler
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Step 6: Process first 3 messages
messages_to_process = dataset.messages[:3]
for i, message in enumerate(messages_to_process):
# Get current message
current_msg_id = queue_manager.get_current_message_id()
assert current_msg_id == message.message_id
# Provide feedback (alternate between correct and incorrect)
if i % 2 == 0:
# Mark as correct
handler.handle_correct_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test_indicator"],
)
else:
# Mark as incorrect with correction
correction = "red" if message.pre_classified_label != "red" else "green"
handler.handle_incorrect_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test_indicator"],
ground_truth_label=correction,
verifier_notes="Test correction",
)
# Step 7: Verify session statistics
stats = handler.get_session_statistics()
assert stats["verified_count"] == 3
assert stats["correct_count"] == 2 # First and third are correct
assert stats["incorrect_count"] == 1 # Second is incorrect
# Step 8: Export to CSV
csv_content = store.export_to_csv(session.session_id)
# Step 9: Verify CSV content
assertion_helpers.assert_csv_has_summary_section(csv_content)
assertion_helpers.assert_csv_contains_columns(
csv_content,
["Patient Message", "Classifier Said", "You Said", "Notes", "Date"]
)
# Verify CSV has correct number of data rows (3 messages + header + summary)
lines = csv_content.split("\n")
assert len(lines) > 5 # Summary + header + at least 3 data rows
# Verify accuracy in CSV
assert "Accuracy %" in csv_content
assert "66" in csv_content or "67" in csv_content # 2/3 ≈ 66.67%
def test_workflow_with_all_correct_feedback(
self, temp_storage_dir, test_data_generator, assertion_helpers
):
"""Test workflow where all feedback is marked as correct."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.HEALTHY_POSITIVE_DATASET
session = test_data_generator.create_verification_session(
session_id="all_correct_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Mark all messages as correct
for message in dataset.messages[:5]:
handler.handle_correct_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.90,
classifier_indicators=["positive"],
)
# Verify all are correct
stats = handler.get_session_statistics()
assert stats["verified_count"] == 5
assert stats["correct_count"] == 5
assert stats["incorrect_count"] == 0
assert stats["accuracy"] == 100.0
# Export and verify
csv_content = store.export_to_csv(session.session_id)
assert "100.0" in csv_content # 100% accuracy
def test_workflow_with_all_incorrect_feedback(
self, temp_storage_dir, test_data_generator, assertion_helpers
):
"""Test workflow where all feedback is marked as incorrect."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.SUICIDAL_IDEATION_DATASET
session = test_data_generator.create_verification_session(
session_id="all_incorrect_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Mark all messages as incorrect (change red to yellow)
for message in dataset.messages[:5]:
handler.handle_incorrect_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.90,
classifier_indicators=["severe"],
ground_truth_label="yellow", # Wrong correction
verifier_notes="Classifier was wrong",
)
# Verify all are incorrect
stats = handler.get_session_statistics()
assert stats["verified_count"] == 5
assert stats["correct_count"] == 0
assert stats["incorrect_count"] == 5
assert stats["accuracy"] == 0.0
# Export and verify
csv_content = store.export_to_csv(session.session_id)
assert "0.0" in csv_content # 0% accuracy
def test_workflow_with_mixed_classifications(
self, temp_storage_dir, test_data_generator, assertion_helpers
):
"""Test workflow with mixed classification types."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.MIXED_SCENARIOS_DATASET
session = test_data_generator.create_verification_session(
session_id="mixed_class_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Process messages and verify accuracy by type
for message in dataset.messages[:6]:
handler.handle_correct_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
)
stats = handler.get_session_statistics()
# Verify accuracy by type is calculated
assert "accuracy_by_type" in stats
assert "green" in stats["accuracy_by_type"]
assert "yellow" in stats["accuracy_by_type"]
assert "red" in stats["accuracy_by_type"]
class TestSessionResumptionWorkflow:
"""Tests for session resumption workflow."""
def test_resume_session_after_partial_verification(
self, temp_storage_dir, test_data_generator
):
"""Test resuming a session after partial verification."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.ANXIETY_WORRY_DATASET
# Create and partially complete a session
session = test_data_generator.create_verification_session(
session_id="resume_test_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Process first 3 messages
for message in dataset.messages[:3]:
handler.handle_correct_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["anxiety"],
)
# Get stats before closing
stats_before = handler.get_session_statistics()
assert stats_before["verified_count"] == 3
# Simulate closing and reopening the session
loaded_session = store.load_session(session.session_id)
assert loaded_session is not None
assert len(loaded_session.verifications) == 3
# Resume with new queue manager and handler
queue_manager_resumed = MessageQueueManager(loaded_session)
queue_manager_resumed.initialize_queue(dataset.messages)
handler_resumed = VerificationFeedbackHandler(
loaded_session, store, queue_manager_resumed
)
# Verify we can continue from where we left off
stats_after = handler_resumed.get_session_statistics()
assert stats_after["verified_count"] == 3
assert stats_after["correct_count"] == 3
# Process more messages
for message in dataset.messages[3:5]:
handler_resumed.handle_correct_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["anxiety"],
)
# Verify total count increased
stats_final = handler_resumed.get_session_statistics()
assert stats_final["verified_count"] == 5
def test_resume_session_preserves_all_data(
self, temp_storage_dir, test_data_generator, assertion_helpers
):
"""Test that resuming a session preserves all verification data."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.MIXED_SCENARIOS_DATASET
session = test_data_generator.create_verification_session(
session_id="preserve_data_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Create records with specific notes
test_notes = [
"First message note",
"Second message note",
"Third message note",
]
for i, message in enumerate(dataset.messages[:3]):
if i == 0:
handler.handle_correct_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
)
else:
handler.handle_incorrect_feedback(
message=message,
classifier_decision=message.pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
ground_truth_label="green" if message.pre_classified_label != "green" else "red",
verifier_notes=test_notes[i],
)
# Load session and verify data is preserved
loaded_session = store.load_session(session.session_id)
assert len(loaded_session.verifications) == 3
assert loaded_session.verifications[0].is_correct is True
assert loaded_session.verifications[1].verifier_notes == test_notes[1]
assert loaded_session.verifications[2].verifier_notes == test_notes[2]
def test_get_last_session_returns_most_recent(
self, temp_storage_dir, test_data_generator
):
"""Test that get_last_session returns the most recently created session."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
# Create multiple sessions
session1 = test_data_generator.create_verification_session(
session_id="session_1",
verifier_name="Verifier 1",
)
store.save_session(session1)
session2 = test_data_generator.create_verification_session(
session_id="session_2",
verifier_name="Verifier 2",
)
store.save_session(session2)
session3 = test_data_generator.create_verification_session(
session_id="session_3",
verifier_name="Verifier 3",
)
store.save_session(session3)
# Get last session
last_session = store.get_last_session()
# Should be session 3 (most recent)
assert last_session is not None
assert last_session.session_id == "session_3"
class TestErrorRecoveryWorkflows:
"""Tests for error recovery workflows."""
def test_recovery_from_failed_feedback_submission(
self, temp_storage_dir, test_data_generator
):
"""Test recovery when feedback submission fails."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.HEALTHY_POSITIVE_DATASET
session = test_data_generator.create_verification_session(
session_id="error_recovery_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Try to handle feedback with missing correction (should fail)
with pytest.raises(Exception):
handler.handle_incorrect_feedback(
message=dataset.messages[0],
classifier_decision=dataset.messages[0].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
ground_truth_label="", # Missing correction
verifier_notes="",
)
# Verify session is still in valid state
loaded_session = store.load_session(session.session_id)
assert len(loaded_session.verifications) == 0 # No records added
# Should be able to retry with valid data
result = handler.handle_correct_feedback(
message=dataset.messages[0],
classifier_decision=dataset.messages[0].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
)
assert result is True
# Verify record was saved on retry
loaded_session = store.load_session(session.session_id)
assert len(loaded_session.verifications) == 1
def test_recovery_from_csv_export_failure(
self, temp_storage_dir, test_data_generator
):
"""Test recovery when CSV export fails."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
session = test_data_generator.create_verification_session(
session_id="csv_error_001",
total_messages=0,
)
store.save_session(session)
# Try to export with no verified messages (should fail)
# The error message is formatted by the error handler
with pytest.raises((ValueError, RuntimeError)):
store.export_to_csv(session.session_id)
# Add some messages and retry
dataset = TestDatasetManager.HEALTHY_POSITIVE_DATASET
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
handler.handle_correct_feedback(
message=dataset.messages[0],
classifier_decision=dataset.messages[0].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
)
# Now export should succeed
csv_content = store.export_to_csv(session.session_id)
assert csv_content is not None
assert len(csv_content) > 0
def test_recovery_from_session_load_failure(
self, temp_storage_dir, test_data_generator
):
"""Test recovery when session load fails."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
# Try to load non-existent session
loaded_session = store.load_session("non_existent_session")
assert loaded_session is None
# Should be able to create new session
session = test_data_generator.create_verification_session(
session_id="recovery_new_session",
)
store.save_session(session)
# Now load should succeed
loaded_session = store.load_session("recovery_new_session")
assert loaded_session is not None
assert loaded_session.session_id == "recovery_new_session"
def test_recovery_from_invalid_correction_selection(
self, temp_storage_dir, test_data_generator
):
"""Test recovery when invalid correction is selected."""
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.ANXIETY_WORRY_DATASET
session = test_data_generator.create_verification_session(
session_id="invalid_correction_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Try with invalid correction
with pytest.raises(Exception):
handler.handle_incorrect_feedback(
message=dataset.messages[0],
classifier_decision=dataset.messages[0].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
ground_truth_label="invalid_option",
verifier_notes="",
)
# Verify session is still valid
loaded_session = store.load_session(session.session_id)
assert len(loaded_session.verifications) == 0
# Should be able to retry with valid correction
result = handler.handle_incorrect_feedback(
message=dataset.messages[0],
classifier_decision=dataset.messages[0].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
ground_truth_label="red",
verifier_notes="",
)
assert result is True
def test_recovery_from_completed_session_modification_attempt(
self, temp_storage_dir, test_data_generator
):
"""Test recovery when attempting to modify a completed session."""
from src.core.verification_feedback_handler import FeedbackValidationError
store = JSONVerificationStore(storage_dir=temp_storage_dir)
dataset = TestDatasetManager.HEALTHY_POSITIVE_DATASET
session = test_data_generator.create_verification_session(
session_id="completed_session_001",
dataset_id=dataset.dataset_id,
dataset_name=dataset.name,
total_messages=len(dataset.messages),
)
store.save_session(session)
queue_manager = MessageQueueManager(session)
queue_manager.initialize_queue(dataset.messages)
handler = VerificationFeedbackHandler(session, store, queue_manager)
# Add some feedback
handler.handle_correct_feedback(
message=dataset.messages[0],
classifier_decision=dataset.messages[0].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
)
# Mark session as complete
store.mark_session_complete(session.session_id)
# Try to add more feedback (should fail with FeedbackValidationError)
with pytest.raises(FeedbackValidationError, match="Cannot modify completed session"):
handler.handle_correct_feedback(
message=dataset.messages[1],
classifier_decision=dataset.messages[1].pre_classified_label,
classifier_confidence=0.85,
classifier_indicators=["test"],
)
# Verify original feedback is still there
loaded_session = store.load_session(session.session_id)
assert len(loaded_session.verifications) == 1
assert loaded_session.is_complete is True
|