Spaces:
Running
Running
File size: 55,764 Bytes
c2ea5ed |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 |
"""
Utility functions for database operations.
"""
import os
import json
import logging
import uuid
from datetime import datetime
from typing import Dict, List, Any, Optional, Union
import hashlib
from sqlalchemy.orm import Session
from sqlalchemy import func
from . import models
from . import get_db, init_db
logger = logging.getLogger(__name__)
def initialize_database(clear_all=False):
"""
Initialize the database and create tables.
Args:
clear_all: If True, drops all existing tables before creating new ones
"""
if clear_all:
from . import reinit_db
reinit_db()
logger.info("Database reinitialized (all previous data cleared)")
else:
from . import init_db
init_db()
logger.info("Database initialized (existing data preserved)")
def get_knowledge_graph(session: Session, filename: str) -> Optional[models.KnowledgeGraph]:
"""Get a knowledge graph by filename."""
return session.query(models.KnowledgeGraph).filter_by(filename=filename).first()
def save_knowledge_graph(
session,
filename,
graph_data,
trace_id=None,
window_index=None,
window_total=None,
window_start_char=None,
window_end_char=None,
is_original=False,
processing_run_id=None
):
"""
Save a knowledge graph to the database.
Args:
session: Database session
filename: Filename to save under
graph_data: Knowledge graph data
trace_id: Optional ID to group knowledge graphs from the same trace
window_index: Optional sequential index of window within a trace
window_total: Optional total number of windows in the trace
window_start_char: Optional starting character position in the original trace
window_end_char: Optional ending character position in the original trace
is_original: Whether this is an original knowledge graph (sets status to "created")
processing_run_id: Optional ID to distinguish multiple processing runs
Returns:
The created KnowledgeGraph object
"""
from backend.database.models import KnowledgeGraph
# Check if the knowledge graph already exists
kg = session.query(KnowledgeGraph).filter(KnowledgeGraph.filename == filename).first()
if kg:
# Update the existing knowledge graph using graph_content to ensure counts are updated
kg.graph_content = graph_data
kg.update_timestamp = datetime.utcnow()
# Update trace information if provided
if trace_id is not None:
kg.trace_id = trace_id
if window_index is not None:
kg.window_index = window_index
if window_total is not None:
kg.window_total = window_total
if window_start_char is not None:
kg.window_start_char = window_start_char
if window_end_char is not None:
kg.window_end_char = window_end_char
if processing_run_id is not None:
kg.processing_run_id = processing_run_id
# Set status if is_original is True
if is_original:
kg.status = "created"
session.add(kg)
session.commit()
return kg
else:
# Create a new knowledge graph
kg = KnowledgeGraph(
filename=filename,
trace_id=trace_id,
window_index=window_index,
window_total=window_total,
window_start_char=window_start_char,
window_end_char=window_end_char,
status="created" if is_original else None,
processing_run_id=processing_run_id
)
# Set graph content after creation to ensure counts are updated
kg.graph_content = graph_data
session.add(kg)
session.commit()
return kg
def update_knowledge_graph_status(session: Session, kg_id: Union[int, str], status: str) -> models.KnowledgeGraph:
"""
Update the status of a knowledge graph.
Args:
session: Database session
kg_id: Knowledge graph ID or filename
status: New status (created, enriched, perturbed, causal)
Returns:
Updated knowledge graph
"""
# Check if kg_id is a filename or an ID
if isinstance(kg_id, str):
kg = session.query(models.KnowledgeGraph).filter_by(filename=kg_id).first()
else:
kg = session.query(models.KnowledgeGraph).filter_by(id=kg_id).first()
if not kg:
raise ValueError(f"Knowledge graph with ID/filename {kg_id} not found")
# Update status
kg.status = status
session.commit()
return kg
def extract_entities_and_relations(session: Session, kg: models.KnowledgeGraph):
"""Extract entities and relations from a knowledge graph and save them to the database."""
# Get the graph data
data = kg.graph_data
# Skip if no data
if not data:
return
# First, delete existing relations and entities for this knowledge graph
# We need to delete relations first due to foreign key constraints
session.query(models.Relation).filter_by(graph_id=kg.id).delete()
session.query(models.Entity).filter_by(graph_id=kg.id).delete()
session.flush()
# Process entities
entity_map = {} # Map entity_id to Entity instance
for entity_data in data.get('entities', []):
try:
# Skip if no id
if 'id' not in entity_data:
continue
entity_id = entity_data.get('id')
# Create entity
entity = models.Entity.from_dict(entity_data, kg.id)
# Add to session
session.add(entity)
session.flush() # Flush to get the ID
# Add to map
entity_map[entity_id] = entity
except Exception as e:
logger.error(f"Error extracting entity {entity_data.get('id')}: {str(e)}")
# Process relations
for relation_data in data.get('relations', []):
try:
# Skip if no id, source, or target
if 'id' not in relation_data or 'source' not in relation_data or 'target' not in relation_data:
continue
source_id = relation_data.get('source')
target_id = relation_data.get('target')
# Get source and target entities
source_entity = entity_map.get(source_id)
target_entity = entity_map.get(target_id)
# Skip if source or target entity not found
if not source_entity or not target_entity:
logger.warning(f"Skipping relation {relation_data.get('id')}: Source or target entity not found")
continue
# Create relation
relation = models.Relation.from_dict(
relation_data,
kg.id,
source_entity,
target_entity
)
# Add to session
session.add(relation)
except Exception as e:
logger.error(f"Error extracting relation {relation_data.get('id')}: {str(e)}")
# Commit the changes
session.commit()
def get_test_result(session: Session, filename: str) -> Optional[Dict[str, Any]]:
"""
Get a test result by filename from the knowledge graph.
This now returns a dictionary with test result data instead of a TestResult model.
"""
# Try to find a knowledge graph with this test result filename
kg = session.query(models.KnowledgeGraph).filter_by(filename=filename).first()
if kg and kg.content:
try:
data = json.loads(kg.content)
if 'test_result' in data:
return data['test_result']
except json.JSONDecodeError:
pass
# Try standard file locations
standard_file_locations = [
f"datasets/test_results/{filename}",
f"datasets/{filename}"
]
for file_path in standard_file_locations:
try:
with open(file_path, 'r') as f:
return json.load(f)
except (FileNotFoundError, json.JSONDecodeError):
continue
return None
def save_test_result(session: Session, filename: str, data: Dict[str, Any]) -> models.KnowledgeGraph:
"""
Save a test result to the database.
Test results are now stored within the KnowledgeGraph content field
rather than as a separate TestResult model.
"""
# Find or create a knowledge graph for this test result
kg = session.query(models.KnowledgeGraph).filter_by(filename=filename).first()
if not kg:
# Create new knowledge graph for this test result
kg = models.KnowledgeGraph()
kg.filename = filename
kg.creation_timestamp = datetime.utcnow()
# Get existing content or initialize empty dict
try:
if kg.content:
content = json.loads(kg.content)
else:
content = {}
except json.JSONDecodeError:
content = {}
# Update test result data
content['test_result'] = data
content['test_timestamp'] = datetime.utcnow().isoformat()
content['model_name'] = data.get('model', '')
content['perturbation_type'] = data.get('perturbation_type', '')
content['completed'] = data.get('completed', False)
# Find the related knowledge graph if referenced
kg_filename = data.get('knowledge_graph_filename')
if kg_filename:
related_kg = session.query(models.KnowledgeGraph).filter_by(filename=kg_filename).first()
if related_kg:
content['knowledge_graph_id'] = related_kg.id
# Save updated content back to knowledge graph
kg.content = json.dumps(content)
# Save to database
session.add(kg)
session.commit()
return kg
def get_test_progress(session: Session, test_filename: str) -> Optional[Dict[str, Any]]:
"""
Get test progress by test filename.
Now returns progress data as a dictionary instead of a TestProgress model.
"""
kg = session.query(models.KnowledgeGraph).filter_by(filename=test_filename).first()
if kg and kg.content:
try:
content = json.loads(kg.content)
if 'test_progress' in content:
return content['test_progress']
except json.JSONDecodeError:
pass
# Try to find a progress file
progress_filename = f"progress_{test_filename}"
progress_path = str(PROJECT_ROOT / 'datasets' / 'test_results' / progress_filename)
if os.path.exists(progress_path):
try:
with open(progress_path, 'r') as f:
return json.load(f)
except (FileNotFoundError, json.JSONDecodeError):
pass
return None
def save_test_progress(session: Session, test_filename: str, data: Dict[str, Any]) -> models.KnowledgeGraph:
"""
Save test progress to the database.
Test progress is now stored within the KnowledgeGraph content field
rather than as a separate TestProgress model.
"""
# Find the knowledge graph for this test
kg = session.query(models.KnowledgeGraph).filter_by(filename=test_filename).first()
if not kg:
# Create new knowledge graph for this test
kg = models.KnowledgeGraph()
kg.filename = test_filename
kg.creation_timestamp = datetime.utcnow()
# Get existing content or initialize empty dict
try:
if kg.content:
content = json.loads(kg.content)
else:
content = {}
except json.JSONDecodeError:
content = {}
# Initialize test_progress if it doesn't exist
if 'test_progress' not in content:
content['test_progress'] = {}
# Update progress data
if 'progress' in data:
progress_data = data['progress']
content['test_progress']['status'] = progress_data.get('status', content['test_progress'].get('status'))
content['test_progress']['current'] = progress_data.get('current', content['test_progress'].get('current'))
content['test_progress']['total'] = progress_data.get('total', content['test_progress'].get('total'))
content['test_progress']['last_tested_relation'] = progress_data.get('last_tested_relation', content['test_progress'].get('last_tested_relation'))
content['test_progress']['overall_progress_percentage'] = progress_data.get('overall_progress_percentage', content['test_progress'].get('overall_progress_percentage'))
content['test_progress']['current_jailbreak'] = progress_data.get('current_jailbreak', content['test_progress'].get('current_jailbreak'))
else:
# Direct update of progress data
for key, value in data.items():
content['test_progress'][key] = value
if 'timestamp' in data:
try:
content['test_progress']['timestamp'] = data['timestamp']
except (ValueError, TypeError):
content['test_progress']['timestamp'] = datetime.utcnow().isoformat()
else:
content['test_progress']['timestamp'] = datetime.utcnow().isoformat()
# Save updated content back to knowledge graph
kg.content = json.dumps(content)
# Save to database
session.add(kg)
session.commit()
# Also save to progress file for backward compatibility
try:
progress_filename = f"progress_{test_filename}"
progress_dir = 'datasets/test_results'
os.makedirs(progress_dir, exist_ok=True)
progress_path = os.path.join(progress_dir, progress_filename)
with open(progress_path, 'w') as f:
json.dump(content['test_progress'], f)
except Exception as e:
logger.warning(f"Failed to save progress file: {str(e)}")
return kg
def get_all_knowledge_graphs(session: Session) -> List[models.KnowledgeGraph]:
"""Get all knowledge graphs."""
return session.query(models.KnowledgeGraph).all()
def get_all_test_results(session: Session) -> List[Dict[str, Any]]:
"""
Get all test results.
Now returns a list of dictionaries containing test result data
extracted from knowledge graphs.
"""
test_results = []
# Get all knowledge graphs that may contain test results
knowledge_graphs = session.query(models.KnowledgeGraph).all()
for kg in knowledge_graphs:
if kg.content:
try:
content = json.loads(kg.content)
if 'test_result' in content:
# Add filename and ID for reference
result = content['test_result'].copy() if isinstance(content['test_result'], dict) else {}
result['filename'] = kg.filename
result['id'] = kg.id
test_results.append(result)
except json.JSONDecodeError:
continue
return test_results
def get_standard_dataset(session: Session, filename: str) -> Optional[Dict[str, Any]]:
"""
Get a standard dataset by filename (e.g., jailbreak techniques).
First attempts to load from the database as a knowledge graph,
then falls back to standard data file locations.
"""
# Try to get from database as a knowledge graph
kg = session.query(models.KnowledgeGraph).filter_by(filename=filename).first()
if kg and kg.content:
try:
return json.loads(kg.content)
except json.JSONDecodeError:
pass
# If not in database, try standard file locations
standard_file_locations = [
f"datasets/{filename}", # Direct in data dir
f"datasets/test_results/{filename}",
f"datasets/knowledge_graphs/{filename}"
]
for file_path in standard_file_locations:
try:
with open(file_path, 'r') as f:
return json.load(f)
except (FileNotFoundError, json.JSONDecodeError):
continue
# Finally try as an absolute path
try:
with open(filename, 'r') as f:
return json.load(f)
except (FileNotFoundError, json.JSONDecodeError):
pass
return None
def find_entity_by_id(session: Session, entity_id: str) -> Optional[models.Entity]:
"""
Find an entity by its ID.
Args:
session: Database session
entity_id: Entity ID to search for
Returns:
Entity or None if not found
"""
query = session.query(models.Entity).filter_by(entity_id=entity_id)
return query.first()
def find_relation_by_id(session: Session, relation_id: str) -> Optional[models.Relation]:
"""
Find a relation by its ID.
Args:
session: Database session
relation_id: Relation ID to search for
Returns:
Relation or None if not found
"""
query = session.query(models.Relation).filter_by(relation_id=relation_id)
return query.first()
def find_entities_by_type(session: Session, entity_type: str) -> List[models.Entity]:
"""
Find entities by type.
Args:
session: Database session
entity_type: Entity type to search for
Returns:
List of entities
"""
query = session.query(models.Entity).filter_by(type=entity_type)
return query.all()
def find_relations_by_type(session: Session, relation_type: str) -> List[models.Relation]:
"""
Find relations by type.
Args:
session: Database session
relation_type: Relation type to search for
Returns:
List of relations
"""
query = session.query(models.Relation).filter_by(type=relation_type)
return query.all()
def merge_knowledge_graphs(session: Session, output_filename: str, input_filenames: List[str]) -> Optional[models.KnowledgeGraph]:
"""
Merge multiple knowledge graphs into a single knowledge graph.
Args:
session: Database session
output_filename: Output filename for the merged knowledge graph
input_filenames: List of filenames of knowledge graphs to merge
Returns:
The merged KnowledgeGraph instance or None if error
"""
# Check if merged graph already exists
existing_kg = get_knowledge_graph(session, output_filename)
if existing_kg:
logger.warning(f"Knowledge graph {output_filename} already exists. Returning existing graph.")
return existing_kg
# Load all input knowledge graphs
knowledge_graphs = []
for filename in input_filenames:
kg = get_knowledge_graph(session, filename)
if not kg:
logger.warning(f"Knowledge graph {filename} not found. Skipping.")
continue
knowledge_graphs.append(kg)
if not knowledge_graphs:
logger.error("No valid knowledge graphs to merge.")
return None
# Create a new merged knowledge graph
merged_data = {
"entities": [],
"relations": [],
"metadata": {
"source_graphs": input_filenames,
"creation_time": datetime.datetime.utcnow().isoformat(),
"merge_method": "concatenate"
}
}
# Keep track of entity and relation IDs to avoid duplicates
entity_ids = set()
relation_ids = set()
# Add entities and relations from each graph
for kg in knowledge_graphs:
graph_data = kg.graph_data
if not graph_data:
logger.warning(f"Knowledge graph {kg.filename} has no data. Skipping.")
continue
# Process entities
for entity in graph_data.get("entities", []):
# Skip if no ID
if "id" not in entity:
continue
# Skip if ID already exists
if entity["id"] in entity_ids:
continue
# Add to merged data
merged_data["entities"].append(entity)
entity_ids.add(entity["id"])
# Process relations
for relation in graph_data.get("relations", []):
# Skip if no ID, source, or target
if "id" not in relation or "source" not in relation or "target" not in relation:
continue
# Skip if ID already exists
if relation["id"] in relation_ids:
continue
# Add to merged data
merged_data["relations"].append(relation)
relation_ids.add(relation["id"])
# Save the merged knowledge graph
return save_knowledge_graph(session, output_filename, merged_data)
def get_knowledge_graph_by_id(session, graph_id):
"""
Get a knowledge graph by its ID or filename
Args:
session: Database session
graph_id: Either an integer ID or a string filename
Returns:
KnowledgeGraph object or None if not found
"""
try:
logger.info(f"Looking up knowledge graph: {graph_id} (type: {type(graph_id)})")
# Special handling for "latest"
if isinstance(graph_id, str) and graph_id.lower() == "latest":
logger.info("Handling 'latest' special case")
kg = session.query(models.KnowledgeGraph).order_by(models.KnowledgeGraph.id.desc()).first()
if kg:
logger.info(f"Found latest knowledge graph with ID {kg.id} and filename {kg.filename}")
return kg
logger.warning("No knowledge graphs found in database")
return None
# Try as integer ID first
if isinstance(graph_id, int) or (isinstance(graph_id, str) and graph_id.isdigit()):
kg_id = int(graph_id)
logger.info(f"Looking up knowledge graph by ID: {kg_id}")
kg = session.query(models.KnowledgeGraph).filter(models.KnowledgeGraph.id == kg_id).first()
if kg:
logger.info(f"Found knowledge graph by ID {kg_id}: {kg.filename}")
return kg
logger.warning(f"Knowledge graph with ID {kg_id} not found")
# If not found by ID or not an integer, try as filename
if isinstance(graph_id, str):
logger.info(f"Looking up knowledge graph by filename: {graph_id}")
kg = session.query(models.KnowledgeGraph).filter(models.KnowledgeGraph.filename == graph_id).first()
if kg:
logger.info(f"Found knowledge graph by filename {graph_id}: ID {kg.id}")
return kg
logger.warning(f"Knowledge graph with filename {graph_id} not found")
logger.error(f"Knowledge graph not found: {graph_id}")
return None
except Exception as e:
logger.error(f"Error retrieving knowledge graph by ID: {str(e)}")
return None
def update_knowledge_graph(session: Session, filename: str, graph_data: dict) -> models.KnowledgeGraph:
"""
Update an existing knowledge graph with new data.
Args:
session: Database session
filename: Filename of the knowledge graph to update
graph_data: New graph data
Returns:
Updated KnowledgeGraph instance
"""
# Get the knowledge graph
kg = get_knowledge_graph(session, filename)
if not kg:
# Create a new knowledge graph if it doesn't exist
logger.info(f"Knowledge graph {filename} not found. Creating a new one.")
return save_knowledge_graph(session, filename, graph_data)
# Update the knowledge graph data
kg.graph_data = graph_data
# Update entity and relation counts
if isinstance(graph_data, dict):
if 'entities' in graph_data and isinstance(graph_data['entities'], list):
kg.entity_count = len(graph_data['entities'])
if 'relations' in graph_data and isinstance(graph_data['relations'], list):
kg.relation_count = len(graph_data['relations'])
# Update last modified timestamp
kg.update_timestamp = datetime.utcnow()
# Save to database
session.add(kg)
session.commit()
logger.info(f"Updated knowledge graph {filename}")
return kg
def delete_knowledge_graph(session: Session, identifier: Union[int, str]) -> bool:
"""
Delete a knowledge graph and all its associated entities and relations.
Args:
session: Database session
identifier: Knowledge graph ID or filename
Returns:
True if deletion was successful, False otherwise
"""
try:
# Find the knowledge graph
if isinstance(identifier, str):
# Identifier is a filename
kg = session.query(models.KnowledgeGraph).filter_by(filename=identifier).first()
else:
# Identifier is an ID
kg = session.query(models.KnowledgeGraph).filter_by(id=identifier).first()
if not kg:
logger.warning(f"Knowledge graph with identifier {identifier} not found")
return False
kg_id = kg.id
filename = kg.filename
# Count associated entities and relations for logging
entity_count = session.query(models.Entity).filter_by(graph_id=kg_id).count()
relation_count = session.query(models.Relation).filter_by(graph_id=kg_id).count()
# Begin transaction
logger.info(f"Deleting knowledge graph {filename} (ID: {kg_id}) with {entity_count} entities and {relation_count} relations")
# Due to the CASCADE setting in the relationships, deleting the knowledge graph
# will automatically delete all associated entities and relations.
# However, we'll delete them explicitly for clarity and to ensure proper cleanup.
# Delete relations first (due to foreign key constraints)
session.query(models.Relation).filter_by(graph_id=kg_id).delete()
# Delete entities
session.query(models.Entity).filter_by(graph_id=kg_id).delete()
# Delete the knowledge graph
session.delete(kg)
# Commit transaction
session.commit()
logger.info(f"Successfully deleted knowledge graph {filename} (ID: {kg_id}) and its associated data")
return True
except Exception as e:
# Rollback on error
session.rollback()
logger.error(f"Error deleting knowledge graph: {str(e)}")
return False
def get_trace(session: Session, trace_id: str) -> Optional[models.Trace]:
"""
Get a trace by its ID or filename.
Args:
session: Database session
trace_id: Either a UUID trace_id or a filename
Returns:
Trace object or None if not found
"""
# Try as UUID trace_id first
trace = session.query(models.Trace).filter_by(trace_id=trace_id).first()
if trace:
return trace
# If not found, try as filename
trace = session.query(models.Trace).filter_by(filename=trace_id).first()
if trace:
return trace
# If not found, try as ID
try:
id_value = int(trace_id)
trace = session.query(models.Trace).filter_by(id=id_value).first()
if trace:
return trace
except (ValueError, TypeError):
pass
return None
def save_trace(
session: Session,
content: str,
filename: Optional[str] = None,
title: Optional[str] = None,
description: Optional[str] = None,
trace_type: Optional[str] = None,
trace_source: str = "user_upload",
uploader: Optional[str] = None,
tags: Optional[List[str]] = None,
trace_metadata: Optional[Dict[str, Any]] = None
) -> models.Trace:
"""
Save a trace to the database.
Args:
session: Database session
content: The content of the trace
filename: Optional filename
title: Optional title
description: Optional description
trace_type: Optional type of trace
trace_source: Source of the trace (default: "user_upload")
uploader: Optional name of the uploader
tags: Optional list of tags
trace_metadata: Optional additional metadata
Returns:
The created or updated Trace object
"""
# Generate content hash for deduplication
content_hash = hashlib.sha256(content.encode('utf-8')).hexdigest()
# Check if trace already exists with this content hash
existing_trace = session.query(models.Trace).filter_by(content_hash=content_hash).first()
if existing_trace:
logger.info(f"Trace with matching content hash already exists (ID: {existing_trace.id})")
# Update fields if provided
if filename:
existing_trace.filename = filename
if title:
existing_trace.title = title
if description:
existing_trace.description = description
if trace_type:
existing_trace.trace_type = trace_type
if uploader:
existing_trace.uploader = uploader
if tags:
existing_trace.tags = tags
if trace_metadata:
# Merge metadata rather than replace
if existing_trace.trace_metadata:
existing_trace.trace_metadata.update(trace_metadata)
else:
existing_trace.trace_metadata = trace_metadata
# Update timestamp
existing_trace.update_timestamp = datetime.utcnow()
session.add(existing_trace)
session.commit()
return existing_trace
# Create new trace
trace = models.Trace.from_content(
content=content,
filename=filename,
title=title,
description=description,
trace_type=trace_type,
trace_source=trace_source,
uploader=uploader,
tags=tags,
trace_metadata=trace_metadata
)
session.add(trace)
session.commit()
logger.info(f"New trace saved to database (ID: {trace.id}, trace_id: {trace.trace_id})")
return trace
def get_all_traces(session: Session) -> List[models.Trace]:
"""
Get all traces from the database.
Args:
session: Database session
Returns:
List of Trace objects
"""
return session.query(models.Trace).order_by(models.Trace.upload_timestamp.desc()).all()
def get_traces_by_status(session: Session, status: str) -> List[models.Trace]:
"""
Get traces by status.
Args:
session: Database session
status: Status to filter by
Returns:
List of Trace objects with the specified status
"""
return session.query(models.Trace).filter_by(status=status).order_by(models.Trace.upload_timestamp.desc()).all()
def update_trace_status(session: Session, trace_id: str, status: str) -> models.Trace:
"""
Update the status of a trace.
Args:
session: Database session
trace_id: ID of the trace to update
status: New status
Returns:
Updated Trace object
"""
trace = get_trace(session, trace_id)
if not trace:
raise ValueError(f"Trace with ID {trace_id} not found")
trace.status = status
trace.update_timestamp = datetime.utcnow()
session.add(trace)
session.commit()
return trace
def update_trace_content(session: Session, trace_id: str, content: str) -> models.Trace:
"""
Update the content of a trace.
Args:
session: Database session
trace_id: ID of the trace to update
content: New content value
Returns:
Updated Trace object
"""
trace = get_trace(session, trace_id)
if not trace:
raise ValueError(f"Trace with ID {trace_id} not found")
trace.content = content
trace.character_count = len(content)
# Recalculate turn count if needed
trace.turn_count = len([line for line in content.split('\n') if line.strip()])
trace.update_timestamp = datetime.utcnow()
session.add(trace)
session.commit()
return trace
def link_knowledge_graph_to_trace(
session: Session,
kg_id: Union[int, str],
trace_id: str,
window_index: Optional[int] = None,
window_total: Optional[int] = None,
window_start_char: Optional[int] = None,
window_end_char: Optional[int] = None
) -> models.KnowledgeGraph:
"""
Link a knowledge graph to a trace.
Args:
session: Database session
kg_id: ID or filename of the knowledge graph
trace_id: ID of the trace
window_index: Optional index of the window within the trace
window_total: Optional total number of windows
window_start_char: Optional start position in the trace
window_end_char: Optional end position in the trace
Returns:
Updated KnowledgeGraph object
"""
# Get the knowledge graph
kg = get_knowledge_graph_by_id(session, kg_id)
if not kg:
raise ValueError(f"Knowledge graph with ID {kg_id} not found")
# Get the trace
trace = get_trace(session, trace_id)
if not trace:
raise ValueError(f"Trace with ID {trace_id} not found")
# Update knowledge graph with trace information
kg.trace_id = trace.trace_id
if window_index is not None:
kg.window_index = window_index
if window_total is not None:
kg.window_total = window_total
if window_start_char is not None:
kg.window_start_char = window_start_char
if window_end_char is not None:
kg.window_end_char = window_end_char
# Update graph metadata to include trace info
graph_data = kg.graph_data or {}
if "metadata" not in graph_data:
graph_data["metadata"] = {}
graph_data["metadata"]["trace_info"] = {
"trace_id": trace.trace_id,
"window_index": window_index,
"window_total": window_total,
"linked_at": datetime.utcnow().isoformat()
}
kg.graph_data = graph_data
session.add(kg)
session.commit()
return kg
def get_knowledge_graphs_for_trace(session: Session, trace_id: str) -> List[models.KnowledgeGraph]:
"""
Get all knowledge graphs associated with a trace.
Args:
session: Database session
trace_id: ID of the trace
Returns:
List of KnowledgeGraph objects linked to the trace
"""
trace = get_trace(session, trace_id)
if not trace:
raise ValueError(f"Trace with ID {trace_id} not found")
return session.query(models.KnowledgeGraph).filter_by(trace_id=trace.trace_id).order_by(
models.KnowledgeGraph.window_index
).all()
def check_knowledge_graph_exists(session: Session, trace_id: str, is_original: bool = None) -> Optional[models.KnowledgeGraph]:
"""
Check if a knowledge graph exists for a trace with specific criteria.
Args:
session: Database session
trace_id: ID of the trace
is_original: If True, only return knowledge graphs with status='created'
If False, only return knowledge graphs with other statuses
Returns:
KnowledgeGraph object if found, None otherwise
"""
query = session.query(models.KnowledgeGraph).filter_by(trace_id=trace_id)
if is_original is True:
# Original KGs have status 'created'
query = query.filter_by(status='created')
elif is_original is False:
# Non-original KGs have other statuses
query = query.filter(models.KnowledgeGraph.status != 'created')
return query.first()
def delete_trace(session: Session, trace_id: str, delete_related_kgs: bool = False) -> bool:
"""
Delete a trace from the database.
Args:
session: Database session
trace_id: ID of the trace to delete
delete_related_kgs: Whether to also delete related knowledge graphs
Returns:
True if successful, False otherwise
"""
trace = get_trace(session, trace_id)
if not trace:
return False
# If requested, delete related knowledge graphs
if delete_related_kgs:
for kg in trace.knowledge_graphs:
session.delete(kg)
else:
# Otherwise, just unlink knowledge graphs from this trace
for kg in trace.knowledge_graphs:
kg.trace_id = None
session.add(kg)
# Delete the trace
session.delete(trace)
session.commit()
return True
def get_prompt_reconstructions_for_kg(session, kg_identifier):
"""
Fetch all prompt reconstructions for a given knowledge graph (by ID or filename).
Returns a dict mapping relation_id to reconstructed_prompt.
"""
from backend.database.models import KnowledgeGraph, PromptReconstruction
if isinstance(kg_identifier, int):
kg = session.query(KnowledgeGraph).filter_by(id=kg_identifier).first()
else:
kg = session.query(KnowledgeGraph).filter_by(filename=kg_identifier).first()
if not kg:
return {}
prompt_reconstructions = session.query(PromptReconstruction).filter_by(knowledge_graph_id=kg.id).all()
return {pr.relation_id: pr.reconstructed_prompt for pr in prompt_reconstructions}
def get_prompt_reconstruction_for_relation(session, kg_identifier, relation_id):
"""
Fetch a single reconstructed prompt for a given knowledge graph and relation_id.
Returns the reconstructed_prompt string or None.
"""
from backend.database.models import KnowledgeGraph, PromptReconstruction
if isinstance(kg_identifier, int):
kg = session.query(KnowledgeGraph).filter_by(id=kg_identifier).first()
else:
kg = session.query(KnowledgeGraph).filter_by(filename=kg_identifier).first()
if not kg:
return None
pr = session.query(PromptReconstruction).filter_by(knowledge_graph_id=kg.id, relation_id=relation_id).first()
return pr.reconstructed_prompt if pr else None
def save_causal_analysis(
session: Session,
knowledge_graph_id: int,
perturbation_set_id: str,
analysis_method: str,
analysis_result: dict = None,
causal_score: float = None,
analysis_metadata: dict = None
):
"""Save a causal analysis result to the database."""
from backend.database import models
causal_analysis = models.CausalAnalysis(
knowledge_graph_id=knowledge_graph_id,
perturbation_set_id=perturbation_set_id,
analysis_method=analysis_method,
analysis_result=analysis_result,
causal_score=causal_score,
analysis_metadata=analysis_metadata
)
session.add(causal_analysis)
session.commit()
session.refresh(causal_analysis)
return causal_analysis
def get_causal_analysis(
session: Session,
knowledge_graph_id: int,
perturbation_set_id: str,
analysis_method: str
) -> Optional[models.CausalAnalysis]:
"""
Get causal analysis results from the database.
Args:
session: Database session
knowledge_graph_id: ID of the knowledge graph
perturbation_set_id: ID of the perturbation set
analysis_method: Method used for analysis
Returns:
CausalAnalysis object or None if not found
"""
return session.query(models.CausalAnalysis).filter_by(
knowledge_graph_id=knowledge_graph_id,
perturbation_set_id=perturbation_set_id,
analysis_method=analysis_method
).first()
def get_all_causal_analyses(
session: Session,
knowledge_graph_id: Optional[int] = None,
perturbation_set_id: Optional[str] = None,
analysis_method: Optional[str] = None
) -> List[models.CausalAnalysis]:
"""
Get all causal analysis results from the database with optional filters.
Args:
session: Database session
knowledge_graph_id: Optional filter by knowledge graph ID
perturbation_set_id: Optional filter by perturbation set ID
analysis_method: Optional filter by analysis method
Returns:
List of CausalAnalysis objects
"""
query = session.query(models.CausalAnalysis)
if knowledge_graph_id is not None:
query = query.filter_by(knowledge_graph_id=knowledge_graph_id)
if perturbation_set_id is not None:
query = query.filter_by(perturbation_set_id=perturbation_set_id)
if analysis_method is not None:
query = query.filter_by(analysis_method=analysis_method)
return query.all()
def get_causal_analysis_for_perturbation(session: Session, perturbation_set_id: str) -> List[Dict[str, Any]]:
"""
Get all causal analysis results for a specific perturbation set.
Args:
session: Database session
perturbation_set_id: ID of the perturbation set
Returns:
List of causal analysis results with their associated data
"""
from backend.database.models import CausalAnalysis, KnowledgeGraph, PromptReconstruction
results = session.query(
CausalAnalysis,
KnowledgeGraph,
PromptReconstruction
).join(
KnowledgeGraph,
CausalAnalysis.knowledge_graph_id == KnowledgeGraph.id
).join(
PromptReconstruction,
CausalAnalysis.prompt_reconstruction_id == PromptReconstruction.id
).filter(
CausalAnalysis.perturbation_set_id == perturbation_set_id
).all()
return [{
'analysis': analysis.to_dict(),
'knowledge_graph': kg.to_dict(),
'prompt_reconstruction': {
'id': pr.id,
'relation_id': pr.relation_id,
'reconstructed_prompt': pr.reconstructed_prompt,
'dependencies': pr.dependencies
}
} for analysis, kg, pr in results]
def get_causal_analysis_by_method(session: Session, knowledge_graph_id: int, method: str) -> List[Dict[str, Any]]:
"""
Get causal analysis results for a specific knowledge graph and analysis method.
Args:
session: Database session
knowledge_graph_id: ID of the knowledge graph
method: Analysis method (e.g., 'graph', 'component', 'dowhy')
Returns:
List of causal analysis results
"""
from backend.database.models import CausalAnalysis, PerturbationTest
results = session.query(
CausalAnalysis,
PerturbationTest
).join(
PerturbationTest,
CausalAnalysis.perturbation_test_id == PerturbationTest.id
).filter(
CausalAnalysis.knowledge_graph_id == knowledge_graph_id,
CausalAnalysis.analysis_method == method
).all()
return [{
'analysis': analysis.to_dict(),
'perturbation_test': {
'id': pt.id,
'perturbation_type': pt.perturbation_type,
'test_result': pt.test_result,
'perturbation_score': pt.perturbation_score
}
} for analysis, pt in results]
def get_causal_analysis_summary(session: Session, knowledge_graph_id: int) -> Dict[str, Any]:
"""
Get a summary of causal analysis results for a knowledge graph.
Args:
session: Database session
knowledge_graph_id: ID of the knowledge graph
Returns:
Dictionary containing summary statistics and results by method
"""
from backend.database.models import CausalAnalysis
from sqlalchemy import func
# Get all analyses for this knowledge graph
analyses = session.query(CausalAnalysis).filter_by(
knowledge_graph_id=knowledge_graph_id
).all()
if not analyses:
return {
'total_analyses': 0,
'methods': {},
'average_scores': {}
}
# Group by method
method_results = {}
for analysis in analyses:
method = analysis.analysis_method
if method not in method_results:
method_results[method] = []
method_results[method].append(analysis)
# Calculate statistics
summary = {
'total_analyses': len(analyses),
'methods': {},
'average_scores': {}
}
for method, results in method_results.items():
scores = [r.causal_score for r in results if r.causal_score is not None]
summary['methods'][method] = {
'count': len(results),
'average_score': sum(scores) / len(scores) if scores else None,
'min_score': min(scores) if scores else None,
'max_score': max(scores) if scores else None
}
return summary
# Add these functions to handle perturbation tests
def save_perturbation_test(session,
knowledge_graph_id: int,
prompt_reconstruction_id: int,
relation_id: str,
perturbation_type: str,
perturbation_set_id: str,
test_result: dict = None,
perturbation_score: float = None,
test_metadata: dict = None) -> int:
"""
Save a perturbation test to the database.
Args:
session: Database session
knowledge_graph_id: ID of the knowledge graph
prompt_reconstruction_id: ID of the prompt reconstruction
relation_id: ID of the relation
perturbation_type: Type of perturbation
perturbation_set_id: ID of the perturbation set
test_result: Test result dictionary
perturbation_score: Perturbation score
test_metadata: Test metadata dictionary
Returns:
int: ID of the saved perturbation test
"""
from backend.database.models import PerturbationTest
# Create new perturbation test
test = PerturbationTest(
knowledge_graph_id=knowledge_graph_id,
prompt_reconstruction_id=prompt_reconstruction_id,
relation_id=relation_id,
perturbation_type=perturbation_type,
perturbation_set_id=perturbation_set_id,
test_result=test_result or {},
perturbation_score=perturbation_score,
test_metadata=test_metadata or {}
)
# Add to session and commit
session.add(test)
session.commit()
return test.id
def delete_perturbation_test(session, test_id: int) -> bool:
"""
Delete a perturbation test from the database.
Args:
session: Database session
test_id: ID of the perturbation test to delete
Returns:
bool: True if successful, False otherwise
"""
from backend.database.models import PerturbationTest
# Query the test
test = session.query(PerturbationTest).filter_by(id=test_id).first()
if test:
# Delete and commit
session.delete(test)
session.commit()
return True
return False
def delete_perturbation_tests_by_set(session, perturbation_set_id: str) -> int:
"""
Delete all perturbation tests in a set.
Args:
session: Database session
perturbation_set_id: ID of the perturbation set
Returns:
int: Number of tests deleted
"""
from backend.database.models import PerturbationTest
# Query all tests in the set
tests = session.query(PerturbationTest).filter_by(perturbation_set_id=perturbation_set_id).all()
# Delete all tests
deleted_count = 0
for test in tests:
session.delete(test)
deleted_count += 1
# Commit changes
session.commit()
return deleted_count
def get_context_document_stats(session: Session, trace_id: str) -> Dict[str, Any]:
"""
Get statistics about context documents for a trace.
Args:
session: Database session
trace_id: Trace ID to get stats for
Returns:
Dictionary with context document statistics
"""
trace = get_trace(session, trace_id)
if not trace or not trace.trace_metadata or "context_documents" not in trace.trace_metadata:
return {"total_count": 0, "active_count": 0, "by_type": {}}
docs = trace.trace_metadata["context_documents"]
# Count by type
by_type = {}
active_count = 0
for doc in docs:
doc_type = doc.get("document_type", "unknown")
if doc_type not in by_type:
by_type[doc_type] = 0
by_type[doc_type] += 1
if doc.get("is_active", True):
active_count += 1
return {
"total_count": len(docs),
"active_count": active_count,
"by_type": by_type
}
def get_context_documents_from_trace(session: Session, trace_id: str) -> List[Dict[str, Any]]:
"""
Get context documents from a trace's metadata.
Args:
session: Database session
trace_id: ID of the trace
Returns:
List of context documents, or empty list if none found
"""
trace = get_trace(session, trace_id)
if not trace or not trace.trace_metadata or "context_documents" not in trace.trace_metadata:
return []
docs = trace.trace_metadata["context_documents"]
# Filter to only active documents
active_docs = [doc for doc in docs if doc.get("is_active", True)]
return active_docs
def get_temporal_windows_by_trace_id(session: Session, trace_id: str, processing_run_id: Optional[str] = None) -> Dict[str, Any]:
"""
Get all knowledge graph windows for a specific trace, ordered by window_index.
Also returns the full/merged version if available.
Used for temporal force-directed graph visualization.
This function handles cases where KGs exist with trace_id but no trace record exists.
Args:
session: Database session
trace_id: ID of the trace
processing_run_id: Optional ID to filter by specific processing run
Returns:
Dict containing windowed KGs and full KG information
"""
logger.info(f"Looking up temporal windows for trace_id: {trace_id}")
if processing_run_id:
logger.info(f"Filtering by processing_run_id: {processing_run_id}")
# First check if we can find the trace
trace = get_trace(session, trace_id)
# Get all knowledge graphs for this trace_id (even if trace record doesn't exist)
query = session.query(models.KnowledgeGraph).filter(models.KnowledgeGraph.trace_id == trace_id)
# Filter by processing_run_id if provided
if processing_run_id:
query = query.filter(models.KnowledgeGraph.processing_run_id == processing_run_id)
all_kgs = query.all()
logger.info(f"Found {len(all_kgs)} total knowledge graphs for trace_id {trace_id}")
if processing_run_id:
logger.info(f"(filtered by processing_run_id: {processing_run_id})")
if not all_kgs:
logger.warning(f"No knowledge graphs found for trace_id: {trace_id}")
return {"windows": [], "full_kg": None, "trace_info": None}
# If no trace record exists, create minimal trace info from KGs
if not trace:
logger.warning(f"No trace record found for trace_id {trace_id}, but {len(all_kgs)} KGs exist")
trace_info = {
"trace_id": trace_id,
"title": f"Trace {trace_id[:8]}...",
"description": "Knowledge graphs exist but no trace record found",
"upload_timestamp": min([kg.creation_timestamp for kg in all_kgs if kg.creation_timestamp])
}
else:
logger.info(f"Found trace record {trace.trace_id}")
trace_info = {
"trace_id": trace.trace_id,
"title": trace.title,
"description": trace.description,
"upload_timestamp": trace.upload_timestamp
}
# Separate windowed KGs from the full/merged KG
windowed_kgs = []
full_kg = None
for kg in all_kgs:
# Full/merged KG: has window_total but no window_index, null start/end chars
if (kg.window_total is not None and
kg.window_index is None and
kg.window_start_char is None and
kg.window_end_char is None):
full_kg = kg
logger.info(f"Found full/merged KG: {kg.filename} with window_total={kg.window_total}")
# Windowed KG: has window_index
elif kg.window_index is not None:
windowed_kgs.append(kg)
# KG without proper window info - try to assign window_index
else:
logger.info(f"Found KG without proper window info: {kg.filename}")
# If we don't have a full KG yet and this looks like it could be one
if (full_kg is None and
kg.window_total is None and
kg.window_start_char is None and
kg.window_end_char is None):
# Check if this KG has significantly more entities than others (indicating it's merged)
if kg.graph_data and len(kg.graph_data.get("entities", [])) > 10:
kg.window_total = len(windowed_kgs) + 1 # Set based on current windowed KGs
full_kg = kg
session.add(kg)
logger.info(f"Assigned {kg.filename} as full KG with window_total={kg.window_total}")
# If we have windowed KGs but some are missing window_index, assign them
kgs_without_index = [kg for kg in all_kgs if kg.window_index is None and kg != full_kg]
if kgs_without_index and windowed_kgs:
logger.info("Assigning window_index to knowledge graphs based on creation order")
# Sort by creation timestamp and assign window_index starting from the highest existing + 1
max_window_index = max([kg.window_index for kg in windowed_kgs], default=-1)
kgs_without_index.sort(key=lambda kg: kg.creation_timestamp or datetime.utcnow())
for i, kg in enumerate(kgs_without_index):
kg.window_index = max_window_index + 1 + i
session.add(kg)
windowed_kgs.append(kg)
session.commit()
logger.info(f"Assigned window_index to {len(kgs_without_index)} knowledge graphs")
# Sort windowed KGs by window_index
windowed_kgs.sort(key=lambda kg: kg.window_index)
logger.info(f"Found {len(windowed_kgs)} windowed KGs and {'1' if full_kg else '0'} full KG for trace_id {trace_id}")
# Update entity_count and relation_count if they're 0 or None but graph_data has content
updated_count = 0
for kg in windowed_kgs + ([full_kg] if full_kg else []):
if kg and kg.graph_data:
needs_update = False
if kg.entity_count is None or kg.entity_count == 0:
entities = kg.graph_data.get("entities", [])
if entities:
kg.entity_count = len(entities)
needs_update = True
if kg.relation_count is None or kg.relation_count == 0:
relations = kg.graph_data.get("relations", [])
if relations:
kg.relation_count = len(relations)
needs_update = True
if needs_update:
session.add(kg)
updated_count += 1
if updated_count > 0:
session.commit()
logger.info(f"Updated entity/relation counts for {updated_count} knowledge graphs")
return {
"windows": windowed_kgs,
"full_kg": full_kg,
"trace_info": trace_info
} |