Spaces:
Running
Running
File size: 38,632 Bytes
a55ea57 6663ad4 a55ea57 6663ad4 a55ea57 6663ad4 a55ea57 6663ad4 a55ea57 6663ad4 a55ea57 6663ad4 2f1e7ff 6663ad4 a55ea57 6663ad4 a55ea57 6663ad4 a55ea57 | 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 | """
File: /data.py
Project: postgresql
Created Date: Sunday May 24th 2026
Author: Christian Nonis <alch.infoemail@gmail.com>
"""
from __future__ import annotations
from contextlib import contextmanager
import json
import threading
from typing import Any, Iterator, List, Literal, Optional, Tuple
import psycopg2
import psycopg2.extras
from src.adapters.interfaces.data import DataClient, SearchResult
from src.config import SEARCH_FTS_REGCONFIGS, config
from src.services.brain_identity import BrainAlreadyExistsError, validate_brain_id
from src.constants.data import (
Brain,
KGChanges,
KGChangesType,
Observation,
StructuredData,
TextChunk,
Workspace,
)
from ._provisioning import (
borrow,
ensure_brain_database,
ensure_database_exists,
get_brain_pool,
get_system_pool,
clear_brain_database,
drop_brain_database,
require_registered_brain,
)
_SYSTEM_DDL = """
CREATE TABLE IF NOT EXISTS data_brains (
id TEXT PRIMARY KEY,
name_key TEXT UNIQUE NOT NULL,
pat TEXT NOT NULL,
document JSONB NOT NULL,
created_at TIMESTAMPTZ NOT NULL DEFAULT NOW()
);
CREATE TABLE IF NOT EXISTS data_workspaces (
id TEXT PRIMARY KEY,
brain_id TEXT UNIQUE NOT NULL,
slug TEXT UNIQUE NOT NULL,
document JSONB NOT NULL,
created_at TIMESTAMPTZ NOT NULL DEFAULT NOW()
);
"""
_BRAIN_DDL = """
CREATE TABLE IF NOT EXISTS data_text_chunks (
id TEXT PRIMARY KEY,
text TEXT,
metadata JSONB,
inserted_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
document JSONB NOT NULL
);
CREATE INDEX IF NOT EXISTS idx_data_text_chunks_inserted_at
ON data_text_chunks (inserted_at DESC);
CREATE TABLE IF NOT EXISTS data_observations (
id TEXT PRIMARY KEY,
text TEXT,
resource_id TEXT,
metadata JSONB,
inserted_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
document JSONB NOT NULL
);
CREATE INDEX IF NOT EXISTS idx_data_observations_resource
ON data_observations (resource_id);
CREATE INDEX IF NOT EXISTS idx_data_observations_labels
ON data_observations USING gin ((metadata -> 'labels'));
CREATE TABLE IF NOT EXISTS data_structured_data (
id TEXT PRIMARY KEY,
data JSONB,
types TEXT[],
metadata JSONB,
inserted_at TIMESTAMPTZ NOT NULL DEFAULT NOW(),
document JSONB NOT NULL
);
CREATE INDEX IF NOT EXISTS idx_data_structured_data_types
ON data_structured_data USING gin (types);
CREATE INDEX IF NOT EXISTS idx_data_structured_data_inserted_at
ON data_structured_data (inserted_at DESC);
CREATE TABLE IF NOT EXISTS data_kg_changes (
id TEXT PRIMARY KEY,
type TEXT,
change JSONB,
timestamp TIMESTAMPTZ NOT NULL DEFAULT NOW(),
document JSONB NOT NULL
);
CREATE INDEX IF NOT EXISTS idx_data_kg_changes_timestamp
ON data_kg_changes (timestamp DESC);
"""
_SEARCH_DDL = """
ALTER TABLE data_text_chunks
ADD COLUMN IF NOT EXISTS search_tsv tsvector
GENERATED ALWAYS AS (to_tsvector('english', coalesce(text, ''))) STORED;
ALTER TABLE data_text_chunks
ADD COLUMN IF NOT EXISTS search_len integer
GENERATED ALWAYS AS (
GREATEST(length(to_tsvector('english', coalesce(text, ''))), 0)
) STORED;
CREATE INDEX IF NOT EXISTS idx_data_text_chunks_search_tsv
ON data_text_chunks USING gin (search_tsv);
"""
def _search_alt_ddl(regconfig: str) -> str:
if regconfig not in SEARCH_FTS_REGCONFIGS:
raise ValueError(f"Unsupported FTS regconfig {regconfig!r}")
return f"""
ALTER TABLE data_text_chunks
ADD COLUMN IF NOT EXISTS search_tsv_alt tsvector
GENERATED ALWAYS AS (to_tsvector('{regconfig}', coalesce(text, ''))) STORED;
ALTER TABLE data_text_chunks
ADD COLUMN IF NOT EXISTS search_len_alt integer
GENERATED ALWAYS AS (
GREATEST(length(to_tsvector('{regconfig}', coalesce(text, ''))), 0)
) STORED;
CREATE INDEX IF NOT EXISTS idx_data_text_chunks_search_tsv_alt
ON data_text_chunks USING gin (search_tsv_alt);
"""
def _bm25_sql(
regconfig: str,
tsv_col: str,
len_col: str,
*,
query_match: str | None = None,
) -> str:
if tsv_col not in {"search_tsv", "search_tsv_alt"}:
raise ValueError(f"Unsupported tsv column {tsv_col!r}")
if len_col not in {"search_len", "search_len_alt"}:
raise ValueError(f"Unsupported len column {len_col!r}")
if regconfig not in SEARCH_FTS_REGCONFIGS and regconfig != "english":
raise ValueError(f"Unsupported FTS regconfig {regconfig!r}")
if query_match is None:
query_match = "or" if tsv_col == "search_tsv_alt" else "and"
if query_match not in {"and", "or"}:
raise ValueError(f"Unsupported BM25 query_match {query_match!r}")
if query_match == "and":
match_sql = f"c.{tsv_col} @@ plainto_tsquery('{regconfig}', %s)"
else:
match_sql = "TRUE"
return f"""
WITH query_lexemes AS (
SELECT unnest(tsvector_to_array(to_tsvector('{regconfig}', %s)))
AS lexeme
),
coll AS (
SELECT
COUNT(*)::double precision AS n_docs,
COALESCE(
AVG(GREATEST({len_col}, 1)),
1
)::double precision AS avgdl
FROM data_text_chunks
),
idf AS (
SELECT
s.word AS lexeme,
ln(
1.0 + (coll.n_docs - s.ndoc + 0.5)
/ (s.ndoc + 0.5)
) AS idf
FROM ts_stat(
'SELECT {tsv_col} FROM data_text_chunks'
) AS s
CROSS JOIN coll
WHERE s.word IN (SELECT lexeme FROM query_lexemes)
),
docs AS (
SELECT
c.id,
c.document,
c.text,
GREATEST(c.{len_col}, 1)::double precision AS dl,
lex.lexeme,
COALESCE(
array_length(lex.positions, 1),
1
)::double precision AS tf
FROM data_text_chunks c
CROSS JOIN LATERAL unnest(c.{tsv_col})
AS lex(lexeme, positions)
WHERE {match_sql}
AND lex.lexeme IN (SELECT lexeme FROM query_lexemes)
)
SELECT
d.id,
d.document,
d.text,
SUM(
COALESCE(i.idf, 0)
* (d.tf * (%s + 1.0))
/ (
d.tf
+ %s * (
1.0 - %s
+ %s * d.dl
/ NULLIF(c.avgdl, 0)
)
)
) AS bm25
FROM docs d
CROSS JOIN coll c
LEFT JOIN idf i ON i.lexeme = d.lexeme
GROUP BY d.id, d.document, d.text
ORDER BY bm25 DESC, d.id ASC
LIMIT %s
"""
def _ilike_pattern(query_text: str) -> str:
escaped = query_text.replace("\\", "\\\\").replace("%", "\\%").replace("_", "\\_")
return f"%{escaped}%"
def _validate_changelog(document: dict) -> Optional[KGChanges]:
if "change" in document and isinstance(document["change"], dict):
change_type = document["change"].get("type")
if isinstance(change_type, str):
document["change"]["type"] = KGChangesType(change_type)
if isinstance(document.get("type"), str):
document["type"] = KGChangesType(document["type"])
return KGChanges.model_validate(document)
class PostgreSQLDataClient(DataClient):
"""
Data client backed by PostgreSQL with JSONB-stored documents.
Mirrors the Mongo client's `database = brain_id` layout: a single registry
database (``config.postgresql.system_database``) holds the ``data_brains``
catalog, and every brain owns its own Postgres database (named
``brain_<sanitized_id>``) containing all of its tables.
"""
def __init__(self):
self._lock = threading.Lock()
self._system_initialized = False
self._initialized_brains: set[str] = set()
self._search_ready_brains: set[str] = set()
self._search_alt_brains: set[str] = set()
@contextmanager
def _system_connection(self) -> Iterator[psycopg2.extensions.connection]:
self._ensure_system_schema()
with borrow(get_system_pool()) as conn:
yield conn
@contextmanager
def _brain_connection(
self, brain_id: str
) -> Iterator[psycopg2.extensions.connection]:
# Existing connection only: provisioning belongs to create_brain().
with borrow(get_brain_pool(brain_id)) as conn:
yield conn
def _ensure_system_schema(self) -> None:
if self._system_initialized:
return
with self._lock:
if self._system_initialized:
return
ensure_database_exists(config.postgresql.system_database)
with borrow(get_system_pool()) as conn:
with conn.cursor() as cur:
cur.execute(_SYSTEM_DDL)
conn.commit()
self._system_initialized = True
def _ensure_brain_schema(self, brain_id: str) -> None:
require_registered_brain(brain_id)
alt_regconfig = config.search_fts_regconfig_for_brain(brain_id)
needs_base = brain_id not in self._initialized_brains
needs_search = (
config.search_enabled and brain_id not in self._search_ready_brains
)
needs_alt = (
config.search_enabled
and alt_regconfig is not None
and brain_id not in self._search_alt_brains
)
if not needs_base and not needs_search and not needs_alt:
return
with self._lock:
alt_regconfig = config.search_fts_regconfig_for_brain(brain_id)
needs_base = brain_id not in self._initialized_brains
needs_search = (
config.search_enabled and brain_id not in self._search_ready_brains
)
needs_alt = (
config.search_enabled
and alt_regconfig is not None
and brain_id not in self._search_alt_brains
)
if not needs_base and not needs_search and not needs_alt:
return
ensure_brain_database(brain_id)
with borrow(get_brain_pool(brain_id)) as conn:
with conn.cursor() as cur:
if needs_base:
cur.execute(_BRAIN_DDL)
if needs_search:
cur.execute(_SEARCH_DDL)
if needs_alt and alt_regconfig is not None:
cur.execute(_search_alt_ddl(alt_regconfig))
conn.commit()
if needs_base:
self._initialized_brains.add(brain_id)
if needs_search:
self._search_ready_brains.add(brain_id)
if needs_alt:
self._search_alt_brains.add(brain_id)
def save_text_chunk(self, text_chunk: TextChunk, brain_id: str) -> TextChunk:
document = text_chunk.model_dump(mode="json")
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
INSERT INTO data_text_chunks
(id, text, metadata, inserted_at, document)
VALUES (%s, %s, %s::jsonb, %s, %s::jsonb)
ON CONFLICT (id) DO UPDATE SET
text = EXCLUDED.text,
metadata = EXCLUDED.metadata,
inserted_at = EXCLUDED.inserted_at,
document = EXCLUDED.document
""",
(
text_chunk.id,
text_chunk.text,
json.dumps(text_chunk.metadata or {}, default=str),
text_chunk.inserted_at,
json.dumps(document, default=str),
),
)
conn.commit()
return text_chunk
def save_observations(
self, observations: List[Observation], brain_id: str
) -> List[Observation]:
if not observations:
return observations
rows = [
(
observation.id,
observation.text,
observation.resource_id,
json.dumps(observation.metadata or {}, default=str),
observation.inserted_at,
json.dumps(observation.model_dump(mode="json"), default=str),
)
for observation in observations
]
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
psycopg2.extras.execute_values(
cur,
"""
INSERT INTO data_observations
(id, text, resource_id, metadata, inserted_at, document)
VALUES %s
ON CONFLICT (id) DO UPDATE SET
text = EXCLUDED.text,
resource_id = EXCLUDED.resource_id,
metadata = EXCLUDED.metadata,
inserted_at = EXCLUDED.inserted_at,
document = EXCLUDED.document
""",
rows,
template="(%s, %s, %s, %s::jsonb, %s, %s::jsonb)",
)
conn.commit()
return observations
def search(
self,
text: str,
brain_id: str,
collection: str = "text_chunks",
limit: int = 10,
) -> SearchResult:
table = (
"data_text_chunks"
if collection in ("*", "text_chunks")
else f"data_{collection}"
)
pattern = _ilike_pattern(text)
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
f"""
SELECT document
FROM {table}
WHERE text ILIKE %s
LIMIT %s
""",
(pattern, limit),
)
text_rows = cur.fetchall()
text_chunks = [
TextChunk.model_validate(row["document"]) for row in text_rows
]
ids = [chunk.id for chunk in text_chunks]
observations: list[Observation] = []
if ids:
cur.execute(
"""
SELECT document
FROM data_observations
WHERE resource_id = ANY(%s)
""",
(ids,),
)
observations = [
Observation.model_validate(row["document"])
for row in cur.fetchall()
]
return SearchResult(text_chunks=text_chunks, observations=observations)
def search_bm25(
self,
text: str,
brain_id: str,
collection: str = "text_chunks",
limit: int = 10,
) -> List[Tuple[TextChunk, float]]:
if not text or not text.strip() or limit <= 0:
return []
if collection not in ("*", "text_chunks"):
return []
k1 = float(config.search_bm25_k1)
b = float(config.search_bm25_b)
alt_regconfig = config.search_fts_regconfig_for_brain(brain_id)
if alt_regconfig:
sql = _bm25_sql(alt_regconfig, "search_tsv_alt", "search_len_alt")
params = (text, k1, k1, b, b, limit)
else:
sql = _bm25_sql("english", "search_tsv", "search_len")
params = (text, text, k1, k1, b, b, limit)
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
sql,
params,
)
rows = cur.fetchall()
results: List[Tuple[TextChunk, float]] = []
for row in rows:
chunk = TextChunk.model_validate(row["document"])
results.append((chunk, float(row["bm25"] or 0.0)))
return results
def get_text_chunks_by_ids(
self, ids: List[str], with_observations: bool = False, brain_id: str = "default"
) -> Tuple[List[TextChunk], List[Observation]]:
if not ids:
return ([], [])
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"""
SELECT document FROM data_text_chunks
WHERE id = ANY(%s)
""",
(ids,),
)
text_chunks = [
TextChunk.model_validate(row["document"]) for row in cur.fetchall()
]
observations: list[Observation] = []
if with_observations:
cur.execute(
"""
SELECT document FROM data_observations
WHERE resource_id = ANY(%s)
""",
(ids,),
)
observations = [
Observation.model_validate(row["document"])
for row in cur.fetchall()
]
return (text_chunks, observations)
def save_structured_data(
self, structured_data: StructuredData, brain_id: str
) -> StructuredData:
document = structured_data.model_dump(mode="json")
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
INSERT INTO data_structured_data
(id, data, types, metadata, inserted_at, document)
VALUES (%s, %s::jsonb, %s, %s::jsonb, %s, %s::jsonb)
ON CONFLICT (id) DO UPDATE SET
data = EXCLUDED.data,
types = EXCLUDED.types,
metadata = EXCLUDED.metadata,
inserted_at = EXCLUDED.inserted_at,
document = EXCLUDED.document
""",
(
structured_data.id,
json.dumps(structured_data.data or {}, default=str),
list(structured_data.types or []),
json.dumps(structured_data.metadata or {}, default=str),
structured_data.inserted_at,
json.dumps(document, default=str),
),
)
conn.commit()
return structured_data
def create_brain(self, name_key: str) -> Brain:
name_key = validate_brain_id(name_key)
brain = Brain(name_key=name_key)
document = brain.model_dump(mode="json")
try:
with self._system_connection() as conn:
with conn.cursor() as cur:
cur.execute(
"""
INSERT INTO data_brains (id, name_key, pat, document)
VALUES (%s, %s, %s, %s::jsonb)
""",
(
brain.id,
brain.name_key,
brain.pat or "",
json.dumps(document, default=str),
),
)
conn.commit()
except psycopg2.errors.UniqueViolation:
raise BrainAlreadyExistsError("Brain already exists") from None
self._ensure_brain_schema(name_key)
return brain
def get_brain(self, name_key: str) -> Optional[Brain]:
with self._system_connection() as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"SELECT document FROM data_brains WHERE name_key = %s",
(name_key,),
)
row = cur.fetchone()
if not row:
return None
return Brain.model_validate(row["document"])
def get_brain_by_pat(self, pat: str) -> Optional[Brain]:
with self._system_connection() as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"SELECT document FROM data_brains WHERE pat = %s",
(pat,),
)
row = cur.fetchone()
if not row:
return None
return Brain.model_validate(row["document"])
def get_brains_list(self) -> List[Brain]:
with self._system_connection() as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute("SELECT document FROM data_brains")
rows = cur.fetchall()
return sorted(
(Brain.model_validate(row["document"]) for row in rows),
key=lambda brain: brain.name_key,
)
def create_workspace(self, workspace: Workspace) -> Workspace:
with self._system_connection() as conn:
with conn.cursor() as cur:
cur.execute(
"INSERT INTO data_workspaces (id, brain_id, slug, document) VALUES (%s, %s, %s, %s::jsonb) ON CONFLICT (brain_id) DO NOTHING",
(workspace.id, workspace.brain_id, workspace.slug, json.dumps(workspace.model_dump(mode="json"))),
)
conn.commit()
return self.get_workspace_by_brain_id(workspace.brain_id) or workspace
def _get_workspace(self, column: str, value: str) -> Workspace | None:
if column not in {"slug", "brain_id"}:
raise ValueError("invalid workspace lookup")
with self._system_connection() as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(f"SELECT document FROM data_workspaces WHERE {column} = %s", (value,))
row = cur.fetchone()
return Workspace.model_validate(row["document"]) if row else None
def get_workspace(self, slug: str) -> Workspace | None:
return self._get_workspace("slug", slug)
def get_workspace_by_brain_id(self, brain_id: str) -> Workspace | None:
return self._get_workspace("brain_id", brain_id)
def get_workspaces(self) -> list[Workspace]:
with self._system_connection() as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute("SELECT document FROM data_workspaces ORDER BY created_at")
rows = cur.fetchall()
return [Workspace.model_validate(row["document"]) for row in rows]
def update_workspace(self, workspace: Workspace) -> Workspace:
with self._system_connection() as conn:
with conn.cursor() as cur:
cur.execute(
"UPDATE data_workspaces SET document = %s::jsonb WHERE id = %s",
(json.dumps(workspace.model_dump(mode="json")), workspace.id),
)
conn.commit()
return workspace
def clear_brain_data(self, brain_id: str) -> None:
clear_brain_database(brain_id)
def delete_brain_storage(self, brain_id: str) -> None:
drop_brain_database(brain_id)
self._initialized_brains.discard(brain_id)
self._search_ready_brains.discard(brain_id)
self._search_alt_brains.discard(brain_id)
def delete_brain_registry(self, brain_id: str) -> bool:
with self._system_connection() as conn:
with conn.cursor() as cur:
cur.execute(
"DELETE FROM data_brains WHERE name_key = %s",
(brain_id,),
)
deleted = cur.rowcount > 0
conn.commit()
return deleted
def save_kg_changes(self, kg_changes: KGChanges, brain_id: str) -> KGChanges:
document = kg_changes.model_dump(mode="json")
type_value = (
kg_changes.type.value
if isinstance(kg_changes.type, KGChangesType)
else kg_changes.type
)
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
INSERT INTO data_kg_changes
(id, type, change, timestamp, document)
VALUES (%s, %s, %s::jsonb, %s, %s::jsonb)
ON CONFLICT (id) DO UPDATE SET
type = EXCLUDED.type,
change = EXCLUDED.change,
timestamp = EXCLUDED.timestamp,
document = EXCLUDED.document
""",
(
kg_changes.id,
type_value,
json.dumps(document.get("change") or {}, default=str),
kg_changes.timestamp,
json.dumps(document, default=str),
),
)
conn.commit()
return kg_changes
def get_structured_data_by_id(
self, id: str, brain_id: str
) -> Optional[StructuredData]:
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"""
SELECT document FROM data_structured_data
WHERE id = %s
""",
(id,),
)
row = cur.fetchone()
if not row:
return None
return StructuredData.model_validate(row["document"])
def get_structured_data_list(
self,
brain_id: str,
limit: int = 10,
skip: int = 0,
types: list[str] = None,
query_text: str = None,
) -> Tuple[list[StructuredData], int]:
clauses: list[str] = []
params: list[Any] = []
if types:
clauses.append("types && %s")
params.append(list(types))
if query_text:
pattern = _ilike_pattern(query_text)
clauses.append(
"(data::text ILIKE %s OR metadata::text ILIKE %s "
"OR EXISTS (SELECT 1 FROM unnest(types) t WHERE t ILIKE %s))"
)
params.extend([pattern, pattern, pattern])
where_sql = f"WHERE {' AND '.join(clauses)}" if clauses else ""
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
f"""
SELECT document FROM data_structured_data
{where_sql}
ORDER BY inserted_at DESC
OFFSET %s LIMIT %s
""",
(*params, skip, limit),
)
rows = cur.fetchall()
cur.execute(
f"SELECT COUNT(*) AS total FROM data_structured_data {where_sql}",
tuple(params),
)
total = cur.fetchone()["total"]
return (
[StructuredData.model_validate(row["document"]) for row in rows],
int(total or 0),
)
def get_text_chunks(
self,
brain_id: str,
limit: int = 10,
skip: int = 0,
query_text: str = None,
metadata_eq: dict[str, str] | None = None,
order: Literal["asc", "desc"] = "desc",
) -> Tuple[List[TextChunk], int]:
clauses: list[str] = []
params: list[Any] = []
if query_text:
pattern = _ilike_pattern(query_text)
clauses.append("(text ILIKE %s OR metadata::text ILIKE %s)")
params.extend([pattern, pattern])
if metadata_eq:
for key, value in metadata_eq.items():
clauses.append("metadata ->> %s = %s")
params.extend([key, value])
where_sql = f"WHERE {' AND '.join(clauses)}" if clauses else ""
order_sql = "ASC" if order == "asc" else "DESC"
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
f"""
SELECT document FROM data_text_chunks
{where_sql}
ORDER BY inserted_at {order_sql}
OFFSET %s LIMIT %s
""",
(*params, skip, limit),
)
rows = cur.fetchall()
cur.execute(
f"SELECT COUNT(*) AS total FROM data_text_chunks {where_sql}",
tuple(params),
)
total = cur.fetchone()["total"]
return (
[TextChunk.model_validate(row["document"]) for row in rows],
int(total or 0),
)
def get_structured_data_types(self, brain_id: str) -> list[str]:
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
SELECT DISTINCT unnest(types) AS t
FROM data_structured_data
WHERE types IS NOT NULL
ORDER BY t
"""
)
return [row[0] for row in cur.fetchall() if row[0] is not None]
def get_observation_by_id(self, id: str, brain_id: str) -> Optional[Observation]:
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"""
SELECT document FROM data_observations
WHERE id = %s
""",
(id,),
)
row = cur.fetchone()
if not row:
return None
return Observation.model_validate(row["document"])
def get_observations_list(
self,
brain_id: str,
limit: int = 10,
skip: int = 0,
resource_id: str = None,
labels: list[str] = None,
query_text: str = None,
) -> list[Observation]:
clauses: list[str] = []
params: list[Any] = []
if resource_id:
clauses.append("resource_id = %s")
params.append(resource_id)
if labels:
clauses.append("metadata -> 'labels' ?| %s")
params.append(list(labels))
if query_text:
pattern = _ilike_pattern(query_text)
clauses.append(
"(text ILIKE %s OR resource_id ILIKE %s OR metadata::text ILIKE %s)"
)
params.extend([pattern, pattern, pattern])
where_sql = f"WHERE {' AND '.join(clauses)}" if clauses else ""
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
f"""
SELECT document FROM data_observations
{where_sql}
ORDER BY inserted_at DESC
OFFSET %s LIMIT %s
""",
(*params, skip, limit),
)
rows = cur.fetchall()
return [Observation.model_validate(row["document"]) for row in rows]
def get_observation_labels(self, brain_id: str) -> list[str]:
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
SELECT DISTINCT jsonb_array_elements_text(metadata -> 'labels') AS label
FROM data_observations
WHERE metadata ? 'labels'
AND jsonb_typeof(metadata -> 'labels') = 'array'
ORDER BY label
"""
)
return [row[0] for row in cur.fetchall() if row[0] is not None]
def get_changelog_by_id(self, id: str, brain_id: str) -> Optional[KGChanges]:
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"""
SELECT document FROM data_kg_changes
WHERE id = %s
""",
(id,),
)
row = cur.fetchone()
if not row:
return None
return _validate_changelog(dict(row["document"]))
def get_changelogs_list(
self,
brain_id: str,
limit: int = 10,
skip: int = 0,
types: list[str] = None,
query_text: str = None,
) -> list[KGChanges]:
clauses: list[str] = []
params: list[Any] = []
if types:
clauses.append("type = ANY(%s)")
params.append(list(types))
if query_text:
pattern = _ilike_pattern(query_text)
clauses.append("(change::text ILIKE %s OR type ILIKE %s)")
params.extend([pattern, pattern])
where_sql = f"WHERE {' AND '.join(clauses)}" if clauses else ""
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
f"""
SELECT document FROM data_kg_changes
{where_sql}
ORDER BY timestamp DESC
OFFSET %s LIMIT %s
""",
(*params, skip, limit),
)
rows = cur.fetchall()
changelogs: list[KGChanges] = []
for row in rows:
try:
changelogs.append(_validate_changelog(dict(row["document"])))
except Exception as exc:
print(f"Error parsing changelog {row['document'].get('id')}: {exc}")
continue
return changelogs
def get_changelog_types(self, brain_id: str) -> list[str]:
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
SELECT DISTINCT type FROM data_kg_changes
WHERE type IS NOT NULL
ORDER BY type
"""
)
return [row[0] for row in cur.fetchall()]
def update_structured_data(
self, structured_data: StructuredData, brain_id: str
) -> StructuredData:
document = structured_data.model_dump(mode="json")
with self._brain_connection(brain_id) as conn:
with conn.cursor() as cur:
cur.execute(
"""
UPDATE data_structured_data SET
data = %s::jsonb,
types = %s,
metadata = %s::jsonb,
inserted_at = %s,
document = %s::jsonb
WHERE id = %s
""",
(
json.dumps(structured_data.data or {}, default=str),
list(structured_data.types or []),
json.dumps(structured_data.metadata or {}, default=str),
structured_data.inserted_at,
json.dumps(document, default=str),
structured_data.id,
),
)
conn.commit()
return structured_data
def get_last_text_chunks(self, brain_id: str, limit: int = 10) -> list[TextChunk]:
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"""
SELECT document FROM data_text_chunks
ORDER BY inserted_at DESC
LIMIT %s
""",
(limit,),
)
rows = cur.fetchall()
return [TextChunk.model_validate(row["document"]) for row in rows]
def get_last_structured_data(
self, brain_id: str, limit: int = 10
) -> list[StructuredData]:
with self._brain_connection(brain_id) as conn:
with conn.cursor(cursor_factory=psycopg2.extras.RealDictCursor) as cur:
cur.execute(
"""
SELECT document FROM data_structured_data
ORDER BY inserted_at DESC
LIMIT %s
""",
(limit,),
)
rows = cur.fetchall()
return [StructuredData.model_validate(row["document"]) for row in rows]
_postgresql_data_client: Optional[PostgreSQLDataClient] = None
def get_postgresql_data_client() -> PostgreSQLDataClient:
global _postgresql_data_client
if _postgresql_data_client is None:
_postgresql_data_client = PostgreSQLDataClient()
return _postgresql_data_client
|