Spaces:
Running
Running
File size: 13,805 Bytes
79b0bef | 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 | """
src/db/repository.py
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
Repository layer β all database queries live here.
The repository pattern keeps SQL out of the API routes and
business logic. Routes call repository functions; they never
build queries directly. This makes the API easy to test
(swap in a fake repository) and easy to change (queries are
in one place, not scattered across the codebase).
One class per table, following a consistent interface:
get(id) β single record or None
list(filters) β list of records
create(data) β new record
update(id, data) β modified record
delete(id) β bool (True if found and deleted)
All methods accept a Session argument so the caller controls
the transaction boundary β the repository never opens its own
session. This is intentional: it lets the API layer wrap
multiple repository calls in a single transaction.
ββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
"""
from __future__ import annotations
from sqlalchemy import func, select
from sqlalchemy.orm import Session, joinedload
from src.db.models import ClinicalNote, Entity, ICD10Mapping, ModelRun
from src.utils.logger import get_logger
logger = get_logger(__name__)
class NoteRepository:
"""Query interface for the clinical_notes table."""
@staticmethod
def get(session: Session, note_id: int) -> ClinicalNote | None:
"""Fetch a single note by primary key.
Args:
session: Active SQLAlchemy session.
note_id: Primary key of the note.
Returns:
The matching :class:`ClinicalNote`, or None if not found.
"""
return session.get(ClinicalNote, note_id)
@staticmethod
def get_by_source_id(
session: Session, source_id: str
) -> ClinicalNote | None:
"""Fetch a note by its source dataset identifier.
Used by the ETL loader to check for duplicates before inserting.
Args:
session: Active SQLAlchemy session.
source_id: The stable identifier from the source dataset.
Returns:
Matching note or None.
"""
return (
session.query(ClinicalNote)
.filter_by(source_id=source_id)
.first()
)
@staticmethod
def list_by_specialty(
session: Session,
specialty: str,
limit: int = 100,
offset: int = 0,
) -> list[ClinicalNote]:
"""Return notes filtered by medical specialty.
Args:
session: Active SQLAlchemy session.
specialty: Specialty string to filter on (case-insensitive).
limit: Maximum records to return.
offset: Number of records to skip (for pagination).
Returns:
List of :class:`ClinicalNote` objects.
"""
return (
session.query(ClinicalNote)
.filter(
func.lower(ClinicalNote.specialty) == specialty.lower()
)
.order_by(ClinicalNote.id)
.limit(limit)
.offset(offset)
.all()
)
@staticmethod
def list_by_severity(
session: Session,
severity: str,
limit: int = 100,
offset: int = 0,
) -> list[ClinicalNote]:
"""Return notes filtered by severity label.
Args:
session: Active SQLAlchemy session.
severity: One of ``"routine"``, ``"urgent"``, ``"critical"``.
limit: Maximum records to return.
offset: Records to skip.
Returns:
List of :class:`ClinicalNote` objects.
"""
return (
session.query(ClinicalNote)
.filter_by(severity=severity)
.order_by(ClinicalNote.id)
.limit(limit)
.offset(offset)
.all()
)
@staticmethod
def count_by_severity(session: Session) -> dict[str, int]:
"""Return the count of notes per severity label.
Used by the dashboard to show distribution charts.
Args:
session: Active SQLAlchemy session.
Returns:
Dict mapping severity label to count,
e.g. ``{"routine": 3200, "urgent": 1400, "critical": 399}``.
"""
rows = (
session.query(ClinicalNote.severity, func.count(ClinicalNote.id))
.group_by(ClinicalNote.severity)
.all()
)
return {row[0]: row[1] for row in rows if row[0]}
@staticmethod
def count_by_specialty(session: Session) -> dict[str, int]:
"""Return the count of notes per medical specialty.
Args:
session: Active SQLAlchemy session.
Returns:
Dict mapping specialty to note count, sorted descending.
"""
rows = (
session.query(ClinicalNote.specialty, func.count(ClinicalNote.id))
.group_by(ClinicalNote.specialty)
.order_by(func.count(ClinicalNote.id).desc())
.all()
)
return {row[0]: row[1] for row in rows if row[0]}
@staticmethod
def total(session: Session) -> int:
"""Return the total number of notes in the database.
Args:
session: Active SQLAlchemy session.
Returns:
Integer count.
"""
return session.query(func.count(ClinicalNote.id)).scalar() or 0
class EntityRepository:
"""Query interface for the entities table."""
@staticmethod
def get(session: Session, entity_id: int) -> Entity | None:
"""Fetch a single entity by primary key.
Args:
session: Active SQLAlchemy session.
entity_id: Primary key.
Returns:
The matching :class:`Entity`, or None.
"""
return session.get(Entity, entity_id)
@staticmethod
def list_by_note(
session: Session,
note_id: int,
) -> list[Entity]:
"""Return all entities extracted from a specific note.
Eager-loads ICD-10 mappings so callers don't trigger
additional queries per entity.
Args:
session: Active SQLAlchemy session.
note_id: Primary key of the parent note.
Returns:
List of :class:`Entity` objects with mappings pre-loaded.
"""
return (
session.query(Entity)
.options(joinedload(Entity.icd10_mappings))
.filter_by(note_id=note_id)
.order_by(Entity.start_char)
.all()
)
@staticmethod
def list_by_label(
session: Session,
label: str,
limit: int = 500,
) -> list[Entity]:
"""Return entities filtered by entity type label.
Args:
session: Active SQLAlchemy session.
label: Entity type to filter on (e.g. ``"DISEASE"``).
limit: Maximum records to return.
Returns:
List of :class:`Entity` objects.
"""
return (
session.query(Entity)
.filter(func.upper(Entity.label) == label.upper())
.limit(limit)
.all()
)
@staticmethod
def top_entities(
session: Session,
label: str | None = None,
limit: int = 20,
) -> list[tuple[str, int]]:
"""Return the most frequently extracted entity texts.
Used by the dashboard frequency chart.
Args:
session: Active SQLAlchemy session.
label: If provided, filter to this entity type only.
limit: Number of top entities to return.
Returns:
List of ``(entity_text, count)`` tuples, sorted by count
descending.
Example::
top = EntityRepository.top_entities(session, label="DISEASE", limit=10)
# β [("hypertension", 412), ("diabetes", 389), ...]
"""
q = (
session.query(Entity.text, func.count(Entity.id).label("n"))
.group_by(Entity.text)
.order_by(func.count(Entity.id).desc())
)
if label:
q = q.filter(func.upper(Entity.label) == label.upper())
return [(row.text, row.n) for row in q.limit(limit).all()]
@staticmethod
def cooccurrence_pairs(
session: Session,
label: str = "DISEASE",
min_count: int = 5,
limit: int = 200,
) -> list[tuple[str, str, int]]:
"""Return entity pairs that appear together in the same note.
Used to build the co-occurrence network graph.
Args:
session: Active SQLAlchemy session.
label: Entity type to consider.
min_count: Minimum co-occurrence count to include a pair.
limit: Maximum number of pairs to return.
Returns:
List of ``(entity_a, entity_b, count)`` tuples.
"""
# Self-join on note_id to find pairs within the same note
e1 = Entity.__table__.alias("e1")
e2 = Entity.__table__.alias("e2")
stmt = (
select(
e1.c.text.label("entity_a"),
e2.c.text.label("entity_b"),
func.count().label("n"),
)
.select_from(e1)
.join(e2, (e1.c.note_id == e2.c.note_id) & (e1.c.text < e2.c.text))
.where(
func.upper(e1.c.label) == label.upper(),
func.upper(e2.c.label) == label.upper(),
)
.group_by(e1.c.text, e2.c.text)
.having(func.count() >= min_count)
.order_by(func.count().desc())
.limit(limit)
)
rows = session.execute(stmt).fetchall()
return [(r.entity_a, r.entity_b, r.n) for r in rows]
class ICD10Repository:
"""Query interface for the icd10_mappings table."""
@staticmethod
def list_by_entity(
session: Session,
entity_id: int,
) -> list[ICD10Mapping]:
"""Return all ICD-10 mappings for a given entity.
Results are ordered by rank (best match first).
Args:
session: Active SQLAlchemy session.
entity_id: Primary key of the entity.
Returns:
List of :class:`ICD10Mapping` objects.
"""
return (
session.query(ICD10Mapping)
.filter_by(entity_id=entity_id)
.order_by(ICD10Mapping.rank)
.all()
)
@staticmethod
def top_codes(
session: Session,
limit: int = 20,
) -> list[tuple[str, str, int]]:
"""Return the most frequently assigned ICD-10 codes.
Used by the dashboard to show which diagnoses appear most
often across the dataset.
Args:
session: Active SQLAlchemy session.
limit: Number of top codes to return.
Returns:
List of ``(icd10_code, description, count)`` tuples.
"""
rows = (
session.query(
ICD10Mapping.icd10_code,
ICD10Mapping.description,
func.count(ICD10Mapping.id).label("n"),
)
.group_by(ICD10Mapping.icd10_code, ICD10Mapping.description)
.order_by(func.count(ICD10Mapping.id).desc())
.limit(limit)
.all()
)
return [(r.icd10_code, r.description, r.n) for r in rows]
class ModelRunRepository:
"""Query interface for the model_runs table."""
@staticmethod
def create(
session: Session,
model_name: str,
task: str,
**kwargs,
) -> ModelRun:
"""Record a new classifier training run.
Args:
session: Active SQLAlchemy session.
model_name: Base model used for fine-tuning.
task: Classification task name.
**kwargs: Optional columns β ``val_accuracy``, ``val_f1``,
``training_samples``, ``epochs``, ``run_notes``.
Returns:
The newly created :class:`ModelRun` record.
"""
run = ModelRun(model_name=model_name, task=task, **kwargs)
session.add(run)
session.flush() # populate run.id without committing
logger.info(
"Model run recorded: %s / %s (id=%d)", model_name, task, run.id
)
return run
@staticmethod
def get_deployed(session: Session, task: str) -> ModelRun | None:
"""Return the currently deployed model run for a task.
Args:
session: Active SQLAlchemy session.
task: Task name to look up (e.g. ``"severity"``).
Returns:
The deployed :class:`ModelRun`, or None if none is marked.
"""
return (
session.query(ModelRun)
.filter_by(task=task, is_deployed=True)
.order_by(ModelRun.created_at.desc())
.first()
)
@staticmethod
def list_by_task(
session: Session,
task: str,
limit: int = 10,
) -> list[ModelRun]:
"""Return recent training runs for a task, newest first.
Args:
session: Active SQLAlchemy session.
task: Task name.
limit: Maximum runs to return.
Returns:
List of :class:`ModelRun` objects.
"""
return (
session.query(ModelRun)
.filter_by(task=task)
.order_by(ModelRun.created_at.desc())
.limit(limit)
.all()
)
|