Merge branch 'main' of gitlab.prodcontest.com:team-39/backend

* 'main' of gitlab.prodcontest.com:team-39/backend:
  fix(): fix e2e ml tests, handle  no entries for vacancies

compose.yaml

@@ -83,14 +83,12 @@ services:
     profiles:
       - ml
     ports:
-      - 8081:8080
-    # ports:
-    #   - name: web
-    #     target: 8081
-    #     published: 13562
-    #     host_ip: 127.0.0.1
-    #     protocol: tcp
-        # app_protocol: http
+      - name: web
+        target: 8081
+        published: 13562
+        host_ip: 127.0.0.1
+        protocol: tcp
+        app_protocol: http
     restart: unless-stopped
     shm_size: 4mb
     volumes:

pyproject.toml

@@ -233,4 +233,7 @@ omit = [
     '*/__about__.py',
     '*/__main__.py',
     '*/__init__.py',
+    'src/dataset/*',
+    'src/template_project/ml/*',
+    'src/template_project/application/resume/interactors/predict_model.py',
 ]

src/template_project/adapters/data_gateways/tables.py

@@ -22,7 +22,7 @@ from sqlalchemy.orm import registry
 
 from template_project.application.access_token.entity import AccessToken
 from template_project.application.auth_identity.entity import AuthIdentity, AuthMethod
-from template_project.application.common.enums import EducationGrade, ExperienceType, ExperienceType
+from template_project.application.common.enums import EducationGrade, ExperienceType
 from template_project.application.notification_device.entity import NotificationDevice
 from template_project.application.resume.entity import (
     Resume,

src/template_project/adapters/ml_api_gateway.py

@@ -57,6 +57,7 @@ class MlApiGateway:
             timeout=100,
         )
 
+        response.raise_for_status()
         response_json = response.json()
         return GenerateResumePredictionResponse(
             salary_from=Decimal(str(response_json["salary_from"])),

src/template_project/application/resume/interactors/predict_model.py

@@ -46,7 +46,7 @@ class PredictModelInteractor:
 
     def _predict_salary(self, vacancies: list[VacancyInput], resume_skills: list[str]) -> tuple[Decimal, Decimal]:
         if not vacancies:
-            return Decimal(50000), Decimal(80000)
+            return Decimal(0), Decimal(0)
 
         vacancy_weights: list[float] = []
         for vacancy in vacancies:
@@ -56,7 +56,7 @@ class PredictModelInteractor:
 
         total_weight = sum(vacancy_weights)
         if total_weight == 0:
-            return Decimal(50000), Decimal(80000)
+            return Decimal(0), Decimal(0)
 
         weighted_from_sum = Decimal(0)
         weighted_to_sum = Decimal(0)
@@ -143,6 +143,9 @@ class PredictModelInteractor:
             if skill in candidate_skills
         }
 
+        if not candidate_skills:
+            return []
+
         frequencies = [skill_frequencies[skill] for skill in candidate_skills]
         avg_salaries = [float(skill_avg_salaries[skill]) for skill in candidate_skills]
 

src/template_project/ml/entry_point.py

@@ -3,6 +3,8 @@ import asyncio
 import logging
 import os
 import sys
+from collections.abc import AsyncIterator
+from contextlib import asynccontextmanager
 from pathlib import Path
 from typing import Final
 
@@ -12,6 +14,7 @@ from dishka.integrations.fastapi import setup_dishka
 from fastapi import FastAPI
 from fastapi.middleware.cors import CORSMiddleware
 from prometheus_fastapi_instrumentator import Instrumentator
+from sentence_transformers import SentenceTransformer
 
 from template_project.ml.configuration import load_configuration
 from template_project.ml.ioc.make import make_ioc
@@ -38,10 +41,18 @@ LOG_CONFIG: Final = {
 }
 
 
+@asynccontextmanager
+async def lifespan(app: FastAPI) -> AsyncIterator[None]:
+    await app.state.dishka_container.get(SentenceTransformer)
+    yield
+    await app.state.dishka_container.close()
+
+
 def make_asgi_application(
     ioc: AsyncContainer,
 ) -> FastAPI:
     app = FastAPI(
+        lifespan=lifespan,
         docs_url="/docs",
         title="ML Service",
         description="ML Service API",
@@ -83,10 +94,7 @@ async def _main(
         access_log=configuration.server.access_log,
     )
     server = uvicorn.Server(config)
-    try:
     await server.serve()
-    finally:
-        await ioc.close()
 
 
 def main() -> None:

src/template_project/ml/routes/predict.py

@@ -42,7 +42,7 @@ class PredictSalaryRequestModel(BaseModel):
     key_skills: list[str] = Field(
         description="List of key skills from resume", examples=[["Python", "FastAPI", "PostgreSQL"]]
     )
-    vacancies: list[VacancyInputModel] = Field(description="List of relevant vacancies", examples=[[]])
+    vacancies: list[VacancyInputModel] = Field(description="List of relevant vacancies", examples=[[]], min_length=0)
 
     model_config = {
         "json_schema_extra": {

tests/web_api/e2e/test_resume.py

@@ -1,7 +1,6 @@
 from typing import Final
 
-import pytest
-from dirty_equals import IsDict, IsPartialDict, IsUUID
+from dirty_equals import IsDict, IsOneOf, IsPartialDict, IsUUID
 from dishka import FromDishka
 from uuid_utils.compat import uuid7
 
@@ -17,7 +16,6 @@ from tests.web_api.test_api_gateway import TestApiGateway
 DEFAULT_PASSWORD: Final = "Sup3rSecret"  # noqa: S105
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_success_add_resume(
     unique_email: str,
@@ -43,7 +41,6 @@ async def test_success_add_resume(
     )
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_unauthorized_add_resume(
     unique_email: str,
@@ -63,7 +60,6 @@ async def test_unauthorized_add_resume(
     assert is_unauthorized_response(response)
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_success_get_resume(
     unique_email: str,
@@ -89,7 +85,8 @@ async def test_success_get_resume(
         resume_id=response.json()["resume_id"],
     )
     assert is_success_response(response)
-    assert response.json() == IsPartialDict(
+    json_response = response.json()
+    assert json_response == IsPartialDict(
         position="Position",
         location="Moscow",
         about_me="About me",
@@ -98,11 +95,17 @@ async def test_success_get_resume(
         experience=[],
         education=[],
         projects=[],
-        prediction=None,
+    )
+    assert json_response["prediction"] == IsOneOf(
+        None,
+        IsPartialDict(
+            from_salary="0",
+            to_salary="0",
+            recommended_skills=[],
+        ),
     )
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_unauthorized_get_resume(
     unique_email: str,
@@ -148,7 +151,6 @@ async def test_not_found_get_resume(
     assert is_not_found_response(response)
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_success_edit_resume(
     unique_email: str,
@@ -191,7 +193,6 @@ async def test_success_edit_resume(
     )
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_unauthorized_edit_resume(
     unique_email: str,
@@ -245,7 +246,6 @@ async def test_not_found_edit_resume(
     assert is_not_found_response(response)
 
 
-@pytest.mark.skip(reason="Requires ML service")
 @inject
 async def test_forbidden_edit_resume(
     unique_email: str,