Merge pull request #15 from prentice-jobs/feat/integration-ml

Feat: RecSys integration to Backend

.env.example

@@ -7,4 +7,7 @@ POSTGRES_DB_NAME=
 POSTGRES_DB_USER=
 POSTGRES_DB_PASSWORD=
 
-GCS_BUCKET_OFFER_LETTER=
+# CLOUD STORAGE
+GCS_BUCKET_OFFER_LETTER=
+GCS_BUCKET_STOPWORDS=
+GCS_BUCKET_RECSYS=

alembic/versions/0b7939b453d5_create_recsys_sim_scores_and_.py

@@ -0,0 +1,69 @@
+"""create recsys sim_scores and recommendation cache tables
+
+Revision ID: 0b7939b453d5
+Revises: fa73e6dafaad
+Create Date: 2024-06-18 13:34:44.532365
+
+"""
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+
+# revision identifiers, used by Alembic.
+revision: str = '0b7939b453d5'
+down_revision: Union[str, None] = 'fa73e6dafaad'
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    op.create_table(
+        'user_review_similarity_scores',
+        sa.Column('id', sa.UUID(), autoincrement=False),
+        sa.Column('created_at', sa.TIMESTAMP(timezone=True)),
+        sa.Column('updated_at', sa.TIMESTAMP(timezone=True)),
+        sa.Column('is_deleted', sa.Boolean()),
+
+        sa.Column('user_id', sa.UUID()),
+        sa.Column('review_id', sa.UUID()),
+        sa.Column('sim_score', sa.Float(precision=3)),
+
+        sa.PrimaryKeyConstraint('id'),
+        sa.ForeignKeyConstraint(
+            ["user_id"], ["users.id"],
+            name="fk_sim_scores_user",
+        ),
+        sa.ForeignKeyConstraint(
+            ["review_id"], ["company_reviews.id"],
+            name="fk_sim_scores_review",
+        )
+    )
+
+    op.create_table(
+        'user_review_recommendations_cache',
+        sa.Column('id', sa.UUID(), autoincrement=False),
+        sa.Column('created_at', sa.TIMESTAMP(timezone=True)),
+        sa.Column('updated_at', sa.TIMESTAMP(timezone=True)),
+        sa.Column('is_deleted', sa.Boolean()),
+
+        sa.Column('user_id', sa.UUID()),
+        sa.Column('review_id', sa.UUID()),
+        sa.Column('sim_score', sa.Float(precision=3)),
+
+        sa.PrimaryKeyConstraint('id'),
+        sa.ForeignKeyConstraint(
+            ["user_id"], ["users.id"],
+            name="fk_sim_scores_user",
+        ),
+        sa.ForeignKeyConstraint(
+            ["review_id"], ["company_reviews.id"],
+            name="fk_sim_scores_review",
+        )
+    )
+
+
+def downgrade() -> None:
+    op.drop_table('user_review_recommendations_cache')
+    op.drop_table('user_review_similarity_scores')

alembic/versions/61c33e0f05b6_add_user_and_userpreferences_model.py

@@ -43,9 +43,9 @@ def upgrade() -> None:
         sa.Column('updated_at', sa.TIMESTAMP(timezone=True)),
         sa.Column('is_deleted', sa.Boolean()),
 
-        sa.Column('role', sa.String(255), nullable=True),
-        sa.Column('industry', sa.String(255), nullable=True),
-        sa.Column('location', sa.String(255), nullable=True),
+        sa.Column('role', sa.String(255)),
+        sa.Column('industry', sa.String(255)),
+        sa.Column('location', sa.String(255)),
 
         sa.Column('user_id', sa.UUID(), nullable=False),
 

alembic/versions/d5d068b44817_add_company_reviews_table.py

@@ -44,6 +44,8 @@ def upgrade() -> None:
         sa.Column('offer_letter_url', sa.String()),
         sa.Column('annual_salary', sa.BigInteger()),
         sa.Column('salary_currency', sa.String(3)),
+
+        sa.PrimaryKeyConstraint('id'),
     )
 
 

alembic/versions/fa73e6dafaad_create_review_sentiments_review_.py

@@ -29,6 +29,8 @@ def upgrade() -> None:
         sa.Column('review_id', sa.UUID()),
         sa.Column('sentiment_score', sa.SmallInteger()),
         sa.Column('sentiment', sa.String(30)),
+
+        sa.PrimaryKeyConstraint('id'),
     )
 
     op.create_table(
@@ -41,7 +43,9 @@ def upgrade() -> None:
         sa.Column('review_id', sa.UUID()),
         sa.Column('author_id', sa.UUID()),
         sa.Column('likes_count', sa.Integer()),
-        sa.Column('content', sa.String(1000))
+        sa.Column('content', sa.String(1000)),
+
+        sa.PrimaryKeyConstraint('id'),
     )
 
     op.create_table(
@@ -53,6 +57,8 @@ def upgrade() -> None:
 
         sa.Column('review_comment_id', sa.UUID()),
         sa.Column('liker_id', sa.UUID()),
+
+        sa.PrimaryKeyConstraint('id'),
     )
 
 

requirements.txt

@@ -32,13 +32,16 @@ httptools==0.6.1
 httpx==0.27.0
 idna==3.7
 Jinja2==3.1.4
+joblib==1.4.2
 logger==1.4
 Mako==1.3.5
 markdown-it-py==3.0.0
 MarkupSafe==2.1.5
 mdurl==0.1.2
 msgpack==1.0.8
+numpy==2.0.0
 orjson==3.10.3
+pandas==2.2.2
 proto-plus==1.23.0
 protobuf==4.25.3
 psycopg2==2.9.9
@@ -50,19 +53,25 @@ pydantic_core==2.18.2
 Pygments==2.18.0
 PyJWT==2.8.0
 pyparsing==3.1.2
+python-dateutil==2.9.0.post0
 python-dotenv==1.0.1
 python-multipart==0.0.9
 pytz==2024.1
 PyYAML==6.0.1
 requests==2.32.3
 rich==13.7.1
 rsa==4.9
+scikit-learn==1.5.0
+scipy==1.13.1
 shellingham==1.5.4
+six==1.16.0
 sniffio==1.3.1
 SQLAlchemy==2.0.30
 starlette==0.37.2
+threadpoolctl==3.5.0
 typer==0.12.3
 typing_extensions==4.11.0
+tzdata==2024.1
 ujson==5.10.0
 uritemplate==4.1.1
 urllib3==2.2.1

src/account/constants/messages.py

@@ -1,4 +1,5 @@
-SERVICE_CREATE_USER_SUCCESS = "Successfully created user."
+CREATE_NEW_USER_SUCCESS = "Successfully created user."
+SAVE_USER_PREFERENCES_SUCCESS = "Successfully saved user preferences."
 
 UNAUTHORIZED_ACTION = "Unauthorized: Failed to perform this operation."
 UNAUTHORIZED_ACTION_RECOMMENDATION = "You are not logged in. Try logging in with the required permissions."

src/account/controller.py

@@ -24,13 +24,16 @@
 from src.account.schema import  (
     CheckUserRegisteredSchema,
     RegisterSchema,
+    UserPreferencesSchema,
 )
 
 from src.account.service import AccountService
 from src.account.security import get_current_user
 from src.account.exceptions import (
     UserAlreadyExistsException,
-    RegistrationFailedException
+    RegistrationFailedException,
+    SavePreferencesFailedException,
+    UserPreferencesAlreadyExistsException,
 )
 
 VERSION = "v1"
@@ -44,7 +47,7 @@
 @account_router.post("/exists", status_code=HTTPStatus.OK, response_model=bool)
 def check_user_registered(
     payload: CheckUserRegisteredSchema = Body(),
-    session = Depends(get_db),
+    session: Session = Depends(get_db),
 ):
     is_registered = AccountService.check_user_is_registered(
         session=session, 
@@ -56,7 +59,7 @@ def check_user_registered(
 @account_router.post("/register", status_code=HTTPStatus.CREATED, response_model=GenericAPIResponseModel)
 def register(
     payload: RegisterSchema = Body(),
-    session = Depends(get_db),
+    session: Session = Depends(get_db),
 ):
     try:
         response: GenericAPIResponseModel = AccountService.register_user(session=session, payload=payload)
@@ -81,6 +84,37 @@ def register(
 
         return build_api_response(response)
 
+@account_router.post("/preferences", status_code=HTTPStatus.OK, response_model=GenericAPIResponseModel)
+def save_user_preferences(
+    payload: UserPreferencesSchema = Body(),
+    session: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+):
+    try:
+        response: GenericAPIResponseModel = AccountService.save_user_preferences(
+            payload=payload,
+            session=session,
+            user=user,
+        )
+
+        return build_api_response(response)
+    except UserPreferencesAlreadyExistsException as err:
+        response = GenericAPIResponseModel(
+            status=HTTPStatus.CONFLICT,
+            message=err.message,
+            error=err.message,
+        )
+
+        return build_api_response(response)
+    except SavePreferencesFailedException as err:
+        response = GenericAPIResponseModel(
+            status=HTTPStatus.INTERNAL_SERVER_ERROR,
+            message=err.message,
+            error=err.message,
+        )
+
+        return build_api_response(response)
+
 @account_router.get("/", status_code=HTTPStatus.OK)
 def fetch_user_info(
     prentice_user: User = Depends(get_current_user),

src/account/exceptions.py

@@ -21,6 +21,16 @@ def __init__(self, message="Error while registering new user"):
         super().__init__(message)
         self.message = message
 
+class UserPreferencesAlreadyExistsException(Exception):
+    def __init__(self, message="User already has a preference setting!"):
+        super().__init__(message)
+        self.message = message
+
+class SavePreferencesFailedException(Exception):
+    def __init__(self, message="Error while saving user preferences"):
+        super().__init__(message)
+        self.message = message
+
 class FirebaseTokenVerificationException(HTTPException):
     def __init__(
         self, 

src/account/model.py

@@ -41,10 +41,9 @@ class User(PrenticeBaseModel):
 class UserPreferences(PrenticeBaseModel):
     __tablename__ = "user_preferences"
 
-    role = Column(String(255), nullable=True)
-    industry = Column(String(255), nullable=True)
-    location = Column(String(255), nullable=True)
-    is_active = Column(Boolean)
+    role = Column(String(255))
+    industry = Column(String(255))
+    location = Column(String(255))
 
     user_id = Column(UUID(as_uuid=True), ForeignKey("users.id"))
     user = relationship(

src/account/schema.py

@@ -1,9 +1,11 @@
 from datetime import datetime
 from typing import Optional
+from typing_extensions import Annotated
 from pydantic import (
     UUID4,
     BaseModel,
-    EmailStr
+    EmailStr,
+    Field,
 )
 
 from src.core.schema import (
@@ -20,6 +22,16 @@ class RegisterSchema(BaseModel):
     photo_url: Optional[str]
     email_verified: bool
 
+class UserPreferencesSchema(BaseModel):
+    # https://stackoverflow.com/questions/61326020/how-can-i-set-max-string-field-length-constraint-in-pydantic
+    # Setting max length to avoid DB insertion errors
+    role: str = Field(..., max_length=255) 
+    industry: str = Field(..., max_length=255)
+    location: str = Field(..., max_length=255)
+
+class UserPreferencesResponseSchema(UserPreferencesSchema, PrenticeBaseSchema):
+    user_id: UUID4
+
 class RegisterResponseSchema(BaseModel):
     email: EmailStr
     created_at: datetime