Add authenticate function to clustering API

.gitignore

@@ -11,6 +11,8 @@ postgres_data/
 *.key
 *.pub
 
+*/**/.rpc_secret
+
 # Sqlite extensions
 *.sqlite
 *.sqlite-shm

docker-compose.yaml

@@ -2,6 +2,7 @@ version: '3.7'
 
 services:
   app: 
+    # platform: linux/x86_64
     build: service_app/
     depends_on:
       - db

service_app/Dockerfile

@@ -8,18 +8,21 @@ ENV PORT=4000
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
-FROM python:3.10-alpine AS build-image
+FROM python:3.10-slim-buster AS build-image
 WORKDIR /app
 COPY --from=compile-image /opt/venv /opt/venv
-RUN apk upgrade --no-cache && \
-    apk add --no-cache postgresql-client bash openssl libgcc libstdc++ ncurses-libs
-# COPY ./src .
+# RUN apk upgrade --no-cache && \
+#     apk add --no-cache postgresql-client bash openssl libgcc libstdc++ ncurses-libs
 # COPY .env .
 
 # Make sure we use the virtualenv:
 ENV PATH="/opt/venv/bin:$PATH"
 # ENV DATABASE_URL="sqlite://db.sqlite"
 # EXPOSE 4000
 #
-CMD uvicorn app.api:app --reload --host 0.0.0.0 --port $PORT
-
+# CMD uvicorn app.api:app --reload --host 0.0.0.0 --port $PORT
+#
+# FROM python:3.10
+# COPY requirements.txt .
+# RUN pip install --no-cache-dir -r requirements.txt
+# COPY ./src .

service_app/requirements.txt

@@ -14,3 +14,6 @@ python-multipart
 facebook-scraper
 firebase-admin<7
 pyrebase4
+joblib<=1.1.0
+bertopic
+textherox==1.2.0

service_app/src/app/api.py

@@ -21,7 +21,7 @@
     DatapointOut_Pydantic,
     )
 
-from .db.schemas import DualResourceActionResponse, SingleResourceActionResponse, CreateDatasetBody, SignUpRequestBody, SignInRequestBody
+from .db.schemas import DualResourceActionResponse, SingleResourceActionResponse, CreateDatasetBody, SignUpRequestBody, SignInRequestBody, LineByLineTextInput, VerifyTokenInput
 from .auth.security import  get_password_hash
 from tortoise.contrib.fastapi import register_tortoise
 from dotenv import load_dotenv
@@ -34,15 +34,19 @@
 import pyrebase
 import json
 
-from typing import List
+from typing import List, Dict
+from bertopic import BERTopic
+import texthero as hero
+import pandas as pd
 
 load_dotenv()  # take environment variables from .env.
 
 app = FastAPI()
 
 # Allowed origins
 origins = [
-    "http://localhost:3000",
+    "http://127.0.0.1:3333",
+    "http://localhost:3333",
 ]
 
 app.add_middleware(
@@ -64,6 +68,11 @@
 
 # Setup supabase authentication
 
+def api_authen(token):
+    user = supabase.auth.api.get_user(jwt=token)
+    if user.aud and user.aud == "authenticated":
+        return True
+    return False
 
 # signup endpoint
 @app.post("/signup")
@@ -89,6 +98,31 @@ async def signup(req: SignUpRequestBody):
     except Exception as e:
         return HTTPException(detail={'message': str(e)}, status_code=400)
 
+@app.post("/api/v1/user")
+async def get_user_base_on_token(req: VerifyTokenInput):
+    token: str = req.token
+    user =  supabase.auth.api.get_user(jwt=token)
+    return {"aud": user.aud, "email": user.email }  
+
+    # Only required email for signup at this time
+    # password: str = 
+    #
+    # if email is None or password is None:
+    #    return HTTPException(detail={'message': 'Error! Missing Email or Password'}, status_code=400)
+    # try:
+    #    user = auth.create_user(
+    #        email=email,
+    #        password=password
+    #    )
+    #    return JSONResponse(content={'message': f'Successfully created user {user.uid}'}, status_code=200)    
+    # except Exception as e:
+    #     return HTTPException(detail={'message': str(e)}, status_code=400)
+
+    try:
+        user: Dict[str, Any] = supabase.auth.sign_up(email=email)
+        return JSONResponse(content={'message': f'Successfully created user {user.uid}'}, status_code=200)    
+    except Exception as e:
+        return HTTPException(detail={'message': str(e)}, status_code=400)
 
 @app.post("/signin")
 async def signin(request: SignInRequestBody):
@@ -107,13 +141,10 @@ async def signin(request: SignInRequestBody):
     except Exception as e:
         return HTTPException(detail={'message': str(e)}, status_code=400)
 
-@app.post("/ping")
-async def validate(request: Request):
-   headers = request.headers
-   jwt = headers.get('authorization')
-   print(f"jwt:{jwt}")
-   user = auth.verify_id_token(jwt)
-   return user["uid"]
+@app.get("/api/v1/ping")
+async def ping(token: str = Depends(oauth2_scheme)):
+    return {"result": "Cool"}
+
 
 @app.post("/login", include_in_schema=False)
 async def login(request: SignUpRequestBody):
@@ -133,11 +164,10 @@ async def read_root():
     return "Please go to /docs to read the documentation"
 
 @app.post("/api/v1/workflows", response_model = WorkflowOut_Pydantic)
-async def create_workflow(workflow: WorkflowIn_Pydantic, token: str = Depends(check_auth)):
+async def create_workflow(workflow: WorkflowIn_Pydantic):
     """
     Create a new Workflow
     """
-    supabase.
     workflow_obj = await Workflows.create(**workflow.dict(exclude_unset=True))
     return await WorkflowOut_Pydantic.from_tortoise_orm(workflow_obj)
 
@@ -265,6 +295,47 @@ async def list_datapoints_from_dataset(dataset_id, token: str = Depends(oauth2_s
     return queryset
 
 
+@app.post("/api/v1/apps/topic_model/process")
+async def run_process_clustering(req: VerifyTokenInput, payload: LineByLineTextInput):
+    """
+    Process a request data
+    """
+    authenticated = api_authen(req.token)
+    if authenticated:
+        topic_model = BERTopic(embedding_model="all-MiniLM-L6-v2")
+        docs = payload.text_data.split("\n")
+
+        # Preprocess
+        clean_docs = pd.Series(docs).pipe(hero.clean)
+
+        topics, probs = topic_model.fit_transform(clean_docs)
+
+        num_topics = max(topics)
+        sentences_by_topics = {topic_id: {"data":[],"topic": []} for topic_id in range(num_topics+1)}
+
+        for doc, topic_id in zip(docs, topics):
+            # topic "-1" are stopwords
+            if topic_id == -1: continue
+            sentences_by_topics[topic_id]["data"].append(doc)
+
+        for topic_id in range(num_topics+1):
+            # Get keywords
+            #keywords = topic_model.get_topic_info(topic_id).Name.values[0].split("_")[1:]
+            topic = topic_model.get_topic(topic_id)
+
+            # convert to percentage
+            topic = [(t[0].capitalize(), round(t[1]*100,2)) for t in topic]
+            sentences_by_topics[topic_id]["topic"] = topic
+            sentences_by_topics[topic_id]["count"] = len(sentences_by_topics[topic_id]["data"]) 
+
+
+        # Change to list
+        sentences_by_topics = [topic_items for (_, topic_items) in sentences_by_topics.items()]
+
+        return sentences_by_topics 
+    return []
+
+
     # return await Datapoints.all().prefetch_related(Prefetch(dataset, queryset))
 # 
 #### POSTPONED UNTIL HAVE USERS MANAGEMENT

service_app/src/app/db/schemas.py

@@ -25,3 +25,13 @@ class SignInRequestBody(BaseModel):
     email: str
     password: str
 
+class LineByLineTextInput(BaseModel):
+    """
+    This is first line\nThis is second line
+
+    """
+    text_data: str
+
+class VerifyTokenInput(BaseModel):
+    token: str
+

service_auth/.rpc_secret

@@ -1 +1 @@
-hevenuzchkgl
+bhnfcbucmxox

service_auth/quorum.txt

@@ -1 +1 @@
-43635c40-6ba3-4f7b-ad96-c31f15ce69af
+ecaedbb5-7b61-4955-8c44-89beaccb1273