추가: ocr 분석 준비 및 배포

.github/workflows/deploy.yml

@@ -1,5 +1,6 @@
 name: CI/CD Pipeline
 
+# Trigger the workflow on push events to the main branch
 on:
   push:
     branches:
@@ -10,37 +11,50 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-    - name: Checkout code
-      uses: actions/checkout@v2
-
-    - name: Set up Docker Buildx
-      uses: docker/setup-buildx-action@v1
-
-    - name: Login to Amazon ECR
-      id: login-ecr
-      uses: aws-actions/amazon-ecr-login@v1
-      with:
-        region: ap-northeast-2
-        access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
-        secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
-
-    - name: Build and push Docker image
-      id: build-image
-      run: |
-        IMAGE_TAG=$(date +%Y%m%d%H%M)
-        docker build -t ${{ secrets.ECR_REPOSITORY }}:$IMAGE_TAG .
-        docker tag ${{ secrets.ECR_REPOSITORY }}:$IMAGE_TAG ${{ secrets.ECR_REPOSITORY }}:latest
-        docker push ${{ secrets.ECR_REPOSITORY }}:$IMAGE_TAG
-        docker push ${{ secrets.ECR_REPOSITORY }}:latest
-      env:
-        DATABASE_URL: ${{ secrets.DATABASE_URL }}
-
-    - name: Deploy to EC2
-      run: |
-        ssh -o StrictHostKeyChecking=no -i ${{ secrets.EC2_SSH_KEY }} ec2-user@${{ secrets.EC2_INSTANCE_IP }} << 'EOF'
-        docker pull ${{ secrets.ECR_REPOSITORY }}:latest
-        docker-compose down
-        docker-compose up -d
-        EOF
-      env:
-        DATABASE_URL: ${{ secrets.DATABASE_URL }}
+      # Step 1: Check out the code from the repository
+      - name: Checkout code
+        uses: actions/checkout@v2
+
+      # Step 2: Set up Docker Buildx
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v1
+
+      # Step 3: Configure AWS credentials for the workflow
+      - name: Configure AWS credentials
+        uses: aws-actions/configure-aws-credentials@v1
+        with:
+          aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+          aws-region: ap-northeast-2
+
+      # Step 4: Log in to Amazon ECR (Elastic Container Registry)
+      - name: Login to Amazon ECR
+        run: |
+          aws ecr-public get-login-password --region us-east-1 | docker login --username AWS --password-stdin public.ecr.aws
+
+      # Step 5: Build and push Docker image to Amazon ECR
+      - name: Build and push Docker image
+        id: build-image
+        run: |
+          IMAGE_TAG=$(date +%Y%m%d%H%M)
+          docker build -t public.ecr.aws/h7p2f6d8/ticats_ai:$IMAGE_TAG .
+          docker tag public.ecr.aws/h7p2f6d8/ticats_ai:$IMAGE_TAG public.ecr.aws/h7p2f6d8/ticats_ai:latest
+          docker push public.ecr.aws/h7p2f6d8/ticats_ai:$IMAGE_TAG
+          docker push public.ecr.aws/h7p2f6d8/ticats_ai:latest
+        env:
+          DATABASE_URL: ${{ secrets.DATABASE_URL }}
+
+      # Step 6: Deploy to EC2 instance
+      - name: Run SSH Command
+        uses: appleboy/ssh-action@master
+        with:
+          host: ${{ secrets.EC2_INSTANCE_IP }}
+          username: ec2-user
+          key: ${{ secrets.EC2_SSH_KEY }}
+          port: 22
+          script: |
+            docker stop ticats_ai || true
+            docker rm ticats_ai || true
+            docker image rm public.ecr.aws/h7p2f6d8/ticats_ai:latest || true
+            docker pull public.ecr.aws/h7p2f6d8/ticats_ai:latest
+            docker run -d -p 8000:8000 --name ticats_ai -e DATABASE_URL="${{ secrets.DATABASE_URL }}" public.ecr.aws/h7p2f6d8/ticats_ai:latest

Dockerfile

@@ -1,7 +1,15 @@
 FROM python:3.9
 
+# 작업 디렉토리 설정
 WORKDIR /code
 
+# 종속성 파일 복사 및 설치
+COPY ./requirements.txt /code/
+RUN pip install --no-cache-dir --upgrade -r requirements.txt
+
+
+# 애플리케이션 코드 복사
 COPY ./ /code/
 
-RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+ENTRYPOINT ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000", "--reload"]
+

README.md

@@ -60,8 +60,16 @@
 │  └─routers
 ```
 
-## Git Convention
+## AI 협업 방법
+1. 가상환경 활성화(3.9버전)
+2. pip install -r requirements.txt 
+3. uvicorn main:app --reload
+4. 127.0.0.1:8000/docs 열기
+5. ocr 분석 실행
+6. run_ocr 잘 되는지 확인
+7. git push 또는 pr 올리기
 
+## Git Convention
 - github-flow를 따름
 - branch Convention
   ```angular2html

app/database.py

@@ -13,4 +13,10 @@
 Base = declarative_base()
 
 # 읽기 전용 세션 설정 (autocommit 제거)
-SessionReadOnly = sessionmaker(bind=engine, autoflush=False)
+SessionReadOnly = sessionmaker(bind=engine, autoflush=False)
+def get_db():
+    db = SessionReadOnly()
+    try:
+        yield db
+    finally:
+        db.close()

app/routers/__init__.py

@@ -0,0 +1 @@
+from app.services.ocr_service import OcrService

app/routers/ocr_router.py

@@ -1,11 +1,19 @@
-# Fast Api
-from fastapi import APIRouter
+from fastapi import APIRouter, Depends
+from pydantic import BaseModel
 
-# Schemas
-from app.schemas.ocr_schema import OcrRequestSchema, OcrResponseSchema
+from app.services.ocr_service import OcrService
+
+
+# Dependency injection function
+def get_ocr_service():
+    return OcrService()
+
+
+# Response schema definition
+class OcrResponseSchema(BaseModel):
+    result: str
+    title: str
 
-# Services
-from app.services import ocr_service
 
 # Router
 router = APIRouter(
@@ -14,7 +22,7 @@
 )
 
 
-# Main Section
 @router.post("", response_model=OcrResponseSchema, summary="OCR 분석", description="")
-async def upload_ocr_photo(ocr_photo: OcrRequestSchema):
-    return ocr_service.get_ocr_values(ocr_photo=ocr_photo)
+async def upload_ocr_photo(ocr_service: OcrService = Depends(get_ocr_service)):
+    ocr_result = ocr_service.run_ocr()
+    return {"result": "hello", "title": "OCR Analysis Result"}

app/routers/recommend_router.py

@@ -5,10 +5,10 @@
 from fastapi import APIRouter
 from fastapi.params import Query
 from typing import List
-from app.dependencies import get_db
 from fastapi import Depends
 from sqlalchemy.orm import Session
 
+from app.database import get_db
 from app.services.recommend_service import RecommendService
 
 # Router

app/services/keyword_service.py

@@ -6,13 +6,10 @@
 from torch.autograd.profiler import record_function
 
 np.bool = np.bool_
-from collections import Counter
 
 import warnings;
 
 warnings.filterwarnings('ignore')
-from sklearn.metrics.pairwise import cosine_similarity
-from sklearn.feature_extraction.text import CountVectorizer
 from torch.utils.data import Dataset
 from constants import STOP_WORDS
 import numpy as np
@@ -137,43 +134,6 @@ def united_Processor(self, target: str) -> (str, str):
 
         return topics, sentiments
 
-    def content_recommender(self, base_df, content_id_lst,
-                            return_type='return_id'):  # 다중 인풋 처리용, id를 리스트 형태로 넣으면 됨(=content_id_lst).
-        count_vect = CountVectorizer(min_df=0, ngram_range=(1, 2), lowercase=False)  # 유사도 측정을 위한 피처 백터화
-        genre_mat = count_vect.fit_transform(base_df['sentiment'])  # standard 인자 추가 가능
-
-        genre_sim = cosine_similarity(genre_mat, genre_mat)  # 코사인 유사도 측정
-        genre_sim_sorted_idx = genre_sim.argsort()[:, ::-1]  # 유사도가 높은 순으로 인덱스 나열(각 인덱스(작품)별로)
-
-        sentiment_candidates = []
-        for content_id in content_id_lst:
-            content_idx = base_df[base_df['culturalEventId'] == content_id].index.values
-            similar_indexes = genre_sim_sorted_idx[content_idx, :4]  # top3 만큼 가져옴(기준 인덱스 포함하여 4)
-            similar_indexes = similar_indexes[similar_indexes != content_idx].reshape(-1)
-            for idx in similar_indexes:
-                sentiment_candidates.append(base_df.loc[idx, 'sentiment'])  # 감정 추출
-
-        sentiment_candidates = ', '.join(sentiment_candidates).split(',')
-        counter = Counter(sentiment_candidates)
-        recommend_sentiment = ', '.join([item[0].strip() for item in counter.most_common(2)])  # 빈도 수 높은 감정 상위 2개 추출
-
-        base_content = base_df[base_df['sentiment'] == recommend_sentiment].sample(
-            n=1).index.values  # 같은 감정을 가진 작품 랜덤으로 선택
-        similar_indexes = genre_sim_sorted_idx[base_content, :]  # base와 비슷한 작품 추천
-        similar_indexes = similar_indexes[similar_indexes != content_idx].reshape(-1)
-
-        if return_type == 'return_id':  # id
-            idx_to_id_lst = []
-            for idx in similar_indexes:
-                idx_to_id_lst.append(base_df['culturalEventId'].iloc[idx])
-            return idx_to_id_lst
-
-        elif return_type == 'return_idx':  # index
-            return similar_indexes
-
-        else:
-            print('return_type error')
-
     def sentiment_prediction(self, target_sentence, tok, vocab, model, device):
         start_time = time.time()  # 시작 시간 기록
 

app/services/ocr_service.py

@@ -1,10 +1,8 @@
-# Fast-app
-from app.schemas.ocr_schema import OcrRequestSchema
-
-
 # Main Section
-def get_ocr_values(ocr_photo: OcrRequestSchema):
-    file = ocr_photo.file
+class OcrService:
+    def run_ocr(self):
+        self.other_function()
+        return "hello"
 
-    values = {"title": "title example"}
-    return values
+    def other_function(self):
+        pass