Milestone 4 - PR to develop from back-end #240610

.github/workflows/gradle.yml

@@ -54,6 +54,20 @@ jobs:
           echo "${{ secrets.AWS_PROPERTIES }}" > ./application.yml
         shell: bash
 
+<<<<<<< HEAD
+      # Google Translation key 설정
+      - name: Create Google Translation JSON file from secret
+        uses: jsdaniell/[email protected]
+        with:
+          name: "./back-end/src/main/resources/service-account-file.json"
+          json: ${{ secrets.GOOGLE_TRANSLATION_KEY }}
+
+      - name: Verify JSON file creation
+        run: cat ./back-end/src/main/resources/service-account-file.json
+
+
+=======
+>>>>>>> develop
       # gradlew 파일 실행권한 설정
       - name: Grant execute permission for gradlew
         run: |

back-end/.gitignore

@@ -38,4 +38,11 @@ out/
 
 ### *.yml / *.properties ###
 *.yml
-application.properties
+<<<<<<< HEAD
+application.properties
+
+### google cloud translation api key ###
+service-account-file.json
+=======
+application.properties
+>>>>>>> develop

back-end/build.gradle

@@ -76,6 +76,16 @@ dependencies {
 
     // WebSocket
     implementation 'org.springframework.boot:spring-boot-starter-websocket'
+<<<<<<< HEAD
+
+    // ChatGpt
+    implementation 'com.squareup.okhttp3:okhttp:4.9.1'
+    implementation 'org.apache.commons:commons-io:1.3.2'
+
+    // Google Cloud Translation
+    implementation 'com.google.cloud:google-cloud-translate:2.6.0'
+=======
+>>>>>>> develop
 }
 
 tasks.named('test') {

back-end/lambda/EmbeddingCategory.py

@@ -0,0 +1,70 @@
+import os
+import json
+import urllib3
+from openai import OpenAI
+from pymongo.mongo_client import MongoClient
+from pymongo.server_api import ServerApi
+
+mongo_uri = os.environ.get("MONGO_URI")
+
+def lambda_handler(event, context):
+    # Fetch user's profile data
+    request_event = json.loads(event["body"])
+    category_id = request_event["category_id"]
+    category = request_event["category"]
+    prompt = request_event["prompt"]
+
+    # Create a new client and connect to the server
+    client = MongoClient(mongo_uri, server_api=ServerApi('1'))
+
+    # Name of database
+    db = client.CategoryEmbedding
+
+    # Name of collection
+    collection = db.Embedding
+
+    embedding = getEmbedding(prompt)
+    document = {"category": category, "prompt": prompt, "embedding": embedding}
+
+    # Insert document
+    result = collection.update_one({"category_id": category_id}, {"$set": document}, upsert=True)
+
+    if result.acknowledged:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 200,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "Successfully added the category."
+        }
+    else:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 502,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "There was error during the insertion."
+        }
+
+
+def getEmbedding(document):
+    open_ai_url = os.environ.get("OPENAI_URI")
+    open_ai_key = os.environ.get("OPENAI_KEY")
+
+    header = {
+        'Content-Type': 'application/json',
+        'Authorization': f'Bearer {open_ai_key}'
+    }
+
+    body = json.dumps({
+        'input': f'{document}',
+        'model': "text-embedding-3-large",
+        'format': "float"
+    })
+
+    http = urllib3.PoolManager()
+    response = http.request('POST', open_ai_url, body=body, headers=header)
+
+    data = json.loads(response.data)
+
+    return data['data'][0]['embedding']

back-end/lambda/EmbeddingMeeting.py

@@ -0,0 +1,74 @@
+import os
+import json
+import urllib3
+from openai import OpenAI
+from pymongo.mongo_client import MongoClient
+from pymongo.server_api import ServerApi
+
+mongo_uri = os.environ.get("MONGO_URI")
+
+def lambda_handler(event, context):
+    # Fetch user's profile data
+    event_body = json.loads(event["body"])
+    moim_id = event_body['moimId']
+    moimname = event_body['moimTitle']
+    prompt = event_body['prompt']
+
+    # Create a new client and connect to the server
+    client = MongoClient(mongo_uri, server_api=ServerApi('1'))
+
+    # Name of database
+    db = client.MoimEmbedding
+
+    # Name of collection
+    collection = db.Embedding
+
+    embedding = getEmbedding(prompt)
+    document = {
+        "moim_id": moim_id,
+        "moim_title": moimname,
+        "prompt": prompt, 
+        "embedding": embedding}
+
+    # Insert document
+    result = collection.update_one({"moim_id": moim_id}, {"$set": document}, upsert=True)
+
+    if result.acknowledged:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 200,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "Success"
+        }
+    else:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 400,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "Failed"
+        }
+
+
+def getEmbedding(document):
+    open_ai_url = os.environ.get("OPENAI_URI")
+    open_ai_key = os.environ.get("OPENAI_KEY")
+
+    header = {
+        'Content-Type': 'application/json',
+        'Authorization': f'Bearer {open_ai_key}'
+    }
+
+    body = json.dumps({
+        'input': f'{document}',
+        'model': "text-embedding-3-large",
+        'format': "float"
+    })
+
+    http = urllib3.PoolManager()
+    response = http.request('POST', open_ai_url, body=body, headers=header)
+
+    data = json.loads(response.data)
+
+    return data['data'][0]['embedding']

back-end/lambda/EmbeddingSearchCategory.py

@@ -0,0 +1,78 @@
+import os
+import json
+import urllib3
+from openai import OpenAI
+from pymongo.mongo_client import MongoClient
+from pymongo.server_api import ServerApi
+
+mongo_uri = os.environ.get("MONGO_URI")
+
+def lambda_handler(event, context):
+    # Fetch user's profile data
+    user_id = int(event['queryStringParameters']['userID'])
+
+    # Create a new client and connect to the server
+    client = MongoClient(mongo_uri, server_api=ServerApi('1'))
+
+    # Name of database
+    db = client.UserEmbedding
+
+    # Name of collection
+    collection = db.Embedding
+
+    query = { "user_id": user_id }
+    projection = { "embedding": 1 }
+
+    # Insert document
+    result = collection.find_one(query, projection)
+    if result == None:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 400,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "There's no such user in DB."
+        }
+
+    query_embedding = result['embedding']
+
+    pipeline = [
+        {
+            '$vectorSearch': {
+                'index': 'CategoryVectorIndex',
+                'path': 'embedding',
+                'queryVector': query_embedding,
+                'numCandidates': 100,
+                'limit': 100
+            }
+        }, {
+            '$project': {
+                '_id': 0,
+                'prompt': 0,
+                'embedding': 0
+            }    
+        }
+    ]
+
+    results = client['CategoryEmbedding']['Embedding'].aggregate(pipeline)
+
+    moim_list = []
+    for search_result in results:
+        moim_list.append(search_result['category_id'])
+
+    if (len(moim_list) > 0):
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 200,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "[" + ','.join(str(e) for e in moim_list) + "]"
+        }
+    else:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 502,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "Error occured during insertion."
+        }

back-end/lambda/EmbeddingSearchMeeting.py

@@ -0,0 +1,69 @@
+import os
+import json
+import urllib3
+from openai import OpenAI
+from pymongo.mongo_client import MongoClient
+from pymongo.server_api import ServerApi
+
+mongo_uri = os.environ.get("MONGO_URI")
+
+def lambda_handler(event, context):
+    # Fetch user's profile data
+    user_id = int(event['queryStringParameters']['userID'])
+
+    # Create a new client and connect to the server
+    client = MongoClient(mongo_uri, server_api=ServerApi('1'))
+
+    # Name of database
+    db = client.UserEmbedding
+
+    # Name of collection
+    collection = db.Embedding
+
+    query = { "user_id": user_id }
+    projection = { "embedding": 1 }
+
+    # Insert document
+    result = collection.find_one(query, projection)
+
+    query_embedding = result['embedding']
+
+    pipeline = [
+        {
+            '$vectorSearch': {
+                'index': 'MoimVectorIndex',
+                'path': 'embedding',
+                'queryVector': query_embedding,
+                'numCandidates': 100,
+                'limit': 50
+            }
+        }, {
+            '$project': {
+                '_id': 0,
+                'embedding': 0
+            }    
+        }
+    ]
+
+    results = client['MoimEmbedding']['Embedding'].aggregate(pipeline)
+
+    moim_list = []
+    for search_result in results:
+        moim_list.append(search_result['moim_id'])
+
+    if (len(moim_list) > 0):
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 200,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "[" + ','.join(str(e) for e in moim_list) + "]"
+        }
+    else:
+        return {
+            "isBase64Encoded": False,
+            "statusCode": 400,
+            "headers": {},
+            "multiValueHeaders": {},
+            "body": "Failed"
+        }