Update app.py
Browse files
app.py
CHANGED
|
@@ -0,0 +1,100 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from __future__ import print_function
|
| 2 |
+
from config import *
|
| 3 |
+
|
| 4 |
+
import tiktoken
|
| 5 |
+
import pinecone
|
| 6 |
+
import uuid
|
| 7 |
+
import sys
|
| 8 |
+
import logging
|
| 9 |
+
|
| 10 |
+
from flask import Flask, jsonify
|
| 11 |
+
from flask_cors import CORS, cross_origin
|
| 12 |
+
from flask import request
|
| 13 |
+
|
| 14 |
+
from handle_file import handle_file
|
| 15 |
+
from answer_question import get_answer_from_files
|
| 16 |
+
|
| 17 |
+
logging.basicConfig(
|
| 18 |
+
level=logging.INFO,
|
| 19 |
+
format="%(asctime)s [%(levelname)s] %(message)s",
|
| 20 |
+
handlers=[
|
| 21 |
+
logging.FileHandler("debug.log"),
|
| 22 |
+
logging.StreamHandler(sys.stdout)
|
| 23 |
+
]
|
| 24 |
+
)
|
| 25 |
+
|
| 26 |
+
logging.basicConfig(
|
| 27 |
+
level=logging.INFO,
|
| 28 |
+
format="%(asctime)s [%(levelname)s] %(message)s",
|
| 29 |
+
handlers=[
|
| 30 |
+
logging.FileHandler("debug.log"),
|
| 31 |
+
logging.StreamHandler(sys.stdout)
|
| 32 |
+
]
|
| 33 |
+
)
|
| 34 |
+
|
| 35 |
+
def load_pinecone_index() -> pinecone.Index:
|
| 36 |
+
"""
|
| 37 |
+
Load index from Pinecone, raise error if the index can't be found.
|
| 38 |
+
"""
|
| 39 |
+
pinecone.init(
|
| 40 |
+
api_key=PINECONE_API_KEY,
|
| 41 |
+
environment=PINECONE_ENV,
|
| 42 |
+
)
|
| 43 |
+
index_name = PINECONE_INDEX
|
| 44 |
+
if not index_name in pinecone.list_indexes():
|
| 45 |
+
print(pinecone.list_indexes())
|
| 46 |
+
raise KeyError(f"Index '{index_name}' does not exist.")
|
| 47 |
+
index = pinecone.Index(index_name)
|
| 48 |
+
|
| 49 |
+
return index
|
| 50 |
+
|
| 51 |
+
def create_app():
|
| 52 |
+
pinecone_index = load_pinecone_index()
|
| 53 |
+
tokenizer = tiktoken.get_encoding("gpt2")
|
| 54 |
+
session_id = str(uuid.uuid4().hex)
|
| 55 |
+
app = Flask(__name__)
|
| 56 |
+
app.pinecone_index = pinecone_index
|
| 57 |
+
app.tokenizer = tokenizer
|
| 58 |
+
app.session_id = session_id
|
| 59 |
+
# log session id
|
| 60 |
+
logging.info(f"session_id: {session_id}")
|
| 61 |
+
app.config["file_text_dict"] = {}
|
| 62 |
+
CORS(app, supports_credentials=True)
|
| 63 |
+
|
| 64 |
+
return app
|
| 65 |
+
|
| 66 |
+
app = create_app()
|
| 67 |
+
|
| 68 |
+
@app.route(f"/process_file", methods=["POST"])
|
| 69 |
+
@cross_origin(supports_credentials=True)
|
| 70 |
+
def process_file():
|
| 71 |
+
try:
|
| 72 |
+
file = request.files['file']
|
| 73 |
+
logging.info(str(file))
|
| 74 |
+
handle_file(
|
| 75 |
+
file, app.session_id, app.pinecone_index, app.tokenizer)
|
| 76 |
+
return jsonify({"success": True})
|
| 77 |
+
except Exception as e:
|
| 78 |
+
logging.error(str(e))
|
| 79 |
+
return jsonify({"success": False})
|
| 80 |
+
|
| 81 |
+
@app.route(f"/answer_question", methods=["POST"])
|
| 82 |
+
@cross_origin(supports_credentials=True)
|
| 83 |
+
def answer_question():
|
| 84 |
+
try:
|
| 85 |
+
params = request.get_json()
|
| 86 |
+
question = params["question"]
|
| 87 |
+
|
| 88 |
+
answer_question_response = get_answer_from_files(
|
| 89 |
+
question, app.session_id, app.pinecone_index)
|
| 90 |
+
return answer_question_response
|
| 91 |
+
except Exception as e:
|
| 92 |
+
return str(e)
|
| 93 |
+
|
| 94 |
+
@app.route("/healthcheck", methods=["GET"])
|
| 95 |
+
@cross_origin(supports_credentials=True)
|
| 96 |
+
def healthcheck():
|
| 97 |
+
return "OK"
|
| 98 |
+
|
| 99 |
+
if __name__ == "__main__":
|
| 100 |
+
app.run(debug=True, port=SERVER_PORT, threaded=True)
|