Greg Thompson
Update nlu endpoint with additionally logging around alternative decode process
8f01606
raw
history blame
No virus
10.4 kB
"""FastAPI endpoint
To run locally use 'uvicorn app:app --host localhost --port 7860'
or
`python -m uvicorn app:app --reload --host localhost --port 7860`
"""
import ast
import json
from json import JSONDecodeError
from logging import getLogger
import mathactive.microlessons.num_one as num_one_quiz
import os
import sentry_sdk
from fastapi import FastAPI, Request
from fastapi.responses import JSONResponse
from fastapi.staticfiles import StaticFiles
from fastapi.templating import Jinja2Templates
# from mathtext.sentiment import sentiment
from mathtext.text2int import text2int
from mathtext_fastapi.logging import prepare_message_data_for_logging
from mathtext_fastapi.conversation_manager import manage_conversation_response
from mathtext_fastapi.v2_conversation_manager import manage_conversation_response
from mathtext_fastapi.nlu import evaluate_message_with_nlu
from mathtext_fastapi.nlu import run_intent_classification
from pydantic import BaseModel
from dotenv import load_dotenv
load_dotenv()
log = getLogger(__name__)
sentry_sdk.init(
dsn=os.environ.get('SENTRY_DSN'),
# Set traces_sample_rate to 1.0 to capture 100%
# of transactions for performance monitoring.
# We recommend adjusting this value in production,
traces_sample_rate=1.0,
)
app = FastAPI()
app.mount("/static", StaticFiles(directory="static"), name="static")
templates = Jinja2Templates(directory="templates")
class Text(BaseModel):
content: str = ""
@app.get("/")
def home(request: Request):
return templates.TemplateResponse("home.html", {"request": request})
@app.get("/sentry-debug")
async def trigger_error():
division_by_zero = 1 / 0
@app.post("/hello")
def hello(content: Text = None):
content = {"message": f"Hello {content.content}!"}
return JSONResponse(content=content)
# @app.post("/sentiment-analysis")
# def sentiment_analysis_ep(content: Text = None):
# ml_response = sentiment(content.content)
# content = {"message": ml_response}
# return JSONResponse(content=content)
@app.post("/text2int")
def text2int_ep(content: Text = None):
ml_response = text2int(content.content)
content = {"message": ml_response}
return JSONResponse(content=content)
@app.post("/v1/manager")
async def programmatic_message_manager(request: Request):
"""
Calls conversation management function to determine the next state
Input
request.body: dict - message data for the most recent user response
{
"author_id": "+47897891",
"contact_uuid": "j43hk26-2hjl-43jk-hnk2-k4ljl46j0ds09",
"author_type": "OWNER",
"message_body": "a test message",
"message_direction": "inbound",
"message_id": "ABJAK64jlk3-agjkl2QHFAFH",
"message_inserted_at": "2022-07-05T04:00:34.03352Z",
"message_updated_at": "2023-02-14T03:54:19.342950Z",
}
Output
context: dict - the information for the current state
{
"user": "47897891",
"state": "welcome-message-state",
"bot_message": "Welcome to Rori!",
"user_message": "",
"type": "ask"
}
"""
data_dict = await request.json()
context = manage_conversation_response(data_dict)
return JSONResponse(context)
@app.post("/v2/manager")
async def programmatic_message_manager(request: Request):
"""
Calls conversation management function to determine the next state
Input
request.body: dict - message data for the most recent user response
{
"author_id": "+47897891",
"contact_uuid": "j43hk26-2hjl-43jk-hnk2-k4ljl46j0ds09",
"author_type": "OWNER",
"message_body": "a test message",
"message_direction": "inbound",
"message_id": "ABJAK64jlk3-agjkl2QHFAFH",
"message_inserted_at": "2022-07-05T04:00:34.03352Z",
"message_updated_at": "2023-02-14T03:54:19.342950Z",
}
Output
context: dict - the information for the current state
{
"user": "47897891",
"state": "welcome-message-state",
"bot_message": "Welcome to Rori!",
"user_message": "",
"type": "ask"
}
"""
data_dict = await request.json()
context = manage_conversation_response(data_dict)
return JSONResponse(context)
@app.post("/intent-classification")
def intent_classification_ep(content: Text = None):
ml_response = run_intent_classification(content.content)
content = {"message": ml_response}
return JSONResponse(content=content)
@app.post("/nlu")
async def evaluate_user_message_with_nlu_api(request: Request):
""" Calls nlu evaluation and returns the nlu_response
Input
- request.body: json - message data for the most recent user response
Output
- int_data_dict or sent_data_dict: dict - the type of NLU run and result
{'type':'integer', 'data': '8', 'confidence': 0}
{'type':'sentiment', 'data': 'negative', 'confidence': 0.99}
"""
log.info(f'Received request: {request}')
log.info(f'Request header: {request.headers}')
request_body = await request.body()
log.info(f'Request body: {request_body}')
request_body_str = request_body.decode()
log.info(f'Request_body_str: {request_body_str}')
try:
data_dict = await request.json()
except JSONDecodeError:
log.error(f'Request.json failed: {dir(request)}')
data_dict = {}
message_data = data_dict.get('message_data')
if not message_data:
log.error(f'Data_dict: {data_dict}')
message_data = data_dict.get('message', {})
nlu_response = evaluate_message_with_nlu(message_data)
return JSONResponse(content=nlu_response)
@app.post("/num_one")
async def num_one(request: Request):
"""
Input:
{
"user_id": 1,
"message_text": 5,
}
Output:
{
'messages':
["Let's", 'practice', 'counting', '', '', '46...', '47...', '48...', '49', '', '', 'After', '49,', 'what', 'is', 'the', 'next', 'number', 'you', 'will', 'count?\n46,', '47,', '48,', '49'],
'input_prompt': '50',
'state': 'question'
}
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
user_id = message_data['user_id']
message_text = message_data['message_text']
return num_one_quiz.process_user_message(user_id, message_text)
@app.post("/start")
async def ask_math_question(request: Request):
"""Generate a question data
Input
{
'difficulty': 0.1,
'do_increase': True | False
}
Output
{
'text': 'What is 1+2?',
'difficulty': 0.2,
'question_numbers': [3, 1, 4]
}
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
difficulty = message_data['difficulty']
do_increase = message_data['do_increase']
return JSONResponse(generators.start_interactive_math(difficulty, do_increase))
@app.post("/hint")
async def get_hint(request: Request):
"""Generate a hint data
Input
{
'start': 5,
'step': 1,
'difficulty': 0.1
}
Output
{
'text': 'What number is greater than 4 and less than 6?',
'difficulty': 0.1,
'question_numbers': [5, 1, 6]
}
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
start = message_data['start']
step = message_data['step']
difficulty = message_data['difficulty']
return JSONResponse(hints.generate_hint(start, step, difficulty))
@app.post("/question")
async def ask_math_question(request: Request):
"""Generate a question data
Input
{
'start': 5,
'step': 1,
'question_num': 1 # optional
}
Output
{
'question': 'What is 1+2?',
'start': 5,
'step': 1,
'answer': 6
}
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
start = message_data['start']
step = message_data['step']
arg_tuple = (start, step)
try:
question_num = message_data['question_num']
arg_tuple += (question_num,)
except KeyError:
pass
return JSONResponse(questions.generate_question_data(*arg_tuple))
@app.post("/difficulty")
async def get_hint(request: Request):
"""Generate a number matching difficulty
Input
{
'difficulty': 0.01,
'do_increase': True
}
Output - value from 0.01 to 0.99 inclusively:
0.09
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
difficulty = message_data['difficulty']
do_increase = message_data['do_increase']
return JSONResponse(utils.get_next_difficulty(difficulty, do_increase))
@app.post("/start_step")
async def get_hint(request: Request):
"""Generate a start and step values
Input
{
'difficulty': 0.01,
'path_to_csv_file': 'scripts/quiz/data.csv' # optional
}
Output - tuple (start, step):
(5, 1)
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
difficulty = message_data['difficulty']
arg_tuple = (difficulty,)
try:
path_to_csv_file = message_data['path_to_csv_file']
arg_tuple += (path_to_csv_file,)
except KeyError:
pass
return JSONResponse(utils.get_next_difficulty(*arg_tuple))
@app.post("/sequence")
async def generate_question(request: Request):
"""Generate a sequence from start, step and optional separator parameter
Input
{
'start': 5,
'step': 1,
'sep': ', ' # optional
}
Output
5, 6, 7
"""
data_dict = await request.json()
message_data = ast.literal_eval(data_dict.get('message_data', '').get('message_body', ''))
start = message_data['start']
step = message_data['step']
arg_tuple = (start, step)
try:
sep = message_data['sep']
arg_tuple += (sep,)
except KeyError:
pass
return JSONResponse(utils.convert_sequence_to_string(*arg_tuple))