Python Script alteration & Open ai API

I am looking for a Python developer to make alterations to my existing script and integrate it with the OpenAI ChatGPT API. The specific requirements for the project are as follows:

- Alteration of the input/output format in the Python script

- Integration of new functionalities as per the provided requirements and examples

- Optimization of the existing code for improved performance

Ideal skills and experience for this job include:

- Proficiency in Python programming

- Experience with OpenAI APIs, specifically ChatGPT

- Strong understanding of input/output formatting in Python scripts

- Ability to optimize code for improved performance

Here is the script from [login to view URL] - [login to view URL]

I want to bring the results for this into a text file instead sending it to the front end

from serpapi import GoogleSearch

import os

import requests

# import random

from fastapi import FastAPI

from dotenv import load_dotenv

from serpapi import GoogleSearch

from [login to view URL] import ThreadPoolExecutor

import asyncio

import newspaper

from bs4 import BeautifulSoup

from [login to view URL] import stopwords

from [login to view URL] import word_tokenize

from [login to view URL] import bigrams, trigrams

import nltk

from .[login to view URL] import add_post, save_links, get_post, get_all_post, remove_post, update_post

# Load env

load_dotenv()

SERPAPI_KEY = [login to view URL]('c0797abfd638908bf4cadfb3c2522a2865d2d39f33d0d627411dc858332f6329')

app = FastAPI()

# Setup Cors

origins = [

"http://localhost",

"http://localhost:3000",

]

app.add_middleware(

CORSMiddleware,

allow_origins=origins,

# allow_origins=["*"],

allow_credentials=True,

allow_methods=["*"],

allow_headers=["*"],

)

[login to view URL]('punkt')

[login to view URL]('stopwords')

# API Routes

@[login to view URL]("/")

def read_root():

return {"Hello": "World"}

@[login to view URL]('/search/')

def search(search: dict):

# 1. Get google search API

googleSearch = GoogleSearch({

"api_key": SERPAPI_KEY,

"engine": "google",

"num": 20,

"q": search['botox melbourne'],

"hl": search['english'],

"gl": search['melbourne'],

"google_domain": search['[login to view URL]'],

# "location": location,

})

result = googleSearch.get_dict()

#2. Get google autocomplete API

autocompleteSearch = GoogleSearch({

"api_key": SERPAPI_KEY,

"engine": "google_autocomplete",

"q": search['botox melbourne']

})

autocompleteResult = autocompleteSearch.get_dict()

result['autocomplete'] = autocompleteResult['suggestions']

post = add_post(

title = search['keyword'],

search_query = {

"q": search['keyword'],

"hl": search['searchLang'],

"gl": search['searchLocation'],

"googleDomain": search['googleDomain'],

},

search_result = result

)

return {

"post_id": [login to view URL],

"search_result": result

}

@[login to view URL]('/links/')

def saveLinks(linkData: dict):

post = save_links(

post_id = linkData['postId'],

choosen_links = linkData['choosenLinks']

)

return post

@[login to view URL]('/posts')

def getPosts():

posts = get_all_post()

return posts

@[login to view URL]('/posts/{post_id}')

def getPost(post_id: str):

post = get_post(post_id)

return post

@[login to view URL]('/posts/{post_id}')

def updatePost(post_id: str, data: dict):

post = update_post(post_id, data)

return post

@[login to view URL]('/posts/{post_id}')

def deletePost(post_id: str):

post = remove_post(post_id)

return post

@[login to view URL]('/scrape/{post_id}')

async def scrape(post_id: str):

post = get_post(post_id)

# Return early if exists

if post.links_scrape_result:

return {

"status": "success",

"data": post.links_scrape_result

}

# Scrape links if new

links = post.choosen_links

lang = post.search_query['hl']

contentInfo = []

allKeywords = []

with ThreadPoolExecutor() as executor:

tasks = [asyncio.get_running_loop().run_in_executor(executor, _scrape_article, link, lang) for link in links]

results = await [login to view URL](*tasks)

for result in results:

if result:

[login to view URL](result)

[login to view URL](result['keywords'])

topKeywords = _getTopKeywords(allKeywords)

averageWords = sum([content['totalWords'] for content in contentInfo]) / len(contentInfo)

links_scrape_result = {

"contentInfo": contentInfo,

"topKeywords": topKeywords,

"averageWords": averageWords

}

update_post(post_id, {

"links_scrape_result": links_scrape_result

})

return {

"status": "success",

"data": links_scrape_result

}

# Currently lang get from hl query parameter

# There's no guarantee it match available language at nltk

# [login to view URL]

# Newspaper Config

# When using proxy

# config = [login to view URL]()

# [login to view URL] = {

# 'http': '[login to view URL]' # sample only

# }

# config.request_timeout = 20

def _scrape_article(link, lang):

# Add config=config in paramter if adding proxy/user_agents

# article = [login to view URL](link, keep_article_html=True, config=config)

article = [login to view URL](link, keep_article_html=True)

content = None

try:

[login to view URL]()

[login to view URL]()

content = [login to view URL]

content_html = article.article_html

content_title = [login to view URL]

# Current logic to detect if it's blocked

# is by check if text results is very short

if len(content) < 100:

print("Scraping might be blocked! In case expection block not catching up")

print('---- link ---- \n' + link)

# Scrape from golang endpoint

endpoint = "http://api-golang:8080/scrape?link=" + link

response = [login to view URL](endpoint)

if response.status_code == 200:

data = [login to view URL]()

content = data['content']

content_html = data['content_html']

else:

print(f"Error {response.status_code}: {[login to view URL]}")

soup = BeautifulSoup(content_html, '[login to view URL]')

headings = []

for heading in soup.find_all(['h1', 'h2', 'h3']):

[login to view URL]({

"text": [login to view URL],

"tag": [login to view URL]

})

except:

print("-----------")

print("Error scrape. Todo: run manual fetch ", link)

print("-----------")

return None

if content == None:

return None

_lang = 'english'

if lang == 'id':

_lang = 'english'

# later add more language, mapping hl query -> nltk language

# [login to view URL]

text = [login to view URL]()

word_tokens = word_tokenize(text)

stop_words = set([login to view URL](_lang))

filtered_text = [word for word in word_tokens if [login to view URL]() and word not in stop_words]

single_freq_dist = [login to view URL](filtered_text)

bigram_freq_dist = [login to view URL](bigrams(filtered_text))

trigram_freq_dist = [login to view URL](trigrams(filtered_text))

# combine all frequency distribution

freq_dist = single_freq_dist + bigram_freq_dist + trigram_freq_dist

keywords = []

for word, frequency in freq_dist.most_common(10):

[login to view URL]({

"word": word,

"frequency": frequency

})

totalWords = len([login to view URL]())

return {

"link": link,

"totalWords": totalWords,

"keywords": keywords,

"title": content_title,

"headings": headings,

}

def _getTopKeywords(allKeywords):

MAX_KEYWORD = 15

topKeywords = {}

for word_dict in allKeywords:

# if word_dict['word'] is tuple, convert it to string

_word = word_dict['word']

if isinstance(_word, tuple):

_word = ' '.join(word_dict['word'])

word = [login to view URL]()

if word in topKeywords:

topKeywords[word] += word_dict['frequency']

else:

topKeywords[word] = word_dict['frequency']

# sort the dictionary by the frequency in descending order and get the first 10 items

sorted_topKeywords = sorted([login to view URL](), key=[login to view URL](1), reverse=True)[:MAX_KEYWORD]

return sorted_topKeywords

______

We need to use the [login to view URL] app on git hub

utilise the results of 2 aspects and store them in text file

And then be able to use those results in another python script.

Can the team help thanks

Python GitHub SQL

Nº del proyecto: #37631624

Sobre el proyecto