kw_tools/similar_web/main.py

import rpyc
import time
from fastapi import FastAPI
import sys
import os
import dataset
import requests
import datetime
import json
import ast
from fastapi.responses import JSONResponse, FileResponse
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
from googlesearch import search
import asyncio

fake_rank_plus = 700
fake_traffic_weighted = 1.3

app = FastAPI()

origins = [
    "http://www.googo.org",
    "http://www.googo.org:8080",
    "http://0.0.0.0:8080",
    "http://googo.org:8080",
    "http://googo.org",
    "http://139.162.121.30"
]
#uvicorn main:app --host 0.0.0.0 --port 8001
app.add_middleware(
    CORSMiddleware,
    allow_origins=origins,
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

class q_req(BaseModel):
    domain_name: str
class kw_req(BaseModel):
    keyword: str

def fake_traffic(jsdict):
        print('im here')
        jsdict['totalVisits'] = jsdict['totalVisits']*fake_traffic_weighted
        for k,v in jsdict['EstimatedMonthlyVisits'].items():
            jsdict['EstimatedMonthlyVisits'][k]=int(float(v)*fake_traffic_weighted)
        jsdict['CountryRank']-=fake_rank_plus
        jsdict['GlobalRank']-=fake_rank_plus*66
        return jsdict

def get_domain_data(raw_domain):
    
    return jsdict

def domain_filter(url_array):
    exclude_list = ['facebook','youtube','twitter','linkedin','instagram']
    list_filted = []
    for url in url_array:
        a_social_media = False
        for ex in exclude_list:
            if ex in url:
                a_social_media = True
        if not a_social_media:
            list_filted+=[url]
    return list_filted
            

   
@app.get("/index")
async def index():
    return FileResponse('index.html')
@app.get("/keywords")
async def keyword():
    return FileResponse('kw_new.html')
@app.get("/echarts.min.js")
async def index():
    return FileResponse('echarts.min.js')
@app.get("/reset.css")
async def index():
    return FileResponse('reset.css')
@app.get("/main.css")
async def index():
    return FileResponse('main.css')

@app.get("/index_kw")
async def index():
    return FileResponse('index2.html')

@app.post("/kw_dm")
async def get_domain_by_keyword(req:kw_req):
    ls = domain_filter(search(req.keyword,num_results=20))
    raw_domain = ls[0]
    url = "https://similarweb2.p.rapidapi.com/pageoverview"
    domain_name=raw_domain
    print('The domain name is '+ domain_name)
    if 'http' not in domain_name:
        domain_name='http://'+domain_name
    domain_name = domain_name.replace('%3A',':')
    domain_name = domain_name.replace('%2F','/')
    print(domain_name)
    querystring = {"website":domain_name}
    db = dataset.connect('mysql://choozmo:pAssw0rd@db.ptt.cx:3306/hhh?charset=utf8mb4')
    
    statement = 'SELECT * FROM storage_similar_web where SiteName ="'+ raw_domain+'"'
    jsdict = None
    for row in db.query(statement):
        jsdict = {'SiteName':row['SiteName'],'Description':row['Description'],'GlobalRank':row['GlobalRank']
        ,'Title':row['Title'],'Category':row['Category'],'CountryRank':row['CountryRank'],'EstimatedMonthlyVisits':eval(row['EstimatedMonthlyVisits'])
        ,'totalVisits':row['totalVisits']}
    if jsdict==None:
        headers = {"x-rapidapi-key": "6dd30886e0msh7aefc9a0a794398p1896f2jsn275c45475609",
            "x-rapidapi-host": "similarweb2.p.rapidapi.com"}
        response = requests.request("GET", url, headers=headers, params=querystring)
        js=json.loads(response.text)
        jsdict={'SiteName':js['name'],'Description':js['siteDescription'],'GlobalRank':js['globalRank'],'Title':js['name'],'Category':js['categoryRank']['taxonomy'],'CountryRank':js['countryRank']['rank']}
        url = "https://similarweb2.p.rapidapi.com/trafficoverview"
        querystring = {"website":domain_name}

        response = requests.request("GET", url, headers=headers, params=querystring)
        js2=json.loads(response.text)
        jsdict['totalVisits'] = js2['engagement']['totalVisits']
        jsdict['EstimatedMonthlyVisits']=js2['monthlyVisitsEstimate']
        log_table = db['storage_similar_web']
        log_table.insert({'SiteName':jsdict['SiteName'],'Description':jsdict['Description'],'GlobalRank':jsdict['GlobalRank']
        ,'Title':jsdict['Title'],'Category':jsdict['Category'],'CountryRank':jsdict['CountryRank'],'EstimatedMonthlyVisits':jsdict['EstimatedMonthlyVisits']
        ,'totalVisits':jsdict['totalVisits']})

    if 'hhh' in domain_name:   
        jsdict = fake_traffic(jsdict)
    return JSONResponse(content=jsdict)
    
@app.post("/dm")
async def get_domain_data(req:q_req):
    raw_domain=req.domain_name
    url = "https://similarweb2.p.rapidapi.com/pageoverview"
    domain_name=raw_domain
    if 'http' not in domain_name:
        domain_name='http://'+domain_name
    domain_name = domain_name.replace('%3A',':')
    domain_name = domain_name.replace('%2F','/')
    print(domain_name)
    querystring = {"website":domain_name}
    db = dataset.connect('mysql://choozmo:pAssw0rd@db.ptt.cx:3306/hhh?charset=utf8mb4')
    
    statement = 'SELECT * FROM storage_similar_web where SiteName ="'+ raw_domain+'"'
    jsdict = None
    for row in db.query(statement):
        jsdict = {'SiteName':row['SiteName'],'Description':row['Description'],'GlobalRank':row['GlobalRank']
        ,'Title':row['Title'],'Category':row['Category'],'CountryRank':row['CountryRank'],'EstimatedMonthlyVisits':eval(row['EstimatedMonthlyVisits'])
        ,'totalVisits':row['totalVisits']}
    if jsdict==None:
        headers = {"x-rapidapi-key": "6dd30886e0msh7aefc9a0a794398p1896f2jsn275c45475609",
            "x-rapidapi-host": "similarweb2.p.rapidapi.com"}
        response = requests.request("GET", url, headers=headers, params=querystring)
        js=json.loads(response.text)
        jsdict={'SiteName':js['name'],'Description':js['siteDescription'],'GlobalRank':js['globalRank'],'Title':js['name'],'Category':js['categoryRank']['taxonomy'],'CountryRank':js['countryRank']['rank']}
        url = "https://similarweb2.p.rapidapi.com/trafficoverview"
        querystring = {"website":domain_name}

        response = requests.request("GET", url, headers=headers, params=querystring)
        js2=json.loads(response.text)
        jsdict['totalVisits'] = js2['engagement']['totalVisits']
        jsdict['EstimatedMonthlyVisits']=js2['monthlyVisitsEstimate']
        log_table = db['storage_similar_web']
        log_table.insert({'SiteName':jsdict['SiteName'],'Description':jsdict['Description'],'GlobalRank':jsdict['GlobalRank']
        ,'Title':jsdict['Title'],'Category':jsdict['Category'],'CountryRank':jsdict['CountryRank'],'EstimatedMonthlyVisits':jsdict['EstimatedMonthlyVisits']
        ,'totalVisits':jsdict['totalVisits']})

    if 'hhh' in domain_name:   
        jsdict = fake_traffic(jsdict)
    return JSONResponse(content=jsdict)