(TSRT) Source Generator THBwiki

a.py

'''
Touhou Music To Address From Thwiki
'''

import re
import urllib.request
import sys  
import json

def getURL(url):
    # print("访问开始:"+url)
    request=urllib.request.Request(url)
    response=urllib.request.urlopen(request)
    html=response.read()
    return html.decode('utf-8')

cate="N/A"

with open("test.txt","r",encoding="UTF-8") as f:
    with open("out.txt","w",encoding="UTF-8") as of:
        read=f.readlines()
        count=1

        x=int(input("Start position:"))

        for line in read:

            line=line.strip()
            if '/' in line:
                
                if count<x:
                    count+=1
                    continue

                name=line.split("/")[0].strip()

                print(count,"/",len(read),"-",cate,name)

                url="https://thwiki.cc/"+urllib.parse.quote(name.replace(" ","_"))
                res=getURL(url)
                res2=re.search('"https:\\/\\/upload\\.thwiki\\.cc\\/.\\/..\\/.+\\.((ogg)|(mp3))"',res)

                if res2==None:
                    print("[WARNING] NO LINK FOUND")
                else:
                    add=res[(res2.span()[0]+1):(res2.span()[1]-1)]
                    of.write(name+"!!!"+cate+"!!!"+add+"\n")
                    of.flush()
                
            else:
                cate=line
            
            count+=1