#!/usr/bin/env python

import scholarly
from scholarly import ProxyGenerator
from scholarly import scholarly

# pickle提供了一个简单的持久化功能。可以将对象以文件的形式存放在磁盘上
# python中几乎所有的数据类型（列表，字典，集合，类等）都可以用pickle来序列化
import pickle

# Set up a ProxyGenerator object to use free proxies
# This needs to be done only once per session
# need to register and got the API
pg = ProxyGenerator()
success = pg.ScraperAPI("YOUR API")

# Now search Google Scholar from behind a proxy
scholarly.use_proxy(pg)


search_query = scholarly.search_author('AUTHOR NAME')

first_author_result = next(search_query)
author = scholarly.fill(first_author_result)
first_publication = author['publications'][1]

first_publication_filled = scholarly.fill(first_publication)

# Which papers cited that publication?
citations_all = [citation for citation in scholarly.citedby(first_publication_filled)]

# Save all files into a file
file = open("dictionary_data.pkl", "wb")
pickle.dump(citations_all, file)
file.close()

import pickle
import pandas as pd

file = open("dictionary_data.pkl", "rb")
citations_all = pickle.load(file)

# 提取每个引用字典中的所需信息，构建一个列表
data = []
for citation in citations_all:
    bib = citation['bib']
    
    if 'citedby_url' in citation:
        citedby_url = citation['citedby_url']
    else:
        citedby_url = 'NA'
        
    if 'pub_url' in citation:
        pub_url = citation['pub_url']
    else:
        pub_url = 'NA'
        
    if 'eprint_url' in citation:
        eprint_url = citation['eprint_url']
    else:
        eprint_url = 'NA'
    
    # else
    if 'num_citations' in citation:
        num_citations = citation['num_citations']
    else:
        num_citations = 'NA'
    
    data.append([
        bib['title'],
        ', '.join(bib['author']),
        bib['venue'],
        bib['pub_year'],
        ', '.join(citation['author_id']),
        citedby_url,
        pub_url,
        eprint_url,
        num_citations
    ])

# 构建 DataFrame
columns = ['title', 'author', 'venue', 'pub_year', 'author_id', 'citedby_url', 'pub_url', 'eprint_url', 'num_citations']
df = pd.DataFrame(data, columns=columns)

# 将 DataFrame 保存为 CSV 文件
df.to_csv('citations.csv', index=False)