import os
import sys
import csv
import subprocess
import time
import random
import re
import json
import click
from urllib.parse import urlparse
from s2 import SemanticScholarAPI
from util import *

s2 = SemanticScholarAPI()

@click.command()
@click.option('--fn', '-i', default='db_paper_doi.csv', help='Filename of CSV (id, url,)')
def fetch_doi_list(fn):
  lines = read_csv(fn, keys=False)
  domains = []
  for line in lines:
    paper_id, url = line
    if url:
      domain = fetch_doi(paper_id, url)
      print(domain)
  print("{} papers processed".format(len(lines)))

def fetch_doi(paper_id, url, replace=False):
  os.makedirs(make_doi_path(paper_id), exist_ok=True)
  doi_fn = make_doi_fn(paper_id)
  url_fn = make_url_fn(paper_id)
  txt_fn = make_txt_fn(paper_id)
  if replace and os.path.exists(doi_fn):
    os.rename(doi_fn, old_doi_fn(paper_id))
    os.rename(url_fn, old_url_fn(paper_id))
  if os.path.exists(doi_fn) or os.path.exists(txt_fn):
    # return read_json(doi_fn)
    return None, None
  size, final_url = s2.fetch_doi(url, doi_fn)
  if size is None:
    print("{} empty?".format(paper_id))
    time.sleep(random.randint(2, 5))
    return None, None
  print("{} {} kb".format(paper_id, int(size / 1024)))
  domain = urlparse(final_url).netloc
  write_json(url_fn, {
    'paper_id': paper_id,
    'domain': domain
  })
  time.sleep(random.randint(2, 5))
  return domain
  # return paper

def make_doi_path(paper_id):
  return './datasets/s2/doi/{}/{}'.format(paper_id[0:2], paper_id)
def make_doi_fn(paper_id):
  return './datasets/s2/doi/{}/{}/paper.doi'.format(paper_id[0:2], paper_id)
def make_url_fn(paper_id):
  return './datasets/s2/doi/{}/{}/paper.url'.format(paper_id[0:2], paper_id)
def make_txt_fn(paper_id):
  return './datasets/s2/pdf/{}/{}/paper.txt'.format(paper_id[0:2], paper_id)

def old_doi_fn(paper_id):
  return './datasets/s2/doi/{}/{}/paper.doi2'.format(paper_id[0:2], paper_id)
def old_url_fn(paper_id):
  return './datasets/s2/doi/{}/{}/paper.url2'.format(paper_id[0:2], paper_id)

if __name__ == '__main__':
  fetch_doi_list()