main.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
r"""
Life's pathetic, have fun ("▔□▔)/hi~♡ Nasy.

Excited without bugs::

    |             *         *
    |                  .                .
    |           .
    |     *                      ,
    |                   .
    |
    |                               *
    |          |\___/|
    |          )    -(             .              '
    |         =\  -  /=
    |           )===(       *
    |          /   - \
    |          |-    |
    |         /   -   \     0.|.0
    |  NASY___\__( (__/_____(\=/)__+1s____________
    |  ______|____) )______|______|______|______|_
    |  ___|______( (____|______|______|______|____
    |  ______|____\_|______|______|______|______|_
    |  ___|______|______|______|______|______|____
    |  ______|______|______|______|______|______|_
    |  ___|______|______|______|______|______|____

* author: Nasy https://nasy.moe <Nasy>
* date: Apr 2, 2018
* email: echo bmFzeXh4QGdtYWlsLmNvbQo= | base64 -D
* filename: main.py
* Last modified time: Apr 2, 2018
* license: MIT

A crawler that crawl the data from http://apps.webofknowledge.com .

There are more things in heaven and earth, Horatio, than are dreamt.
 --  From "Hamlet"
"""
import re

import bs4
import requests as req

JURNAL_LIST = re.findall(r"\n\([0-9]+\)\. (.+)", """
(1). Annals of Statistics
(2). Journal of the American Statistical Association
(3). Journal of the Royal Statistical Society: Series B
(4). Biometrika
(5). Biometrics
(6). Journal of the Royal Statistical Society: Series A
(7). Econometrica
(8). Journal of Computational and Graphical Statistics
(9). Journal of Multivariate Analysis
(10). Journal of the Royal Statistical Society: Series C
(11). Journal of Statistical Software
(12). Computational Statistics & Data Analysis
(13). Journal of Business & Economic Statistics
(14). Journal of Econometrics
""")
JURNALS = tuple(enumerate(JURNAL_LIST, 1))

BASEURL = "http://apps.webofknowledge.com"
URL = (
    "http://apps.webofknowledge.com/summary.do?product=UA&parentProduct=UA"
    f"&search_mode=GeneralSearch&parentQid=&qid={qid}&SID=7BTagtKnmliEPhll9jz&"
    f"&update_back2search_link_param=yes&page={page}")