Disclaimer

I am not responsible for the illegitimate use of this tool. For example, the download of non-open-access papers or even those if this method is not allowed by the editors.

Download

Download pdfs from Scihub via DOI.

Easy to use.
Easy to deal with captcha.
Easy to update Scihub newest domains.

Install

$ pip3 install -U scidownl

Usage

Command line

$ scidownl -h
usage: Command line tool to download pdf via DOI from Scihub.
       [-h] [-c CHOOSE] [-D DOI] [-o OUTPUT] [-u] [-l]

optional arguments:
  -h, --help            show this help message and exit
  -c CHOOSE, --choose CHOOSE
                        choose scihub url by index
  -D DOI, --DOI DOI     the DOI number of the paper
  -o OUTPUT, --output OUTPUT
                        directory to download the pdf
  -u, --update          update available Scihub links
  -l, --list            list current saved sichub urls

Examples

# Update available links of Scihub
$ scidownl -u
[INFO] Updating links ...
[INFO] https://sci-hub.ren
[INFO] http://sci-hub.ren
[INFO] http://sci-hub.red
[INFO] http://sci-hub.se
[INFO] https://sci-hub.se
[INFO] http://sci-hub.tw

# Choose scihub url by the index.
$ scidownl -c 5
Current scihub url: http://sci-hub.tw

# List available links of Scihub. You can see the current scihub url is pointing to the 5th scihub url.
$ scidownl -l
  [0] https://sci-hub.ren
  [1] http://sci-hub.ren
  [2] http://sci-hub.red
  [3] http://sci-hub.se
  [4] https://sci-hub.se
* [5] http://sci-hub.tw

# Download to the current directory
$ scidownl -D 10.1021/ol9910114
$ scidownl -D 10.1021/ol9910114 -o .

# Download to the specified directory, ie. '-o paper' for downloading to paper directory.
$ scidownl -D 10.1021/ol9910114 -o paper

# if 'PermessionError' shows, just use sudo. ie:
$ sudo scidownl -u

Module

If you have a list of DOIs, using scidownl in your python scripts for downloading all of the papers is recommended.

Download single paper via DOI.

from scidownl.scihub import *

DOI = "10.1021/ol9910114"
out = 'paper'
sci = SciHub(DOI, out).download(choose_scihub_url_index=3)

Dowloading a list of DOIS by simply using a for loop.

from scidownl.scihub import *

DOIs = ['doi1','doi2','doi3','doi4','doi5']
out = 'paper'
for doi in DOIs:
  SciHub(doi, out).download(choose_scihub_url_index=3)

from scidownl.scihub import *

a_file = open("doilist.txt", "r")

DOIs = []
for line in a_file:
  DOIs.append(line)
a_file.close()

out = 'paper'
for doi in DOIs:
  print(doi)
  SciHub(doi, out).download(choose_scihub_url_index=1)

Update available Scihub links.

from scidownl.update_link import *

# Use crawling method to update available Scihub links.
update_link(mod='c')
# Use brute force search method to update available Scihub links.
update_link(mod='b')

RELEASE

v0.1.0: First release.
v0.2.0:
- Optimized the download speed.
- Optimized the captcha processment.
v0.2.1:
- Applied stream download.
- Display of download progress is added.
- Fixed bugs of invalid scihub links.
v0.2.2:
- Add new source website.
- Add -l/--list argument in command line tool.
v0.2.3:
- Fix bugs of empty filename and wrong scidhub urls.
- Fix bugs in the brute-force method of updating scihub urls.
V0.2.4:
- Fix #2.
- Fix bugs of error: file name too long.
V0.2.5:
- Reconstruct code.
- Fix 'no content-length' error.
- Add -c/--choose argument for manually choosing scihub url used.
V0.2.6:
- Fix bug where retry time too long.
V0.2.7:
- Add -b/--brute-update argument for updating scihub urls by brute-force search method.
V0.2.8:
- Replace the mspider with qspider in brute-force search.

LICENSE

Licensed under the MIT License.

0xFORK/doi2pdf