phyks/bmc
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

another fix for sciencedirect.com

Browse Source
This commit is contained in:
Bryan Bishop 2013-01-20 22:21:44 -06:00
parent a89129b424
commit eba857dd7e
1 changed files with 4 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
modules/papers.py
View File

@ -172,6 +172,10 @@ def download_url(url):
if "sciencedirect.com" in url: if "sciencedirect.com" in url:
title = tree.xpath("//h1[@class='svTitle']")[0].text title = tree.xpath("//h1[@class='svTitle']")[0].text
pdf_url = tree.xpath("//a[@id='pdfLink']/@href")[0] pdf_url = tree.xpath("//a[@id='pdfLink']/@href")[0]
response = requests.get(pdf_url, headers={"User-Agent": "sdf-macross"})
content = response.content
if "pdf" in response.headers["content-type"]:
extension = ".pdf"
elif "h1 class=\"articleTitle" in content: elif "h1 class=\"articleTitle" in content:
try: try:
title = tree.xpath("//h1[@class='articleTitle']")[0].text title = tree.xpath("//h1[@class='articleTitle']")[0].text