def upsert_paper(info, author_id):
try:
with db_session:
publisher_id = info['publisher_id']
if publisher_id:
publisher = Publisher.get(publisher_id=publisher_id)
if publisher:
pass
else:
publisher = Publisher(publisher_id=publisher_id)
publisher.name = info['publishername']
except:
pass
with db_session:
p = Paper.get(paper_id=info['paper_id'])
if p:
logger.debug('paper has existed, paper_id=%s', info['paper_id'])
return
paper = Paper(paper_id=info['paper_id'])
paper.title = info['title']
paper.abstract = info['abstract']
paper.cite_num = info['cite_num']
paper.cited_num = info['cited_num']
publisher_id = info['publisher_id']
if publisher_id:
publisher = Publisher.get(publisher_id=publisher_id)
if publisher:
paper.publisher = publisher
if author_id is None:
return
a = Author.get(author_id=author_id)
if a:
paper.authors.add(a)
else:
a_info = api.get_author(author_id)
author = Author(author_id=a_info['author_id'])
author.name = a_info['name']
author.image_url = a_info['image_url']
author.organization = a_info['organization']
author.home_page = a_info['home_page']
author.paper_count = a_info['paper_count']
author.citied_count = a_info['cited_count']
paper.authors.add(author)
def add(search_query, author, title):
fl = [
'id', 'author', 'first_author', 'bibcode', 'id', 'year', 'title',
'abstract', 'doi', 'pubdate', "pub", "keyword", "doctype",
"identifier", "links_data"
]
if author:
search_query += "author:" + author
if title:
search_query += "title:" + title
papers = list(ads.SearchQuery(q=search_query, fl=fl))
if len(papers) == 0:
selection = ads.search.Article
exit()
elif len(papers) == 1:
selection = papers[0] # type:ads.search.Article
else:
# first_ten = itertools.islice(papers, 10)
first_ten = papers[:10]
single_paper: ads.search.Article
for index, single_paper in enumerate(first_ten):
print(index, single_paper.title[0], single_paper.first_author)
selected_index = click.prompt('select paper', type=int)
selection = papers[selected_index] # type:ads.search.Article
assert len(selection.doi) == 1
doi = selection.doi[0]
try:
paper = Paper.get(Paper.doi == doi)
print("this paper has already been added")
exit(1)
except peewee.DoesNotExist:
pass
print("fetching bibcode")
q = ads.ExportQuery([selection.bibcode])
bibtex = q.execute()
print("saving in db")
paper = Paper()
assert len(selection.title) == 1
paper.doi = doi
paper.title = selection.title[0]
paper.abstract = selection.abstract
paper.bibcode = selection.bibcode
paper.year = selection.year
paper.pubdate = selection.pubdate
paper.pdf_downloaded = False
paper.first_author = Author.get_or_create(name=selection.first_author)[0]
paper.publication = Publication.get_or_create(name=selection.pub)[0]
paper.doctype = Doctype.get_or_create(name=selection.doctype)[0]
paper.arxiv_identifier = [
ident for ident in selection.identifier if "arXiv:" in ident
][0].split("arXiv:")[-1]
paper.bibtex = bibtex
links = [json.loads(string) for string in selection.links_data]
print(links)
paper.save()
authors = [Author.get_or_create(name=name)[0] for name in selection.author]
for author in db.batch_commit(authors, 100):
PaperAuthors.create(author=author, paper=paper)
keywords = [
Keyword.get_or_create(keyword=keyword)[0]
for keyword in selection.keyword
]
for keyword in db.batch_commit(keywords, 100):
PaperKeywords.create(keyword=keyword, paper=paper)
print("fetching PDF")
arxiv_url = "https://arxiv.org/pdf/{id}".format(id=paper.arxiv_identifier)
r = requests.get(arxiv_url, stream=True)
print(arxiv_url)
with open('library/{filename}.pdf'.format(filename=paper.id), 'wb') as f:
chunk_size = 1024 # bytes
file_size = int(r.headers.get('content-length', 0))
progress_length = math.ceil(file_size // chunk_size)
with click.progressbar(r.iter_content(chunk_size=20),
length=progress_length) as progress_chunks:
for chunk in progress_chunks:
f.write(chunk)
paper.pdf_downloaded = True
paper.save()
def get_references_citations_by_id(profile_id):
if isinstance(profile_id, dict):
profile_id = profile_id.get('profile_id')
if MONGO:
if data_collection.find({"id": profile_id}).count() > 0:
# 说明这个数据已经被爬取过了
return []
print('func2')
if not profile_id:
return -1
headers = {
'user-agent':
'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/77.0.3865.75 Safari/537.36',
'accept-language': 'zh-CN,zh;q=0.9'
}
session = requests.Session()
while True:
try:
response = session.get(
'https://cn.bing.com/academic/profile?id={}&encoded=0&v=paper_preview&mkt=zh-cn'
.format(profile_id),
headers=headers)
response.raise_for_status()
response.encoding = 'utf-8'
break
except KeyboardInterrupt:
raise KeyboardInterrupt
except Exception as e:
time.sleep(3.0)
print(e)
result = re.search(r'IG:"(.*?)"', response.text)
if result:
ig = result.group(1)
result = re.search(
r'被 引 量</span></span><span class="aca_content"><div>(\d*)</div>',
response.text)
if result:
citation_num = result.group(1)
html = etree.HTML(response.text)
paper = Paper(save2mongo=MONGO)
try:
paper.title = html.xpath('//li[@class="aca_title"]/text()')[0]
paper.id = profile_id
paper.citation_num = citation_num
result = re.search(
r'<span class="aca_label">DOI</span></span><span class="aca_content"><div>(.*?)</div>',
response.text)
if result:
paper.doi = result.group(1)
paper.authors = html.xpath(
'//div[@class="aca_desc b_snippet"]/span//a/text()')
paper.abstract = html.xpath(
'//div[@class="aca_desc b_snippet"]/span[1]//text()')[-1]
result = re.search(
r'<span class="aca_label">发表日期</span></span><span class="aca_content"><div>(\d*)</div>',
response.text)
if result:
paper.publish_year = result.group(1)
base_url = 'https://cn.bing.com/academic/papers?ajax=scroll&infscroll=1&id={id}&encoded=0&v=paper_preview&mkt=zh-cn&first={first}&count={count}&IG={ig}&IID=morepage.{num}&SFX={num}&rt={rt}'
count = 9
citation_links = list()
for i in range(1, int(citation_num) // count):
ajax_url = base_url.format(id=profile_id,
first=i * (count + 1),
count=count + 1,
ig=ig,
num=i,
rt='2')
while True:
try:
response = session.get(ajax_url, headers=headers)
response.raise_for_status()
response.encoding = 'utf-8'
break
except KeyboardInterrupt:
raise KeyboardInterrupt
except Exception as e:
time.sleep(3.0)
print(e)
html = etree.HTML(response.text)
citation_links.extend(html.xpath('//a[@target="_blank"]/@href'))
print('number of citation_links', len(citation_links), 'citation_num',
citation_num)
if len(citation_links) >= 0:
for i, citation_link in enumerate(citation_links):
profile_id = get_profile_id(citation_link)
if profile_id.get('title', False):
paper.citations.append(profile_id)
print('get_profile_id: {}/{}\r'.format(i + 1,
len(citation_links)),
end='')
print('\nnumber of ids:', len(paper.citations))
except KeyboardInterrupt:
raise KeyboardInterrupt
except Exception as e:
print(e)
paper.save()
# for profile_id in paper.citations:
# get_references_citations_by_id(profile_id)
return paper.citations