def test_author(self):
with db_session:
a = Author(author_id='3333')
a.name = 'given'
p = Paper(paper_id='qqqq')
p.title = 'titttttttle'
a.papers.add(p)
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 write_db(self):
print "len of entry list " + str(len(self.entry_list))
for entry in self.entry_list:
paper = Paper()
if entry.has_key("id"):
paper.id = entry["id"]
if entry.has_key("type"):
paper.type = entry["type"]
if entry.has_key("title"):
paper.title = entry["title"]
if entry.has_key("author"):
paper.authors = entry["author"]
if entry.has_key("year"):
paper.year = int(entry["year"])
if entry.has_key("journal"):
paper.journal = entry["journal"]
if entry.has_key("booktitle"):
paper.book_title = entry["booktitle"]
if entry.has_key("publisher"):
paper.publisher = entry["publisher"]
if entry.has_key("institution"):
paper.institution = entry["institution"]
if entry.has_key("volume"):
paper.volume = int(entry["volume"])
if entry.has_key("number"):
paper.number = int(entry["number"])
if entry.has_key("pages"):
paper.pages = entry["pages"]
if entry.has_key("url"):
paper.url = entry["url"]
if entry.has_key("doi"):
paper.doi = entry["doi"]
if entry.has_key("isbn"):
paper.isbn = entry["isbn"]
paper.save()
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 post(self, request):
param = QueryDict(request.body)
uuid = param.get('uuid')
title = param.get('title')
time = param.get('time')
origin = param.get('origin')
_authors = param.getlist('authors')
link = param.get('link')
_tags = param.getlist('tags')
content = param.get('content')
refer_to = param.getlist('reference')
score = param.get('score')
try:
year, month = time.split('-')
year, month = int(year), int(month)
publish_time = datetime.date(year, month, 1)
except Exception as e:
logger.error(traceback.format_exc(e))
return JsonResponse({'msg': '提供的日期{}有误'.format(time)}, status=500)
for _tag in _tags:
try:
_tag = int(_tag)
_ = ResearchTag.objects.get(research_tag_id=_tag)
except Exception as e:
logger.error(traceback.format_exc(e))
return JsonResponse({'msg': '错误的标签{}'.format(_tag)},
status=500)
tags = ResearchTag.objects.filter(
research_tag_id__in=[int(_t) for _t in _tags])
author_ids = []
for _author in _authors:
if _author.isdigit():
author_ids.append(int(_author))
elif Author.objects.filter(name=_author).exists():
a = Author.objects.get(name=_author).author_id
author_ids.append(a)
else:
a = Author(name=_author)
a.save()
author_ids.append(a.author_id)
authors = Author.objects.filter(author_id__in=author_ids)
try:
score = int(score)
except Exception as e:
logger.error(traceback.format_exc(e))
return JsonResponse({'msg': '错误的评分分数格式'}, status=500)
if not Paper.objects.filter(paper_uuid=uuid).exists():
# 新建的场合
try:
comment = PaperComment(content=content)
comment.save()
paper = Paper(paper_uuid=uuid,
title=title,
publish_origin=origin,
publish_time=publish_time,
author=authors,
link=link,
tag=tags,
comment=comment,
self_score=score)
paper.save()
redis.set(self.LATEST_KEY, str(uuid_gen.uuid4()))
except Exception as e:
logger.error(traceback.format_exc(e))
return JsonResponse({'msg': '保存失败'}, status=500)
else:
return JsonResponse({
'next':
reverse('paperdb.detail',
kwargs={'paper_uuid': paper.paper_uuid})
})
try:
# 编辑的场合
paper = Paper.objects.get(paper_uuid=uuid)
except Exception as e:
logger.error(traceback.format_exc(e))
return JsonResponse({'msg': '错误的uuid/未找到相关论文记录'}, status=404)
else:
paper.title = title
paper.publish_time = publish_time
paper.publish_origin = origin
paper.author = authors
paper.link = paper.link
paper.tag = tags
paper.self_score = score
try:
paper.save()
except Exception as e:
logger.error(traceback.format_exc(e))
return JsonResponse({'msg': '保存失败'}, status=500)
if paper.comment is None:
if content != '':
comment = PaperComment(content=content)
comment.save()
paper.comment = comment
paper.save()
elif content != paper.comment.content.replace(
'\r\n', '\n'): # traditional下的换行符出入
paper.comment.content = content
paper.comment.save()
for refer_to_paper in Paper.objects.filter(paper_uuid__in=refer_to):
if not Reference.objects.filter(
reference_src=paper,
reference_trg=refer_to_paper).exists():
reference = Reference(reference_src=paper,
reference_trg=refer_to_paper)
reference.save()
return JsonResponse({
'next':
reverse('paperdb.detail', kwargs={'paper_uuid': paper.paper_uuid})
})
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
def test_paper(self):
with db_session:
p = Paper(paper_id='666aaaaa66')
p.title = 'hehehehhehe'
