def deal_captcha(self, url, outdir, filename, chunk_size):
"""deal with the captcha
"""
soup = WR.get_soup(url)
img = soup.select_one('#captcha').attrs['src']
img_url = url.rsplit('/', 3)[0] + img
# print(img_url)
self.logger.warning(f'need to type in the captcha: {img_url}')
if os.getenv('DISPLAY'):
self.logger.info(
f'failed to open the picture, you can open it in your browser: {img_url}'
)
else:
content = WR.get_response(img_url, max_try=1).content
im = Image.open(io.BytesIO(content))
# im.show() # this will block the program
im.save('tmp.png')
# ****************************
# ***** non-blocked mode *****
# ****************************
pylab.ion()
img = pylab.imread('tmp.png')
pylab.imshow(img)
pylab.show()
while True:
answer = click.prompt('please input the captcha')
if answer == 'new':
return self.download(url,
outdir=outdir,
filename=filename,
chunk_size=chunk_size)
payload = {
'id': img_url.split('/')[-1].split('.')[0],
'answer': answer
}
# payload = {'id': '6058249282282', 'answer': 'manila'}
self.logger.debug(payload)
resp = WR.get_response(url,
method='POST',
stream=True,
data=payload)
if resp.headers['Content-Type'] == 'application/pdf':
pylab.close()
return resp
self.logger.warning('bad captcha, try again!')
def list_provinces(self):
"""省份列表
"""
self.logger.debug('list provinces ...')
soup = WR.get_soup(self.index_url)
province_list = [
each.attrs['value']
for each in soup.select('#province_main option[value!=""]')
]
return province_list
def search(self, term, max_try=3):
"""
term: URL, PMID, DOI or search string
return: the url of pdf
"""
soup = WR.get_soup(self.url)
form = soup.select_one('form[method="POST"]')
post_url = self.url if form.attrs['action'] == '/' else form.attrs[
'action']
payload = {'sci-hub-plugin-check': '', 'request': term}
self.logger.debug(f'search pdf url for: {term}')
while max_try:
max_try -= 1
soup = WR.get_soup(post_url,
method='POST',
data=payload,
timeout=self.timeout)
pdf = soup.select_one('#pdf')
if 'article not found' in soup.text:
self.logger.warning(f'article not found [{term}]')
return
elif not pdf:
# print(soup.select('title'))
continue
pdf_url = pdf.attrs['src']
if pdf_url.startswith('//'):
pdf_url = post_url.split(':')[0] + f':{pdf_url}'
self.logger.info(f'pdf url: {pdf_url}')
return pdf_url
self.logger.error(
f'your searching has no result, please check! [{term}]')
def check_host():
"""
checking available urls of Sci-Hub
"""
url = 'https://lovescihub.wordpress.com/'
soup = WR.get_soup(url)
text = soup.select_one('.entry-content p:nth-child(2)').text
update_text = soup.select_one('.entry-title').text
update_time = date_parse(
re.findall(r'Last check time:(.+?)\)', update_text)[0])
hosts = dict(re.findall(r'(http.+?)\s+(.+?s)', text))
return hosts, update_time
def list_support_types(self):
"""项目类别列表
Bug: 网页显示：应急管理项目
实际应该：科学部主任基金项目/应急管理项目
"""
self.logger.debug('list support types ...')
soup = WR.get_soup(self.index_url)
subcategory_list = []
for option in soup.select('select#subcategory option')[1:]:
if option.text == '应急管理项目':
text = '科学部主任基金项目/应急管理项目'
else:
text = option.text
subcategory_list.append(text)
return subcategory_list
def search_page(self, params, payload):
"""查询页面
"""
self.logger.debug(
f'searching for: {payload} [page: {params["currentpage"]}]')
while True:
soup = WR.get_soup(self.search_url,
method='POST',
params=params,
data=payload)
if not soup.select_one('#dict div b'):
self.logger.warning(f'{soup.text}')
if '需要先注册登录' in soup.text:
exit()
time.sleep(30)
continue
time.sleep(random.randint(5, 10))
return soup
def fetch_factor(*kws):
for kw in kws:
url = BASE_URL + kw
resp = WebRequest.get_response(url)
soup = WebRequest.get_soup(resp)
context = {}
trs = soup.select('table tr')
if len(trs) > 2:
print('multiple result for kw: {}'.format(kw))
elif len(trs) < 2:
print('no result for kw: {}'.format(kw))
else:
title = [th.text for th in trs[0].find_all('th')[2:]]
values = [td.text for td in trs[1].find_all('td')[2:]]
if values[-1]:
context['factor_history'] = json.dumps(dict(zip(title,
values)))
context['factor'] = values[-1]
context['kw'] = kw
if context:
return context
def get_soup(self, url):
soup = WR.get_soup(url)
return soup
def list_projects(cls):
soup = WR.get_soup(cls.url)
for box in soup.select('.input-area .ms-checkbox input'):
yield box.attrs['value']
def get_detail(cls, url):
soup = WR.get_soup(url)
for column in soup.select('.journal-content .journal-content-column'):
key = column.select_one('.column-label').text
value = column.select_one('.font-black').text.strip()
yield key, value
def search(cls,
page=1,
txtitle='',
project_classname_list='',
date_begin='',
date_end='',
**kwargs):
params = {
'txtitle': txtitle,
'page': page,
'project_classname_list': project_classname_list,
'cost_begin': '',
'cost_end': '',
'date_begin': date_begin,
'date_end': date_end,
'sort_type': '3',
}
soup = WR.get_soup(cls.url, params=params)
total_count = int(
re.findall(r'\d+',
soup.select_one('.list-result').text)[0])
total_page = math.ceil(total_count / 15.)
click.secho(f'total page: {total_page}, total count: {total_count}',
err=True,
fg='yellow')
if total_count == 500:
click.secho(
f'too many results: {params}, searching by each project ...',
err=True,
fg='yellow')
for project in cls.list_projects():
if params['project_classname_list']:
click.secho(f'still too many results: {params} ...',
err=True,
fg='red')
exit(1)
params['project_classname_list'] = project
yield from cls.search(**params)
for page in range(1, total_page + 1):
click.secho(f'>>> crawling page: {page}/{total_page}',
err=True,
fg='green')
params['page'] = page
soup = WR.get_soup(cls.url, params=params)
for a in soup.select('#journalList .journal-item strong a'):
click.secho(str(a), err=True, fg='white')
context = {}
href = a.attrs['href']
data = dict(list(cls.get_detail(href)))
context['title'] = data['项目名称']
context['project_id'] = data['项目批准号']
context['project_type'] = data['资助类型']
context['person'] = data['负责人']
context['institution'] = data['依托单位']
context['money'] = data['批准金额'].strip('万元')
context['approval_year'] = data['批准年份']
context['subject_code'] = data['学科分类'].split()[0]
context['start_time'], context['end_time'] = data[
'起止时间'].split('-')
yield context
