myned/modufur
1
0
Fork 0
mirror of https://github.com/myned/modufur.git synced 2024-11-01 13:02:38 +00:00
Code Issues Activity

Fix Kheina parsing and eval

Browse source
This commit is contained in:
Myned 2019-09-24 19:13:10 -04:00
parent 1262dc2ba7
commit 5987b46925
No known key found for this signature in database
GPG key ID: BC58C09870A63E59
1 changed files with 18 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

27
src/utils/scraper.py
View file

@ -12,7 +12,7 @@ from utils import utils as u
# async def get_harry(url): # async def get_harry(url):
# content = await u.fetch('https://iqdb.harry.lu', params={'url': url}) # content = await u.fetch('https://iqdb.harry.lu', params={'url': url})
# soup = BeautifulSoup(content, 'html.parser') # soup = BeautifulSoup(content, 'html5lib')
# #
# if soup.find('div', id='show1').string is 'Not the right one? ': # if soup.find('div', id='show1').string is 'Not the right one? ':
# parent = soup.find('th', string='Probable match:').parent.parent # parent = soup.find('th', string='Probable match:').parent.parent
@ -41,17 +41,25 @@ from utils import utils as u
async def get_kheina(url): async def get_kheina(url):
content = await u.fetch('https://kheina.com', params={'url': url}) content = await u.fetch('https://kheina.com', params={'url': url}, text=True)
soup = BeautifulSoup(content, 'html.parser')
results = ast.literal_eval(soup.find('data', id='results').string)[-1] content = content.replace('"', 'quot;').replace(''', 'apos;')
iqdbdata = ast.literal_eval(soup.find('data', id='iqdbdata').string)[0] soup = BeautifulSoup(content, 'html5lib')
results = soup.find('data', id='results').string.replace('quot;', '"').replace('apos;', ''')
results = ast.literal_eval(results)
iqdbdata = soup.find('data', id='iqdbdata').string
iqdbdata = ast.literal_eval(iqdbdata)
for e in results:
if iqdbdata[0]['iqdbid'] in e:
match = e
break
result = { result = {
'source': results[3], 'source': match[3],
'artist': results[4], 'artist': match[4],
'thumbnail': f'https://f002.backblazeb2.com/file/kheinacom/{results[1]}.jpg', 'thumbnail': f'https://f002.backblazeb2.com/file/kheinacom/{match[1]}.jpg',
'similarity': str(int(float(iqdbdata['similarity']))), 'similarity': str(int(float(iqdbdata[0]['similarity']))),
'database': 'Kheina' 'database': 'Kheina'
} }
@ -63,6 +71,7 @@ async def get_saucenao(url):
'https://saucenao.com/search.php', 'https://saucenao.com/search.php',
params={'url': url, 'api_key': u.config['saucenao_api'], 'output_type': 2}, params={'url': url, 'api_key': u.config['saucenao_api'], 'output_type': 2},
json=True) json=True)
results = content['results'][0] results = content['results'][0]
for i in range(len(content['results'])): for i in range(len(content['results'])):
if 'e621' in content['results'][i]['header']['index_name']: if 'e621' in content['results'][i]['header']['index_name']: