xevion/the-office
1
0
Fork 0
mirror of https://github.com/Xevion/the-office.git synced 2025-12-10 16:08:52 -06:00
Code Issues Packages Projects Releases Wiki Activity

attempt at fixing newpeat quotes breaking database building

Browse Source
This commit is contained in:
Xevion
2020-01-21 03:13:18 -06:00
parent 8a1b8060c0
commit 7cb1121cd0
1 changed files with 15 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
app/models.py
View File

@@ -71,21 +71,26 @@ class Episode(db.Model):
"""downloads, processes, and automatically creates Sections and Quotes""" """downloads, processes, and automatically creates Sections and Quotes"""
link = f'http://officequotes.net/no{self.season_id}-{str(self.number).zfill(2)}.php' link = f'http://officequotes.net/no{self.season_id}-{str(self.number).zfill(2)}.php'
data = requests.get(link).text data = requests.get(link).text
open('test.html', 'w+', encoding='utf-8').write(data)
soup = BeautifulSoup(data, 'html.parser') soup = BeautifulSoup(data, 'html.parser')
sections = soup.find_all(attrs={'class' : 'quote'}) sections = soup.find_all(attrs={'class' : 'quote'})
deleted = 0 deleted = 0
for section in sections: for section in sections:
try: isNewpeat = False
isNewpeat = False quotes = []
quotes = [] for quote in section.find_all('b'):
for quote in section.find_all('b'): if 'Newpeat' in quote.string:
print(quote.string.lower()) quote = quote.next_sibling
if 'newpeat' in quote.string.lower(): isNewpeat = True
isNewPeat = True if quote is None or quote.next_sibling is None:
else: print('Quote is None or next sibling is None')
quotes.append(quote.string + quote.next_sibling.string) continue
quotes.append(quote.string + quote.next_sibling.string)
if len(quotes) == 0:
print(f'Section found with Zero quotes. Newpeat: {isNewpeat}')
continue
isDeletedScene = quotes[0].lower().startswith('deleted scene') isDeletedScene = quotes[0].lower().startswith('deleted scene')
if isDeletedScene: if isDeletedScene:
deleted += 1 deleted += 1