research/stock
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

modify some scripts.

Browse Source
This commit is contained in:
oscar
2025-03-09 16:52:32 +08:00
parent d9fc5be383
commit 4d5f9521ef
1 changed files with 7 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
scripts/iafd/src/iafd_scraper.py
View File

@ -298,6 +298,8 @@ def parse_credits_table(table, distributor_list):
# rows = table.find_all('tr', class_='we') # rows = table.find_all('tr', class_='we')
for row in rows: for row in rows:
#tr_class = row.get('class', '') # 获取 class 属性，如果没有则返回空字符串
tr_class = ' '.join(row.get('class', [])) # 获取 class 属性，如果没有则返回空字符串
cols = row.find_all('td') cols = row.find_all('td')
if len(cols) >= 6: if len(cols) >= 6:
title = cols[0].text.strip() title = cols[0].text.strip()
@ -305,6 +307,8 @@ def parse_credits_table(table, distributor_list):
href = href_a['href'] if href_a else '' href = href_a['href'] if href_a else ''
year = cols[1].text.strip() year = cols[1].text.strip()
distributor = cols[2].text.strip().lower() distributor = cols[2].text.strip().lower()
href_d = cols[2].find('a')
href_dist = host_url + href_d['href'] if href_d else ''
notes = cols[3].text.strip() notes = cols[3].text.strip()
rev = cols[4].text.strip() rev = cols[4].text.strip()
formats = cols[5].text.strip() formats = cols[5].text.strip()
@ -318,9 +322,11 @@ def parse_credits_table(table, distributor_list):
'href' : href, 'href' : href,
'year': year, 'year': year,
'distributor': distributor, 'distributor': distributor,
'distributor_href': href_dist,
'notes': notes, 'notes': notes,
'rev': rev, 'rev': rev,
'formats': formats 'formats': formats,
'tr_class': tr_class
}) })
return movies, distributor_count return movies, distributor_count