Handle newlines in title HTML.
This commit is contained in:
parent
2f4f1481c7
commit
52dfe9e42c
4
check.py
4
check.py
|
@ -9,7 +9,7 @@ from email.utils import formatdate, make_msgid
|
|||
|
||||
import requests
|
||||
|
||||
re_title = re.compile("<title>(.*?)</title>")
|
||||
re_title = re.compile("<title>(.*?)</title>", re.DOTALL)
|
||||
|
||||
AGENT = "Mozilla/5.0 (Windows NT 6.1) Gecko/20100101 Firefox/29.0"
|
||||
headers = {"User-Agent": AGENT, "Accept": "text/html"}
|
||||
|
@ -56,7 +56,7 @@ def find_not_here_message(html: str) -> str | None:
|
|||
def get_title(html: str) -> str:
|
||||
"""Title from web page."""
|
||||
m = re_title.search(html)
|
||||
return m.group(1) if m and m.group(1) else "no title"
|
||||
return m.group(1).strip() if m and m.group(1) else "no title"
|
||||
|
||||
|
||||
def check_conference(name: str, url: str) -> tuple[bool, str]:
|
||||
|
|
Loading…
Reference in a new issue