Befunde der Abschlussprüfung behoben
This commit is contained in:
23 files changed
+651
-44
No files matched your search
@@ -8,6 +8,7 @@ mailbox credentials come from the environment.
|
||||
|
||||
import datetime
|
||||
import email
|
||||
import html
|
||||
import imaplib
|
||||
import json
|
||||
import logging
|
||||
@@ -54,6 +55,8 @@ UNSUBSCRIBE_PATTERN = re.compile(
|
||||
)
|
||||
URL_PATTERN = re.compile(r"https?://[^\s<>\"']+")
|
||||
UID_PATTERN = re.compile(r"^\d+$")
|
||||
# Characters that end a sentence or close a quote or bracket in running text, not a URL.
|
||||
TRAILING_PUNCTUATION = ".,;:!?)\"'\u201d\u2019\u00bb\u203a"
|
||||
|
||||
|
||||
def is_safe_link(url: str) -> bool:
|
||||
@@ -89,7 +92,10 @@ def find_unsubscribe_links(text: str) -> list[str]:
|
||||
if UNSUBSCRIBE_PATTERN.search(str(anchor)) and is_safe_link(href) and href not in links:
|
||||
links.append(href)
|
||||
for url in URL_PATTERN.findall(text):
|
||||
if UNSUBSCRIBE_PATTERN.search(url) and url not in links:
|
||||
# the text search also runs over raw HTML: unescape (&) like the parser does for hrefs, so the same
|
||||
# link is not taken twice, and drop punctuation that follows a link in running text
|
||||
url = html.unescape(url).rstrip(TRAILING_PUNCTUATION)
|
||||
if UNSUBSCRIBE_PATTERN.search(url) and is_safe_link(url) and url not in links:
|
||||
links.append(url)
|
||||
return links
|
||||
|
||||
@@ -168,6 +174,7 @@ if STORE:
|
||||
inbox=EMAIL_FOLDER, done_folder=DONE_FOLDER,
|
||||
retention_days=int(os.environ.get("RETENTION_DAYS", "30")),
|
||||
max_cases=int(os.environ.get("AGENT_MAX_CASES", "10")),
|
||||
max_deferrals=int(os.environ.get("AGENT_MAX_DEFERRALS", "12")),
|
||||
allowed_domains=_env_list("AGENT_ALLOWED_DOMAINS"),
|
||||
allowed_senders=_env_list("AGENT_ALLOWED_SENDERS"),
|
||||
)
|
||||
|
||||
Reference in new issue
Block a user