indymedia
/
epicyon
mirror of https://gitlab.com/bashrc2/epicyon
7
4
Fork 1
Code Issues 14 Releases Wiki Activity

Remove zero length strings from feeds

merge-requests/30/head
Bob Mottram 2023-11-29 11:37:44 +00:00
parent 29f665a116
commit dd55ad92e6
2 changed files with 8 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
newswire.py
View File

@ -19,6 +19,7 @@ from datetime import timezone
from collections import OrderedDict from collections import OrderedDict
from utils import valid_post_date from utils import valid_post_date
from categories import set_hashtag_category from categories import set_hashtag_category
from utils import remove_zero_length_strings
from utils import date_from_string_format from utils import date_from_string_format
from utils import acct_handle_dir from utils import acct_handle_dir
from utils import remove_eol from utils import remove_eol
@ -1373,6 +1374,7 @@ def get_rss(base_dir: str, domain: str, session, url: str,
timeout=timeout_sec, timeout=timeout_sec,
allow_redirects=True) allow_redirects=True)
if result: if result:
result.text = remove_zero_length_strings(result.text)
if int(len(result.text) / 1024) >= max_feed_size_kb: if int(len(result.text) / 1024) >= max_feed_size_kb:
print('WARN: feed is too large: ' + url) print('WARN: feed is too large: ' + url)
elif not contains_invalid_chars(result.text): elif not contains_invalid_chars(result.text):

6
utils.py
View File

@ -46,6 +46,12 @@ INVALID_ACTOR_URL_CHARACTERS = (
) )
def remove_zero_length_strings(text: str) -> str:
"""removes zero length strings from text
"""
return text.replace('​', '')
def _utc_mktime(utc_tuple): def _utc_mktime(utc_tuple):
"""Returns number of seconds elapsed since epoch """Returns number of seconds elapsed since epoch
Note that no timezone are taken into consideration. Note that no timezone are taken into consideration.