Merge pull request #5 from abslimit/master

fixed issues with big non-space text chunks and media links at the st…
This commit is contained in:
ChronosX88 2020-09-27 10:46:26 +04:00 committed by GitHub
commit e12db036e0
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23

View File

@ -19,8 +19,7 @@ class BridgeBot:
def __init__(self, cfg: dict):
# RegExps
self.re_md_links = re.compile(r'\[([\w\d\s\-,?!.]+)\]\((https?:\/\/[\w\d./?=#&-]+)\)')
self.re_md_media_link = re.compile(r'^\[\]\(https?:\/\/telegra.ph[\w\d./?=#&-]+\)')
self.re_md_links = re.compile(r'\[(.*?)\]\((https?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\(\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+)\)')
# Config init
self.config = cfg
self.mastodon_clients = {}
@ -69,14 +68,14 @@ class BridgeBot:
logging.info(f"Catched new post from telegram channel {channel.username}")
# Common Mastodon message limit size. Change if you increased this limit.
mstdn_post_limit = 500
#full_text = event.message.raw_text
full_text = event.message.text
full_text = re.sub(self.re_md_media_link, '', full_text)
full_text = re.sub(self.re_md_links, r'\g<1> \g<2>', full_text)
# Uncomment for debug
#print(full_text)
full_text = full_text.replace('**', '')
full_text = full_text.replace('__', '')
full_text = full_text.replace('~~', '')
full_text = full_text.replace('`', '')
full_text = re.sub(self.re_md_links, r'\g<1> \g<2> ', full_text)
# URL of Telegram message
tg_message_url = f"[https://t.me/{channel.username}/" + str(event.message.id) + "]\n\n"
if event.message.file and not (event.message.photo or event.message.video or event.message.gif):
@ -94,7 +93,7 @@ class BridgeBot:
# Set reply_start to non zero for future chunking and make mstdn post with continuaniton note
logging.debug("full_text_size: " + str(full_text_size))
if full_text_size > mstdn_post_size:
long_post_tail = "\n\n[Откройте оригинальный пост по ссылке, либо прочитайте продолжение в ответах к посту.]"
long_post_tail = "\n\n[Откройте пост по ссылке или прочитайте продолжение в обсуждении]"
mstdn_post_size = mstdn_post_size - len(long_post_tail)
reply_start = full_text.rfind(' ', reply_start, mstdn_post_size)
post_text: str = tg_message_url + full_text[0:reply_start] + long_post_tail
@ -120,19 +119,23 @@ class BridgeBot:
else:
mstdn_media_meta = None
# First root mstdn post
reply_to = current_mastodon_client.status_post(post_text, media_ids=[mstdn_media_meta], visibility=current_mstdn_acc_visibility)
tg_message_url = f"[Продолжение https://t.me/{channel.username}/" + str(event.message.id) + "]\n\n"
# Chunking post into mstdn limit chunks and reply to root post
i = 0
emergency_break = 0
emergency_break_limit = 14
while reply_start + mstdn_post_limit < full_text_size:
logging.debug("while reply_start:" + str(reply_start))
reply_end = full_text.rfind(' ', reply_start, reply_start + mstdn_post_limit - (len(tg_message_url)*2))
reply_end = full_text.rfind(' ', reply_start, reply_start + mstdn_post_limit - len(tg_message_url)*2)
if reply_end == reply_start:
reply_end = reply_start + mstdn_post_limit - len(tg_message_url)*2
logging.debug("while reply_end:" + str(reply_end))
post_text: str = tg_message_url + full_text[reply_start+1:reply_end]
reply_to = current_mastodon_client.status_post(post_text, in_reply_to_id=reply_to, visibility=current_mstdn_acc_visibility)
reply_start = reply_end
i = i + 1
if i > 15:
# Emergency break for long or endlessly looped posts
emergency_break = emergency_break + 1
if emergency_break >= emergency_break_limit:
logging.debug("Breaking very long reply thread")
break
# Final chunk to reply to root post