liz/PluralKit
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add proxy permission errors, refactor proxy matching, change member name length error handling. Closes #25.

Browse Source
This commit is contained in:
Ske
2018-11-30 21:42:01 +01:00
parent 5d3cb7b6bf
commit c36a054519
7 changed files with 149 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files

41
src/pluralkit/system.py
View File

@@ -1,9 +1,10 @@
import random
import re
import string
from datetime import datetime
from collections.__init__ import namedtuple
from typing import Optional, List
from typing import Optional, List, Tuple
from pluralkit import db, errors
from pluralkit.member import Member
@@ -63,11 +64,6 @@ class System(namedtuple("System", ["id", "hid", "name", "description", "tag", "a
if contains_custom_emoji(new_tag):
raise errors.CustomEmojiError()
# Check name+tag length for all members
members_exceeding = await db.get_members_exceeding(conn, system_id=self.id, length=32 - len(new_tag) - 1)
if len(members_exceeding) > 0:
raise errors.TagTooLongWithMembersError([member.name for member in members_exceeding])
await db.update_system_field(conn, self.id, "tag", new_tag)
async def set_avatar(self, conn, new_avatar_url: Optional[str]):
@@ -133,12 +129,43 @@ class System(namedtuple("System", ["id", "hid", "name", "description", "tag", "a
return None
def get_member_name_limit(self) -> int:
"""Returns the maximum length a member's name or nickname is allowed to be. Depends on the system tag."""
"""Returns the maximum length a member's name or nickname is allowed to be in order for the member to be proxied. Depends on the system tag."""
if self.tag:
return 32 - len(self.tag) - 1
else:
return 32
async def match_proxy(self, conn, message: str) -> Optional[Tuple[Member, str]]:
"""Tries to find a member with proxy tags matching the given message. Returns the member and the inner contents."""
members = await db.get_all_members(conn, self.id)
# Sort by specificity (members with both prefix and suffix defined go higher)
# This will make sure more "precise" proxy tags get tried first and match properly
members = sorted(members, key=lambda x: int(bool(x.prefix)) + int(bool(x.suffix)), reverse=True)
for member in members:
proxy_prefix = member.prefix or ""
proxy_suffix = member.suffix or ""
# Check if the message matches these tags
if message.startswith(proxy_prefix) and message.endswith(proxy_suffix):
# If the message starts with a mention, "separate" that and match the bit after
mention_match = re.match(r"^(<(@|@!|#|@&|a?:\w+:)\d+>\s*)+", message)
leading_mentions = ""
if mention_match:
message = message[mention_match.span(0)[1]:].strip()
leading_mentions = mention_match.group(0)
# Extract the inner message (special case because -0 is invalid as an end slice)
if len(proxy_suffix) == 0:
inner_message = message[len(proxy_prefix):]
else:
inner_message = message[len(proxy_prefix):-len(proxy_suffix)]
# Add the stripped mentions back if there are any
inner_message = leading_mentions + inner_message
return member, inner_message
def to_json(self):
return {
"id": self.hid,