wikilink with semantic syntax :)

This commit is contained in:
sneakers-the-rat 2023-05-26 01:18:54 -07:00
parent 4f239ea0d7
commit 01b1860e4d
10 changed files with 1188 additions and 574 deletions

View File

@ -4,3 +4,45 @@ Bot to add tweets using an extended wikilink syntax to the wiki
Starting with twitter, but then will add masto
# Mediawiki
https://www.mediawiki.org/wiki/Manual:Creating_a_bot
- Go to `Special:BotPasswords`
- Permissions
- Edit existing pages
- Create, edit, and move pages
-
- save to `mediawiki_creds.json` like
```json
{
"user": "Jonny@wikibot",
"password": "<THE BOT PASSWORD>"
}
```
# Slack
Using the `python-slack-sdk` https://slack.dev/python-slack-sdk/
## Make App
https://slack.dev/python-slack-sdk/socket-mode/index.html
- Make slack app - https://api.slack.com/apis/connections/socket#setup
- Assign to workplace
- Permissions
- `channels:history`
- `channels:read`
- `app_mentions:read`
- `chat:write`
- `links:read`
- `reactions:read`
- `reactions:write`
- `users:read`
- Create App-level token with `connections:write`
- Configure app
- Enable socket mode - https://api.slack.com/apis/connections/socket#toggling
- Enable Events - `message.channels`
-

1238
poetry.lock generated

File diff suppressed because it is too large Load Diff

View File

@ -10,8 +10,10 @@ packages = [
[tool.poetry.scripts]
wikipostbot = "wiki_postbot.main:main"
discord_bot = "wiki_postbot.clients.discord_client:main"
discord_bot = "wiki_postbot.clients.discord:main"
install_discord_bot = "wiki_postbot.service:main"
slack_bot = "wiki_postbot.clients.slack:main"
install_slack_bot = "wiki_postbot.service:main_slack"
[tool.poetry.dependencies]
python = "^3.9"
@ -20,10 +22,21 @@ rich = "^12.4.4"
parse = "^1.19.0"
pywikibot = "^7.7.0"
pyparsing = "^3.0.9"
"discord.py" = "^2.0.1"
wikitextparser = "^0.51.1"
certifi = "^2022.9.24"
[tool.poetry.group.discord]
optional = true
[tool.poetry.group.discord.dependencies]
"discord.py" = "^2.0.1"
[tool.poetry.group.slack]
optional = true
[tool.poetry.group.slack.dependencies]
slack-sdk = "^3.21.3"
[tool.poetry.dev-dependencies]
pytest = "^7.1.2"
Faker = "^15.1.0"

View File

@ -0,0 +1,91 @@
from abc import ABC, abstractmethod
from pathlib import Path
from wiki_postbot.interfaces.mediawiki import Wiki
from wiki_postbot.logger import init_logger
from wiki_postbot.patterns.wikilink import Wikilink
from typing import TYPE_CHECKING, List
if TYPE_CHECKING:
from logging import Logger
class Client(ABC):
"""
Metaclass for different clients (things that receive inputs like discord/slack messages)
pending some actual formalization of the concepts in this library (like idk what do we
call the wiki, a sink? idk)
Subclasses must implement all the abstract methods (that is, i guess, the definition
of abstract methods) but may do whatever other logic they need to do to handle messages,
eg. discord bot subclasses
"""
def __init__(
self,
wiki:Wiki,
name: str = "wikibot_client",
log_dir: Path = Path('/var/log/wikibot'),
):
super(Client, self).__init__()
self._wiki = None
self.wiki = wiki
self.name = name
self.log_dir = Path(log_dir)
self.logger = self._init_logger(name=self.name, log_dir=self.log_dir)
# --------------------------------------------------
# abstract methods
# --------------------------------------------------
# @abstractmethod
# def _react_progress(self, message):
# """React to a message indicating that we are processing it"""
# pass
# @abstractmethod
# def _react_complete(self, message):
# """React to a message indicating that we have completed processing it"""
# pass
# @abstractmethod
# def _react_error(self, message):
# """React to a message indicating there was some error"""
@abstractmethod
def parse_wikilinks(self, message) -> List[Wikilink]:
"""Parse a given message, returning the wikilinks it contains"""
@abstractmethod
def handle_wikilinks(self, message, wl: List[Wikilink]):
"""Handle the message with wikilinks, most likely by posting it to the wiki!"""
# --------------------------------------------------
# Properties
# --------------------------------------------------
@property
def wiki(self) -> Wiki:
return self._wiki
@wiki.setter
def wiki(self, wiki: Wiki):
if wiki.sess is None:
raise RuntimeError("Wiki client is not logged in! Login before passing to client")
self._wiki = wiki
# --------------------------------------------------
# Private methods
# --------------------------------------------------
def _init_logger(self, name:str, log_dir:Path) -> 'Logger':
# Try and make log directory, if we cant, it should fail.
log_dir.mkdir(exist_ok=True)
logger = init_logger(name=name, basedir=log_dir)
return logger

View File

@ -8,42 +8,40 @@ from wiki_postbot.creds import Discord_Creds, Mediawiki_Creds
from wiki_postbot.patterns.wikilink import Wikilink
from wiki_postbot.interfaces.mediawiki import Wiki
from wiki_postbot.logger import init_logger
from wiki_postbot.clients.client import Client as Wikibot_Client
from discord.ext import commands
from discord import Emoji
import pdb
class DiscordClient(Client):
class DiscordClient(Client, Wikibot_Client):
def __init__(
self,
wiki:Wiki,
name:str='discord_bot',
intents=None,
debug:bool=False,
reply_channel:str="wikibot",
log_dir:Path=Path('/var/log/wikibot'),
**kwargs
):
Wikibot_Client.__init__(
self,
wiki=wiki,
name=name,
log_dir=log_dir)
if intents is None:
intents = Intents.default()
intents.message_content = True
self.wiki = wiki
if self.wiki.sess is None:
raise RuntimeError("Wiki client is not logged in! Login before passing to discord client")
Client.__init__(self, intents=intents, **kwargs)
self.debug = debug
self.reply_channel_name = reply_channel
self.reply_channel = None # type: Optional[discord.TextChannel]
self.log_dir = Path(log_dir)
# Try and make log directory, if we cant, it should fail.
self.log_dir.mkdir(exist_ok=True)
self.logger = init_logger(name="discord_bot", basedir=self.log_dir)
super(DiscordClient, self).__init__(intents=intents, **kwargs)
async def get_channel(self, channel_name:str) -> discord.TextChannel:
channel = discord.utils.get(self.get_all_channels(), name=channel_name)
self.logger.debug(f"Got channel {channel}")
@ -69,15 +67,15 @@ class DiscordClient(Client):
await message.add_reaction("❤️‍🔥")
try:
wl = Wikilink.parse(message.content)
wl = self.parse_wikilinks(message)
self.logger.debug(f"Parsed wikilinks: {wl}")
if len(wl)>0:
await self.handle_wikilink(message, wl)
await self.handle_wikilinks(message, wl)
except Exception as e:
self.logger.exception(f"Error parsing wikilink! got exception: ")
async def handle_wikilink(self, message:discord.message.Message, wl:List[Wikilink]):
async def handle_wikilinks(self, message:discord.message.Message, wl:List[Wikilink]):
log_msg = f"Wikilinks detected: \n" + '\n'.join([str(l) for l in wl])
self.logger.info(log_msg)
if self.debug:
@ -109,6 +107,10 @@ class DiscordClient(Client):
# TODO: Logging!
def parse_wikilinks(self, message) -> List[Wikilink]:
wikilinks = Wikilink.parse(message.content)
return wikilinks
# def add_links(self, links:Wikilink, msg:discord.message.Message):
# if 'testing links' in message.content:
# await message.channel.send(embed=Embed().add_field(name="Links", value="there are [links](https://example.com)"))

View File

@ -0,0 +1,264 @@
import pdb
from pprint import pformat
import argparse
from pathlib import Path
from typing import List, Optional, Tuple, Union, Dict
from threading import Event
from dataclasses import dataclass
from datetime import datetime
from wiki_postbot.clients.client import Client
from wiki_postbot.creds import Mediawiki_Creds, Slack_Creds
from wiki_postbot.interfaces.mediawiki import Wiki
from wiki_postbot.patterns.wikilink import Wikilink
from slack_sdk.web import WebClient
from slack_sdk.socket_mode import SocketModeClient
from slack_sdk.socket_mode.response import SocketModeResponse
from slack_sdk.socket_mode.request import SocketModeRequest
class SlackClient(Client):
def __init__(
self,
creds: Slack_Creds,
wiki:Wiki,
name:str="wikibot_slack",
reply_channel="wikibot",
log_dir:Path = Path('/var/log/wikibot'),
):
"""Wikibot but for slack!"""
super(SlackClient, self).__init__(wiki=wiki, name=name, log_dir=log_dir)
self.creds = creds
self._initialized = False
self.web_client = None # type: Optional[WebClient]
self.socket_client = None # type: Optional[SocketModeClient]
self.reply_channel_name = reply_channel
self._reply_channel = None
self._channel_inverse = None
# self.web_client, self.socket_client = self._init_client(self.creds)
@property
def reply_channel(self) -> Union[str, None]:
if self._reply_channel is None:
self.logger.debug(f"Getting channel named {self.reply_channel_name}")
channels = self.web_client.conversations_list()
channel = [c for c in channels['channels'] if c['name'] == self.reply_channel_name]
# self.logger.debug(channel)
if len(channel) == 1:
self._reply_channel = channel[0]['id']
elif len(channel) > 1:
self.logger.exception(f"Got too many channels to reply to!")
self._reply_channel = None
else:
self.logger("Reply channel not found!")
self._reply_channel = None
return self._reply_channel
@property
def channel_inverse(self) -> Dict[str,str]:
"""Maps channel IDs to channel names"""
if self._channel_inverse is None:
self.logger.debug("Getting inverse channel map")
channels = self.web_client.conversations_list()
self._channel_inverse = {c['id']: c['name'] for c in channels['channels']}
return self._channel_inverse
def _init_client(self, creds: Slack_Creds) -> Tuple[WebClient, SocketModeClient]:
web_client = WebClient(creds.bot_token)
socket_client = SocketModeClient(app_token=creds.app_token, web_client=web_client)
return web_client, socket_client
def handle_event(self, client:SocketModeClient, req: SocketModeRequest):
self.logger.debug(f"type: {req.type}, payload_type: {req.payload['type']}\n{pformat(req.payload)}")
if req.type == "events_api":
# acknowledge we got it so it don't get resent
response = SocketModeResponse(envelope_id=req.envelope_id)
client.send_socket_mode_response(response)
else:
self.logger.debug(f'Unhandled event type: {req.type}')
return
if req.type == "events_api" and \
req.payload['type'] == 'event_callback' and \
req.payload['event'].get('type', False) == 'message':
# Handle messages
self.logger.debug(f"Handling message")
message_text = req.payload['event']['text']
if 'good bot' in message_text.lower():
self.good_bot(client, req)
try:
wl = self.parse_wikilinks(message_text)
if len(wl) > 0:
self.logger.debug(f"Parsed wikilinks: {wl}")
self.handle_wikilinks(req, wl, client)
else:
self.logger.debug("No wikilinks found")
except Exception:
self.logger.exception("Error parsing wikilinks! got exception...")
else:
self.logger.debug(f"Was event, but not a message. Payload type: {req.payload['type']}, Event type: {req.payload.get('event', {}).get('type', 'unknown')}")
return
# pdb.set_trace()
# if
def parse_wikilinks(self, message) -> List[Wikilink]:
wikilinks = Wikilink.parse(message)
return wikilinks
def handle_wikilinks(self, message, wl: List[Wikilink], client:SocketModeClient):
self.react('hourglass', client, message)
try:
# expand fields in message
channel_id = message.payload['event']['channel']
channel_name = self.channel_inverse[channel_id]
msg = SlackMessage(
content = message.payload['event']['text'],
user_id = message.payload['event']['user'],
channel_id=channel_id,
channel = channel_name,
timestamp = message.payload['event']['ts']
)
msg.complete(client.web_client)
self.logger.debug(f"Posting message:\n{msg}")
result = self.wiki.handle_slack(msg)
ok = result.ok
except Exception:
self.logger.exception("Error handling slack message")
result = None
ok = False
self.react('hourglass', client, message, remove=True)
if ok:
self.react('white_check_mark', client, message)
else:
self.react('x', client, message)
if result and result.reply:
if self.reply_channel is None:
self.logger.exception(f"Do not have channel to reply to!")
else:
# await self.reply_channel.send(embed=Embed().add_field(name="WikiLinks", value=result.reply))
self.logger.debug('TODO: should reply here!')
def good_bot(self, client:SocketModeClient, req: SocketModeRequest):
self.logger.info('Got told we are a good bot ^_^')
self.react('heart', client, req)
self.react('heavy_plus_sign', client, req)
self.react('fire', client, req)
self.react('heavy_equals_sign', client, req)
self.react('heart_on_fire', client, req)
def react(self, emoji:str, client:SocketModeClient, req: SocketModeRequest, remove:bool=False):
if remove:
client.web_client.reactions_remove(
name=emoji,
channel=req.payload["event"]["channel"],
timestamp=req.payload["event"]["ts"],
)
else:
client.web_client.reactions_add(
name=emoji,
channel=req.payload["event"]["channel"],
timestamp=req.payload["event"]["ts"],
)
def run(self, creds: Optional[Slack_Creds] = None):
if creds is None:
creds = self.creds
else:
self.creds = creds
self.web_client, self.socket_client = self._init_client(creds)
self.socket_client.socket_mode_request_listeners.append(self.handle_event)
self.logger.debug("Connecting...")
self.socket_client.connect()
self.logger.debug(f"Got Reply Channel ID {self.reply_channel} for {self.reply_channel_name}")
try:
self.logger.info("Slack Client Listening")
Event().wait()
except KeyboardInterrupt:
self.logger.info("Quitting Slack client!")
@dataclass
class SlackMessage:
content: str # the actual content of the message
user_id: str
channel_id: str
channel: str # currently expected to be passed at instantiation because the client keeps a reverse index. bad information hiding i know.
timestamp: str
"""The unix epoch string timestamp stored as a slack event's ts attribute"""
# these need to be filled in after instantiation by passing a webclient to the completion methods
avatar: str = '' # URL of image
permalink: str = ''
author:str = '' # display name
_complete:bool=False
def get_permalink(self, client: WebClient):
permalink = client.chat_getPermalink(channel=self.channel_id, message_ts=self.timestamp)
self.permalink = permalink['permalink']
def get_user(self, client:WebClient):
user_info = client.users_info(user=self.user_id)
self.avatar = user_info['user']['profile']['image_192']
self.author = user_info['user']['profile']['display_name']
def complete(self, client:WebClient):
self.get_permalink(client)
self.get_user(client)
self._complete = True
@property
def date_sent(self) -> datetime:
return datetime.fromtimestamp(float(self.timestamp))
def argparser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(
prog="slack_bot",
description="A slack bot for posting messages with wikilinks to an associated mediawiki wiki"
)
parser.add_argument('-d', '--directory', default='/etc/wikibot/', type=Path,
help="Directory that stores credential files and logs")
parser.add_argument('-w', '--wiki', help="URL of wiki", type=str)
return parser
def main():
parser = argparser()
args = parser.parse_args()
directory = Path(args.directory)
log_dir = directory / "logs"
slack_creds = Slack_Creds.from_json(directory / 'slack_creds.json')
wiki_creds = Mediawiki_Creds.from_json(directory / 'mediawiki_creds.json')
wiki = Wiki(url=args.wiki, log_dir=log_dir, creds=wiki_creds)
wiki.login(wiki_creds)
client = SlackClient(creds=slack_creds, wiki=wiki, log_dir=log_dir)
client.run()
if __name__ == "__main__":
main()

View File

@ -52,4 +52,17 @@ class Mediawiki_Creds:
def from_json(cls, path:Path) -> 'Mediawiki_Creds':
with open(path, 'r') as jfile:
creds = json.load(jfile)
return Mediawiki_Creds(**creds)
return Mediawiki_Creds(**creds)
@dataclass
class Slack_Creds:
app_token:str
bot_token:str
@classmethod
@classmethod
def from_json(cls, path:Path) -> 'Slack_Creds':
"""jesus christ this package is so sloppy"""
with open(path, 'r') as jfile:
creds = json.load(jfile)
return Slack_Creds(**creds)

View File

@ -1,7 +1,7 @@
"""
Class for interfacing with mediawiki
"""
from typing import List, Optional
from typing import List, Optional, TYPE_CHECKING
from pathlib import Path
from urllib.parse import urljoin
from dataclasses import dataclass
@ -16,6 +16,9 @@ import requests
from discord.message import Message, Embed
import pdb
if TYPE_CHECKING:
from wiki_postbot.clients.slack import SlackMessage
# creds = Mediawiki_Creds.from_json('mediawiki_creds.json')
@ -170,6 +173,39 @@ class Wiki:
else:
return Result(ok=False, log=f"Got exceptions: {errored_pages}")
def handle_slack(self, msg:'SlackMessage') -> Result:
"""
Brooooo it's getting laaaaaaaaaate and I think it's obvious how this should be refactored
but I want to play some zeldaaaaaaaaaaaa so I am copy and pasting for now
"""
self.login(self.creds)
# Get message in mediawiki template formatting
template_str = TemplateMessage.format_slack(msg)
# parse wikilinks, add to each page
wikilinks = Wikilink.parse(msg.content)
errored_pages = []
for link in wikilinks:
if link.section is None:
section = "Slack"
else:
section = link.section
res = self.insert_text(link.link, section, template_str)
if res.json()['edit']['result'] != 'Success':
errored_pages.append(res.json())
# Add to index page (only once)
self.add_to_index(template_str)
if len(errored_pages) == 0:
# gather links for a reply
reply = '\n'.join([f"[{l.link}]({urljoin(self.url, l.link.replace(' ', '_'))})" for l in wikilinks])
return Result(ok=True, log=f"Successfully posted message to {[l.link for l in wikilinks]}", reply=reply)
else:
return Result(ok=False, log=f"Got exceptions: {errored_pages}")
def add_to_index(self, message):
section = datetime.today().strftime("%y-%m-%d")
self.insert_text(page=self.index_page, section=section, text=message)

View File

@ -64,3 +64,11 @@ def main():
service = Service(wiki_url=args.wiki, directory=args.directory)
service.install_service()
def main_slack():
parser = argparser()
args = parser.parse_args()
service = Service(wiki_url=args.wiki, directory=args.directory)
service.bot_script = shutil.which('slack_bot')
service.install_service()

View File

@ -1,12 +1,15 @@
"""
Templates for representing different kinds of messages on mediawiki
"""
from typing import TYPE_CHECKING
from wiki_postbot.formats.wiki import WikiPage
from abc import abstractmethod
from discord.message import Message
import warnings
if TYPE_CHECKING:
from wiki_postbot.clients.slack import SlackMessage
class WikiTemplate(WikiPage):
@abstractmethod
@ -37,6 +40,18 @@ class TemplateMessage(WikiTemplate):
"}}"
)
@classmethod
def format_slack(cls, msg:'SlackMessage') -> str:
return (
"{{Message\n"
f"|Author={msg.author}\n"
f"|Avatar={msg.avatar}\n"
f"|Date Sent={msg.date_sent.strftime('%y-%m-%d %H:%M:%S')}\n"
f"|Channel={msg.channel}\n"
f"|Text={msg.content}\n"
f"|Link={msg.permalink}\n"
"}}"
)
#
# template_message = TemplateMessage.from_source(