Optional command prefix #2
Conversation
|
in my testing this seems to interfere with the apologies for the slow response here, i was having trouble with testing due to server issues. |
|
at this point this conflicts with other changes that have happened to the code as well, so it would need to be resolved. |
There was a problem hiding this comment.
I'm also really interested by this feature and hope this could be merged soon. That way, the plugin can be integrated to a more global bot.
Exepect the comments I placed, I don't see any conflict reason with the !clear. Maybe an empty command_prefix, so an empty command ? This case could be handled with a default value of gpt.
| @@ -154,7 +162,7 @@ async def _call_gpt(self, prompt): | |||
| #self.log.debug(content) | |||
| return content | |||
|
|
|||
| @command.new(name='gpt', help='control chatGPT functionality', require_subcommand=True) | |||
| @command.new(name=f"self.config['command_prefix']", help='control chatGPT functionality', require_subcommand=True) | |||
There was a problem hiding this comment.
The f-string here is wrong and unecessary. It should just be self.config['command_prefix']
| return | ||
|
|
||
| if len(command_prefix) > 0 and not event.content.body.startswith(f"!{command_prefix}"): | ||
| return # ignore a comand does not match the required prefix | ||
| if len(command_prefix) == 0 and event.content.body.startswith(f'!'): |
There was a problem hiding this comment.
All of this could be replaced by
if not command_prefix or not event.content.body.startswith(f'!{command_prefix}'):
return| @@ -45,6 +47,7 @@ async def on_message(self, event: MessageEvent) -> None: | |||
| user = '' | |||
| content = '' | |||
| timestamp = datetime.today().strftime('%Y-%m-%d %H:%M:%S') | |||
| command_prefix = self.config['command_prefix'] | |||
There was a problem hiding this comment.
It could be interesting to strip the config value to avoid "strange" behavior.
Adds the option to use the !gpt prefix to get a response from the bot. Note the bot still records all messages to later use as context when prompted.