Regexp filter for message handler only match if message is 2 words

[100nandoo]

I want to create regexp filter for message handler but only if message is exactly 2 words?

Example message:
BTC USDT

[evgfilim1]

Hello, @100nandoo. You can use this regexp for your handler: ^(\w+) (\w+)$

Explanation

• ^ asserts the start of the line
• (...) starts capturing group, so you can use the result of this capture later
• \w is the same as [a-zA-Z0-9_]
  • [...] is a choice list, so it matches exactly one character from this list
  • [a-z] matches everything from a (ASCII codepoint 97) to z (ASCII codepoint 122)
  • [a-z0-9] helps you to match either [a-z] or [0-9]
  • so [a-zA-Z0-9_] is one of [a-z], [A-Z], [0-9] or _.
• + is the same as {1,}
  • In common, {a,b} means "match previous expression (\w in this case) from a to b times"
  • If a is missing, it is considered equal 0
  • If b is missing, it is considered maximum
• matches Space literally
• $ asserts the end of the line

Example

@dp.message_handler(regexp=r"^(\w+) (\w+)$")
async def words(message: types.Message, match: typing.Match[str]):
    await message.reply(f"Convert {match[1]} to {match[2]}")  # Note that capturing groups are counted from 1 as 0 is reserved for full match

In your case it would reply "Convert BTC to USDT"

See also

• https://regex101.com/ — very useful webapp to help learn regexps, debug them and find new ones
• https://docs.python.org/3/library/re.html — Python documentation for re module, which is used for regexps

• aiogram/aiogram/dispatcher/filters/builtin.py

  Line 403 in d5a4c0c

  class Regexp(Filter):

[100nandoo]

Thanks for the thorough explanation. I found another issue, while I was using this regex. My other message handler that accept more than 2 words is never trigger anymore.

Is there a way to trigger it if only the word is exactly 2?

[Olegt0rr]

It's not a school of coding.
Please use google first.

Just check len(message.text.split()) == 2 in the filter.

[100nandoo]

Understood. I didn't know there is filter method inside Regexp. Thanks