Special chars validations

General
1

Hello team, i am working on a bot where i need to validate agaisnt some special characters, some of the special characters are behaving incorrectly in certain combinations, for example:

for ^:
^hello, is transforming it to hello
^ is showing an error related to
^^ is showing me the correct error, but asking the user to put a valid value again, its not taking the next input, i have to do it again to “reset” the damage that ^^ did.

for \ is also failing on the inputs.

below is the regexp that is beign used to match, and give a positive or negative validation:

/^[^0-9{[]\}?!-@.¿#_¡’=$%,^|&*()><;(“.+”)":"’]+$/;

So if any of these characters is present on the input, it should give me an error, and ask the user for input again.

2

Can you expand on your requirement please?

The overall domain for bots is conversation, and so punctuation and special characters have some distinct handling, for example in how an utterance is tokenized into words and sentences, which then leads on to spell correction. Therefore for an input of “^hello”, the caret is dropped because that is not ordinarily part of any standard word, leaving just “hello” to be able to be matched. (Though “hello” itself is a special case and is transformed again!).

Note that all entities are processed on a word basis, and as such the custom regex entity has implicit \b tokens added to expression. Sentence markers like ^ and $ make no sense.

3