Fixed parsing of tags

[riccardoferretti]

fixes problem referred to in #302 (see when reopened)

[jmg-duarte]

I think the regex could be more restrictive as to avoid this check

[jmg-duarte]

This should work const HASHTAG_REGEX = /(^|[ ])#([a-zA-Z][\w_-]+\b)/gm;

[riccardoferretti]

I agree that more regex magic could do the trick, but after spending so much time on it I got pragmatic (tired) and settled here.

This is not the best tag parsing code out there, I have seen something by twitter https://github.com/twitter/twitter-text/ but it's much more (I have also pulled out their tag code, it just feels a bit heavier and doesn't quite match the tags cases we want).
I have not been able to find a simple tag extraction library/gist/file out there, would love pointers if people are aware of some.

[jmg-duarte]

The addition I made is simple enough to work in "every" case, you just force the first character to be a letter

[riccardoferretti]

thing is, it can start with a number, it just can't be all numbers:

      62 |     expect(
      63 |       extractHashtags('this #123 tag should be ignore, but not #123four')
    > 64 |     ).toEqual(new Set(['123four']));

[jmg-duarte]

This was merged, have you tried with the latest version?

[bserrao]

Yes, using v0.77. The code is a bit different know.. But i don't know if i'm doing something wrong...

export const extractHashtags = (text: string): Set<string> => {
  return isSome(text)
    ? new Set(Array.from(text.matchAll(HASHTAG_REGEX), m => m[2].trim()))
    : new Set();
};

Vscode highlight it as a tag but it's not recognized as one in tag explorer. Maybe the problem is with tag explorer itself?

[jmg-duarte]

@riccardoferretti do you have any idea what this might be?

[jmg-duarte]

In the meantime, please submit an issue @bserrao

[riccardoferretti]

interesting.. I reproduced it as well, that's odd. yes let's start by filing an issue