Trying to add & detect custom prefixes

[Goatform]

Hey, I recently ran into this plugin and started using it. I'm trying to find a way to detect and linkify custom prefixes ... e.g. UID-123123. Also, I did look into plugins such as mention and hashtag but it still wasn't completely clear to me as there didn't seem to be a way to add a custom prefix, be it a single or multiple prefix check, in a plugin. What would be the best way to detect if the string contains the UID- prefix if so, because, either I am missing something in documentation or there's really no way (currently) to do such a thing. I'd appreciate any suggestion regarding this if so, please and thank you.

[robertn702]

I'm also looking for the exact same thing. An easy to way to customize prefixes or more documentation around writing custom plugins would be really helpful for this use case.

[WIStudent]

After reading https://github.com/Hypercontext/linkifyjs/issues/167#issuecomment-1018668663 and looking at the existing plugins I tried to create a custom plugin that looks for something like UID-123 but I failed:

import { createTokenClass, registerTokenPlugin, registerPlugin, tokenize, stringToArray} from 'linkifyjs';
import linkifyStr from 'linkify-string';

const PrefixToken = createTokenClass('prefix', {
  isLink: true,
  toHref() {
    return `https://example.com/${this.toString()}`
  }
})

registerTokenPlugin('prefix', ({scanner}) => {
  scanner.start.ts(stringToArray('UID'), 'UID');
})

registerPlugin('uid', ({parser, scanner}) => {
  const {HYPHEN, NUM, WORD} = scanner.tokens
  // @ts-ignore
  parser.start.tt('UID').tt(HYPHEN).tt(NUM, PrefixToken);
})

const str = 'UID-123 UI-123 abc https://google.com https://example.com/UID-123 defg';
console.log(tokenize(str));
console.log(linkifyStr(str))

The returned tokens look like this:

UID was not recognized as a UID token but as a WORD token, so I think I am doing something wrong in the registerTokenPlugin part.

Using

registerPlugin('uid', ({parser, scanner}) => {
  const {HYPHEN, NUM, WORD} = scanner.tokens
  // @ts-ignore
  parser.start.tt(WORD).tt(HYPHEN).tt(NUM, PrefixToken);
})

instead works, but now any WORD HYPHEN NUM sequence will be recognized, not only parts starting with UID:

[nfrasser]

@WIStudent I think the problem is that linkify is not implemented with the case sensitivity, which makes plug-in registration finicky sometimes. Try replacing the registerTokenPlugin call with this:

registerTokenPlugin('prefix', ({scanner}) => {
  scanner.start.ts(stringToArray('uid'), 'UID', {ascii: true});
})