How does .match() differ from simple .replace(); How to use match option list with phrases

Trying to parse transcription results for instances when the phrase representing the punctuation and replace it with the correct formatting.

I am wondering how .match() differs from a simple search and replace

This is an example at an attempt to use match.

  javascript
  _punctType(results) {
    switch (results.match('[<punctuation>(period|comma|exclamation mark|question mark)]', 'punctuation').text()) {
    case 'period':
      return periodRegex;
    case 'comma':
      return commaRegex;
    case 'exclamation mark':
      return exclamationRegex;
    case 'question mark':
      return questionRegex;
    default:
      break;
    }
  }

In this example use .match() I am not able to get multi-word phrases to be correctly recognized.

[spencermountain]

ohhh, sorry i misunderstood. yeah, the multiple-word inside of ( ) is pretty awkward right now. you may have better luck with .lookup():

let doc=nlp('i used a question mark and nothing else, period.')
doc.lookup(['comma','question mark','period']).out('array')

https://runkit.com/spencermountain/5e612505ecbc3d00138ed3ad cheers

[spencermountain]

no problem. lookup checks-against the parsed and normalized form of the terms, so it will guard against punctuation, contractions and other shenanigans in the text. it can also be much faster, when working with large # of lookups or texts.

but yeah, if you write some safe regular expressions, you can use a simple javascript replace too