Why can I not match a sentence with comma

[trsh]

let doc = nlp(`You can eat, sugar and water`)
    let m = doc.match('You can eat, [*]')
    console.log((<View>m.groups('0')).out('array'));

If I remove , in sentence and match, all works, but I need it.

Update:

let doc = nlp(`You can eat sugar, cold stuff and water`)
    let m = doc.match('You can eat [*], [*] and [*]')
    console.log((<View>m.groups('0')).out('array'));
    console.log((<View>m.groups('1')).out('array'));
    console.log((<View>m.groups('2')).out('array'));

I want to extract sugar, cold stuff, water. But I can't. When I use comma in match, nothing is matched. If I dont use it, the results are incorrect.

Update2:

Another example:

let doc = nlp(`Eat sugar! and melon`);
    let m = doc.match('Eat [*] and [*]')
    console.log((<View>m.groups('0')).out('array'));
    console.log((<View>m.groups('1')).out('array'));

It doesn't work, unless I remove "!" from text. Why?

[spencermountain]

hey Janis, good questions.
yeah - the match statements are not meant to be natural language in, with punctuation and stuff. That's an interesting thought - how come you need the comma in the match statement?

[*] is greedy, and i would recommend [.] to match one term, instead.

In the last one - with the exclamation, it is splitting into two sentences. It is not possible to match between two sentences, but that would be a neat feature to add.
cheers

[trsh]

To answer you question look at my example: "You can eat sugar, cold stuff and water", where I want to extract sugar, cold stuff and water. To do this correctly I must take everything with and and , in between, after You can eat. And that is the reason.
However I went for a splitting strategy with regexes.