Cannot parse "rest of line" as a token without matching the entire line

[whweaver]

I have a grammar in which I have lines like this:

    key: value

I need to extract the indent, key, and value from here. The indent is the number of spaces divided by 4, the key is any string that does not include ":", and the value is everything from the end of the whitespace following ":" until the end of the line, including any subsequent colons. My initial attempt at this was to make the following rules:

rule(indent: /(    )*/).as { |str| str.length / 4 }
rule(key: /[^:\s][^:]*/)
rule(":" => /:\s*/)
rule(value: /\S.*/)
rule(:line) do |r|
  r[:indent, :key, ":", :value]
end

The problem is that the :value rule matches the entire "key: value" sequence, making it the longest match, so whittle throws an error when it finds a :value after the :indent instead of a :key. Does whittle have a way to represent this kind of grammar?

[booch]

I think you want to use a rule that matches the end-of-line explicitly. Like the example at https://github.com/d11wtq/whittle#skipping-whitespace-and-comments, so you'd have:

rule(value: /\S.*$/m)