0

How would I create a parser that allows a character which also happens to be the same as the begin/end character. Using the following example:

'Isn't it hot'

The second single-quote should be accepted as part of the content that is between the beginning and ending single-quote. I created a parser like this:

char("'").seq((word()|char("'")|whitespace()).plus()).seq(char("'"))

but it fails as:

Failure[1:15]: "'" expected

If I use "any()|char("'") then it greedily consumes the ending single-quote causing an error as well.

Would I need to create an actual Grammar class? I have attempted to create one but can't figure out how to make a Parser that doesn't try to consume the end marker greedily.

Lukas Renggli
  • 8,754
  • 23
  • 46
EndlessLife
  • 319
  • 3
  • 13
  • I would say that the general solution to this problem is to require quoting of the single-quote in the middle. See for instance c#'s verbatim string syntax: https://msdn.microsoft.com/en-us/library/aa691090(v=vs.71).aspx – Alexander Torstling Nov 23 '16 at 21:36

1 Answers1

2

The problem is that plus() is greedy and blind. This means the repetition consumes as much input as possible, but does not consider what comes afterwards. In your example, everything up to the end of the input is consumed, but then the last quote in the sequence cannot be matched anymore.

You can solve the problem by using the non-blind variation plusGreedy(Parser) instead:

char("'")
  .seq((word() | char("'") | whitespace()).plusGreedy(char("'")))
  .seq(char("'"));

This consumes the input as long as there is still a char("'") left that can be consumed afterwards.

Lukas Renggli
  • 8,754
  • 23
  • 46