Problem with JavaTokenParsers...

[Mattias Wallman]

I have some problems with a simple parser that basically is defined as:

object MyParser extends JavaTokenParsers {

  def test: Parser[String] = string | decimalNumber

  def string: Parser[String] = """[\w-]+"""

  def main(args: Array[String]) {

    println(parseAll(test, "3e9ad088-5d08-4e61-9864-a2a56db93db3"))

    println(parseAll(test, "123.33"))

  }

}

This will produce the output 

[1.37] parsed: 3e9ad088-5d08-4e61-9864-a2a56db93db3

[1.4] failure: string matching regex `\z' expected but `.' found

123.33

   ^

Then I tried to modify  the test rule:

object MyParser extends JavaTokenParsers {

  def test: Parser[String] =   decimalNumber | string

  def string: Parser[String] = """[\w-]+""".r

  def main(args: Array[String]) {

    println(parseAll(test, "3e9ad088-5d08-4e61-9864-a2a56db93db3"))

    println(parseAll(test, "123.33"))

  }

}

This will produce the output:

[1.2] failure: string matching regex `\z' expected but `e' found

3e9ad088-5d08-4e61-9864-a2a56db93db3

 ^

[1.7] parsed: 123.33

So, the question is: why does it matter in what order the decimalNumber and string rule is declared in the test rule, and how can I do to fix this little problem, so I can parse my decimal number and string properly?

Regards

Mattias

[Daniel Sobral]

The | operator is not backtracking: if the left side succeeds, the right side will never be tried. If something fails latter on, there's no going back.

You may use ||| to pick the longest match instead, but, generally speaking, you should strive to avoid such ambiguities.

--
Daniel C. Sobral

I travel to the future all the time.

[Mattias Wallman]

Ok, the ||| seems to work, thank you!

But if that is an undesirable construction, is there a better alternative?

[Daniel Sobral]

On Thu, Nov 8, 2012 at 3:45 PM, Mattias Wallman <mattias...@gmail.com> wrote:

Ok, the ||| seems to work, thank you!

But if that is an undesirable construction, is there a better alternative?

That can only be answered by looking at the whole grammar.