How to solve an error related to creating parser f

2019-09-06 01:03发布

问题:

I am writing a parser using StandardTokenParsers in Scala. Need to create a regex parser to parse a path. I have tested the regex works fine but sending it to a function to parse it, the program gives an error that I am not able to figure it out! a part of code that is related to this parser is as follow:

 class InfixToPostfix extends StandardTokenParsers {
 import scala.util.matching.Regex
 import lexical.StringLit
//parsing the path
 def regexStringLit(r: Regex): Parser[String] =
 acceptMatch( "string literal matching regex " + r,{ case  StringLit(s) if r.unapplySeq(s).isDefined => s })
// Regex for path
 val pathIdent ="""/hdfs://[\d.]+:\d+/[\w/]+/\w+([.+]\w+)+""".r   
 def pathIdente: Parser[String] =regexStringLit(pathIdent)

 lexical.delimiters ++= List("+","-","*","/", "^","(",")",",")
 def value :Parser[Expr] = numericLit ^^ { s => Number(s) }
 def variable:Parser[Expr] =  pathIdente ^^ { s => Variable(s) }
 def parens:Parser[Expr] = "(" ~> expr <~ ")"

 def argument:Parser[Expr] = expr <~ (","?)
 def func:Parser[Expr] = ( pathIdente ~ "(" ~ (argument+) ~ ")" ^^ { case f ~ _ ~ e ~ _ => Function(f, e) })
//and the rest of the code ....

This parser is going to parse arithmetic operations. I use args(0) to send my input to the program which is : "/hdfs://111.33.55.2:8888/folder1/p.a3d+1"

and I get the following error:

[1.1] failure: string literal matching regex /hdfs://([\d\.]+):(\d+)/([\w/]+/(\w+\.\w+)) expected

 /hdfs://111.33.55.2:8888/folder1/p.a3d
 ^

Couldn't figure out how to solve it!

FYI: The part for "+1" is going to handle by the parser in the code so the part "pathIdent" is only for the path and that is the part causing the trouble. This is also good :

  """/hdfs://\d+(\.\d+){3}:\d+(/(\w+([.+]\w+)*))+""".r 

it works fine outside of the code checking it in : regexpal.com but still same error using it inside the program.

I am wondering if StringLit is the one that doesn't contain some of the characters and causing the error. Is there anything else other than StringLit that I can use here?

回答1:

The failure to match will be because the matcher is greedy. This is a common problem with regular expression matching (and hence lexical analysis) in several languages.

The greedy matching catches you at the end of the expression.

You have ([\w/]+/(\w+\.\w+)) but this will fail to match because the word p matched by the \w represented by the input text folder1/p is swallowed up by the piece ([\w/]+. It stops at the period .. There is therefore no word before the dot to permit (\w+\.\w+) to ever match.

You'll have to rethink your regular expression and make each path fragment terminate at a solidus / rather than make it part of a set.

Do you see?

To make this work you need to express in the following way:

"""/hdfs://[\d.]+:\d+/(\w/)+\w+([.+]\w+)+""".r 

Where I replaced [\w/]+/ by (\w/)+. This now specifies the ordering of the words and slashes and leaves a word unmatched for the following pattern to succeed.