在NearleyJS中解析字符串中的字符串

Question

嗨，我在CFG中很幼稚，我正在尝试为我的表达式编写一个语法，但我在解析nearley输入值中的字符串时遇到了一个问题。

下面是我的表达式

"(I have a person: where "1+1=2")"

下面是我的语法：

main -> statement {% id %}
statement -> "(" constline optional:? ")" 

constline -> "I have a person" {% id %}
optional -> ":where" _ "\"" whereclause "\"" _ {%d=>d[2]%}
whereclause -> _ expression _ compare _ expression {%d=> d[0]+d[1]+d[3]%}
expression -> numbers "+" numbers {% d=>d[0]+d[1]+d[2]%}
           | numbers "-" numbers  {% d=>d[0]+d[1]+d[2]%}
numbers -> digit
        | digit numbers {% d=> d[0]+d[1]%}
digit -> [0-9]
_ -> [\t]:*

现在，在解析时，我得到了下面的错误：:where "1+1=2")

Unexpected "\"". Instead, I was expecting to see one of the following:
A character matching /[0-9]/ based on:
   digit -> /[0-9]/
   numbers-> digit 

请帮帮忙

Answer 1

您的语法缺少使lexer正常工作的正则表达式，因此它无法理解大部分语法。此外，您还遗漏了比较闭包，我相信在我们的例子中它的意思是'=‘。我想这个应该可以了。

​

@{%
const moo = require('moo');
let lexer = moo .compile({
    //string: /"(?:\\["bfnrt\/\\]|\\u[a-fA-F0-9]{4}|[^"\\])*"/,
    numbers: /[0-9]+/,
    '"': '"',
    'I have a person': 'I have a person',
    '(': '(',
    ')': ')',
    ':': ':',
    '+': '+',
    '=': '=',
    '=': '=',
    ws: /[ \t]/,
    'where': 'where',
    
});
console.clear();

%}

@lexer lexer

main -> "\"" statement "\"" {% d => d[1] %}
statement -> "(" constline optional:? ")" 

constline -> "I have a person" {% id %}
optional -> ":" _ "where" _ expressionString _ {%d=>d[3]%}
whereclause -> _ expressionString _ compare _ expressionString {%d=> d[0]+d[1]+d[3]%}
expressionString -> "\"" expression "\"" {% d => d[1] %}
expression -> numbers "+" numbers "=" numbers {% d=>d[0]+d[1]+d[2]%}
           | numbers "-" numbers "=" numbers {% d=>d[0]+d[1]+d[2]%}
numbers -> digit {% id %}
        | digit numbers {% d=> d[0]+d[1]%}
digit -> [0-9]
_ -> [ \t]:*

​

如果这对你有效，请告诉我。