如何从转义字符中排除转义字符

Question

I have a Java string

String t = "Region S\u00FCdost SER";

where \ü is a replacement for the unicode character "ü"

If i add a new escape char to the above string, i would still want my below function to escape other chars excluding the current .

For example, the below function on re running would return the result as "Region S\\\üdost SER" and "Region S\\\\\\\üdost SER" on subsequent iterations.

How do we prevent this?

public static String escapeString(String str)
    {
        StringBuffer result = new StringBuffer();

        // char is 16 bits long and can hold an UTF-16 code        
        // i iterate on chars and not on code points
        // i guess this will be enough until we need to support surrogate pairs 
        for (int i = 0; i < str.length(); i++)
        {
            char c = str.charAt(i);
            switch (c) {

            case '"':
                result.append("\\\""); //$NON-NLS-1$
                break;
            case '\b':
                result.append("\\b"); //$NON-NLS-1$
                break;
            case '\t':
                result.append("\\t"); //$NON-NLS-1$
                break;
            case '\n':
                result.append("\\n"); //$NON-NLS-1$
                break;
            case '\f':
                result.append("\\f"); //$NON-NLS-1$
                break;
            case '\r':
                result.append("\\r"); //$NON-NLS-1$
                break;
            case '\'':
                   result.append("\\'"); //$NON-NLS-1$   

                break;
            case '\\':

                result.append("\\\\"); //$NON-NLS-1$

                break;

            default:
                if (c < 128)
                {
                    //is ascii
                    result.append(c);
                }
                else
                {
                    result.append(
                            String.format("\\u%04X", (int) c)); //$NON-NLS-1$
                }
            }
        }

        return result.toString();
    }
}

Answer 1

You can do:

case '\\':
    if(str.charAt(i+1)!='u')
        result.append("\\\\");
    else 
        result.append("\\");
break;

Assuming that \\u\u003c/code> will always denote a unicode character sequence in your string.

Answer 2

When you write a Java string literal as "Region S\üdost SER" , the Java compiler will interpret that as the string value Region Südost SER , which is what the escape() method will see when called on t .

If you wanted the string Region S\üdost SER , you should have escaped the \\ , ie "Region S\\\üdost SER" .

If you keep running the escape() method, I believe you'll see what you want.

String s = "Region S\u00FCdost SER";
System.out.println(s); // print original text
for (int i = 0; i < 4; i++) {
    s = escapeString(s);
    System.out.println(s);
}

Output:

Region Südost SER                           <-- original text
Region S\u00FCdost SER
Region S\\u00FCdost SER
Region S\\\\u00FCdost SER
Region S\\\\\\\\u00FCdost SER

If you change input to "He'd say: \\"Bitte schön\\"" , you get:

He'd say: "Bitte schön"                     <-- original text
He\'d say: \"Bitte sch\u00F6n\"
He\\\'d say: \\\"Bitte sch\\u00F6n\\\"
He\\\\\\\'d say: \\\\\\\"Bitte sch\\\\u00F6n\\\\\\\"
He\\\\\\\\\\\\\\\'d say: \\\\\\\\\\\\\\\"Bitte sch\\\\\\\\u00F6n\\\\\\\\\\\\\\\"

I mean, this is what you wanted, right? If not, please clarify question by actually showing example output of what you want.