Overview

When using the "yen mark (half-width ¥)" on Android Because the character code has been treated as "backslash (half-width /)" in Windows We have summarized the difference in character code points between the yen mark and backslash and this measure.

Java is used for all source code.

Differences in code points in each environment (circle code / backslash)

With UTF-8 in each environment The following code points are used for the yen code and backslash.

Windows Web Yen mark: 5C Backslash: 5C

Mac / iOS / Android Web Yen mark: A5 Backslash: 5C

iOS / Android native app Yen mark: C2A5 Backslash: 5C

[Reference URL] Character code of yen mark and backslash

Countermeasures

As a countermeasure, all the code points of the yen mark, This is a countermeasure for Windows Web (code point: A5 or C2A5 ⇒ C5).

A5 ⇒ C5
C2A5 ⇒ C5

Source code creation

I created a class (CodePointConversion.java) that performs codepoint conversion. Converts the target code point (A5 or C2A5) to the code point (C5).

`CodePointConversion.java`


import java.util.Map;
import java.util.HashMap;
import java.lang.StringBuilder;

/**
*Code point conversion class
* @author HogeHoge
*/
public class CodePointConversion {
	//Code point conversion table. Convert from KEY to VALUE code point.
	private static Map<Integer, Integer> conversion_map = new HashMap<Integer, Integer>() {
		// ￥ → \
		{put(0xA5, 0x5C);
		 put(0xC2A5, 0x5C);}
	};
	
	/**
     *Perform code point conversion
     * @param str Code point conversion string
     * @return String after code point conversion
     */
     public static String convertCordPoint(String str) {
     	//null check
		if (str == null) {
			return str;
		}
		StringBuilder sb = new StringBuilder(str);
		
		//Get character loop
		for(int i = 0; i < sb.length(); i++) {
			//Get the code point of the acquired character
			int code_point = sb.codePointAt(i);
			
			for (Map.Entry<Integer, Integer> entry : conversion_map.entrySet()) {
		    	if (code_point == entry.getKey()) {
		    		//If it is a code point conversion target, perform code point conversion.
		    		String converted_char = new String(Character.toChars(entry.getValue()));
					sb.replace(i, i+1, converted_char);
				}
			}
     	}
     	
     	return sb.toString();
    }
}

Test code creation

This is a test code for checking the operation. For the input value and output value, check the execution result.

`CodePointConversionTest.java`


import org.junit.jupiter.api.Test;
import static org.junit.Assert.*;
import static org.hamcrest.CoreMatchers.*;

class CodePointConversionTest {

	@Test
	void testConvertCordPoint001() {
		//[Input value] \\ * \ is a half-width character.
		String input_str = new String(Character.toChars(0xA5)) + new String(Character.toChars(0xC2A5));
		//【Expected value】\\※the first\Is an escape character.
		String expect_str = "\\\\";
		
		System.out.println("【Input value】\n" + input_str);
		System.out.println("[Input code point]\n" + Integer.toHexString(input_str.codePointAt(0)) + "\n" + Integer.toHexString(input_str.codePointAt(1)));
		
		//Code point conversion. ¥ →\Is converted to.
		String result_str = CodePointConversion.convertCordPoint(input_str);
		
		//Output value
		System.out.println("\n [Output value]\n" + result_str);
		System.out.println("[Output code point]\n" + Integer.toHexString(result_str.codePointAt(0)) + "\n" + Integer.toHexString(result_str.codePointAt(1)));
		
		//Check the output result.
		assertThat(result_str, is(expect_str));
	}

}

The following execution results.

【Input value】
\?
[Input code point]
a5
c2a5

【Output value】
\\
[Output code point]
5c
5c

The above input value is "?", But it is a problem on the screen display of the Windows terminal. Actually, the character string "Yen mark: A5 Yen mark: C2A5" is included as an input value.

How to treat backslash as a circle code on Android

The methods described so far have changed the yen mark to a backslash (code point: A5 or C2A5 ⇒ C5). I described how to convert code points, but On Android, there may be cases where you want to treat it as a backslash ⇒ yen mark (code point: C5 ⇒ A5).

In that case, please handle in the form of reversing the KEY and VALUE of conversion_map (code point conversion table).

(Example) How to treat the code point of C5 as A5.
	//Code point conversion table. Convert from KEY to VALUE code point.
	private static Map<Integer, Integer> conversion_map = new HashMap<Integer, Integer>() {
		// \ → ￥
		{put(0x5C, 0xA5);}
	};

Whether to treat the C5 code point as A5 or C2A5 should be changed depending on each environment.

At the end

This time, I described how to perform code point conversion (code point: A5 or C2A5 ⇔ C5) of yen mark and backslash. Just add code points to conversion_map (code point conversion table) It is possible to convert various code points.

If you like, please use it when performing code point conversion in Java.

[Android → WEB] Yen mark / backslash code point conversion (code point: A5 or C2A5 ⇔ C5)