RED-5562 japanse space characters
This commit is contained in:
parent
5feb6891e2
commit
1ce47b7fbc
@ -1,16 +1,25 @@
|
|||||||
package com.iqser.red.service.redaction.v1.server.redaction.utils;
|
package com.iqser.red.service.redaction.v1.server.redaction.utils;
|
||||||
|
|
||||||
|
import lombok.extern.slf4j.Slf4j;
|
||||||
|
|
||||||
import java.util.Set;
|
import java.util.Set;
|
||||||
import java.util.regex.Pattern;
|
import java.util.regex.Pattern;
|
||||||
|
|
||||||
|
@Slf4j
|
||||||
public class SeparatorUtils {
|
public class SeparatorUtils {
|
||||||
|
|
||||||
private final static Set<Character> quotes = Set.of('\'', '\u0022', '\u00AB', '\u00BB', '\u2018', '\u2019', '\u201A', '\u201C', '\u201D', '\u201E', '\u2039', '\u203A');
|
private final static Set<Character> quotes = Set.of('\'', '\u0022', '\u00AB', '\u00BB', '\u2018', '\u2019', '\u201A', '\u201C', '\u201D', '\u201E'
|
||||||
|
, '\u2039', '\u203A');
|
||||||
|
|
||||||
|
private final static Set<Integer> japaneseAltPunctuationMarks = Set.of(65288, 65289, 65294, 65339, 65341, 65371, 65373, 65375, 65376, 12443, 12444, 65309);
|
||||||
|
|
||||||
public static boolean isSeparator(char c) {
|
public static boolean isSeparator(char c) {
|
||||||
|
return Character.isWhitespace(c) || Pattern.matches("\\p{Punct}", String.valueOf(c)) || quotes.contains(c) || isJapaneseSeparator(c);
|
||||||
|
}
|
||||||
|
|
||||||
return Character.isWhitespace(c) || Pattern.matches("\\p{Punct}", String.valueOf(c)) || quotes.contains(c);
|
public static boolean isJapaneseSeparator(char c) {
|
||||||
|
var intValue = (int) c;
|
||||||
|
return (intValue >= 12288 && intValue <= 12336) || japaneseAltPunctuationMarks.contains(intValue);
|
||||||
}
|
}
|
||||||
|
|
||||||
}
|
}
|
||||||
|
|||||||
Loading…
x
Reference in New Issue
Block a user