com.day.text.ISO9075 Maven / Gradle / Ivy
/*************************************************************************
*
* ADOBE CONFIDENTIAL
* __________________
*
* Copyright 2012 Adobe Systems Incorporated
* All Rights Reserved.
*
* NOTICE: All information contained herein is, and remains
* the property of Adobe Systems Incorporated and its suppliers,
* if any. The intellectual and technical concepts contained
* herein are proprietary to Adobe Systems Incorporated and its
* suppliers and are protected by trade secret or copyright law.
* Dissemination of this information or reproduction of this material
* is strictly forbidden unless prior written permission is obtained
* from Adobe Systems Incorporated.
**************************************************************************/
package com.day.text;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
* Implements the encode and decode routines as specified for XML name to SQL
* identifier conversion in ISO 9075-14:2003.
* If a character c is not valid at a certain position in an XML 1.0
* NCName it is encoded in the form: '_x' + hexValueOf(c) + '_'
*
*/
public class ISO9075 {
/** Hidden constructor. */
private ISO9075() { }
/** Pattern on an encoded character */
private static final Pattern ENCODE_PATTERN = Pattern.compile("_x\\p{XDigit}{4}_");
/** Padding characters */
private static final char[] PADDING = new char[] {'0', '0', '0'};
/** All the possible hex digits */
private static final String HEX_DIGITS = "0123456789abcdefABCDEF";
/**
* Encodes name as specified in ISO 9075.
* @param name the String to encode.
* @return the encoded String or name if it does
* not need encoding.
*/
public static String encode(String name) {
// quick check for root node name
if (name.length() == 0) {
return name;
}
if (XMLChar.isValidName(name) && name.indexOf("_x") < 0) {
// already valid
return name;
} else {
// encode
StringBuffer encoded = new StringBuffer();
for (int i = 0; i < name.length(); i++) {
if (i == 0) {
// first character of name
if (XMLChar.isNameStart(name.charAt(i))) {
if (needsEscaping(name, i)) {
// '_x' must be encoded
encode('_', encoded);
} else {
encoded.append(name.charAt(i));
}
} else {
// not valid as first character -> encode
encode(name.charAt(i), encoded);
}
} else if (!XMLChar.isName(name.charAt(i))) {
encode(name.charAt(i), encoded);
} else {
if (needsEscaping(name, i)) {
// '_x' must be encoded
encode('_', encoded);
} else {
encoded.append(name.charAt(i));
}
}
}
return encoded.toString();
}
}
/**
* Encodes path as specified in ISO 9075. Please note that
* the character '[' is not encoded but rather interpreted as
* the start of an index in a path segment.
*
* @param path the String to encode.
* @return the encoded String.
*/
public static String encodePath(String path) {
String[] names = Text.explode(path, '/', true);
StringBuffer encoded = new StringBuffer(path.length());
for (int i = 0; i < names.length; i++) {
// detect index
String index = null;
int idx = names[i].indexOf('[');
if (idx != -1) {
index = names[i].substring(idx);
names[i] = names[i].substring(0, idx);
}
encoded.append(encode(names[i]));
if (index != null) {
encoded.append(index);
}
if (i < names.length - 1) {
encoded.append('/');
}
}
return encoded.toString();
}
/**
* Decodes the name.
* @param name the String to decode.
* @return the decoded String.
*/
public static String decode(String name) {
// quick check
if (name.indexOf("_x") < 0) {
// not encoded
return name;
}
StringBuffer decoded = new StringBuffer();
Matcher m = ENCODE_PATTERN.matcher(name);
while (m.find()) {
char ch = (char) Integer.parseInt(m.group().substring(2, 6), 16);
if (ch == '$' || ch == '\\') {
m.appendReplacement(decoded, "\\" + ch);
} else {
m.appendReplacement(decoded, Character.toString(ch));
}
}
m.appendTail(decoded);
return decoded.toString();
}
//-------------------------< internal >-------------------------------------
/**
* Encodes the character c as a String in the following form:
* "_x" + hex value of c + "_". Where the hex value has
* four digits if the character with possibly leading zeros.
*
* Example: ' ' (the space character) is encoded to: _x0020_
* @param c the character to encode
* @param b the encoded character is appended to StringBuffer
* b.
*/
private static void encode(char c, StringBuffer b) {
b.append("_x");
String hex = Integer.toHexString(c);
b.append(PADDING, 0, 4 - hex.length());
b.append(hex);
b.append("_");
}
/**
* Returns true if name.charAt(location) is the underscore
* character and the following character sequence is 'xHHHH_' where H
* is a hex digit.
* @param name the name to check.
* @param location the location to look at.
* @throws ArrayIndexOutOfBoundsException if location > name.length()
*/
private static boolean needsEscaping(String name, int location)
throws ArrayIndexOutOfBoundsException {
if (name.charAt(location) == '_' && name.length() >= location + 6) {
return name.charAt(location + 1) == 'x'
&& HEX_DIGITS.indexOf(name.charAt(location + 2)) != -1
&& HEX_DIGITS.indexOf(name.charAt(location + 3)) != -1
&& HEX_DIGITS.indexOf(name.charAt(location + 4)) != -1
&& HEX_DIGITS.indexOf(name.charAt(location + 5)) != -1;
} else {
return false;
}
}
}