/*
* =============================================================================
*
* Copyright (c) 2014, The UNBESCAPE team (http://www.unbescape.org)
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
* =============================================================================
*/
package org.zkoss.lang;
import java.io.IOException;
import java.io.Writer;
import java.util.Arrays;
/**
* <p>
* Internal class in charge of performing the real escape/unescape operations.
* </p>
*
* @author Daniel Fernández
*
* @since 1.0.0
*
*/
/**package*/ final class JavaScriptEscapeUtil {
/*
* JAVASCRIPT ESCAPE/UNESCAPE OPERATIONS
* -------------------------------------
*
* See: http://www.ecmascript.org/docs.php
* http://mathiasbynens.be/notes/javascript-escapes
*
* (Note that, in the following examples, and in order to avoid escape problems during the compilation
* of this class, the backslash symbol is replaced by '%')
*
* - SINGLE ESCAPE CHARACTERS (SECs):
* U+0000 -> %0
* U+0008 -> %b
* U+0009 -> %t
* U+000A -> %n
* U+000B -> %v [NOT USED IN ESCAPE - Not supported by Internet Explorer < 9]
* U+000C -> %f
* U+000D -> %r
* U+0022 -> %"
* U+0027 -> %'
* U+005C -> %%
* U+002F -> %/ [ONLY USED WHEN / APPEARS IN </, IN ORDER TO AVOID ISSUES INSIDE <script> TAGS]
* - HEXADECIMAL ESCAPE [XHEXA] (only for characters <= U+00FF): %x??
* - UNICODE ESCAPE [UHEXA] (also hexadecimal)
* Characters <= U+FFFF: %u????
* Characters > U+FFFF : %u????%u???? (surrogate character pair)
* %u{?*} [NOT USED - Possible syntax for ECMAScript 6]
* - OCTAL ESCAPE: %377 [NOT USED IN ESCAPE - Deprecated in ECMAScript v5]
* - GENERAL ESCAPE: %* -> * ('%a' -> 'a')
* (except the [%,%n] sequence, which is not an escape sequence but a line continuation)
*
*/
/*
* Prefixes defined for use in escape and unescape operations
*/
private static final char ESCAPE_PREFIX = '\\';
private static final char ESCAPE_XHEXA_PREFIX2 = 'x';
private static final char ESCAPE_UHEXA_PREFIX2 = 'u';
private static final char[] ESCAPE_XHEXA_PREFIX = "\\x".toCharArray();
private static final char[] ESCAPE_UHEXA_PREFIX = "\\u".toCharArray();
/*
* Small utility char arrays for hexadecimal conversion.
*/
private static char[] HEXA_CHARS_UPPER = "0123456789ABCDEF".toCharArray();
private static char[] HEXA_CHARS_LOWER = "0123456789abcdef".toCharArray();
/*
* Structures for holding the Single Escape Characters
*/
private static int SEC_CHARS_LEN = '\\' + 1; // 0x5C + 1 = 0x5D
private static char SEC_CHARS_NO_SEC = '*';
private static char[] SEC_CHARS;
/*
* Structured for holding the 'escape level' assigned to chars (not codepoints) up to ESCAPE_LEVELS_LEN.
* - The last position of the ESCAPE_LEVELS array will be used for determining the level of all
* codepoints >= (ESCAPE_LEVELS_LEN - 1)
*/
private static final char ESCAPE_LEVELS_LEN = 0x9f + 2; // Last relevant char to be indexed is 0x9f
private static final byte[] ESCAPE_LEVELS;
static {
/*
* Initialize Single Escape Characters
*/
SEC_CHARS = new char[SEC_CHARS_LEN];
Arrays.fill(SEC_CHARS,SEC_CHARS_NO_SEC);
SEC_CHARS[0x00] = '0';
SEC_CHARS[0x08] = 'b';
SEC_CHARS[0x09] = 't';
SEC_CHARS[0x0A] = 'n';
SEC_CHARS[0x0C] = 'f';
SEC_CHARS[0x0D] = 'r';
SEC_CHARS[0x22] = '"';
SEC_CHARS[0x27] = '\'';
SEC_CHARS[0x5C] = '\\';
// slash (solidus) character: will only be escaped if in '</'
SEC_CHARS[0x2F] = '/';
/*
* Initialization of escape levels.
* Defined levels :
*
* - Level 1 : Basic escape set
* - Level 2 : Basic escape set plus all non-ASCII
* - Level 3 : All non-alphanumeric characters
* - Level 4 : All characters
*
*/
ESCAPE_LEVELS = new byte[ESCAPE_LEVELS_LEN];
/*
* Everything is level 3 unless contrary indication.
*/
Arrays.fill(ESCAPE_LEVELS, (byte)3);
/*
* Everything non-ASCII is level 2 unless contrary indication.
*/
for (char c = 0x80; c < ESCAPE_LEVELS_LEN; c++) {
ESCAPE_LEVELS[c] = 2;
}
/*
* Alphanumeric characters are level 4.
*/
for (char c = 'A'; c <= 'Z'; c++) {
ESCAPE_LEVELS[c] = 4;
}
for (char c = 'a'; c <= 'z'; c++) {
ESCAPE_LEVELS[c] = 4;
}
for (char c = '0'; c <= '9'; c++) {
ESCAPE_LEVELS[c] = 4;
}
/*
* Simple Escape Character will be level 1 (always escaped)
*/
ESCAPE_LEVELS[0x00] = 1;
ESCAPE_LEVELS[0x08] = 1;
ESCAPE_LEVELS[0x09] = 1;
ESCAPE_LEVELS[0x0A] = 1;
ESCAPE_LEVELS[0x0C] = 1;
ESCAPE_LEVELS[0x0D] = 1;
ESCAPE_LEVELS[0x22] = 1;
ESCAPE_LEVELS[0x27] = 1;
ESCAPE_LEVELS[0x5C] = 1;
// slash (solidus) character: will only be escaped if in '</', but we signal it as level 1 anyway
ESCAPE_LEVELS[0x2F] = 1;
/*
* JavaScript defines two ranges of non-displayable, control characters (some of which are already part of the
* Single Escape Characters list): U+0001 to U+001F and U+007F to U+009F.
*/
for (char c = 0x01; c <= 0x1F; c++) {
ESCAPE_LEVELS[c] = 1;
}
for (char c = 0x7F; c <= 0x9F; c++) {
ESCAPE_LEVELS[c] = 1;
}
}
private JavaScriptEscapeUtil() {
super();
}
static char[] toXHexa(final int codepoint) {
final char[] result = new char[2];
result[1] = HEXA_CHARS_UPPER[codepoint % 0x10];
result[0] = HEXA_CHARS_UPPER[(codepoint >>> 4) % 0x10];
return result;
}
static char[] toUHexa(final int codepoint) {
final char[] result = new char[4];
result[3] = HEXA_CHARS_UPPER[codepoint % 0x10];
result[2] = HEXA_CHARS_UPPER[(codepoint >>> 4) % 0x10];
result[1] = HEXA_CHARS_UPPER[(codepoint >>> 8) % 0x10];
result[0] = HEXA_CHARS_UPPER[(codepoint >>> 12) % 0x10];
return result;
}
/*
* Perform an escape operation, based on String, according to the specified level and type.
*/
static String escape(final String text,
final JavaScriptEscapeType escapeType, final JavaScriptEscapeLevel escapeLevel) {
if (text == null) {
return null;
}
final int level = escapeLevel.getEscapeLevel();
final boolean useSECs = escapeType.getUseSECs();
final boolean useXHexa = escapeType.getUseXHexa();
StringBuilder strBuilder = null;
final int offset = 0;
final int max = text.length();
int readOffset = offset;
for (int i = offset; i < max; i++) {
final int codepoint = Character.codePointAt(text, i);
/*
* Shortcut: most characters will be ASCII/Alphanumeric, and we won't need to do anything at
* all for them
*/
if (codepoint <= (ESCAPE_LEVELS_LEN - 2) && level < ESCAPE_LEVELS[codepoint]) {
continue;
}
/*
* Check whether the character is a slash (solidus). In such case, only escape if it
* appears after a '<' ('</') or level >= 3 (non alphanumeric)
*/
if (codepoint == '/' && level < 3 && (i == 0 || text.charAt(i - 1) != '<')) {
continue;
}
/*
* Shortcut: we might not want to escape non-ASCII chars at all either.
* We also check we are not dealing with U+2028 or U+2029, which the JavaScript spec considers
* LineTerminators and therefore should be escaped always.
*/
if (codepoint > (ESCAPE_LEVELS_LEN - 2) && level < ESCAPE_LEVELS[ESCAPE_LEVELS_LEN - 1] &&
codepoint != '\u2028' && codepoint != '\u2029') {
if (Character.charCount(codepoint) > 1) {
// This is to compensate that we are actually escaping two char[] positions with a single codepoint.
i++;
}
continue;
}
/*
* At this point we know for sure we will need some kind of escape, so we
* can increase the offset and initialize the string builder if needed, along with
* copying to it all the contents pending up to this point.
*/
if (strBuilder == null) {
strBuilder = new StringBuilder(max + 20);
}
if (i - readOffset > 0) {
strBuilder.append(text, readOffset, i);
}
if (Character.charCount(codepoint) > 1) {
// This is to compensate that we are actually reading two char[] positions with a single codepoint.
i++;
}
readOffset = i + 1;
/*
* -----------------------------------------------------------------------------------------
*
* Peform the real escape, attending the different combinations of SECs, XHEXA and UHEXA
*
* -----------------------------------------------------------------------------------------
*/
if (useSECs && codepoint < SEC_CHARS_LEN) {
// We will try to use a SEC
final char sec = SEC_CHARS[codepoint];
if (sec != SEC_CHARS_NO_SEC) {
// SEC found! just write it and go for the next char
strBuilder.append(ESCAPE_PREFIX);
strBuilder.append(sec);
continue;
}
}
/*
* No SEC-escape was possible, so we need xhexa/uhexa escape.
*/
if (useXHexa && codepoint <= 0xFF) {
// Codepoint is <= 0xFF, so we can use XHEXA escapes
strBuilder.append(ESCAPE_XHEXA_PREFIX);
strBuilder.append(toXHexa(codepoint));
continue;
}
if (Character.charCount(codepoint) > 1) {
final char[] codepointChars = Character.toChars(codepoint);
strBuilder.append(ESCAPE_UHEXA_PREFIX);
strBuilder.append(toUHexa(codepointChars[0]));
strBuilder.append(ESCAPE_UHEXA_PREFIX);
strBuilder.append(toUHexa(codepointChars[1]));
continue;
}
strBuilder.append(ESCAPE_UHEXA_PREFIX);
strBuilder.append(toUHexa(codepoint));
}
/*
* -----------------------------------------------------------------------------------------------
* Final cleaning: return the original String object if no escape was actually needed. Otherwise
* append the remaining unescaped text to the string builder and return.
* -----------------------------------------------------------------------------------------------
*/
if (strBuilder == null) {
return text;
}
if (max - readOffset > 0) {
strBuilder.append(text, readOffset, max);
}
return strBuilder.toString();
}
/*
* Perform an escape operation, based on char[], according to the specified level and type.
*/
static void escape(final char[] text, final int offset, final int len, final Writer writer,
final JavaScriptEscapeType escapeType, final JavaScriptEscapeLevel escapeLevel)
throws IOException {
if (text == null || text.length == 0) {
return;
}
final int level = escapeLevel.getEscapeLevel();
final boolean useSECs = escapeType.getUseSECs();
final boolean useXHexa = escapeType.getUseXHexa();
final int max = (offset + len);
int readOffset = offset;
for (int i = offset; i < max; i++) {
final int codepoint = Character.codePointAt(text, i);
/*
* Shortcut: most characters will be ASCII/Alphanumeric, and we won't need to do anything at
* all for them
*/
if (codepoint <= (ESCAPE_LEVELS_LEN - 2) && level < ESCAPE_LEVELS[codepoint]) {
continue;
}
/*
* Check whether the character is a slash (solidus). In such case, only escape if it
* appears after a '<' ('</') or level >= 3 (non alphanumeric)
*/
if (codepoint == '/' && level < 3 && (i == 0 || text[i - 1] != '<')) {
continue;
}
/*
* Shortcut: we might not want to escape non-ASCII chars at all either.
* We also check we are not dealing with U+2028 or U+2029, which the JavaScript spec considers
* LineTerminators and therefore should be escaped always.
*/
if (codepoint > (ESCAPE_LEVELS_LEN - 2) && level < ESCAPE_LEVELS[ESCAPE_LEVELS_LEN - 1] &&
codepoint != '\u2028' && codepoint != '\u2029') {
if (Character.charCount(codepoint) > 1) {
// This is to compensate that we are actually escaping two char[] positions with a single codepoint.
i++;
}
continue;
}
/*
* At this point we know for sure we will need some kind of escape, so we
* can write all the contents pending up to this point.
*/
if (i - readOffset > 0) {
writer.write(text, readOffset, (i - readOffset));
}
if (Character.charCount(codepoint) > 1) {
// This is to compensate that we are actually reading two char[] positions with a single codepoint.
i++;
}
readOffset = i + 1;
/*
* -----------------------------------------------------------------------------------------
*
* Peform the real escape, attending the different combinations of SECs, XHEXA and UHEXA
*
* -----------------------------------------------------------------------------------------
*/
if (useSECs && codepoint < SEC_CHARS_LEN) {
// We will try to use a SEC
final char sec = SEC_CHARS[codepoint];
if (sec != SEC_CHARS_NO_SEC) {
// SEC found! just write it and go for the next char
writer.write(ESCAPE_PREFIX);
writer.write(sec);
continue;
}
}
/*
* No SEC-escape was possible, so we need xhexa/uhexa escape.
*/
if (useXHexa && codepoint <= 0xFF) {
// Codepoint is <= 0xFF, so we can use XHEXA escapes
writer.write(ESCAPE_XHEXA_PREFIX);
writer.write(toXHexa(codepoint));
continue;
}
if (Character.charCount(codepoint) > 1) {
final char[] codepointChars = Character.toChars(codepoint);
writer.write(ESCAPE_UHEXA_PREFIX);
writer.write(toUHexa(codepointChars[0]));
writer.write(ESCAPE_UHEXA_PREFIX);
writer.write(toUHexa(codepointChars[1]));
continue;
}
writer.write(ESCAPE_UHEXA_PREFIX);
writer.write(toUHexa(codepoint));
}
/*
* -----------------------------------------------------------------------------------------------
* Final cleaning: return the original String object if no escape was actually needed. Otherwise
* append the remaining unescaped text to the string builder and return.
* -----------------------------------------------------------------------------------------------
*/
if (max - readOffset > 0) {
writer.write(text, readOffset, (max - readOffset));
}
}
/*
* This methods (the two versions) are used instead of Integer.parseInt(str,radix) in order to avoid the need
* to create substrings of the text being unescaped to feed such method.
* - No need to check all chars are within the radix limits - reference parsing code will already have done so.
*/
static int parseIntFromReference(final String text, final int start, final int end, final int radix) {
int result = 0;
for (int i = start; i < end; i++) {
final char c = text.charAt(i);
int n = -1;
for (int j = 0; j < HEXA_CHARS_UPPER.length; j++) {
if (c == HEXA_CHARS_UPPER[j] || c == HEXA_CHARS_LOWER[j]) {
n = j;
break;
}
}
result = (radix * result) + n;
}
return result;
}
static int parseIntFromReference(final char[] text, final int start, final int end, final int radix) {
int result = 0;
for (int i = start; i < end; i++) {
final char c = text[i];
int n = -1;
for (int j = 0; j < HEXA_CHARS_UPPER.length; j++) {
if (c == HEXA_CHARS_UPPER[j] || c == HEXA_CHARS_LOWER[j]) {
n = j;
break;
}
}
result = (radix * result) + n;
}
return result;
}
static boolean isOctalEscape(final String text, final int start, final int end) {
if (start >= end) {
return false;
}
final char c1 = text.charAt(start);
if (c1 < '0' || c1 > '7') {
return false;
}
if (start + 1 >= end) {
return (c1 != '0'); // It would not be an octal escape, but the U+0000 escape sequence.
}
final char c2 = text.charAt(start + 1);
if (c2 < '0' || c2 > '7') {
return (c1 != '0'); // It would not be an octal escape, but the U+0000 escape sequence.
}
if (start + 2 >= end) {
return (c1 != '0' || c2 != '0'); // It would not be an octal escape, but the U+0000 escape sequence + '0'.
}
final char c3 = text.charAt(start + 2);
if (c3 < '0' || c3 > '7') {
return (c1 != '0' || c2 != '0'); // It would not be an octal escape, but the U+0000 escape sequence + '0'.
}
return (c1 != '0' || c2 != '0' || c3 != '0'); // Check it's not U+0000 (escaped) + '00'
}
static boolean isOctalEscape(final char[] text, final int start, final int end) {
if (start >= end) {
return false;
}
final char c1 = text[start];
if (c1 < '0' || c1 > '7') {
return false;
}
if (start + 1 >= end) {
return (c1 != '0'); // It would not be an octal escape, but the U+0000 escape sequence.
}
final char c2 = text[start + 1];
if (c2 < '0' || c2 > '7') {
return (c1 != '0'); // It would not be an octal escape, but the U+0000 escape sequence.
}
if (start + 2 >= end) {
return (c1 != '0' || c2 != '0'); // It would not be an octal escape, but the U+0000 escape sequence + '0'.
}
final char c3 = text[start + 2];
if (c3 < '0' || c3 > '7') {
return (c1 != '0' || c2 != '0'); // It would not be an octal escape, but the U+0000 escape sequence + '0'.
}
return (c1 != '0' || c2 != '0' || c3 != '0'); // Check it's not U+0000 (escaped) + '00'
}
/*
* Perform an unescape operation based on String.
*/
static String unescape(final String text) {
if (text == null) {
return null;
}
StringBuilder strBuilder = null;
final int offset = 0;
final int max = text.length();
int readOffset = offset;
int referenceOffset = offset;
for (int i = offset; i < max; i++) {
final char c = text.charAt(i);
/*
* Check the need for an unescape operation at this point
*/
if (c != ESCAPE_PREFIX || (i + 1) >= max) {
continue;
}
int codepoint = -1;
if (c == ESCAPE_PREFIX) {
final char c1 = text.charAt(i + 1);
switch (c1) {
case '0': if (!isOctalEscape(text,i + 1,max)) { codepoint = 0x00; referenceOffset = i + 1; }; break;
case 'b': codepoint = 0x08; referenceOffset = i + 1; break;
case 't': codepoint = 0x09; referenceOffset = i + 1; break;
case 'n': codepoint = 0x0A; referenceOffset = i + 1; break;
case 'v': codepoint = 0x0B; referenceOffset = i + 1; break;
case 'f': codepoint = 0x0C; referenceOffset = i + 1; break;
case 'r': codepoint = 0x0D; referenceOffset = i + 1; break;
case '"': codepoint = 0x22; referenceOffset = i + 1; break;
case '\'': codepoint = 0x27; referenceOffset = i + 1; break;
case '\\': codepoint = 0x5C; referenceOffset = i + 1; break;
case '/': codepoint = 0x2F; referenceOffset = i + 1; break;
}
if (codepoint == -1) {
if (c1 == ESCAPE_XHEXA_PREFIX2) {
// This can be a xhexa escape, we need exactly two more characters
int f = i + 2;
while (f < (i + 4) && f < max) {
final char cf = text.charAt(f);
if (!((cf >= '0' && cf <= '9') || (cf >= 'A' && cf <= 'F') || (cf >= 'a' && cf <= 'f'))) {
break;
}
f++;
}
if ((f - (i + 2)) < 2) {
// We weren't able to consume the required two hexa chars, leave it as slash+'x', which
// is invalid, and let the corresponding JavaScript engine fail.
i++;
continue;
}
codepoint = parseIntFromReference(text, i + 2, f, 16);
// Fast-forward to the first char after the parsed codepoint
referenceOffset = f - 1;
// Don't continue here, just let the unescape code below do its job
} else if (c1 == ESCAPE_UHEXA_PREFIX2) {
// This can be a uhexa escape, we need exactly four more characters
int f = i + 2;
while (f < (i + 6) && f < max) {
final char cf = text.charAt(f);
if (!((cf >= '0' && cf <= '9') || (cf >= 'A' && cf <= 'F') || (cf >= 'a' && cf <= 'f'))) {
break;
}
f++;
}
if ((f - (i + 2)) < 4) {
// We weren't able to consume the required four hexa chars, leave it as slash+'u', which
// is invalid, and let the corresponding JavaScript engine fail.
i++;
continue;
}
codepoint = parseIntFromReference(text, i + 2, f, 16);
// Fast-forward to the first char after the parsed codepoint
referenceOffset = f - 1;
// Don't continue here, just let the unescape code below do its job
} else if (c1 >= '0' && c1 <= '7') {
// This can be a octal escape, we need at least 1 more char, and up to 3 more.
int f = i + 2;
while (f < (i + 4) && f < max) { // We need only a max of two more chars
final char cf = text.charAt(f);
if (!(cf >= '0' && cf <= '7')) {
break;
}
f++;
}
codepoint = parseIntFromReference(text, i + 1, f, 8);
if (codepoint > 0xFF) {
// Maximum octal escape char is FF. Ignore the last digit
codepoint = parseIntFromReference(text, i + 1, f - 1, 8);
referenceOffset = f - 2;
} else {
referenceOffset = f - 1;
}
// Don't continue here, just let the unescape code below do its job
} else if (c1 == '8' || c1 == '9' || c1 == '\n' || c1 == '\r' || c1 == '\u2028' || c1 == '\u2029') {
// '8' and '9' are not valid octal escape sequences, and the other four characters
// are LineTerminators, which are not allowed as escape sequences. So we leave it as is
// and expect the corresponding JavaScript engine to fail (except in the case of slash + '\n',
// which is considered a LineContinuator).
i++;
continue;
} else {
// We weren't able to consume any valid escape chars, just consider it a normal char,
// which is allowed by the JavaScript specification (NonEscapeCharacter)
codepoint = (int) c1;
referenceOffset = i + 1;
}
}
}
/*
* At this point we know for sure we will need some kind of unescape, so we
* can increase the offset and initialize the string builder if needed, along with
* copying to it all the contents pending up to this point.
*/
if (strBuilder == null) {
strBuilder = new StringBuilder(max + 5);
}
if (i - readOffset > 0) {
strBuilder.append(text, readOffset, i);
}
i = referenceOffset;
readOffset = i + 1;
/*
* --------------------------
*
* Peform the real unescape
*
* --------------------------
*/
if (codepoint > '\uFFFF') {
strBuilder.append(Character.toChars(codepoint));
} else {
strBuilder.append((char)codepoint);
}
}
/*
* -----------------------------------------------------------------------------------------------
* Final cleaning: return the original String object if no unescape was actually needed. Otherwise
* append the remaining escaped text to the string builder and return.
* -----------------------------------------------------------------------------------------------
*/
if (strBuilder == null) {
return text;
}
if (max - readOffset > 0) {
strBuilder.append(text, readOffset, max);
}
return strBuilder.toString();
}
/*
* Perform an unescape operation based on char[].
*/
static void unescape(final char[] text, final int offset, final int len, final Writer writer)
throws IOException {
if (text == null) {
return;
}
final int max = (offset + len);
int readOffset = offset;
int referenceOffset = offset;
for (int i = offset; i < max; i++) {
final char c = text[i];
/*
* Check the need for an unescape operation at this point
*/
if (c != ESCAPE_PREFIX || (i + 1) >= max) {
continue;
}
int codepoint = -1;
if (c == ESCAPE_PREFIX) {
final char c1 = text[i + 1];
switch (c1) {
case '0': if (!isOctalEscape(text,i + 1,max)) { codepoint = 0x00; referenceOffset = i + 1; }; break;
case 'b': codepoint = 0x08; referenceOffset = i + 1; break;
case 't': codepoint = 0x09; referenceOffset = i + 1; break;
case 'n': codepoint = 0x0A; referenceOffset = i + 1; break;
case 'v': codepoint = 0x0B; referenceOffset = i + 1; break;
case 'f': codepoint = 0x0C; referenceOffset = i + 1; break;
case 'r': codepoint = 0x0D; referenceOffset = i + 1; break;
case '"': codepoint = 0x22; referenceOffset = i + 1; break;
case '\'': codepoint = 0x27; referenceOffset = i + 1; break;
case '\\': codepoint = 0x5C; referenceOffset = i + 1; break;
case '/': codepoint = 0x2F; referenceOffset = i + 1; break;
}
if (codepoint == -1) {
if (c1 == ESCAPE_XHEXA_PREFIX2) {
// This can be a xhexa escape, we need exactly two more characters
int f = i + 2;
while (f < (i + 4) && f < max) {
final char cf = text[f];
if (!((cf >= '0' && cf <= '9') || (cf >= 'A' && cf <= 'F') || (cf >= 'a' && cf <= 'f'))) {
break;
}
f++;
}
if ((f - (i + 2)) < 2) {
// We weren't able to consume the required two hexa chars, leave it as slash+'x', which
// is invalid, and let the corresponding JavaScript engine fail.
i++;
continue;
}
codepoint = parseIntFromReference(text, i + 2, f, 16);
// Fast-forward to the first char after the parsed codepoint
referenceOffset = f - 1;
// Don't continue here, just let the unescape code below do its job
} else if (c1 == ESCAPE_UHEXA_PREFIX2) {
// This can be a uhexa escape, we need exactly four more characters
int f = i + 2;
while (f < (i + 6) && f < max) {
final char cf = text[f];
if (!((cf >= '0' && cf <= '9') || (cf >= 'A' && cf <= 'F') || (cf >= 'a' && cf <= 'f'))) {
break;
}
f++;
}
if ((f - (i + 2)) < 4) {
// We weren't able to consume the required four hexa chars, leave it as slash+'u', which
// is invalid, and let the corresponding JavaScript engine fail.
i++;
continue;
}
codepoint = parseIntFromReference(text, i + 2, f, 16);
// Fast-forward to the first char after the parsed codepoint
referenceOffset = f - 1;
// Don't continue here, just let the unescape code below do its job
} else if (c1 >= '0' && c1 <= '7') {
// This can be a octal escape, we need at least 1 more char, and up to 3 more.
int f = i + 2;
while (f < (i + 4) && f < max) { // We need only a max of two more chars
final char cf = text[f];
if (!(cf >= '0' && cf <= '7')) {
break;
}
f++;
}
codepoint = parseIntFromReference(text, i + 1, f, 8);
if (codepoint > 0xFF) {
// Maximum octal escape char is FF. Ignore the last digit
codepoint = parseIntFromReference(text, i + 1, f - 1, 8);
referenceOffset = f - 2;
} else {
referenceOffset = f - 1;
}
// Don't continue here, just let the unescape code below do its job
} else if (c1 == '8' || c1 == '9' || c1 == '\n' || c1 == '\r' || c1 == '\u2028' || c1 == '\u2029') {
// '8' and '9' are not valid octal escape sequences, and the other four characters
// are LineTerminators, which are not allowed as escape sequences. So we leave it as is
// and expect the corresponding JavaScript engine to fail (except in the case of slash + '\n',
// which is considered a LineContinuator).
i++;
continue;
} else {
// We weren't able to consume any valid escape chars, just consider it a normal char,
// which is allowed by the JavaScript specification (NonEscapeCharacter)
codepoint = (int) c1;
referenceOffset = i + 1;
}
}
}
/*
* At this point we know for sure we will need some kind of unescape, so we
* write all the contents pending up to this point.
*/
if (i - readOffset > 0) {
writer.write(text, readOffset, (i - readOffset));
}
i = referenceOffset;
readOffset = i + 1;
/*
* --------------------------
*
* Peform the real unescape
*
* --------------------------
*/
if (codepoint > '\uFFFF') {
writer.write(Character.toChars(codepoint));
} else {
writer.write((char)codepoint);
}
}
/*
* -----------------------------------------------------------------------------------------------
* Final cleaning: writer the remaining escaped text and return.
* -----------------------------------------------------------------------------------------------
*/
if (max - readOffset > 0) {
writer.write(text, readOffset, (max - readOffset));
}
}
}