au.id.jericho.lib.html.CharacterReference Maven / Gradle / Ivy
// Jericho HTML Parser - Java based library for analysing and manipulating HTML
// Version 2.3
// Copyright (C) 2006 Martin Jericho
// http://sourceforge.net/projects/jerichohtml/
//
// This library is free software; you can redistribute it and/or
// modify it under the terms of the GNU Lesser General Public
// License as published by the Free Software Foundation; either
// version 2.1 of the License, or (at your option) any later version.
// http://www.gnu.org/copyleft/lesser.html
//
// This library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
// Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public
// License along with this library; if not, write to the Free Software
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
package au.id.jericho.lib.html;
import java.util.*;
/**
* Represents an HTML Character Reference,
* implemented by the subclasses {@link CharacterEntityReference} and {@link NumericCharacterReference}.
*
* This class, together with its subclasses, contains static methods to perform most required operations
* without having to instantiate an object.
*
* Instances of this class are useful when the positions of character references in a source document are required,
* or to replace the found character references with customised text.
*
* CharacterReference instances are obtained using one of the following methods:
*
* - {@link CharacterReference#parse(CharSequence characterReferenceText)}
*
- {@link Source#findNextCharacterReference(int pos)}
*
- {@link Source#findPreviousCharacterReference(int pos)}
*
- {@link Segment#findAllCharacterReferences()}
*
*/
public abstract class CharacterReference extends Segment {
int codePoint;
/**
* Represents an invalid unicode code point.
*
* This can be the result of parsing a numeric character reference outside of the valid unicode range of 0x000000-0x10FFFF, or any other invalid character reference.
*/
public static final int INVALID_CODE_POINT=-1;
/**
* The maximum codepoint allowed by unicode, 0x10FFFF (decimal 1114111).
* This can be replaced by Character.MAX_CODE_POINT in java 1.5
*/
static final int MAX_CODE_POINT=0x10FFFF;
/** The number of spaces used to simulate a tab when {@linkplain #encodeWithWhiteSpaceFormatting encoding with white space formatting}. */
private static final int TAB_LENGTH=4;
CharacterReference(final Source source, final int begin, final int end, final int codePoint) {
super(source,begin,end);
this.codePoint=codePoint;
}
/**
* Returns the unicode code point represented by this character reference.
* @return the unicode code point represented by this character reference.
*/
public int getCodePoint() {
return codePoint;
}
/**
* Returns the character represented by this character reference.
*
* If this character reference represents a unicode
* supplimentary code point,
* any bits outside of the least significant 16 bits of the code point are truncated, yielding an incorrect result.
*
* @return the character represented by this character reference.
*/
public char getChar() {
return (char)codePoint;
}
/**
* Indicates whether this character reference is terminated by a semicolon (;).
*
* Conversely, this library defines an unterminated character reference as one which does
* not end with a semicolon.
*
* The SGML specification allows unterminated character references in some circumstances, and because the
* HTML 4.01 specification states simply that
* "authors may use SGML character references",
* it follows that they are also valid in HTML documents, although their use is strongly discouraged.
*
* Unterminated character references are not allowed in XHTML documents.
*
* @return true if this character reference is terminated by a semicolon, otherwise false.
* @see #decode(CharSequence encodedText, boolean insideAttributeValue)
*/
public boolean isTerminated() {
return source.charAt(end-1)==';';
}
/**
* Encodes the specified text, escaping special characters into character references.
*
* Each character is encoded only if the {@link #requiresEncoding(char)} method would return true for that character,
* using its {@link CharacterEntityReference} if available, or a decimal {@link NumericCharacterReference} if its unicode
* code point is greater than U+007F.
*
* The only exception to this is an {@linkplain CharacterEntityReference#_apos apostrophe} (U+0027),
* which depending on the current setting of the static {@link Config#IsApostropheEncoded} property,
* is either left unencoded (default setting), or encoded as the numeric character reference "'".
*
* This method never encodes an apostrophe into its character entity reference {@link CharacterEntityReference#_apos '}
* as this entity is not defined for use in HTML. See the comments in the {@link CharacterEntityReference} class for more information.
*
* To encode text using only numeric character references, use the
* {@link NumericCharacterReference#encode(CharSequence)} method instead.
*
* @param unencodedText the text to encode.
* @return the encoded string.
* @see #decode(CharSequence)
*/
public static String encode(final CharSequence unencodedText) {
if (unencodedText==null) return null;
return appendEncode(new StringBuffer(unencodedText.length()*2),unencodedText,false).toString();
}
/**
* {@linkplain #encode(CharSequence) Encodes} the specified text, preserving line breaks, tabs and spaces for rendering by converting them to markup.
*
* This performs the same encoding as the {@link #encode(CharSequence)} method, but also performs the following conversions:
*
* - Line breaks, being Carriage Return (U+000D) or Line Feed (U+000A) characters, and Form Feed characters (U+000C)
* are converted to "
<br />". CR/LF pairs are treated as a single line break.
* - Multiple consecutive spaces are converted so that every second space is converted to "
"
* while ensuring the last is always a normal space.
* - Tab characters (U+0009) are converted as if they were four consecutive spaces.
*
*
* The conversion of multiple consecutive spaces to alternating space/non-breaking-space allows the correct number of
* spaces to be rendered, but also allows the line to wrap in the middle of it.
*
* Note that zero-width spaces (U+200B) are converted to the numeric character reference
* "​" through the normal encoding process, but IE6 does not render them properly
* either encoded or unencoded.
*
* There is no method provided to reverse this encoding.
*
* @param unencodedText the text to encode.
* @return the encoded string with whitespace formatting converted to markup.
* @see #encode(CharSequence)
*/
public static String encodeWithWhiteSpaceFormatting(final CharSequence unencodedText) {
if (unencodedText==null) return null;
return appendEncode(new StringBuffer(unencodedText.length()*2),unencodedText,true).toString();
}
/**
* Decodes the specified HTML encoded text into normal text.
*
* All {@linkplain CharacterEntityReference character entity references} and {@linkplain NumericCharacterReference numeric character references}
* are converted to their respective characters.
*
* This is equivalent to {@link #decode(CharSequence,boolean) decode(encodedText,false)}.
*
* Unterminated character references are dealt with according to the rules for
* text outside of attribute values in the {@linkplain Config#CurrentCompatibilityMode current compatibility mode}.
*
* Although character entity reference names are case sensitive, and in some cases differ from other entity references only by their case,
* some browsers also recognise them in a case-insensitive way.
* For this reason, all decoding methods in this library recognise character entity reference names even if they are in the wrong case.
*
* @param encodedText the text to decode.
* @return the decoded string.
* @see #encode(CharSequence)
*/
public static String decode(final CharSequence encodedText) {
return decode(encodedText,false);
}
/**
* Decodes the specified HTML encoded text into normal text.
*
* All {@linkplain CharacterEntityReference character entity references} and {@linkplain NumericCharacterReference numeric character references}
* are converted to their respective characters.
*
* Unterminated character references are dealt with according to the
* value of the insideAttributeValue parameter and the
* {@linkplain Config#CurrentCompatibilityMode current compatibility mode}.
*
* Although character entity reference names are case sensitive, and in some cases differ from other entity references only by their case,
* some browsers also recognise them in a case-insensitive way.
* For this reason, all decoding methods in this library recognise character entity reference names even if they are in the wrong case.
*
* @param encodedText the text to decode.
* @param insideAttributeValue specifies whether the encoded text is inside an attribute value.
* @return the decoded string.
* @see #decode(CharSequence)
* @see #encode(CharSequence)
*/
public static String decode(final CharSequence encodedText, final boolean insideAttributeValue) {
if (encodedText==null) return null;
final String encodedString=encodedText.toString();
final int pos=encodedString.indexOf('&');
if (pos==-1) return encodedString;
return appendDecode(new StringBuffer(encodedString.length()),encodedString,pos,insideAttributeValue).toString();
}
/**
* {@linkplain #decode(CharSequence) Decodes} the specified text after collapsing its {@linkplain #isWhiteSpace(char) white space}.
*
* All leading and trailing white space is omitted, and any sections of internal white space are replaced by a single space.
*
* The result is how the text would normally be rendered by a
* user agent,
* assuming it does not contain any tags.
*
* @param text the source text
* @return the decoded text with collapsed white space.
* @see FormControl#getPredefinedValues()
*/
public static String decodeCollapseWhiteSpace(final CharSequence text) {
return decode(appendCollapseWhiteSpace(new StringBuffer(text.length()),text));
}
/**
* Re-encodes the specified text, equivalent to {@linkplain #decode(CharSequence) decoding} and then {@linkplain #encode(CharSequence) encoding} again.
*
* This process ensures that the specified encoded text does not contain any remaining unencoded characters.
*
* IMPLEMENTATION NOTE: At present this method simply calls the {@link #decode(CharSequence) decode} method
* followed by the {@link #encode(CharSequence) encode} method, but a more efficient implementation
* may be used in future.
*
* @param encodedText the text to re-encode.
* @return the re-encoded string.
*/
public static String reencode(final CharSequence encodedText) {
return encode(decode(encodedText,true));
}
/**
* Returns the encoded form of this character reference.
*
* The exact behaviour of this method depends on the class of this object.
* See the {@link CharacterEntityReference#getCharacterReferenceString()} and
* {@link NumericCharacterReference#getCharacterReferenceString()} methods for more details.
*
*
* - Examples:
* CharacterReference.parse(">").getCharacterReferenceString() returns ">"CharacterReference.parse(">").getCharacterReferenceString() returns "e;"
*
* @return the encoded form of this character reference.
* @see #getCharacterReferenceString(int codePoint)
* @see #getDecimalCharacterReferenceString()
*/
public abstract String getCharacterReferenceString();
/**
* Returns the encoded form of the specified unicode code point.
*
* This method returns the {@linkplain CharacterEntityReference#getCharacterReferenceString(int) character entity reference} encoded form of the unicode code point
* if one exists, otherwise it returns the {@linkplain #getDecimalCharacterReferenceString(int) decimal character reference} encoded form.
*
* The only exception to this is an {@linkplain CharacterEntityReference#_apos apostrophe} (U+0027),
* which is encoded as the numeric character reference "'" instead of its character entity reference
* "'".
*
*
* - Examples:
* CharacterReference.getCharacterReferenceString(62) returns ">"CharacterReference.getCharacterReferenceString('>') returns ">"CharacterReference.getCharacterReferenceString('☺') returns "☺"
*
* @param codePoint the unicode code point to encode.
* @return the encoded form of the specified unicode code point.
* @see #getHexadecimalCharacterReferenceString(int codePoint)
*/
public static String getCharacterReferenceString(final int codePoint) {
String characterReferenceString=null;
if (codePoint!=CharacterEntityReference._apos) characterReferenceString=CharacterEntityReference.getCharacterReferenceString(codePoint);
if (characterReferenceString==null) characterReferenceString=NumericCharacterReference.getCharacterReferenceString(codePoint);
return characterReferenceString;
}
/**
* Returns the decimal encoded form of this character reference.
*
* This is equivalent to {@link #getDecimalCharacterReferenceString(int) getDecimalCharacterReferenceString}({@link #getCodePoint()}).
*
*
* - Example:
* CharacterReference.parse(">").getDecimalCharacterReferenceString() returns ">"
*
* @return the decimal encoded form of this character reference.
* @see #getCharacterReferenceString()
* @see #getHexadecimalCharacterReferenceString()
*/
public String getDecimalCharacterReferenceString() {
return getDecimalCharacterReferenceString(codePoint);
}
/**
* Returns the decimal encoded form of the specified unicode code point.
*
*
* - Example:
* CharacterReference.getDecimalCharacterReferenceString('>') returns ">"
*
* @param codePoint the unicode code point to encode.
* @return the decimal encoded form of the specified unicode code point.
* @see #getCharacterReferenceString(int codePoint)
* @see #getHexadecimalCharacterReferenceString(int codePoint)
*/
public static String getDecimalCharacterReferenceString(final int codePoint) {
return appendDecimalCharacterReferenceString(new StringBuffer(),codePoint).toString();
}
/**
* Returns the hexadecimal encoded form of this character reference.
*
* This is equivalent to {@link #getHexadecimalCharacterReferenceString(int) getHexadecimalCharacterReferenceString}({@link #getCodePoint()}).
*
*
* - Example:
* CharacterReference.parse(">").getHexadecimalCharacterReferenceString() returns ">"
*
* @return the hexadecimal encoded form of this character reference.
* @see #getCharacterReferenceString()
* @see #getDecimalCharacterReferenceString()
*/
public String getHexadecimalCharacterReferenceString() {
return getHexadecimalCharacterReferenceString(codePoint);
}
/**
* Returns the hexadecimal encoded form of the specified unicode code point.
*
*
* - Example:
* CharacterReference.getHexadecimalCharacterReferenceString('>') returns ">"
*
* @param codePoint the unicode code point to encode.
* @return the hexadecimal encoded form of the specified unicode code point.
* @see #getCharacterReferenceString(int codePoint)
* @see #getDecimalCharacterReferenceString(int codePoint)
*/
public static String getHexadecimalCharacterReferenceString(final int codePoint) {
return appendHexadecimalCharacterReferenceString(new StringBuffer(),codePoint).toString();
}
/**
* Returns the unicode code point of this character reference in U+ notation.
*
* This is equivalent to {@link #getUnicodeText(int) getUnicodeText(getCodePoint())}.
*
*
* - Example:
* CharacterReference.parse(">").getUnicodeText() returns "U+003E"
*
* @return the unicode code point of this character reference in U+ notation.
* @see #getUnicodeText(int codePoint)
*/
public String getUnicodeText() {
return getUnicodeText(codePoint);
}
/**
* Returns the specified unicode code point in U+ notation.
*
*
* - Example:
* CharacterReference.getUnicodeText('>') returns "U+003E"
*
* @param codePoint the unicode code point.
* @return the specified unicode code point in U+ notation.
*/
public static String getUnicodeText(final int codePoint) {
return appendUnicodeText(new StringBuffer(),codePoint).toString();
}
static final StringBuffer appendUnicodeText(final StringBuffer sb, final int codePoint) {
sb.append("U+");
final String hex=Integer.toString(codePoint,16).toUpperCase();
for (int i=4-hex.length(); i>0; i--) sb.append('0');
sb.append(hex);
return sb;
}
/**
* Parses a single encoded character reference text into a CharacterReference object.
*
* The character reference must be at the start of the given text, but may contain other characters at the end.
* The {@link #getEnd() getEnd()} method can be used on the resulting object to determine at which character position the character reference ended.
*
* If the text does not represent a valid character reference, this method returns null.
*
* Unterminated character references are always accepted, regardless of the settings in the
* {@linkplain Config#CurrentCompatibilityMode current compatibility mode}.
*
* To decode all character references in a given text, use the {@link #decode(CharSequence)} method instead.
*
*
* - Example:
* CharacterReference.parse(">").getChar() returns '>'
*
* @param characterReferenceText the text containing a single encoded character reference.
* @return a CharacterReference object representing the specified text, or null if the text does not represent a valid character reference.
* @see #decode(CharSequence)
*/
public static CharacterReference parse(final CharSequence characterReferenceText) {
return construct(new Source(characterReferenceText.toString()),0,Config.UnterminatedCharacterReferenceSettings.ACCEPT_ALL);
}
/**
* Parses a single encoded character reference text into a unicode code point.
*
* The character reference must be at the start of the given text, but may contain other characters at the end.
*
* If the text does not represent a valid character reference, this method returns {@link #INVALID_CODE_POINT}.
*
* This is equivalent to {@link #parse(CharSequence) parse(characterReferenceText)}.{@link #getCodePoint()},
* except that it returns {@link #INVALID_CODE_POINT} if an invalid character reference is specified instead of throwing a
* NullPointerException.
*
*
* - Example:
* CharacterReference.getCodePointFromCharacterReferenceString(">") returns 38
*
* @param characterReferenceText the text containing a single encoded character reference.
* @return the unicode code point representing representing the specified text, or {@link #INVALID_CODE_POINT} if the text does not represent a valid character reference.
*/
public static int getCodePointFromCharacterReferenceString(final CharSequence characterReferenceText) {
final CharacterReference characterReference=parse(characterReferenceText);
return (characterReference!=null) ? characterReference.getCodePoint() : INVALID_CODE_POINT;
}
/**
* Indicates whether the specified character would need to be encoded in HTML text.
*
* This is the case if a {@linkplain CharacterEntityReference character entity reference} exists for the character, or the unicode code point is greater than U+007F.
*
* The only exception to this is an {@linkplain CharacterEntityReference#_apos apostrophe} (U+0027),
* which only returns true if the static {@link Config#IsApostropheEncoded} property
* is currently set to true.
*
* @param ch the character to test.
* @return true if the specified character would need to be encoded in HTML text, otherwise false.
*/
public static final boolean requiresEncoding(final char ch) {
return ch>127 || (CharacterEntityReference.getName(ch)!=null && (ch!='\'' || Config.IsApostropheEncoded));
}
static StringBuffer appendEncode(final StringBuffer sb, CharSequence unencodedText, final boolean whiteSpaceFormatting) {
if (unencodedText==null) return sb;
int beginPos=0;
int endPos=unencodedText.length();
if (unencodedText instanceof Segment) {
// this might improve performance slightly
final Segment segment=(Segment)unencodedText;
final int segmentOffset=segment.getBegin();
beginPos=segmentOffset;
endPos+=segmentOffset;
unencodedText=segment.source.string;
}
final boolean isApostropheEncoded=Config.IsApostropheEncoded;
for (int i=beginPos; i127) {
appendDecimalCharacterReferenceString(sb,ch);
} else if (!(whiteSpaceFormatting && isWhiteSpace(ch))) {
sb.append(ch);
} else {
// whiteSpaceFormatting tries to simulate the formatting characters by converting them to markup
int spaceCount;
int nexti=i+1;
if (ch!=' ') {
if (ch!='\t') {
// must be line feed, carriage return or form feed, since zero-width space should have been processed as a character reference string
if (ch=='\r' && nexti=2) {
sb.append(" "); // use alternating and spaces to keep original number of spaces
spaceCount-=2;
}
// note that the last character is never a nbsp, so that word wrapping won't result in a nbsp before the first character in a line
i=nexti-1; // minus 1 because top level for loop will add it again
}
}
return sb;
}
static CharacterReference findPreviousOrNext(final Source source, final int pos, final boolean previous) {
return findPreviousOrNext(source,pos,Config.UnterminatedCharacterReferenceSettings.ACCEPT_ALL,previous);
}
private static CharacterReference findPreviousOrNext(final Source source, int pos, final Config.UnterminatedCharacterReferenceSettings unterminatedCharacterReferenceSettings, final boolean previous) {
final ParseText parseText=source.getParseText();
pos=previous ? parseText.lastIndexOf('&',pos) : parseText.indexOf('&',pos);
while (pos!=-1) {
final CharacterReference characterReference=construct(source,pos,unterminatedCharacterReferenceSettings);
if (characterReference!=null) return characterReference;
pos=previous ? parseText.lastIndexOf('&',pos-1) : parseText.indexOf('&',pos+1);
}
return null;
}
static final StringBuffer appendHexadecimalCharacterReferenceString(final StringBuffer sb, final int codePoint) {
return sb.append("&#x").append(Integer.toString(codePoint,16)).append(';');
}
static final StringBuffer appendDecimalCharacterReferenceString(final StringBuffer sb, final int codePoint) {
return sb.append("&#").append(codePoint).append(';');
}
private static CharacterReference construct(final Source source, final int begin, final Config.UnterminatedCharacterReferenceSettings unterminatedCharacterReferenceSettings) {
try {
if (source.getParseText().charAt(begin)!='&') return null;
return (source.getParseText().charAt(begin+1)=='#')
? NumericCharacterReference.construct(source,begin,unterminatedCharacterReferenceSettings)
: CharacterEntityReference.construct(source,begin,unterminatedCharacterReferenceSettings.characterEntityReferenceMaxCodePoint);
} catch (IndexOutOfBoundsException ex) {
return null;
}
}
private static StringBuffer appendDecode(final StringBuffer sb, final String encodedString, int pos, final boolean insideAttributeValue) {
final Config.UnterminatedCharacterReferenceSettings unterminatedCharacterReferenceSettings=Config.CurrentCompatibilityMode.getUnterminatedCharacterReferenceSettings(insideAttributeValue);
int lastEnd=0;
final Source source=new Source(encodedString);
while (true) {
final CharacterReference characterReference=findPreviousOrNext(source,pos,unterminatedCharacterReferenceSettings,false);
if (characterReference==null) break;
if (lastEnd!=characterReference.getBegin()) Util.appendTo(sb,encodedString,lastEnd,characterReference.getBegin());
sb.append((char)characterReference.codePoint);
pos=lastEnd=characterReference.getEnd();
}
if (lastEnd!=encodedString.length()) Util.appendTo(sb,encodedString,lastEnd,encodedString.length());
return sb;
}
}