nyancrimew
/
LatinIME
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Merge RichInputConnection with EditingUtils 

Change-Id: I9982ff325bef56694402caef28a77683c52ccd71
main
Jean Chalard 2012-06-08 19:44:38 +09:00
parent 5475b38328
commit 02308bec63
5 changed files with 239 additions and 270 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

218
java/src/com/android/inputmethod/latin/EditingUtils.java
View File

@ -1,218 +0,0 @@
/*
* Copyright (C) 2009 The Android Open Source Project
*
* 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 com.android.inputmethod.latin;
import android.view.inputmethod.ExtractedText;
import android.view.inputmethod.ExtractedTextRequest;
import android.view.inputmethod.InputConnection;
import java.util.regex.Pattern;
/**
* Utility methods to deal with editing text through an InputConnection.
*/
public class EditingUtils {
/**
* Number of characters we want to look back in order to identify the previous word
*/
// Provision for a long word pair and a separator
private static final int LOOKBACK_CHARACTER_NUM = BinaryDictionary.MAX_WORD_LENGTH * 2 + 1;
private static final int INVALID_CURSOR_POSITION = -1;
private EditingUtils() {
// Unintentional empty constructor for singleton.
}
private static int getCursorPosition(InputConnection connection) {
if (null == connection) return INVALID_CURSOR_POSITION;
final ExtractedText extracted = connection.getExtractedText(new ExtractedTextRequest(), 0);
if (extracted == null) {
return INVALID_CURSOR_POSITION;
}
return extracted.startOffset + extracted.selectionStart;
}
/**
* @param connection connection to the current text field.
* @param separators characters which may separate words
* @return the word that surrounds the cursor, including up to one trailing
* separator. For example, if the field contains "he|llo world", where |
* represents the cursor, then "hello " will be returned.
*/
public static String getWordAtCursor(InputConnection connection, String separators) {
// getWordRangeAtCursor returns null if the connection is null
Range r = getWordRangeAtCursor(connection, separators, 0);
return (r == null) ? null : r.mWord;
}
/**
* Represents a range of text, relative to the current cursor position.
*/
public static class Range {
/** Characters before selection start */
public final int mCharsBefore;
/**
* Characters after selection start, including one trailing word
* separator.
*/
public final int mCharsAfter;
/** The actual characters that make up a word */
public final String mWord;
public Range(int charsBefore, int charsAfter, String word) {
if (charsBefore < 0 || charsAfter < 0) {
throw new IndexOutOfBoundsException();
}
this.mCharsBefore = charsBefore;
this.mCharsAfter = charsAfter;
this.mWord = word;
}
}
/**
* Returns the text surrounding the cursor.
*
* @param connection the InputConnection to the TextView
* @param sep a string of characters that split words.
* @param additionalPrecedingWordsCount the number of words before the current word that should
* be included in the returned range
* @return a range containing the text surrounding the cursor
*/
public static Range getWordRangeAtCursor(InputConnection connection, String sep,
int additionalPrecedingWordsCount) {
if (connection == null || sep == null) {
return null;
}
CharSequence before = connection.getTextBeforeCursor(1000, 0);
CharSequence after = connection.getTextAfterCursor(1000, 0);
if (before == null || after == null) {
return null;
}
// Going backward, alternate skipping non-separators and separators until enough words
// have been read.
int start = before.length();
boolean isStoppingAtWhitespace = true; // toggles to indicate what to stop at
while (true) { // see comments below for why this is guaranteed to halt
while (start > 0) {
final int codePoint = Character.codePointBefore(before, start);
if (isStoppingAtWhitespace == isSeparator(codePoint, sep)) {
break; // inner loop
}
--start;
if (Character.isSupplementaryCodePoint(codePoint)) {
--start;
}
}
// isStoppingAtWhitespace is true every other time through the loop,
// so additionalPrecedingWordsCount is guaranteed to become < 0, which
// guarantees outer loop termination
if (isStoppingAtWhitespace && (--additionalPrecedingWordsCount < 0)) {
break; // outer loop
}
isStoppingAtWhitespace = !isStoppingAtWhitespace;
}
// Find last word separator after the cursor
int end = -1;
while (++end < after.length()) {
final int codePoint = Character.codePointAt(after, end);
if (isSeparator(codePoint, sep)) {
break;
}
if (Character.isSupplementaryCodePoint(codePoint)) {
++end;
}
}
int cursor = getCursorPosition(connection);
if (start >= 0 && cursor + end <= after.length() + before.length()) {
String word = before.toString().substring(start, before.length())
+ after.toString().substring(0, end);
return new Range(before.length() - start, end, word);
}
return null;
}
private static boolean isSeparator(int code, String sep) {
return sep.indexOf(code) != -1;
}
private static final Pattern spaceRegex = Pattern.compile("\\s+");
public static CharSequence getPreviousWord(InputConnection connection,
String sentenceSeperators) {
//TODO: Should fix this. This could be slow!
if (null == connection) return null;
CharSequence prev = connection.getTextBeforeCursor(LOOKBACK_CHARACTER_NUM, 0);
return getPreviousWord(prev, sentenceSeperators);
}
// Get the word before the whitespace preceding the non-whitespace preceding the cursor.
// Also, it won't return words that end in a separator.
// Example :
// "abc def|" -> abc
// "abc def |" -> abc
// "abc def. |" -> abc
// "abc def . |" -> def
// "abc|" -> null
// "abc |" -> null
// "abc. def|" -> null
public static CharSequence getPreviousWord(CharSequence prev, String sentenceSeperators) {
if (prev == null) return null;
String[] w = spaceRegex.split(prev);
// If we can't find two words, or we found an empty word, return null.
if (w.length < 2 || w[w.length - 2].length() <= 0) return null;
// If ends in a separator, return null
char lastChar = w[w.length - 2].charAt(w[w.length - 2].length() - 1);
if (sentenceSeperators.contains(String.valueOf(lastChar))) return null;
return w[w.length - 2];
}
public static CharSequence getThisWord(InputConnection connection, String sentenceSeperators) {
if (null == connection) return null;
final CharSequence prev = connection.getTextBeforeCursor(LOOKBACK_CHARACTER_NUM, 0);
return getThisWord(prev, sentenceSeperators);
}
// Get the word immediately before the cursor, even if there is whitespace between it and
// the cursor - but not if there is punctuation.
// Example :
// "abc def|" -> def
// "abc def |" -> def
// "abc def. |" -> null
// "abc def . |" -> null
public static CharSequence getThisWord(CharSequence prev, String sentenceSeperators) {
if (prev == null) return null;
String[] w = spaceRegex.split(prev);
// No word : return null
if (w.length < 1 || w[w.length - 1].length() <= 0) return null;
// If ends in a separator, return null
char lastChar = w[w.length - 1].charAt(w[w.length - 1].length() - 1);
if (sentenceSeperators.contains(String.valueOf(lastChar))) return null;
return w[w.length - 1];
}
}

21
java/src/com/android/inputmethod/latin/LatinIME.java
View File

@ -783,7 +783,7 @@ public class LatinIME extends InputMethodService implements KeyboardActionListen
ResearchLogger.latinIME_onUpdateSelection(mLastSelectionStart, mLastSelectionEnd,
oldSelStart, oldSelEnd, newSelStart, newSelEnd, composingSpanStart,
composingSpanEnd, mExpectingUpdateSelection,
expectingUpdateSelectionFromLogger, getCurrentInputConnection());
expectingUpdateSelectionFromLogger, mConnection);
if (expectingUpdateSelectionFromLogger) {
return;
}
@ -1760,11 +1760,7 @@ public class LatinIME extends InputMethodService implements KeyboardActionListen
}
// TODO: May need a better way of retrieving previous word
final CharSequence prevWord;
// TODO: move getPreviousWord to AutoInputConnection
prevWord = EditingUtils.getPreviousWord(mConnection.getInputConnection(),
mSettingsValues.mWordSeparators);
final CharSequence prevWord = mConnection.getPreviousWord(mSettingsValues.mWordSeparators);
final CharSequence typedWord = mWordComposer.getTypedWord();
// getSuggestedWords handles gracefully a null value of prevWord
final SuggestedWords suggestedWords = mSuggest.getSuggestedWords(mWordComposer,
@ -1993,8 +1989,7 @@ public class LatinIME extends InputMethodService implements KeyboardActionListen
final SuggestedWords suggestedWords;
if (mCorrectionMode == Suggest.CORRECTION_FULL_BIGRAM) {
final CharSequence prevWord = EditingUtils.getThisWord(getCurrentInputConnection(),
mSettingsValues.mWordSeparators);
final CharSequence prevWord = mConnection.getThisWord(mSettingsValues.mWordSeparators);
if (!TextUtils.isEmpty(prevWord)) {
suggestedWords = mSuggest.getBigramPredictions(prevWord);
} else {
@ -2031,10 +2026,8 @@ public class LatinIME extends InputMethodService implements KeyboardActionListen
}
if (mUserHistoryDictionary != null) {
final CharSequence prevWord;
// TODO: move getPreviousWord to AutoInputConnection
prevWord = EditingUtils.getPreviousWord(mConnection.getInputConnection(),
mSettingsValues.mWordSeparators);
final CharSequence prevWord
= mConnection.getPreviousWord(mSettingsValues.mWordSeparators);
final String secondWord;
if (mWordComposer.isAutoCapitalized() && !mWordComposer.isMostlyCaps()) {
secondWord = suggestion.toString().toLowerCase(
@ -2093,9 +2086,7 @@ public class LatinIME extends InputMethodService implements KeyboardActionListen
// Bail out if word before cursor is 0-length or a single non letter (like an apostrophe)
// Example: " -|" gets rejected here but "e-|" and "e|" are okay
// TODO: move getWordAtCursor inside AutoInputConnection
CharSequence word = EditingUtils.getWordAtCursor(mConnection.getInputConnection(),
mSettingsValues.mWordSeparators);
CharSequence word = mConnection.getWordAtCursor(mSettingsValues.mWordSeparators);
// We don't suggest on leading single quotes, so we have to remove them from the word if
// it starts with single quotes.
while (!TextUtils.isEmpty(word) && Keyboard.CODE_SINGLE_QUOTE == word.charAt(0)) {

5
java/src/com/android/inputmethod/latin/ResearchLogger.java
View File

@ -612,12 +612,13 @@ public class ResearchLogger implements SharedPreferences.OnSharedPreferenceChang
final int lastSelectionEnd, final int oldSelStart, final int oldSelEnd,
final int newSelStart, final int newSelEnd, final int composingSpanStart,
final int composingSpanEnd, final boolean expectingUpdateSelection,
final boolean expectingUpdateSelectionFromLogger, final InputConnection connection) {
final boolean expectingUpdateSelectionFromLogger,
final RichInputConnection connection) {
final Object[] values = {
lastSelectionStart, lastSelectionEnd, oldSelStart, oldSelEnd, newSelStart,
newSelEnd, composingSpanStart, composingSpanEnd, expectingUpdateSelection,
expectingUpdateSelectionFromLogger,
EditingUtils.getWordRangeAtCursor(connection, WHITESPACE_SEPARATORS, 1).mWord
connection.getWordRangeAtCursor(WHITESPACE_SEPARATORS, 1).mWord
};
getInstance().writeEvent(EVENTKEYS_LATINIME_ONUPDATESELECTION, values);
}

188
java/src/com/android/inputmethod/latin/RichInputConnection.java
View File

@ -20,14 +20,23 @@ import android.util.Log;
import android.view.KeyEvent;
import android.view.inputmethod.CompletionInfo;
import android.view.inputmethod.CorrectionInfo;
import android.view.inputmethod.ExtractedText;
import android.view.inputmethod.ExtractedTextRequest;
import android.view.inputmethod.InputConnection;
import java.util.regex.Pattern;
/**
* Wrapper for InputConnection to simplify interaction
*/
public class RichInputConnection {
private static final String TAG = RichInputConnection.class.getSimpleName();
private static final boolean DBG = false;
// Provision for a long word pair and a separator
private static final int LOOKBACK_CHARACTER_NUM = BinaryDictionary.MAX_WORD_LENGTH * 2 + 1;
private static final Pattern spaceRegex = Pattern.compile("\\s+");
private static final int INVALID_CURSOR_POSITION = -1;
InputConnection mIC;
int mNestLevel;
public RichInputConnection() {
@ -35,11 +44,6 @@ public class RichInputConnection {
mNestLevel = 0;
}
// TODO: remove this method - the whole point of this class is void if mIC is escaping
public InputConnection getInputConnection() {
return mIC;
}
public void beginBatchEdit(final InputConnection newInputConnection) {
if (++mNestLevel == 1) {
mIC = newInputConnection;
@ -119,4 +123,178 @@ public class RichInputConnection {
if (mNestLevel <= 0) Log.e(TAG, "Batch edit not in progress!"); // TODO: exception instead
if (null != mIC) mIC.commitCompletion(completionInfo);
}
public CharSequence getPreviousWord(final String sentenceSeperators) {
//TODO: Should fix this. This could be slow!
if (null == mIC) return null;
CharSequence prev = mIC.getTextBeforeCursor(LOOKBACK_CHARACTER_NUM, 0);
return getPreviousWord(prev, sentenceSeperators);
}
/**
* Represents a range of text, relative to the current cursor position.
*/
public static class Range {
/** Characters before selection start */
public final int mCharsBefore;
/**
* Characters after selection start, including one trailing word
* separator.
*/
public final int mCharsAfter;
/** The actual characters that make up a word */
public final String mWord;
public Range(int charsBefore, int charsAfter, String word) {
if (charsBefore < 0 || charsAfter < 0) {
throw new IndexOutOfBoundsException();
}
this.mCharsBefore = charsBefore;
this.mCharsAfter = charsAfter;
this.mWord = word;
}
}
private static boolean isSeparator(int code, String sep) {
return sep.indexOf(code) != -1;
}
// Get the word before the whitespace preceding the non-whitespace preceding the cursor.
// Also, it won't return words that end in a separator.
// Example :
// "abc def|" -> abc
// "abc def |" -> abc
// "abc def. |" -> abc
// "abc def . |" -> def
// "abc|" -> null
// "abc |" -> null
// "abc. def|" -> null
public static CharSequence getPreviousWord(CharSequence prev, String sentenceSeperators) {
if (prev == null) return null;
String[] w = spaceRegex.split(prev);
// If we can't find two words, or we found an empty word, return null.
if (w.length < 2 || w[w.length - 2].length() <= 0) return null;
// If ends in a separator, return null
char lastChar = w[w.length - 2].charAt(w[w.length - 2].length() - 1);
if (sentenceSeperators.contains(String.valueOf(lastChar))) return null;
return w[w.length - 2];
}
public CharSequence getThisWord(String sentenceSeperators) {
if (null == mIC) return null;
final CharSequence prev = mIC.getTextBeforeCursor(LOOKBACK_CHARACTER_NUM, 0);
return getThisWord(prev, sentenceSeperators);
}
// Get the word immediately before the cursor, even if there is whitespace between it and
// the cursor - but not if there is punctuation.
// Example :
// "abc def|" -> def
// "abc def |" -> def
// "abc def. |" -> null
// "abc def . |" -> null
public static CharSequence getThisWord(CharSequence prev, String sentenceSeperators) {
if (prev == null) return null;
String[] w = spaceRegex.split(prev);
// No word : return null
if (w.length < 1 || w[w.length - 1].length() <= 0) return null;
// If ends in a separator, return null
char lastChar = w[w.length - 1].charAt(w[w.length - 1].length() - 1);
if (sentenceSeperators.contains(String.valueOf(lastChar))) return null;
return w[w.length - 1];
}
/**
* @param separators characters which may separate words
* @return the word that surrounds the cursor, including up to one trailing
* separator. For example, if the field contains "he|llo world", where |
* represents the cursor, then "hello " will be returned.
*/
public String getWordAtCursor(String separators) {
// getWordRangeAtCursor returns null if the connection is null
Range r = getWordRangeAtCursor(separators, 0);
return (r == null) ? null : r.mWord;
}
private int getCursorPosition() {
if (null == mIC) return INVALID_CURSOR_POSITION;
final ExtractedText extracted = mIC.getExtractedText(new ExtractedTextRequest(), 0);
if (extracted == null) {
return INVALID_CURSOR_POSITION;
}
return extracted.startOffset + extracted.selectionStart;
}
/**
* Returns the text surrounding the cursor.
*
* @param sep a string of characters that split words.
* @param additionalPrecedingWordsCount the number of words before the current word that should
* be included in the returned range
* @return a range containing the text surrounding the cursor
*/
public Range getWordRangeAtCursor(String sep, int additionalPrecedingWordsCount) {
if (mIC == null || sep == null) {
return null;
}
CharSequence before = mIC.getTextBeforeCursor(1000, 0);
CharSequence after = mIC.getTextAfterCursor(1000, 0);
if (before == null || after == null) {
return null;
}
// Going backward, alternate skipping non-separators and separators until enough words
// have been read.
int start = before.length();
boolean isStoppingAtWhitespace = true; // toggles to indicate what to stop at
while (true) { // see comments below for why this is guaranteed to halt
while (start > 0) {
final int codePoint = Character.codePointBefore(before, start);
if (isStoppingAtWhitespace == isSeparator(codePoint, sep)) {
break; // inner loop
}
--start;
if (Character.isSupplementaryCodePoint(codePoint)) {
--start;
}
}
// isStoppingAtWhitespace is true every other time through the loop,
// so additionalPrecedingWordsCount is guaranteed to become < 0, which
// guarantees outer loop termination
if (isStoppingAtWhitespace && (--additionalPrecedingWordsCount < 0)) {
break; // outer loop
}
isStoppingAtWhitespace = !isStoppingAtWhitespace;
}
// Find last word separator after the cursor
int end = -1;
while (++end < after.length()) {
final int codePoint = Character.codePointAt(after, end);
if (isSeparator(codePoint, sep)) {
break;
}
if (Character.isSupplementaryCodePoint(codePoint)) {
++end;
}
}
int cursor = getCursorPosition();
if (start >= 0 && cursor + end <= after.length() + before.length()) {
String word = before.toString().substring(start, before.length())
+ after.toString().substring(0, end);
return new Range(before.length() - start, end, word);
}
return null;
}
}

77
tests/src/com/android/inputmethod/latin/EditingUtilsTests.java → tests/src/com/android/inputmethod/latin/RichInputConnectionTests.java
View File

@ -22,9 +22,9 @@ import android.view.inputmethod.ExtractedTextRequest;
import android.view.inputmethod.InputConnection;
import android.view.inputmethod.InputConnectionWrapper;
import com.android.inputmethod.latin.EditingUtils.Range;
import com.android.inputmethod.latin.RichInputConnection.Range;
public class EditingUtilsTests extends AndroidTestCase {
public class RichInputConnectionTests extends AndroidTestCase {
// The following is meant to be a reasonable default for
// the "word_separators" resource.
@ -64,12 +64,23 @@ public class EditingUtilsTests extends AndroidTestCase {
}
/* (non-Javadoc)
* @see android.view.inputmethod.InputConnectionWrapper#getExtractedText(ExtractedTextRequest, int)
* @see android.view.inputmethod.InputConnectionWrapper#getExtractedText(
* ExtractedTextRequest, int)
*/
@Override
public ExtractedText getExtractedText(ExtractedTextRequest request, int flags) {
return mExtractedText;
}
@Override
public boolean beginBatchEdit() {
return true;
}
@Override
public boolean endBatchEdit() {
return true;
}
}
/************************** Tests ************************/
@ -79,31 +90,31 @@ public class EditingUtilsTests extends AndroidTestCase {
*/
public void testGetPreviousWord() {
// If one of the following cases breaks, the bigram suggestions won't work.
assertEquals(EditingUtils.getPreviousWord("abc def", sSeparators), "abc");
assertNull(EditingUtils.getPreviousWord("abc", sSeparators));
assertNull(EditingUtils.getPreviousWord("abc. def", sSeparators));
assertEquals(RichInputConnection.getPreviousWord("abc def", sSeparators), "abc");
assertNull(RichInputConnection.getPreviousWord("abc", sSeparators));
assertNull(RichInputConnection.getPreviousWord("abc. def", sSeparators));
// The following tests reflect the current behavior of the function
// EditingUtils#getPreviousWord.
// RichInputConnection#getPreviousWord.
// TODO: However at this time, the code does never go
// into such a path, so it should be safe to change the behavior of
// this function if needed - especially since it does not seem very
// logical. These tests are just there to catch any unintentional
// changes in the behavior of the EditingUtils#getPreviousWord method.
assertEquals(EditingUtils.getPreviousWord("abc def ", sSeparators), "abc");
assertEquals(EditingUtils.getPreviousWord("abc def.", sSeparators), "abc");
assertEquals(EditingUtils.getPreviousWord("abc def .", sSeparators), "def");
assertNull(EditingUtils.getPreviousWord("abc ", sSeparators));
// changes in the behavior of the RichInputConnection#getPreviousWord method.
assertEquals(RichInputConnection.getPreviousWord("abc def ", sSeparators), "abc");
assertEquals(RichInputConnection.getPreviousWord("abc def.", sSeparators), "abc");
assertEquals(RichInputConnection.getPreviousWord("abc def .", sSeparators), "def");
assertNull(RichInputConnection.getPreviousWord("abc ", sSeparators));
}
/**
* Test for getting the word before the cursor (for bigram)
*/
public void testGetThisWord() {
assertEquals(EditingUtils.getThisWord("abc def", sSeparators), "def");
assertEquals(EditingUtils.getThisWord("abc def ", sSeparators), "def");
assertNull(EditingUtils.getThisWord("abc def.", sSeparators));
assertNull(EditingUtils.getThisWord("abc def .", sSeparators));
assertEquals(RichInputConnection.getThisWord("abc def", sSeparators), "def");
assertEquals(RichInputConnection.getThisWord("abc def ", sSeparators), "def");
assertNull(RichInputConnection.getThisWord("abc def.", sSeparators));
assertNull(RichInputConnection.getThisWord("abc def .", sSeparators));
}
/**
@ -111,51 +122,57 @@ public class EditingUtilsTests extends AndroidTestCase {
*/
public void testGetWordRangeAtCursor() {
ExtractedText et = new ExtractedText();
final RichInputConnection ic = new RichInputConnection();
InputConnection mockConnection;
mockConnection = new MockConnection("word wo", "rd", et);
et.startOffset = 0;
et.selectionStart = 7;
Range r;
ic.beginBatchEdit(mockConnection);
// basic case
r = EditingUtils.getWordRangeAtCursor(mockConnection, " ", 0);
r = ic.getWordRangeAtCursor(" ", 0);
assertEquals("word", r.mWord);
r = null;
// more than one word
r = EditingUtils.getWordRangeAtCursor(mockConnection, " ", 1);
r = ic.getWordRangeAtCursor(" ", 1);
assertEquals("word word", r.mWord);
r = null;
ic.endBatchEdit();
// tab character instead of space
mockConnection = new MockConnection("one\tword\two", "rd", et);
r = EditingUtils.getWordRangeAtCursor(mockConnection, "\t", 1);
ic.beginBatchEdit(mockConnection);
r = ic.getWordRangeAtCursor("\t", 1);
ic.endBatchEdit();
assertEquals("word\tword", r.mWord);
r = null;
// only one word doesn't go too far
mockConnection = new MockConnection("one\tword\two", "rd", et);
r = EditingUtils.getWordRangeAtCursor(mockConnection, "\t", 1);
ic.beginBatchEdit(mockConnection);
r = ic.getWordRangeAtCursor("\t", 1);
ic.endBatchEdit();
assertEquals("word\tword", r.mWord);
r = null;
// tab or space
mockConnection = new MockConnection("one word\two", "rd", et);
r = EditingUtils.getWordRangeAtCursor(mockConnection, " \t", 1);
ic.beginBatchEdit(mockConnection);
r = ic.getWordRangeAtCursor(" \t", 1);
ic.endBatchEdit();
assertEquals("word\tword", r.mWord);
r = null;
// tab or space multiword
mockConnection = new MockConnection("one word\two", "rd", et);
r = EditingUtils.getWordRangeAtCursor(mockConnection, " \t", 2);
ic.beginBatchEdit(mockConnection);
r = ic.getWordRangeAtCursor(" \t", 2);
ic.endBatchEdit();
assertEquals("one word\tword", r.mWord);
r = null;
// splitting on supplementary character
final String supplementaryChar = "\uD840\uDC8A";
mockConnection = new MockConnection("one word" + supplementaryChar + "wo", "rd", et);
r = EditingUtils.getWordRangeAtCursor(mockConnection, supplementaryChar, 0);
ic.beginBatchEdit(mockConnection);
r = ic.getWordRangeAtCursor(supplementaryChar, 0);
ic.endBatchEdit();
assertEquals("word", r.mWord);
r = null;
}
}