From 86f36003fd4397143bd37938dda029e5707634af Mon Sep 17 00:00:00 2001 From: Yohei Yukawa Date: Sun, 20 Jul 2014 00:27:27 +0900 Subject: [PATCH] Use CharSequence for spell checker to keep spans preserved This is a ground work to take per word locale information into consideration in the spell checker. This CL is supposed to change no user visible behavior. With this CL, the spell checker session is able to read span information if necessary. BUG: 16029304 Change-Id: Icb1e1ecdf40fe0445e14565b685b1b878b746210 --- .../inputmethod/latin/PrevWordsInfo.java | 12 +++--- .../UserHistoryDictionary.java | 5 ++- .../AndroidSpellCheckerSession.java | 43 +++++++++++-------- .../spellcheck/SentenceLevelAdapter.java | 9 ++-- .../latin/utils/LanguageModelParam.java | 6 +-- 5 files changed, 43 insertions(+), 32 deletions(-) diff --git a/java/src/com/android/inputmethod/latin/PrevWordsInfo.java b/java/src/com/android/inputmethod/latin/PrevWordsInfo.java index f45c73f53..db877ab7a 100644 --- a/java/src/com/android/inputmethod/latin/PrevWordsInfo.java +++ b/java/src/com/android/inputmethod/latin/PrevWordsInfo.java @@ -16,10 +16,12 @@ package com.android.inputmethod.latin; -import java.util.Arrays; +import android.text.TextUtils; import com.android.inputmethod.latin.utils.StringUtils; +import java.util.Arrays; + /** * Class to represent information of previous words. This class is used to add n-gram entries * into binary dictionaries, to get predictions, and to get suggestions. @@ -37,8 +39,8 @@ public class PrevWordsInfo { public static final WordInfo EMPTY_WORD_INFO = new WordInfo(null); public static final WordInfo BEGINNING_OF_SENTENCE = new WordInfo(); - // This is an empty string when mIsBeginningOfSentence is true. - public final String mWord; + // This is an empty char sequence when mIsBeginningOfSentence is true. + public final CharSequence mWord; // TODO: Have sentence separator. // Whether the current context is beginning of sentence or not. This is true when composing // at the beginning of an input field or composing a word after a sentence separator. @@ -50,7 +52,7 @@ public class PrevWordsInfo { mIsBeginningOfSentence = true; } - public WordInfo(final String word) { + public WordInfo(final CharSequence word) { mWord = word; mIsBeginningOfSentence = false; } @@ -73,7 +75,7 @@ public class PrevWordsInfo { return mWord == wordInfo.mWord && mIsBeginningOfSentence == wordInfo.mIsBeginningOfSentence; } - return mWord.equals(wordInfo.mWord) + return TextUtils.equals(mWord, wordInfo.mWord) && mIsBeginningOfSentence == wordInfo.mIsBeginningOfSentence; } } diff --git a/java/src/com/android/inputmethod/latin/personalization/UserHistoryDictionary.java b/java/src/com/android/inputmethod/latin/personalization/UserHistoryDictionary.java index a98b0f156..8e027e4f9 100644 --- a/java/src/com/android/inputmethod/latin/personalization/UserHistoryDictionary.java +++ b/java/src/com/android/inputmethod/latin/personalization/UserHistoryDictionary.java @@ -17,6 +17,7 @@ package com.android.inputmethod.latin.personalization; import android.content.Context; +import android.text.TextUtils; import com.android.inputmethod.annotations.UsedForTesting; import com.android.inputmethod.latin.Constants; @@ -60,7 +61,7 @@ public class UserHistoryDictionary extends DecayingExpandableBinaryDictionaryBas public static void addToDictionary(final ExpandableBinaryDictionary userHistoryDictionary, final PrevWordsInfo prevWordsInfo, final String word, final boolean isValid, final int timestamp, final DistracterFilter distracterFilter) { - final String prevWord = prevWordsInfo.mPrevWordsInfo[0].mWord; + final CharSequence prevWord = prevWordsInfo.mPrevWordsInfo[0].mWord; if (word.length() >= Constants.DICTIONARY_MAX_WORD_LENGTH || (prevWord != null && prevWord.length() >= Constants.DICTIONARY_MAX_WORD_LENGTH)) { return; @@ -71,7 +72,7 @@ public class UserHistoryDictionary extends DecayingExpandableBinaryDictionaryBas null /* shortcutTarget */, 0 /* shortcutFreq */, false /* isNotAWord */, false /* isBlacklisted */, timestamp, distracterFilter); // Do not insert a word as a bigram of itself - if (word.equals(prevWord)) { + if (TextUtils.equals(word, prevWord)) { return; } if (null != prevWord) { diff --git a/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerSession.java b/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerSession.java index 38d720664..14ab2dbbf 100644 --- a/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerSession.java +++ b/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerSession.java @@ -24,7 +24,9 @@ import android.view.textservice.SentenceSuggestionsInfo; import android.view.textservice.SuggestionsInfo; import android.view.textservice.TextInfo; +import com.android.inputmethod.compat.TextInfoCompatUtils; import com.android.inputmethod.latin.PrevWordsInfo; +import com.android.inputmethod.latin.utils.StringUtils; import java.util.ArrayList; import java.util.Locale; @@ -42,15 +44,15 @@ public final class AndroidSpellCheckerSession extends AndroidWordLevelSpellCheck private SentenceSuggestionsInfo fixWronglyInvalidatedWordWithSingleQuote(TextInfo ti, SentenceSuggestionsInfo ssi) { - final String typedText = ti.getText(); - if (!typedText.contains(AndroidSpellCheckerService.SINGLE_QUOTE)) { + final CharSequence typedText = TextInfoCompatUtils.getCharSequenceOrString(ti); + if (!typedText.toString().contains(AndroidSpellCheckerService.SINGLE_QUOTE)) { return null; } final int N = ssi.getSuggestionsCount(); final ArrayList additionalOffsets = new ArrayList<>(); final ArrayList additionalLengths = new ArrayList<>(); final ArrayList additionalSuggestionsInfos = new ArrayList<>(); - String currentWord = null; + CharSequence currentWord = null; for (int i = 0; i < N; ++i) { final SuggestionsInfo si = ssi.getSuggestionsInfoAt(i); final int flags = si.getSuggestionsAttributes(); @@ -59,32 +61,33 @@ public final class AndroidSpellCheckerSession extends AndroidWordLevelSpellCheck } final int offset = ssi.getOffsetAt(i); final int length = ssi.getLengthAt(i); - final String subText = typedText.substring(offset, offset + length); + final CharSequence subText = typedText.subSequence(offset, offset + length); final PrevWordsInfo prevWordsInfo = new PrevWordsInfo(new PrevWordsInfo.WordInfo(currentWord)); currentWord = subText; - if (!subText.contains(AndroidSpellCheckerService.SINGLE_QUOTE)) { + if (!subText.toString().contains(AndroidSpellCheckerService.SINGLE_QUOTE)) { continue; } - final String[] splitTexts = - subText.split(AndroidSpellCheckerService.SINGLE_QUOTE, -1); + final CharSequence[] splitTexts = StringUtils.split(subText, + AndroidSpellCheckerService.SINGLE_QUOTE, + true /* preserveTrailingEmptySegments */ ); if (splitTexts == null || splitTexts.length <= 1) { continue; } final int splitNum = splitTexts.length; for (int j = 0; j < splitNum; ++j) { - final String splitText = splitTexts[j]; + final CharSequence splitText = splitTexts[j]; if (TextUtils.isEmpty(splitText)) { continue; } - if (mSuggestionsCache.getSuggestionsFromCache(splitText, prevWordsInfo) == null) { + if (mSuggestionsCache.getSuggestionsFromCache(splitText.toString(), prevWordsInfo) + == null) { continue; } final int newLength = splitText.length(); // Neither RESULT_ATTR_IN_THE_DICTIONARY nor RESULT_ATTR_LOOKS_LIKE_TYPO final int newFlags = 0; - final SuggestionsInfo newSi = - new SuggestionsInfo(newFlags, EMPTY_STRING_ARRAY); + final SuggestionsInfo newSi = new SuggestionsInfo(newFlags, EMPTY_STRING_ARRAY); newSi.setCookieAndSequence(si.getCookie(), si.getSequence()); if (DBG) { Log.d(TAG, "Override and remove old span over: " + splitText + ", " @@ -194,20 +197,22 @@ public final class AndroidSpellCheckerSession extends AndroidWordLevelSpellCheck final int length = textInfos.length; final SuggestionsInfo[] retval = new SuggestionsInfo[length]; for (int i = 0; i < length; ++i) { - final String prevWord; + final CharSequence prevWord; if (sequentialWords && i > 0) { - final String prevWordCandidate = textInfos[i - 1].getText(); - // Note that an empty string would be used to indicate the initial word - // in the future. - prevWord = TextUtils.isEmpty(prevWordCandidate) ? null : prevWordCandidate; + final TextInfo prevTextInfo = textInfos[i - 1]; + final CharSequence prevWordCandidate = + TextInfoCompatUtils.getCharSequenceOrString(prevTextInfo); + // Note that an empty string would be used to indicate the initial word + // in the future. + prevWord = TextUtils.isEmpty(prevWordCandidate) ? null : prevWordCandidate; } else { prevWord = null; } final PrevWordsInfo prevWordsInfo = new PrevWordsInfo(new PrevWordsInfo.WordInfo(prevWord)); - retval[i] = onGetSuggestionsInternal(textInfos[i], prevWordsInfo, suggestionsLimit); - retval[i].setCookieAndSequence(textInfos[i].getCookie(), - textInfos[i].getSequence()); + final TextInfo textInfo = textInfos[i]; + retval[i] = onGetSuggestionsInternal(textInfo, prevWordsInfo, suggestionsLimit); + retval[i].setCookieAndSequence(textInfo.getCookie(), textInfo.getSequence()); } return retval; } finally { diff --git a/java/src/com/android/inputmethod/latin/spellcheck/SentenceLevelAdapter.java b/java/src/com/android/inputmethod/latin/spellcheck/SentenceLevelAdapter.java index 13352f39e..ae582ea25 100644 --- a/java/src/com/android/inputmethod/latin/spellcheck/SentenceLevelAdapter.java +++ b/java/src/com/android/inputmethod/latin/spellcheck/SentenceLevelAdapter.java @@ -21,6 +21,7 @@ import android.view.textservice.SentenceSuggestionsInfo; import android.view.textservice.SuggestionsInfo; import android.view.textservice.TextInfo; +import com.android.inputmethod.compat.TextInfoCompatUtils; import com.android.inputmethod.latin.Constants; import com.android.inputmethod.latin.settings.SpacingAndPunctuations; import com.android.inputmethod.latin.utils.RunInLocale; @@ -127,7 +128,8 @@ public class SentenceLevelAdapter { public SentenceTextInfoParams getSplitWords(TextInfo originalTextInfo) { final WordIterator wordIterator = mWordIterator; - final CharSequence originalText = originalTextInfo.getText(); + final CharSequence originalText = + TextInfoCompatUtils.getCharSequenceOrString(originalTextInfo); final int cookie = originalTextInfo.getCookie(); final int start = -1; final int end = originalText.length(); @@ -136,8 +138,9 @@ public class SentenceLevelAdapter { int wordEnd = wordIterator.getEndOfWord(originalText, wordStart); while (wordStart <= end && wordEnd != -1 && wordStart != -1) { if (wordEnd >= start && wordEnd > wordStart) { - final String query = originalText.subSequence(wordStart, wordEnd).toString(); - final TextInfo ti = new TextInfo(query, cookie, query.hashCode()); + CharSequence subSequence = originalText.subSequence(wordStart, wordEnd).toString(); + final TextInfo ti = TextInfoCompatUtils.newInstance(subSequence, 0, + subSequence.length(), cookie, subSequence.hashCode()); wordItems.add(new SentenceWordItem(ti, wordStart, wordEnd)); } wordStart = wordIterator.getBeginningOfNextWord(originalText, wordEnd); diff --git a/java/src/com/android/inputmethod/latin/utils/LanguageModelParam.java b/java/src/com/android/inputmethod/latin/utils/LanguageModelParam.java index 9ec19efa8..fbce3f2fd 100644 --- a/java/src/com/android/inputmethod/latin/utils/LanguageModelParam.java +++ b/java/src/com/android/inputmethod/latin/utils/LanguageModelParam.java @@ -43,7 +43,7 @@ public final class LanguageModelParam { private static final int BIGRAM_PROBABILITY_FOR_VALID_WORD = 10; private static final int BIGRAM_PROBABILITY_FOR_OOV_WORD = Dictionary.NOT_A_PROBABILITY; - public final String mTargetWord; + public final CharSequence mTargetWord; public final int[] mWord0; public final int[] mWord1; // TODO: this needs to be a list of shortcuts @@ -57,13 +57,13 @@ public final class LanguageModelParam { public final int mTimestamp; // Constructor for unigram. TODO: support shortcuts - public LanguageModelParam(final String word, final int unigramProbability, + public LanguageModelParam(final CharSequence word, final int unigramProbability, final int timestamp) { this(null /* word0 */, word, unigramProbability, Dictionary.NOT_A_PROBABILITY, timestamp); } // Constructor for unigram and bigram. - public LanguageModelParam(final String word0, final String word1, + public LanguageModelParam(final CharSequence word0, final CharSequence word1, final int unigramProbability, final int bigramProbability, final int timestamp) { mTargetWord = word1;