diff --git a/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerService.java b/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerService.java index 8d495646d..d0316242b 100644 --- a/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerService.java +++ b/java/src/com/android/inputmethod/latin/spellcheck/AndroidSpellCheckerService.java @@ -37,6 +37,7 @@ import com.android.inputmethod.latin.utils.AdditionalSubtypeUtils; import com.android.inputmethod.latin.utils.BinaryDictionaryUtils; import com.android.inputmethod.latin.utils.CollectionUtils; import com.android.inputmethod.latin.utils.LocaleUtils; +import com.android.inputmethod.latin.utils.ScriptUtils; import com.android.inputmethod.latin.utils.StringUtils; import java.lang.ref.WeakReference; @@ -78,40 +79,8 @@ public final class AndroidSpellCheckerService extends SpellCheckerService private final HashSet> mDictionaryCollectionsList = new HashSet<>(); - public static final int SCRIPT_LATIN = 0; - public static final int SCRIPT_CYRILLIC = 1; - public static final int SCRIPT_GREEK = 2; public static final String SINGLE_QUOTE = "\u0027"; public static final String APOSTROPHE = "\u2019"; - private static final TreeMap mLanguageToScript; - static { - // List of the supported languages and their associated script. We won't check - // words written in another script than the selected script, because we know we - // don't have those in our dictionary so we will underline everything and we - // will never have any suggestions, so it makes no sense checking them, and this - // is done in {@link #shouldFilterOut}. Also, the script is used to choose which - // proximity to pass to the dictionary descent algorithm. - // IMPORTANT: this only contains languages - do not write countries in there. - // Only the language is searched from the map. - mLanguageToScript = new TreeMap<>(); - mLanguageToScript.put("cs", SCRIPT_LATIN); - mLanguageToScript.put("da", SCRIPT_LATIN); - mLanguageToScript.put("de", SCRIPT_LATIN); - mLanguageToScript.put("el", SCRIPT_GREEK); - mLanguageToScript.put("en", SCRIPT_LATIN); - mLanguageToScript.put("es", SCRIPT_LATIN); - mLanguageToScript.put("fi", SCRIPT_LATIN); - mLanguageToScript.put("fr", SCRIPT_LATIN); - mLanguageToScript.put("hr", SCRIPT_LATIN); - mLanguageToScript.put("it", SCRIPT_LATIN); - mLanguageToScript.put("lt", SCRIPT_LATIN); - mLanguageToScript.put("lv", SCRIPT_LATIN); - mLanguageToScript.put("nb", SCRIPT_LATIN); - mLanguageToScript.put("nl", SCRIPT_LATIN); - mLanguageToScript.put("pt", SCRIPT_LATIN); - mLanguageToScript.put("sl", SCRIPT_LATIN); - mLanguageToScript.put("ru", SCRIPT_CYRILLIC); - } @Override public void onCreate() { super.onCreate(); @@ -122,22 +91,13 @@ public final class AndroidSpellCheckerService extends SpellCheckerService onSharedPreferenceChanged(prefs, PREF_USE_CONTACTS_KEY); } - public static int getScriptFromLocale(final Locale locale) { - final Integer script = mLanguageToScript.get(locale.getLanguage()); - if (null == script) { - throw new RuntimeException("We have been called with an unsupported language: \"" - + locale.getLanguage() + "\". Framework bug?"); - } - return script; - } - private static String getKeyboardLayoutNameForScript(final int script) { switch (script) { - case AndroidSpellCheckerService.SCRIPT_LATIN: + case ScriptUtils.SCRIPT_LATIN: return "qwerty"; - case AndroidSpellCheckerService.SCRIPT_CYRILLIC: + case ScriptUtils.SCRIPT_CYRILLIC: return "east_slavic"; - case AndroidSpellCheckerService.SCRIPT_GREEK: + case ScriptUtils.SCRIPT_GREEK: return "greek"; default: throw new RuntimeException("Wrong script supplied: " + script); @@ -413,7 +373,7 @@ public final class AndroidSpellCheckerService extends SpellCheckerService } public DictAndKeyboard createDictAndKeyboard(final Locale locale) { - final int script = getScriptFromLocale(locale); + final int script = ScriptUtils.getScriptFromLocale(locale); final String keyboardLayoutName = getKeyboardLayoutNameForScript(script); final InputMethodSubtype subtype = AdditionalSubtypeUtils.createAdditionalSubtype( locale.toString(), keyboardLayoutName, null); diff --git a/java/src/com/android/inputmethod/latin/spellcheck/AndroidWordLevelSpellCheckerSession.java b/java/src/com/android/inputmethod/latin/spellcheck/AndroidWordLevelSpellCheckerSession.java index e9d5a7a95..db15c1cc6 100644 --- a/java/src/com/android/inputmethod/latin/spellcheck/AndroidWordLevelSpellCheckerSession.java +++ b/java/src/com/android/inputmethod/latin/spellcheck/AndroidWordLevelSpellCheckerSession.java @@ -36,6 +36,7 @@ import com.android.inputmethod.latin.WordComposer; import com.android.inputmethod.latin.spellcheck.AndroidSpellCheckerService.SuggestionsGatherer; import com.android.inputmethod.latin.utils.CoordinateUtils; import com.android.inputmethod.latin.utils.LocaleUtils; +import com.android.inputmethod.latin.utils.ScriptUtils; import com.android.inputmethod.latin.utils.StringUtils; import java.util.ArrayList; @@ -116,7 +117,7 @@ public abstract class AndroidWordLevelSpellCheckerSession extends Session { final String localeString = getLocale(); mDictionaryPool = mService.getDictionaryPool(localeString); mLocale = LocaleUtils.constructLocaleFromString(localeString); - mScript = AndroidSpellCheckerService.getScriptFromLocale(mLocale); + mScript = ScriptUtils.getScriptFromLocale(mLocale); } @Override @@ -125,44 +126,6 @@ public abstract class AndroidWordLevelSpellCheckerSession extends Session { cres.unregisterContentObserver(mObserver); } - /* - * Returns whether the code point is a letter that makes sense for the specified - * locale for this spell checker. - * The dictionaries supported by Latin IME are described in res/xml/spellchecker.xml - * and is limited to EFIGS languages and Russian. - * Hence at the moment this explicitly tests for Cyrillic characters or Latin characters - * as appropriate, and explicitly excludes CJK, Arabic and Hebrew characters. - */ - private static boolean isLetterCheckableByLanguage(final int codePoint, - final int script) { - switch (script) { - case AndroidSpellCheckerService.SCRIPT_LATIN: - // Our supported latin script dictionaries (EFIGS) at the moment only include - // characters in the C0, C1, Latin Extended A and B, IPA extensions unicode - // blocks. As it happens, those are back-to-back in the code range 0x40 to 0x2AF, - // so the below is a very efficient way to test for it. As for the 0-0x3F, it's - // excluded from isLetter anyway. - return codePoint <= 0x2AF && Character.isLetter(codePoint); - case AndroidSpellCheckerService.SCRIPT_CYRILLIC: - // All Cyrillic characters are in the 400~52F block. There are some in the upper - // Unicode range, but they are archaic characters that are not used in modern - // Russian and are not used by our dictionary. - return codePoint >= 0x400 && codePoint <= 0x52F && Character.isLetter(codePoint); - case AndroidSpellCheckerService.SCRIPT_GREEK: - // Greek letters are either in the 370~3FF range (Greek & Coptic), or in the - // 1F00~1FFF range (Greek extended). Our dictionary contains both sort of characters. - // Our dictionary also contains a few words with 0xF2; it would be best to check - // if that's correct, but a web search does return results for these words so - // they are probably okay. - return (codePoint >= 0x370 && codePoint <= 0x3FF) - || (codePoint >= 0x1F00 && codePoint <= 0x1FFF) - || codePoint == 0xF2; - default: - // Should never come here - throw new RuntimeException("Impossible value of script: " + script); - } - } - private static final int CHECKABILITY_CHECKABLE = 0; private static final int CHECKABILITY_TOO_MANY_NON_LETTERS = 1; private static final int CHECKABILITY_CONTAINS_PERIOD = 2; @@ -189,7 +152,7 @@ public abstract class AndroidWordLevelSpellCheckerSession extends Session { // Filter by first letter final int firstCodePoint = text.codePointAt(0); // Filter out words that don't start with a letter or an apostrophe - if (!isLetterCheckableByLanguage(firstCodePoint, script) + if (!ScriptUtils.isLetterCheckableByScript(firstCodePoint, script) && '\'' != firstCodePoint) return CHECKABILITY_FIRST_LETTER_UNCHECKABLE; // Filter contents @@ -210,7 +173,7 @@ public abstract class AndroidWordLevelSpellCheckerSession extends Session { if (Constants.CODE_PERIOD == codePoint) { return CHECKABILITY_CONTAINS_PERIOD; } - if (isLetterCheckableByLanguage(codePoint, script)) ++letterCount; + if (ScriptUtils.isLetterCheckableByScript(codePoint, script)) ++letterCount; } // Guestimate heuristic: perform spell checking if at least 3/4 of the characters // in this word are letters diff --git a/java/src/com/android/inputmethod/latin/utils/ScriptUtils.java b/java/src/com/android/inputmethod/latin/utils/ScriptUtils.java new file mode 100644 index 000000000..4dfb38d80 --- /dev/null +++ b/java/src/com/android/inputmethod/latin/utils/ScriptUtils.java @@ -0,0 +1,103 @@ +/* + * Copyright (C) 2012 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.utils; + +import java.util.Locale; +import java.util.TreeMap; + +/** + * A class to help with handling different writing scripts. + */ +public class ScriptUtils { + public static final int SCRIPT_LATIN = 0; + public static final int SCRIPT_CYRILLIC = 1; + public static final int SCRIPT_GREEK = 2; + public static final TreeMap mLanguageToScript; + static { + // List of the supported languages and their associated script. We won't check + // words written in another script than the selected script, because we know we + // don't have those in our dictionary so we will underline everything and we + // will never have any suggestions, so it makes no sense checking them, and this + // is done in {@link #shouldFilterOut}. Also, the script is used to choose which + // proximity to pass to the dictionary descent algorithm. + // IMPORTANT: this only contains languages - do not write countries in there. + // Only the language is searched from the map. + mLanguageToScript = new TreeMap<>(); + mLanguageToScript.put("cs", SCRIPT_LATIN); + mLanguageToScript.put("da", SCRIPT_LATIN); + mLanguageToScript.put("de", SCRIPT_LATIN); + mLanguageToScript.put("el", SCRIPT_GREEK); + mLanguageToScript.put("en", SCRIPT_LATIN); + mLanguageToScript.put("es", SCRIPT_LATIN); + mLanguageToScript.put("fi", SCRIPT_LATIN); + mLanguageToScript.put("fr", SCRIPT_LATIN); + mLanguageToScript.put("hr", SCRIPT_LATIN); + mLanguageToScript.put("it", SCRIPT_LATIN); + mLanguageToScript.put("lt", SCRIPT_LATIN); + mLanguageToScript.put("lv", SCRIPT_LATIN); + mLanguageToScript.put("nb", SCRIPT_LATIN); + mLanguageToScript.put("nl", SCRIPT_LATIN); + mLanguageToScript.put("pt", SCRIPT_LATIN); + mLanguageToScript.put("sl", SCRIPT_LATIN); + mLanguageToScript.put("ru", SCRIPT_CYRILLIC); + } + /* + * Returns whether the code point is a letter that makes sense for the specified + * locale for this spell checker. + * The dictionaries supported by Latin IME are described in res/xml/spellchecker.xml + * and is limited to EFIGS languages and Russian. + * Hence at the moment this explicitly tests for Cyrillic characters or Latin characters + * as appropriate, and explicitly excludes CJK, Arabic and Hebrew characters. + */ + public static boolean isLetterCheckableByScript(final int codePoint, final int script) { + switch (script) { + case SCRIPT_LATIN: + // Our supported latin script dictionaries (EFIGS) at the moment only include + // characters in the C0, C1, Latin Extended A and B, IPA extensions unicode + // blocks. As it happens, those are back-to-back in the code range 0x40 to 0x2AF, + // so the below is a very efficient way to test for it. As for the 0-0x3F, it's + // excluded from isLetter anyway. + return codePoint <= 0x2AF && Character.isLetter(codePoint); + case SCRIPT_CYRILLIC: + // All Cyrillic characters are in the 400~52F block. There are some in the upper + // Unicode range, but they are archaic characters that are not used in modern + // Russian and are not used by our dictionary. + return codePoint >= 0x400 && codePoint <= 0x52F && Character.isLetter(codePoint); + case SCRIPT_GREEK: + // Greek letters are either in the 370~3FF range (Greek & Coptic), or in the + // 1F00~1FFF range (Greek extended). Our dictionary contains both sort of characters. + // Our dictionary also contains a few words with 0xF2; it would be best to check + // if that's correct, but a web search does return results for these words so + // they are probably okay. + return (codePoint >= 0x370 && codePoint <= 0x3FF) + || (codePoint >= 0x1F00 && codePoint <= 0x1FFF) + || codePoint == 0xF2; + default: + // Should never come here + throw new RuntimeException("Impossible value of script: " + script); + } + } + + public static int getScriptFromLocale(final Locale locale) { + final Integer script = mLanguageToScript.get(locale.getLanguage()); + if (null == script) { + throw new RuntimeException("We have been called with an unsupported language: \"" + + locale.getLanguage() + "\". Framework bug?"); + } + return script; + } +}