diff options
Diffstat (limited to 'common/src/org/kelar')
14 files changed, 2145 insertions, 0 deletions
diff --git a/common/src/org/kelar/inputmethod/annotations/ExternallyReferenced.java b/common/src/org/kelar/inputmethod/annotations/ExternallyReferenced.java new file mode 100644 index 000000000..b3dfc06d7 --- /dev/null +++ b/common/src/org/kelar/inputmethod/annotations/ExternallyReferenced.java @@ -0,0 +1,24 @@ +/* + * 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 org.kelar.inputmethod.annotations; + +/** + * Denotes that the class, method or field should not be eliminated by ProGuard, + * because it is externally referenced. (See proguard.flags) + */ +public @interface ExternallyReferenced { +} diff --git a/common/src/org/kelar/inputmethod/annotations/UsedForTesting.java b/common/src/org/kelar/inputmethod/annotations/UsedForTesting.java new file mode 100644 index 000000000..f117f7480 --- /dev/null +++ b/common/src/org/kelar/inputmethod/annotations/UsedForTesting.java @@ -0,0 +1,24 @@ +/* + * 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 org.kelar.inputmethod.annotations; + +/** + * Denotes that the class, method or field should not be eliminated by ProGuard, + * so that unit tests can access it. (See proguard.flags) + */ +public @interface UsedForTesting { +} diff --git a/common/src/org/kelar/inputmethod/latin/common/CodePointUtils.java b/common/src/org/kelar/inputmethod/latin/common/CodePointUtils.java new file mode 100644 index 000000000..9f6970526 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/CodePointUtils.java @@ -0,0 +1,117 @@ +/* + * Copyright (C) 2013 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 org.kelar.inputmethod.latin.common; + +import org.kelar.inputmethod.annotations.UsedForTesting; + +import java.util.Random; + +import javax.annotation.Nonnull; + +// Utility methods related with code points used for tests. +// TODO: Figure out where this class should be. +@UsedForTesting +public class CodePointUtils { + private CodePointUtils() { + // This utility class is not publicly instantiable. + } + + public static final int[] LATIN_ALPHABETS_LOWER = { + 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', + 'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z', + 0x00E0 /* LATIN SMALL LETTER A WITH GRAVE */, + 0x00E1 /* LATIN SMALL LETTER A WITH ACUTE */, + 0x00E2 /* LATIN SMALL LETTER A WITH CIRCUMFLEX */, + 0x00E3 /* LATIN SMALL LETTER A WITH TILDE */, + 0x00E4 /* LATIN SMALL LETTER A WITH DIAERESIS */, + 0x00E5 /* LATIN SMALL LETTER A WITH RING ABOVE */, + 0x00E6 /* LATIN SMALL LETTER AE */, + 0x00E7 /* LATIN SMALL LETTER C WITH CEDILLA */, + 0x00E8 /* LATIN SMALL LETTER E WITH GRAVE */, + 0x00E9 /* LATIN SMALL LETTER E WITH ACUTE */, + 0x00EA /* LATIN SMALL LETTER E WITH CIRCUMFLEX */, + 0x00EB /* LATIN SMALL LETTER E WITH DIAERESIS */, + 0x00EC /* LATIN SMALL LETTER I WITH GRAVE */, + 0x00ED /* LATIN SMALL LETTER I WITH ACUTE */, + 0x00EE /* LATIN SMALL LETTER I WITH CIRCUMFLEX */, + 0x00EF /* LATIN SMALL LETTER I WITH DIAERESIS */, + 0x00F0 /* LATIN SMALL LETTER ETH */, + 0x00F1 /* LATIN SMALL LETTER N WITH TILDE */, + 0x00F2 /* LATIN SMALL LETTER O WITH GRAVE */, + 0x00F3 /* LATIN SMALL LETTER O WITH ACUTE */, + 0x00F4 /* LATIN SMALL LETTER O WITH CIRCUMFLEX */, + 0x00F5 /* LATIN SMALL LETTER O WITH TILDE */, + 0x00F6 /* LATIN SMALL LETTER O WITH DIAERESIS */, + 0x00F7 /* LATIN SMALL LETTER O WITH STROKE */, + 0x00F9 /* LATIN SMALL LETTER U WITH GRAVE */, + 0x00FA /* LATIN SMALL LETTER U WITH ACUTE */, + 0x00FB /* LATIN SMALL LETTER U WITH CIRCUMFLEX */, + 0x00FC /* LATIN SMALL LETTER U WITH DIAERESIS */, + 0x00FD /* LATIN SMALL LETTER Y WITH ACUTE */, + 0x00FE /* LATIN SMALL LETTER THORN */, + 0x00FF /* LATIN SMALL LETTER Y WITH DIAERESIS */ + }; + + @UsedForTesting + @Nonnull + public static int[] generateCodePointSet(final int codePointSetSize, + @Nonnull final Random random) { + final int[] codePointSet = new int[codePointSetSize]; + for (int i = codePointSet.length - 1; i >= 0; ) { + final int r = Math.abs(random.nextInt()); + if (r < 0) { + continue; + } + // Don't insert 0~0x20, but insert any other code point. + // Code points are in the range 0~0x10FFFF. + final int candidateCodePoint = 0x20 + r % (Character.MAX_CODE_POINT - 0x20); + // Code points between MIN_ and MAX_SURROGATE are not valid on their own. + if (candidateCodePoint >= Character.MIN_SURROGATE + && candidateCodePoint <= Character.MAX_SURROGATE) { + continue; + } + codePointSet[i] = candidateCodePoint; + --i; + } + return codePointSet; + } + + /** + * Generates a random word. + */ + @UsedForTesting + @Nonnull + public static String generateWord(@Nonnull final Random random, + @Nonnull final int[] codePointSet) { + final StringBuilder builder = new StringBuilder(); + // 8 * 4 = 32 chars max, but we do it the following way so as to bias the random toward + // longer words. This should be closer to natural language, and more importantly, it will + // exercise the algorithms in dicttool much more. + final int count = 1 + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5) + + (Math.abs(random.nextInt()) % 5); + while (builder.length() < count) { + builder.appendCodePoint(codePointSet[Math.abs(random.nextInt()) % codePointSet.length]); + } + return builder.toString(); + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/CollectionUtils.java b/common/src/org/kelar/inputmethod/latin/common/CollectionUtils.java new file mode 100644 index 000000000..ead46575f --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/CollectionUtils.java @@ -0,0 +1,77 @@ +/* + * 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 org.kelar.inputmethod.latin.common; + +import org.kelar.inputmethod.annotations.UsedForTesting; + +import java.util.ArrayList; +import java.util.Collection; +import java.util.Map; + +import javax.annotation.Nonnull; +import javax.annotation.Nullable; + +/** + * Utility methods for working with collections. + */ +public final class CollectionUtils { + private CollectionUtils() { + // This utility class is not publicly instantiable. + } + + /** + * Converts a sub-range of the given array to an ArrayList of the appropriate type. + * @param array Array to be converted. + * @param start First index inclusive to be converted. + * @param end Last index exclusive to be converted. + * @throws IllegalArgumentException if start or end are out of range or start > end. + */ + @Nonnull + public static <E> ArrayList<E> arrayAsList(@Nonnull final E[] array, final int start, + final int end) { + if (start < 0 || start > end || end > array.length) { + throw new IllegalArgumentException("Invalid start: " + start + " end: " + end + + " with array.length: " + array.length); + } + + final ArrayList<E> list = new ArrayList<>(end - start); + for (int i = start; i < end; i++) { + list.add(array[i]); + } + return list; + } + + /** + * Tests whether c contains no elements, true if c is null or c is empty. + * @param c Collection to test. + * @return Whether c contains no elements. + */ + @UsedForTesting + public static boolean isNullOrEmpty(@Nullable final Collection c) { + return c == null || c.isEmpty(); + } + + /** + * Tests whether map contains no elements, true if map is null or map is empty. + * @param map Map to test. + * @return Whether map contains no elements. + */ + @UsedForTesting + public static boolean isNullOrEmpty(@Nullable final Map map) { + return map == null || map.isEmpty(); + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/ComposedData.java b/common/src/org/kelar/inputmethod/latin/common/ComposedData.java new file mode 100644 index 000000000..69309456f --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/ComposedData.java @@ -0,0 +1,66 @@ +/* + * Copyright (C) 2014 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 org.kelar.inputmethod.latin.common; + +import javax.annotation.Nonnull; + +/** + * An immutable class that encapsulates a snapshot of word composition data. + */ +public class ComposedData { + @Nonnull + public final InputPointers mInputPointers; + public final boolean mIsBatchMode; + @Nonnull + public final String mTypedWord; + + public ComposedData(@Nonnull final InputPointers inputPointers, final boolean isBatchMode, + @Nonnull final String typedWord) { + mInputPointers = inputPointers; + mIsBatchMode = isBatchMode; + mTypedWord = typedWord; + } + + /** + * Copy the code points in the typed word to a destination array of ints. + * + * If the array is too small to hold the code points in the typed word, nothing is copied and + * -1 is returned. + * + * @param destination the array of ints. + * @return the number of copied code points. + */ + public int copyCodePointsExceptTrailingSingleQuotesAndReturnCodePointCount( + @Nonnull final int[] destination) { + // lastIndex is exclusive + final int lastIndex = mTypedWord.length() + - StringUtils.getTrailingSingleQuotesCount(mTypedWord); + if (lastIndex <= 0) { + // The string is empty or contains only single quotes. + return 0; + } + + // The following function counts the number of code points in the text range which begins + // at index 0 and extends to the character at lastIndex. + final int codePointSize = Character.codePointCount(mTypedWord, 0, lastIndex); + if (codePointSize > destination.length) { + return -1; + } + return StringUtils.copyCodePointsAndReturnCodePointCount(destination, mTypedWord, 0, + lastIndex, true /* downCase */); + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/Constants.java b/common/src/org/kelar/inputmethod/latin/common/Constants.java new file mode 100644 index 000000000..9044d1590 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/Constants.java @@ -0,0 +1,339 @@ +/* + * 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 org.kelar.inputmethod.latin.common; + +import org.kelar.inputmethod.annotations.UsedForTesting; +import org.kelar.inputmethod.keyboard.KeyboardActionListener; +import org.kelar.inputmethod.keyboard.internal.BatchInputArbiter; +import org.kelar.inputmethod.keyboard.internal.KeyboardCodesSet; +import org.kelar.inputmethod.latin.settings.SettingsValues; + +import javax.annotation.Nonnull; + +public final class Constants { + + public static final class Color { + /** + * The alpha value for fully opaque. + */ + public final static int ALPHA_OPAQUE = 255; + } + + public static final class ImeOption { + /** + * The private IME option used to indicate that no microphone should be shown for a given + * text field. For instance, this is specified by the search dialog when the dialog is + * already showing a voice search button. + * + * @deprecated Use {@link ImeOption#NO_MICROPHONE} with package name prefixed. + */ + @SuppressWarnings("dep-ann") + public static final String NO_MICROPHONE_COMPAT = "nm"; + + /** + * The private IME option used to indicate that no microphone should be shown for a given + * text field. For instance, this is specified by the search dialog when the dialog is + * already showing a voice search button. + */ + public static final String NO_MICROPHONE = "noMicrophoneKey"; + + /** + * The private IME option used to indicate that no settings key should be shown for a given + * text field. + */ + public static final String NO_SETTINGS_KEY = "noSettingsKey"; + + /** + * The private IME option used to indicate that the given text field needs ASCII code points + * input. + * + * @deprecated Use EditorInfo#IME_FLAG_FORCE_ASCII. + */ + @SuppressWarnings("dep-ann") + public static final String FORCE_ASCII = "forceAscii"; + + /** + * The private IME option used to suppress the floating gesture preview for a given text + * field. This overrides the corresponding keyboard settings preference. + * {@link SettingsValues#mGestureFloatingPreviewTextEnabled} + */ + public static final String NO_FLOATING_GESTURE_PREVIEW = "noGestureFloatingPreview"; + + private ImeOption() { + // This utility class is not publicly instantiable. + } + } + + public static final class Subtype { + /** + * The subtype mode used to indicate that the subtype is a keyboard. + */ + public static final String KEYBOARD_MODE = "keyboard"; + + public static final class ExtraValue { + /** + * The subtype extra value used to indicate that this subtype is capable of + * entering ASCII characters. + */ + public static final String ASCII_CAPABLE = "AsciiCapable"; + + /** + * The subtype extra value used to indicate that this subtype is enabled + * when the default subtype is not marked as ascii capable. + */ + public static final String ENABLED_WHEN_DEFAULT_IS_NOT_ASCII_CAPABLE = + "EnabledWhenDefaultIsNotAsciiCapable"; + + /** + * The subtype extra value used to indicate that this subtype is capable of + * entering emoji characters. + */ + public static final String EMOJI_CAPABLE = "EmojiCapable"; + + /** + * The subtype extra value used to indicate that this subtype requires a network + * connection to work. + */ + public static final String REQ_NETWORK_CONNECTIVITY = "requireNetworkConnectivity"; + + /** + * The subtype extra value used to indicate that the display name of this subtype + * contains a "%s" for printf-like replacement and it should be replaced by + * this extra value. + * This extra value is supported on JellyBean and later. + */ + public static final String UNTRANSLATABLE_STRING_IN_SUBTYPE_NAME = + "UntranslatableReplacementStringInSubtypeName"; + + /** + * The subtype extra value used to indicate this subtype keyboard layout set name. + * This extra value is private to LatinIME. + */ + public static final String KEYBOARD_LAYOUT_SET = "KeyboardLayoutSet"; + + /** + * The subtype extra value used to indicate that this subtype is an additional subtype + * that the user defined. This extra value is private to LatinIME. + */ + public static final String IS_ADDITIONAL_SUBTYPE = "isAdditionalSubtype"; + + /** + * The subtype extra value used to specify the combining rules. + */ + public static final String COMBINING_RULES = "CombiningRules"; + + private ExtraValue() { + // This utility class is not publicly instantiable. + } + } + + private Subtype() { + // This utility class is not publicly instantiable. + } + } + + public static final class TextUtils { + /** + * Capitalization mode for {@link android.text.TextUtils#getCapsMode}: don't capitalize + * characters. This value may be used with + * {@link android.text.TextUtils#CAP_MODE_CHARACTERS}, + * {@link android.text.TextUtils#CAP_MODE_WORDS}, and + * {@link android.text.TextUtils#CAP_MODE_SENTENCES}. + */ + // TODO: Straighten this out. It's bizarre to have to use android.text.TextUtils.CAP_MODE_* + // except for OFF that is in Constants.TextUtils. + public static final int CAP_MODE_OFF = 0; + + private TextUtils() { + // This utility class is not publicly instantiable. + } + } + + public static final int NOT_A_CODE = -1; + public static final int NOT_A_CURSOR_POSITION = -1; + // TODO: replace the following constants with state in InputTransaction? + public static final int NOT_A_COORDINATE = -1; + public static final int SUGGESTION_STRIP_COORDINATE = -2; + public static final int EXTERNAL_KEYBOARD_COORDINATE = -4; + + // A hint on how many characters to cache from the TextView. A good value of this is given by + // how many characters we need to be able to almost always find the caps mode. + public static final int EDITOR_CONTENTS_CACHE_SIZE = 1024; + // How many characters we accept for the recapitalization functionality. This needs to be + // large enough for all reasonable purposes, but avoid purposeful attacks. 100k sounds about + // right for this. + public static final int MAX_CHARACTERS_FOR_RECAPITALIZATION = 1024 * 100; + + // Key events coming any faster than this are long-presses. + public static final int LONG_PRESS_MILLISECONDS = 200; + // TODO: Set this value appropriately. + public static final int GET_SUGGESTED_WORDS_TIMEOUT = 200; + // How many continuous deletes at which to start deleting at a higher speed. + public static final int DELETE_ACCELERATE_AT = 20; + + public static final String WORD_SEPARATOR = " "; + + public static boolean isValidCoordinate(final int coordinate) { + // Detect {@link NOT_A_COORDINATE}, {@link SUGGESTION_STRIP_COORDINATE}, + // and {@link SPELL_CHECKER_COORDINATE}. + return coordinate >= 0; + } + + /** + * Custom request code used in + * {@link KeyboardActionListener#onCustomRequest(int)}. + */ + // The code to show input method picker. + public static final int CUSTOM_CODE_SHOW_INPUT_METHOD_PICKER = 1; + + /** + * Some common keys code. Must be positive. + */ + public static final int CODE_ENTER = '\n'; + public static final int CODE_TAB = '\t'; + public static final int CODE_SPACE = ' '; + public static final int CODE_PERIOD = '.'; + public static final int CODE_COMMA = ','; + public static final int CODE_DASH = '-'; + public static final int CODE_SINGLE_QUOTE = '\''; + public static final int CODE_DOUBLE_QUOTE = '"'; + public static final int CODE_SLASH = '/'; + public static final int CODE_BACKSLASH = '\\'; + public static final int CODE_VERTICAL_BAR = '|'; + public static final int CODE_COMMERCIAL_AT = '@'; + public static final int CODE_PLUS = '+'; + public static final int CODE_PERCENT = '%'; + public static final int CODE_CLOSING_PARENTHESIS = ')'; + public static final int CODE_CLOSING_SQUARE_BRACKET = ']'; + public static final int CODE_CLOSING_CURLY_BRACKET = '}'; + public static final int CODE_CLOSING_ANGLE_BRACKET = '>'; + public static final int CODE_INVERTED_QUESTION_MARK = 0xBF; // ¿ + public static final int CODE_INVERTED_EXCLAMATION_MARK = 0xA1; // ¡ + public static final int CODE_GRAVE_ACCENT = '`'; + public static final int CODE_CIRCUMFLEX_ACCENT = '^'; + public static final int CODE_TILDE = '~'; + + public static final String REGEXP_PERIOD = "\\."; + public static final String STRING_SPACE = " "; + + /** + * Special keys code. Must be negative. + * These should be aligned with constants in + * {@link KeyboardCodesSet}. + */ + public static final int CODE_SHIFT = -1; + public static final int CODE_CAPSLOCK = -2; + public static final int CODE_SWITCH_ALPHA_SYMBOL = -3; + public static final int CODE_OUTPUT_TEXT = -4; + public static final int CODE_DELETE = -5; + public static final int CODE_SETTINGS = -6; + public static final int CODE_SHORTCUT = -7; + public static final int CODE_ACTION_NEXT = -8; + public static final int CODE_ACTION_PREVIOUS = -9; + public static final int CODE_LANGUAGE_SWITCH = -10; + public static final int CODE_EMOJI = -11; + public static final int CODE_SHIFT_ENTER = -12; + public static final int CODE_SYMBOL_SHIFT = -13; + public static final int CODE_ALPHA_FROM_EMOJI = -14; + // Code value representing the code is not specified. + public static final int CODE_UNSPECIFIED = -15; + + public static boolean isLetterCode(final int code) { + return code >= CODE_SPACE; + } + + @Nonnull + public static String printableCode(final int code) { + switch (code) { + case CODE_SHIFT: return "shift"; + case CODE_CAPSLOCK: return "capslock"; + case CODE_SWITCH_ALPHA_SYMBOL: return "symbol"; + case CODE_OUTPUT_TEXT: return "text"; + case CODE_DELETE: return "delete"; + case CODE_SETTINGS: return "settings"; + case CODE_SHORTCUT: return "shortcut"; + case CODE_ACTION_NEXT: return "actionNext"; + case CODE_ACTION_PREVIOUS: return "actionPrevious"; + case CODE_LANGUAGE_SWITCH: return "languageSwitch"; + case CODE_EMOJI: return "emoji"; + case CODE_SHIFT_ENTER: return "shiftEnter"; + case CODE_ALPHA_FROM_EMOJI: return "alpha"; + case CODE_UNSPECIFIED: return "unspec"; + case CODE_TAB: return "tab"; + case CODE_ENTER: return "enter"; + case CODE_SPACE: return "space"; + default: + if (code < CODE_SPACE) return String.format("\\u%02X", code); + if (code < 0x100) return String.format("%c", code); + if (code < 0x10000) return String.format("\\u%04X", code); + return String.format("\\U%05X", code); + } + } + + @Nonnull + public static String printableCodes(@Nonnull final int[] codes) { + final StringBuilder sb = new StringBuilder(); + boolean addDelimiter = false; + for (final int code : codes) { + if (code == NOT_A_CODE) break; + if (addDelimiter) sb.append(", "); + sb.append(printableCode(code)); + addDelimiter = true; + } + return "[" + sb + "]"; + } + + /** + * Screen metrics (a.k.a. Device form factor) constants of + * {@link org.kelar.inputmethod.latin.R.integer#config_screen_metrics}. + */ + public static final int SCREEN_METRICS_SMALL_PHONE = 0; + public static final int SCREEN_METRICS_LARGE_PHONE = 1; + public static final int SCREEN_METRICS_LARGE_TABLET = 2; + public static final int SCREEN_METRICS_SMALL_TABLET = 3; + + @UsedForTesting + public static boolean isPhone(final int screenMetrics) { + return screenMetrics == SCREEN_METRICS_SMALL_PHONE + || screenMetrics == SCREEN_METRICS_LARGE_PHONE; + } + + @UsedForTesting + public static boolean isTablet(final int screenMetrics) { + return screenMetrics == SCREEN_METRICS_SMALL_TABLET + || screenMetrics == SCREEN_METRICS_LARGE_TABLET; + } + + /** + * Default capacity of gesture points container. + * This constant is used by {@link BatchInputArbiter} + * and etc. to preallocate regions that contain gesture event points. + */ + public static final int DEFAULT_GESTURE_POINTS_CAPACITY = 128; + + public static final int MAX_IME_DECODER_RESULTS = 20; + public static final int DECODER_SCORE_SCALAR = 1000000; + public static final int DECODER_MAX_SCORE = 1000000000; + + public static final int EVENT_BACKSPACE = 1; + public static final int EVENT_REJECTION = 2; + public static final int EVENT_REVERT = 3; + + private Constants() { + // This utility class is not publicly instantiable. + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/CoordinateUtils.java b/common/src/org/kelar/inputmethod/latin/common/CoordinateUtils.java new file mode 100644 index 000000000..ec46f743e --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/CoordinateUtils.java @@ -0,0 +1,94 @@ +/* + * 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 org.kelar.inputmethod.latin.common; + +import javax.annotation.Nonnull; + +public final class CoordinateUtils { + private static final int INDEX_X = 0; + private static final int INDEX_Y = 1; + private static final int ELEMENT_SIZE = INDEX_Y + 1; + + private CoordinateUtils() { + // This utility class is not publicly instantiable. + } + + @Nonnull + public static int[] newInstance() { + return new int[ELEMENT_SIZE]; + } + + public static int x(@Nonnull final int[] coords) { + return coords[INDEX_X]; + } + + public static int y(@Nonnull final int[] coords) { + return coords[INDEX_Y]; + } + + public static void set(@Nonnull final int[] coords, final int x, final int y) { + coords[INDEX_X] = x; + coords[INDEX_Y] = y; + } + + public static void copy(@Nonnull final int[] destination, @Nonnull final int[] source) { + destination[INDEX_X] = source[INDEX_X]; + destination[INDEX_Y] = source[INDEX_Y]; + } + + @Nonnull + public static int[] newCoordinateArray(final int arraySize) { + return new int[ELEMENT_SIZE * arraySize]; + } + + @Nonnull + public static int[] newCoordinateArray(final int arraySize, + final int defaultX, final int defaultY) { + final int[] result = new int[ELEMENT_SIZE * arraySize]; + for (int i = 0; i < arraySize; ++i) { + setXYInArray(result, i, defaultX, defaultY); + } + return result; + } + + public static int xFromArray(@Nonnull final int[] coordsArray, final int index) { + return coordsArray[ELEMENT_SIZE * index + INDEX_X]; + } + + public static int yFromArray(@Nonnull final int[] coordsArray, final int index) { + return coordsArray[ELEMENT_SIZE * index + INDEX_Y]; + } + + @Nonnull + public static int[] coordinateFromArray(@Nonnull final int[] coordsArray, final int index) { + final int[] coords = newInstance(); + set(coords, xFromArray(coordsArray, index), yFromArray(coordsArray, index)); + return coords; + } + + public static void setXYInArray(@Nonnull final int[] coordsArray, final int index, + final int x, final int y) { + final int baseIndex = ELEMENT_SIZE * index; + coordsArray[baseIndex + INDEX_X] = x; + coordsArray[baseIndex + INDEX_Y] = y; + } + + public static void setCoordinateInArray(@Nonnull final int[] coordsArray, final int index, + @Nonnull final int[] coords) { + setXYInArray(coordsArray, index, x(coords), y(coords)); + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/FileUtils.java b/common/src/org/kelar/inputmethod/latin/common/FileUtils.java new file mode 100644 index 000000000..8696f26c1 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/FileUtils.java @@ -0,0 +1,61 @@ +/* + * Copyright (C) 2013 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 org.kelar.inputmethod.latin.common; + +import java.io.File; +import java.io.FilenameFilter; + +/** + * A simple class to help with removing directories recursively. + */ +public class FileUtils { + private static final String TAG = "FileUtils"; + + public static boolean deleteRecursively(final File path) { + if (path.isDirectory()) { + final File[] files = path.listFiles(); + if (files != null) { + for (final File child : files) { + deleteRecursively(child); + } + } + } + return path.delete(); + } + + public static boolean deleteFilteredFiles(final File dir, final FilenameFilter fileNameFilter) { + if (!dir.isDirectory()) { + return false; + } + final File[] files = dir.listFiles(fileNameFilter); + if (files == null) { + return false; + } + boolean hasDeletedAllFiles = true; + for (final File file : files) { + if (!deleteRecursively(file)) { + hasDeletedAllFiles = false; + } + } + return hasDeletedAllFiles; + } + + public static boolean renameTo(final File fromFile, final File toFile) { + toFile.delete(); + return fromFile.renameTo(toFile); + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/InputPointers.java b/common/src/org/kelar/inputmethod/latin/common/InputPointers.java new file mode 100644 index 000000000..3ee1dbe92 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/InputPointers.java @@ -0,0 +1,166 @@ +/* + * 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 org.kelar.inputmethod.latin.common; + +import org.kelar.inputmethod.annotations.UsedForTesting; + +import javax.annotation.Nonnull; + +// TODO: This class is not thread-safe. +public final class InputPointers { + private static final boolean DEBUG_TIME = false; + + private final int mDefaultCapacity; + private final ResizableIntArray mXCoordinates; + private final ResizableIntArray mYCoordinates; + private final ResizableIntArray mPointerIds; + private final ResizableIntArray mTimes; + + public InputPointers(final int defaultCapacity) { + mDefaultCapacity = defaultCapacity; + mXCoordinates = new ResizableIntArray(defaultCapacity); + mYCoordinates = new ResizableIntArray(defaultCapacity); + mPointerIds = new ResizableIntArray(defaultCapacity); + mTimes = new ResizableIntArray(defaultCapacity); + } + + private void fillWithLastTimeUntil(final int index) { + final int fromIndex = mTimes.getLength(); + // Fill the gap with the latest time. + // See {@link #getTime(int)} and {@link #isValidTimeStamps()}. + if (fromIndex <= 0) { + return; + } + final int fillLength = index - fromIndex + 1; + if (fillLength <= 0) { + return; + } + final int lastTime = mTimes.get(fromIndex - 1); + mTimes.fill(lastTime, fromIndex, fillLength); + } + + public void addPointerAt(final int index, final int x, final int y, final int pointerId, + final int time) { + mXCoordinates.addAt(index, x); + mYCoordinates.addAt(index, y); + mPointerIds.addAt(index, pointerId); + if (DEBUG_TIME) { + fillWithLastTimeUntil(index); + } + mTimes.addAt(index, time); + } + + @UsedForTesting + public void addPointer(final int x, final int y, final int pointerId, final int time) { + mXCoordinates.add(x); + mYCoordinates.add(y); + mPointerIds.add(pointerId); + mTimes.add(time); + } + + public void set(@Nonnull final InputPointers ip) { + mXCoordinates.set(ip.mXCoordinates); + mYCoordinates.set(ip.mYCoordinates); + mPointerIds.set(ip.mPointerIds); + mTimes.set(ip.mTimes); + } + + public void copy(@Nonnull final InputPointers ip) { + mXCoordinates.copy(ip.mXCoordinates); + mYCoordinates.copy(ip.mYCoordinates); + mPointerIds.copy(ip.mPointerIds); + mTimes.copy(ip.mTimes); + } + + /** + * Append the times, x-coordinates and y-coordinates in the specified {@link ResizableIntArray} + * to the end of this. + * @param pointerId the pointer id of the source. + * @param times the source {@link ResizableIntArray} to read the event times from. + * @param xCoordinates the source {@link ResizableIntArray} to read the x-coordinates from. + * @param yCoordinates the source {@link ResizableIntArray} to read the y-coordinates from. + * @param startPos the starting index of the data in {@code times} and etc. + * @param length the number of data to be appended. + */ + public void append(final int pointerId, @Nonnull final ResizableIntArray times, + @Nonnull final ResizableIntArray xCoordinates, + @Nonnull final ResizableIntArray yCoordinates, final int startPos, final int length) { + if (length == 0) { + return; + } + mXCoordinates.append(xCoordinates, startPos, length); + mYCoordinates.append(yCoordinates, startPos, length); + mPointerIds.fill(pointerId, mPointerIds.getLength(), length); + mTimes.append(times, startPos, length); + } + + /** + * Shift to the left by elementCount, discarding elementCount pointers at the start. + * @param elementCount how many elements to shift. + */ + @UsedForTesting + public void shift(final int elementCount) { + mXCoordinates.shift(elementCount); + mYCoordinates.shift(elementCount); + mPointerIds.shift(elementCount); + mTimes.shift(elementCount); + } + + public void reset() { + final int defaultCapacity = mDefaultCapacity; + mXCoordinates.reset(defaultCapacity); + mYCoordinates.reset(defaultCapacity); + mPointerIds.reset(defaultCapacity); + mTimes.reset(defaultCapacity); + } + + public int getPointerSize() { + return mXCoordinates.getLength(); + } + + @Nonnull + public int[] getXCoordinates() { + return mXCoordinates.getPrimitiveArray(); + } + + @Nonnull + public int[] getYCoordinates() { + return mYCoordinates.getPrimitiveArray(); + } + + @Nonnull + public int[] getPointerIds() { + return mPointerIds.getPrimitiveArray(); + } + + /** + * Gets the time each point was registered, in milliseconds, relative to the first event in the + * sequence. + * @return The time each point was registered, in milliseconds, relative to the first event in + * the sequence. + */ + @Nonnull + public int[] getTimes() { + return mTimes.getPrimitiveArray(); + } + + @Override + public String toString() { + return "size=" + getPointerSize() + " id=" + mPointerIds + " time=" + mTimes + + " x=" + mXCoordinates + " y=" + mYCoordinates; + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/LocaleUtils.java b/common/src/org/kelar/inputmethod/latin/common/LocaleUtils.java new file mode 100644 index 000000000..2ed3f1b85 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/LocaleUtils.java @@ -0,0 +1,210 @@ +/* + * Copyright (C) 2011 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 org.kelar.inputmethod.latin.common; + +import java.util.HashMap; +import java.util.HashSet; +import java.util.Locale; + +import javax.annotation.Nonnull; +import javax.annotation.Nullable; + +/** + * A class to help with handling Locales in string form. + * + * This file has the same meaning and features (and shares all of its code) with the one with the + * same name in Latin IME. They need to be kept synchronized; for any update/bugfix to + * this file, consider also updating/fixing the version in Latin IME. + */ +public final class LocaleUtils { + private LocaleUtils() { + // Intentional empty constructor for utility class. + } + + // Locale match level constants. + // A higher level of match is guaranteed to have a higher numerical value. + // Some room is left within constants to add match cases that may arise necessary + // in the future, for example differentiating between the case where the countries + // are both present and different, and the case where one of the locales does not + // specify the countries. This difference is not needed now. + + // Nothing matches. + public static final int LOCALE_NO_MATCH = 0; + // The languages matches, but the country are different. Or, the reference locale requires a + // country and the tested locale does not have one. + public static final int LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER = 3; + // The languages and country match, but the variants are different. Or, the reference locale + // requires a variant and the tested locale does not have one. + public static final int LOCALE_LANGUAGE_AND_COUNTRY_MATCH_VARIANT_DIFFER = 6; + // The required locale is null or empty so it will accept anything, and the tested locale + // is non-null and non-empty. + public static final int LOCALE_ANY_MATCH = 10; + // The language matches, and the tested locale specifies a country but the reference locale + // does not require one. + public static final int LOCALE_LANGUAGE_MATCH = 15; + // The language and the country match, and the tested locale specifies a variant but the + // reference locale does not require one. + public static final int LOCALE_LANGUAGE_AND_COUNTRY_MATCH = 20; + // The compared locales are fully identical. This is the best match level. + public static final int LOCALE_FULL_MATCH = 30; + + // The level at which a match is "normally" considered a locale match with standard algorithms. + // Don't use this directly, use #isMatch to test. + private static final int LOCALE_MATCH = LOCALE_ANY_MATCH; + + // Make this match the maximum match level. If this evolves to have more than 2 digits + // when written in base 10, also adjust the getMatchLevelSortedString method. + private static final int MATCH_LEVEL_MAX = 30; + + /** + * Return how well a tested locale matches a reference locale. + * + * This will check the tested locale against the reference locale and return a measure of how + * a well it matches the reference. The general idea is that the tested locale has to match + * every specified part of the required locale. A full match occur when they are equal, a + * partial match when the tested locale agrees with the reference locale but is more specific, + * and a difference when the tested locale does not comply with all requirements from the + * reference locale. + * In more detail, if the reference locale specifies at least a language and the testedLocale + * does not specify one, or specifies a different one, LOCALE_NO_MATCH is returned. If the + * reference locale is empty or null, it will match anything - in the form of LOCALE_FULL_MATCH + * if the tested locale is empty or null, and LOCALE_ANY_MATCH otherwise. If the reference and + * tested locale agree on the language, but not on the country, + * LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER is returned if the reference locale specifies a country, + * and LOCALE_LANGUAGE_MATCH otherwise. + * If they agree on both the language and the country, but not on the variant, + * LOCALE_LANGUAGE_AND_COUNTRY_MATCH_VARIANT_DIFFER is returned if the reference locale + * specifies a variant, and LOCALE_LANGUAGE_AND_COUNTRY_MATCH otherwise. If everything matches, + * LOCALE_FULL_MATCH is returned. + * Examples: + * en <=> en_US => LOCALE_LANGUAGE_MATCH + * en_US <=> en => LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER + * en_US_POSIX <=> en_US_Android => LOCALE_LANGUAGE_AND_COUNTRY_MATCH_VARIANT_DIFFER + * en_US <=> en_US_Android => LOCALE_LANGUAGE_AND_COUNTRY_MATCH + * sp_US <=> en_US => LOCALE_NO_MATCH + * de <=> de => LOCALE_FULL_MATCH + * en_US <=> en_US => LOCALE_FULL_MATCH + * "" <=> en_US => LOCALE_ANY_MATCH + * + * @param referenceLocale the reference locale to test against. + * @param testedLocale the locale to test. + * @return a constant that measures how well the tested locale matches the reference locale. + */ + public static int getMatchLevel(@Nullable final String referenceLocale, + @Nullable final String testedLocale) { + if (StringUtils.isEmpty(referenceLocale)) { + return StringUtils.isEmpty(testedLocale) ? LOCALE_FULL_MATCH : LOCALE_ANY_MATCH; + } + if (null == testedLocale) return LOCALE_NO_MATCH; + final String[] referenceParams = referenceLocale.split("_", 3); + final String[] testedParams = testedLocale.split("_", 3); + // By spec of String#split, [0] cannot be null and length cannot be 0. + if (!referenceParams[0].equals(testedParams[0])) return LOCALE_NO_MATCH; + switch (referenceParams.length) { + case 1: + return 1 == testedParams.length ? LOCALE_FULL_MATCH : LOCALE_LANGUAGE_MATCH; + case 2: + if (1 == testedParams.length) return LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER; + if (!referenceParams[1].equals(testedParams[1])) + return LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER; + if (3 == testedParams.length) return LOCALE_LANGUAGE_AND_COUNTRY_MATCH; + return LOCALE_FULL_MATCH; + case 3: + if (1 == testedParams.length) return LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER; + if (!referenceParams[1].equals(testedParams[1])) + return LOCALE_LANGUAGE_MATCH_COUNTRY_DIFFER; + if (2 == testedParams.length) return LOCALE_LANGUAGE_AND_COUNTRY_MATCH_VARIANT_DIFFER; + if (!referenceParams[2].equals(testedParams[2])) + return LOCALE_LANGUAGE_AND_COUNTRY_MATCH_VARIANT_DIFFER; + return LOCALE_FULL_MATCH; + } + // It should be impossible to come here + return LOCALE_NO_MATCH; + } + + /** + * Return a string that represents this match level, with better matches first. + * + * The strings are sorted in lexicographic order: a better match will always be less than + * a worse match when compared together. + */ + public static String getMatchLevelSortedString(final int matchLevel) { + // This works because the match levels are 0~99 (actually 0~30) + // Ideally this should use a number of digits equals to the 1og10 of the greater matchLevel + return String.format(Locale.ROOT, "%02d", MATCH_LEVEL_MAX - matchLevel); + } + + /** + * Find out whether a match level should be considered a match. + * + * This method takes a match level as returned by the #getMatchLevel method, and returns whether + * it should be considered a match in the usual sense with standard Locale functions. + * + * @param level the match level, as returned by getMatchLevel. + * @return whether this is a match or not. + */ + public static boolean isMatch(final int level) { + return LOCALE_MATCH <= level; + } + + private static final HashMap<String, Locale> sLocaleCache = new HashMap<>(); + + /** + * Creates a locale from a string specification. + * @param localeString a string specification of a locale, in a format of "ll_cc_variant" where + * "ll" is a language code, "cc" is a country code. + */ + @Nonnull + public static Locale constructLocaleFromString(@Nonnull final String localeString) { + synchronized (sLocaleCache) { + if (sLocaleCache.containsKey(localeString)) { + return sLocaleCache.get(localeString); + } + final String[] elements = localeString.split("_", 3); + final Locale locale; + if (elements.length == 1) { + locale = new Locale(elements[0] /* language */); + } else if (elements.length == 2) { + locale = new Locale(elements[0] /* language */, elements[1] /* country */); + } else { // localeParams.length == 3 + locale = new Locale(elements[0] /* language */, elements[1] /* country */, + elements[2] /* variant */); + } + sLocaleCache.put(localeString, locale); + return locale; + } + } + + // TODO: Get this information from the framework instead of maintaining here by ourselves. + private static final HashSet<String> sRtlLanguageCodes = new HashSet<>(); + static { + // List of known Right-To-Left language codes. + sRtlLanguageCodes.add("ar"); // Arabic + sRtlLanguageCodes.add("fa"); // Persian + sRtlLanguageCodes.add("iw"); // Hebrew + sRtlLanguageCodes.add("ku"); // Kurdish + sRtlLanguageCodes.add("ps"); // Pashto + sRtlLanguageCodes.add("sd"); // Sindhi + sRtlLanguageCodes.add("ug"); // Uyghur + sRtlLanguageCodes.add("ur"); // Urdu + sRtlLanguageCodes.add("yi"); // Yiddish + } + + public static boolean isRtlLanguage(@Nonnull final Locale locale) { + return sRtlLanguageCodes.contains(locale.getLanguage()); + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/NativeSuggestOptions.java b/common/src/org/kelar/inputmethod/latin/common/NativeSuggestOptions.java new file mode 100644 index 000000000..fe7ed9e8e --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/NativeSuggestOptions.java @@ -0,0 +1,63 @@ +/* + * Copyright (C) 2013 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 org.kelar.inputmethod.latin.common; + +public class NativeSuggestOptions { + // Need to update suggest_options.h when you add, remove or reorder options. + private static final int IS_GESTURE = 0; + private static final int USE_FULL_EDIT_DISTANCE = 1; + private static final int BLOCK_OFFENSIVE_WORDS = 2; + private static final int SPACE_AWARE_GESTURE_ENABLED = 3; + private static final int WEIGHT_FOR_LOCALE_IN_THOUSANDS = 4; + private static final int OPTIONS_SIZE = 5; + + private final int[] mOptions; + + public NativeSuggestOptions() { + mOptions = new int[OPTIONS_SIZE]; + } + + public void setIsGesture(final boolean value) { + setBooleanOption(IS_GESTURE, value); + } + + public void setUseFullEditDistance(final boolean value) { + setBooleanOption(USE_FULL_EDIT_DISTANCE, value); + } + + public void setBlockOffensiveWords(final boolean value) { + setBooleanOption(BLOCK_OFFENSIVE_WORDS, value); + } + + public void setWeightForLocale(final float value) { + // We're passing this option as a fixed point value, in thousands. This is decoded in + // native code by SuggestOptions#weightForLocale(). + setIntegerOption(WEIGHT_FOR_LOCALE_IN_THOUSANDS, (int) (value * 1000)); + } + + public int[] getOptions() { + return mOptions; + } + + private void setBooleanOption(final int key, final boolean value) { + mOptions[key] = value ? 1 : 0; + } + + private void setIntegerOption(final int key, final int value) { + mOptions[key] = value; + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/ResizableIntArray.java b/common/src/org/kelar/inputmethod/latin/common/ResizableIntArray.java new file mode 100644 index 000000000..74a1779fe --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/ResizableIntArray.java @@ -0,0 +1,162 @@ +/* + * 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 org.kelar.inputmethod.latin.common; + +import org.kelar.inputmethod.annotations.UsedForTesting; + +import java.util.Arrays; + +import javax.annotation.Nonnull; + +// TODO: This class is not thread-safe. +public final class ResizableIntArray { + @Nonnull + private int[] mArray; + private int mLength; + + public ResizableIntArray(final int capacity) { + reset(capacity); + } + + public int get(final int index) { + if (index < mLength) { + return mArray[index]; + } + throw new ArrayIndexOutOfBoundsException("length=" + mLength + "; index=" + index); + } + + public void addAt(final int index, final int val) { + if (index < mLength) { + mArray[index] = val; + } else { + mLength = index; + add(val); + } + } + + public void add(final int val) { + final int currentLength = mLength; + ensureCapacity(currentLength + 1); + mArray[currentLength] = val; + mLength = currentLength + 1; + } + + /** + * Calculate the new capacity of {@code mArray}. + * @param minimumCapacity the minimum capacity that the {@code mArray} should have. + * @return the new capacity that the {@code mArray} should have. Returns zero when there is no + * need to expand {@code mArray}. + */ + private int calculateCapacity(final int minimumCapacity) { + final int currentCapcity = mArray.length; + if (currentCapcity < minimumCapacity) { + final int nextCapacity = currentCapcity * 2; + // The following is the same as return Math.max(minimumCapacity, nextCapacity); + return minimumCapacity > nextCapacity ? minimumCapacity : nextCapacity; + } + return 0; + } + + private void ensureCapacity(final int minimumCapacity) { + final int newCapacity = calculateCapacity(minimumCapacity); + if (newCapacity > 0) { + // TODO: Implement primitive array pool. + mArray = Arrays.copyOf(mArray, newCapacity); + } + } + + public int getLength() { + return mLength; + } + + public void setLength(final int newLength) { + ensureCapacity(newLength); + mLength = newLength; + } + + public void reset(final int capacity) { + // TODO: Implement primitive array pool. + mArray = new int[capacity]; + mLength = 0; + } + + @Nonnull + public int[] getPrimitiveArray() { + return mArray; + } + + public void set(@Nonnull final ResizableIntArray ip) { + // TODO: Implement primitive array pool. + mArray = ip.mArray; + mLength = ip.mLength; + } + + public void copy(@Nonnull final ResizableIntArray ip) { + final int newCapacity = calculateCapacity(ip.mLength); + if (newCapacity > 0) { + // TODO: Implement primitive array pool. + mArray = new int[newCapacity]; + } + System.arraycopy(ip.mArray, 0, mArray, 0, ip.mLength); + mLength = ip.mLength; + } + + public void append(@Nonnull final ResizableIntArray src, final int startPos, final int length) { + if (length == 0) { + return; + } + final int currentLength = mLength; + final int newLength = currentLength + length; + ensureCapacity(newLength); + System.arraycopy(src.mArray, startPos, mArray, currentLength, length); + mLength = newLength; + } + + public void fill(final int value, final int startPos, final int length) { + if (startPos < 0 || length < 0) { + throw new IllegalArgumentException("startPos=" + startPos + "; length=" + length); + } + final int endPos = startPos + length; + ensureCapacity(endPos); + Arrays.fill(mArray, startPos, endPos, value); + if (mLength < endPos) { + mLength = endPos; + } + } + + /** + * Shift to the left by elementCount, discarding elementCount pointers at the start. + * @param elementCount how many elements to shift. + */ + @UsedForTesting + public void shift(final int elementCount) { + System.arraycopy(mArray, elementCount, mArray, 0, mLength - elementCount); + mLength -= elementCount; + } + + @Override + public String toString() { + final StringBuilder sb = new StringBuilder(); + for (int i = 0; i < mLength; i++) { + if (i != 0) { + sb.append(","); + } + sb.append(mArray[i]); + } + return "[" + sb + "]"; + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/StringUtils.java b/common/src/org/kelar/inputmethod/latin/common/StringUtils.java new file mode 100644 index 000000000..efbecd328 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/StringUtils.java @@ -0,0 +1,704 @@ +/* + * 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 org.kelar.inputmethod.latin.common; + +import org.kelar.inputmethod.annotations.UsedForTesting; + +import java.util.ArrayList; +import java.util.Arrays; +import java.util.Locale; + +import javax.annotation.Nonnull; +import javax.annotation.Nullable; + +public final class StringUtils { + public static final int CAPITALIZE_NONE = 0; // No caps, or mixed case + public static final int CAPITALIZE_FIRST = 1; // First only + public static final int CAPITALIZE_ALL = 2; // All caps + + @Nonnull + private static final String EMPTY_STRING = ""; + + private static final char CHAR_LINE_FEED = 0X000A; + private static final char CHAR_VERTICAL_TAB = 0X000B; + private static final char CHAR_FORM_FEED = 0X000C; + private static final char CHAR_CARRIAGE_RETURN = 0X000D; + private static final char CHAR_NEXT_LINE = 0X0085; + private static final char CHAR_LINE_SEPARATOR = 0X2028; + private static final char CHAR_PARAGRAPH_SEPARATOR = 0X2029; + + private StringUtils() { + // This utility class is not publicly instantiable. + } + + // Taken from android.text.TextUtils. We are extensively using this method in many places, + // some of which don't have the android libraries available. + /** + * Returns true if the string is null or 0-length. + * @param str the string to be examined + * @return true if str is null or zero length + */ + public static boolean isEmpty(@Nullable final CharSequence str) { + return (str == null || str.length() == 0); + } + + // Taken from android.text.TextUtils to cut the dependency to the Android framework. + /** + * Returns a string containing the tokens joined by delimiters. + * @param delimiter the delimiter + * @param tokens an array objects to be joined. Strings will be formed from + * the objects by calling object.toString(). + */ + @Nonnull + public static String join(@Nonnull final CharSequence delimiter, + @Nonnull final Iterable<?> tokens) { + final StringBuilder sb = new StringBuilder(); + boolean firstTime = true; + for (final Object token: tokens) { + if (firstTime) { + firstTime = false; + } else { + sb.append(delimiter); + } + sb.append(token); + } + return sb.toString(); + } + + // Taken from android.text.TextUtils to cut the dependency to the Android framework. + /** + * Returns true if a and b are equal, including if they are both null. + * <p><i>Note: In platform versions 1.1 and earlier, this method only worked well if + * both the arguments were instances of String.</i></p> + * @param a first CharSequence to check + * @param b second CharSequence to check + * @return true if a and b are equal + */ + public static boolean equals(@Nullable final CharSequence a, @Nullable final CharSequence b) { + if (a == b) { + return true; + } + final int length; + if (a != null && b != null && (length = a.length()) == b.length()) { + if (a instanceof String && b instanceof String) { + return a.equals(b); + } + for (int i = 0; i < length; i++) { + if (a.charAt(i) != b.charAt(i)) { + return false; + } + } + return true; + } + return false; + } + + public static int codePointCount(@Nullable final CharSequence text) { + if (isEmpty(text)) { + return 0; + } + return Character.codePointCount(text, 0, text.length()); + } + + @Nonnull + public static String newSingleCodePointString(final int codePoint) { + if (Character.charCount(codePoint) == 1) { + // Optimization: avoid creating a temporary array for characters that are + // represented by a single char value + return String.valueOf((char) codePoint); + } + // For surrogate pair + return new String(Character.toChars(codePoint)); + } + + public static boolean containsInArray(@Nonnull final String text, + @Nonnull final String[] array) { + for (final String element : array) { + if (text.equals(element)) { + return true; + } + } + return false; + } + + /** + * Comma-Splittable Text is similar to Comma-Separated Values (CSV) but has much simpler syntax. + * Unlike CSV, Comma-Splittable Text has no escaping mechanism, so that the text can't contain + * a comma character in it. + */ + @Nonnull + private static final String SEPARATOR_FOR_COMMA_SPLITTABLE_TEXT = ","; + + public static boolean containsInCommaSplittableText(@Nonnull final String text, + @Nullable final String extraValues) { + if (isEmpty(extraValues)) { + return false; + } + return containsInArray(text, extraValues.split(SEPARATOR_FOR_COMMA_SPLITTABLE_TEXT)); + } + + @Nonnull + public static String removeFromCommaSplittableTextIfExists(@Nonnull final String text, + @Nullable final String extraValues) { + if (isEmpty(extraValues)) { + return EMPTY_STRING; + } + final String[] elements = extraValues.split(SEPARATOR_FOR_COMMA_SPLITTABLE_TEXT); + if (!containsInArray(text, elements)) { + return extraValues; + } + final ArrayList<String> result = new ArrayList<>(elements.length - 1); + for (final String element : elements) { + if (!text.equals(element)) { + result.add(element); + } + } + return join(SEPARATOR_FOR_COMMA_SPLITTABLE_TEXT, result); + } + + /** + * Remove duplicates from an array of strings. + * + * This method will always keep the first occurrence of all strings at their position + * in the array, removing the subsequent ones. + */ + public static void removeDupes(@Nonnull final ArrayList<String> suggestions) { + if (suggestions.size() < 2) { + return; + } + int i = 1; + // Don't cache suggestions.size(), since we may be removing items + while (i < suggestions.size()) { + final String cur = suggestions.get(i); + // Compare each suggestion with each previous suggestion + for (int j = 0; j < i; j++) { + final String previous = suggestions.get(j); + if (equals(cur, previous)) { + suggestions.remove(i); + i--; + break; + } + } + i++; + } + } + + @Nonnull + public static String capitalizeFirstCodePoint(@Nonnull final String s, + @Nonnull final Locale locale) { + if (s.length() <= 1) { + return s.toUpperCase(getLocaleUsedForToTitleCase(locale)); + } + // Please refer to the comment below in + // {@link #capitalizeFirstAndDowncaseRest(String,Locale)} as this has the same shortcomings + final int cutoff = s.offsetByCodePoints(0, 1); + return s.substring(0, cutoff).toUpperCase(getLocaleUsedForToTitleCase(locale)) + + s.substring(cutoff); + } + + @Nonnull + public static String capitalizeFirstAndDowncaseRest(@Nonnull final String s, + @Nonnull final Locale locale) { + if (s.length() <= 1) { + return s.toUpperCase(getLocaleUsedForToTitleCase(locale)); + } + // TODO: fix the bugs below + // - It does not work for Serbian, because it fails to account for the "lj" character, + // which should be "Lj" in title case and "LJ" in upper case. + // - It does not work for Dutch, because it fails to account for the "ij" digraph when it's + // written as two separate code points. They are two different characters but both should + // be capitalized as "IJ" as if they were a single letter in most words (not all). If the + // unicode char for the ligature is used however, it works. + final int cutoff = s.offsetByCodePoints(0, 1); + return s.substring(0, cutoff).toUpperCase(getLocaleUsedForToTitleCase(locale)) + + s.substring(cutoff).toLowerCase(locale); + } + + @Nonnull + public static int[] toCodePointArray(@Nonnull final CharSequence charSequence) { + return toCodePointArray(charSequence, 0, charSequence.length()); + } + + @Nonnull + private static final int[] EMPTY_CODEPOINTS = {}; + + /** + * Converts a range of a string to an array of code points. + * @param charSequence the source string. + * @param startIndex the start index inside the string in java chars, inclusive. + * @param endIndex the end index inside the string in java chars, exclusive. + * @return a new array of code points. At most endIndex - startIndex, but possibly less. + */ + @Nonnull + public static int[] toCodePointArray(@Nonnull final CharSequence charSequence, + final int startIndex, final int endIndex) { + final int length = charSequence.length(); + if (length <= 0) { + return EMPTY_CODEPOINTS; + } + final int[] codePoints = + new int[Character.codePointCount(charSequence, startIndex, endIndex)]; + copyCodePointsAndReturnCodePointCount(codePoints, charSequence, startIndex, endIndex, + false /* downCase */); + return codePoints; + } + + /** + * Copies the codepoints in a CharSequence to an int array. + * + * This method assumes there is enough space in the array to store the code points. The size + * can be measured with Character#codePointCount(CharSequence, int, int) before passing to this + * method. If the int array is too small, an ArrayIndexOutOfBoundsException will be thrown. + * Also, this method makes no effort to be thread-safe. Do not modify the CharSequence while + * this method is running, or the behavior is undefined. + * This method can optionally downcase code points before copying them, but it pays no attention + * to locale while doing so. + * + * @param destination the int array. + * @param charSequence the CharSequence. + * @param startIndex the start index inside the string in java chars, inclusive. + * @param endIndex the end index inside the string in java chars, exclusive. + * @param downCase if this is true, code points will be downcased before being copied. + * @return the number of copied code points. + */ + public static int copyCodePointsAndReturnCodePointCount(@Nonnull final int[] destination, + @Nonnull final CharSequence charSequence, final int startIndex, final int endIndex, + final boolean downCase) { + int destIndex = 0; + for (int index = startIndex; index < endIndex; + index = Character.offsetByCodePoints(charSequence, index, 1)) { + final int codePoint = Character.codePointAt(charSequence, index); + // TODO: stop using this, as it's not aware of the locale and does not always do + // the right thing. + destination[destIndex] = downCase ? Character.toLowerCase(codePoint) : codePoint; + destIndex++; + } + return destIndex; + } + + @Nonnull + public static int[] toSortedCodePointArray(@Nonnull final String string) { + final int[] codePoints = toCodePointArray(string); + Arrays.sort(codePoints); + return codePoints; + } + + /** + * Construct a String from a code point array + * + * @param codePoints a code point array that is null terminated when its logical length is + * shorter than the array length. + * @return a string constructed from the code point array. + */ + @Nonnull + public static String getStringFromNullTerminatedCodePointArray( + @Nonnull final int[] codePoints) { + int stringLength = codePoints.length; + for (int i = 0; i < codePoints.length; i++) { + if (codePoints[i] == 0) { + stringLength = i; + break; + } + } + return new String(codePoints, 0 /* offset */, stringLength); + } + + // This method assumes the text is not null. For the empty string, it returns CAPITALIZE_NONE. + public static int getCapitalizationType(@Nonnull final String text) { + // If the first char is not uppercase, then the word is either all lower case or + // camel case, and in either case we return CAPITALIZE_NONE. + final int len = text.length(); + int index = 0; + for (; index < len; index = text.offsetByCodePoints(index, 1)) { + if (Character.isLetter(text.codePointAt(index))) { + break; + } + } + if (index == len) return CAPITALIZE_NONE; + if (!Character.isUpperCase(text.codePointAt(index))) { + return CAPITALIZE_NONE; + } + int capsCount = 1; + int letterCount = 1; + for (index = text.offsetByCodePoints(index, 1); index < len; + index = text.offsetByCodePoints(index, 1)) { + if (1 != capsCount && letterCount != capsCount) break; + final int codePoint = text.codePointAt(index); + if (Character.isUpperCase(codePoint)) { + ++capsCount; + ++letterCount; + } else if (Character.isLetter(codePoint)) { + // We need to discount non-letters since they may not be upper-case, but may + // still be part of a word (e.g. single quote or dash, as in "IT'S" or "FULL-TIME") + ++letterCount; + } + } + // We know the first char is upper case. So we want to test if either every letter other + // than the first is lower case, or if they are all upper case. If the string is exactly + // one char long, then we will arrive here with letterCount 1, and this is correct, too. + if (1 == capsCount) return CAPITALIZE_FIRST; + return (letterCount == capsCount ? CAPITALIZE_ALL : CAPITALIZE_NONE); + } + + public static boolean isIdenticalAfterUpcase(@Nonnull final String text) { + final int length = text.length(); + int i = 0; + while (i < length) { + final int codePoint = text.codePointAt(i); + if (Character.isLetter(codePoint) && !Character.isUpperCase(codePoint)) { + return false; + } + i += Character.charCount(codePoint); + } + return true; + } + + public static boolean isIdenticalAfterDowncase(@Nonnull final String text) { + final int length = text.length(); + int i = 0; + while (i < length) { + final int codePoint = text.codePointAt(i); + if (Character.isLetter(codePoint) && !Character.isLowerCase(codePoint)) { + return false; + } + i += Character.charCount(codePoint); + } + return true; + } + + public static boolean isIdenticalAfterCapitalizeEachWord(@Nonnull final String text, + @Nonnull final int[] sortedSeparators) { + boolean needsCapsNext = true; + final int len = text.length(); + for (int i = 0; i < len; i = text.offsetByCodePoints(i, 1)) { + final int codePoint = text.codePointAt(i); + if (Character.isLetter(codePoint)) { + if ((needsCapsNext && !Character.isUpperCase(codePoint)) + || (!needsCapsNext && !Character.isLowerCase(codePoint))) { + return false; + } + } + // We need a capital letter next if this is a separator. + needsCapsNext = (Arrays.binarySearch(sortedSeparators, codePoint) >= 0); + } + return true; + } + + // TODO: like capitalizeFirst*, this does not work perfectly for Dutch because of the IJ digraph + // which should be capitalized together in *some* cases. + @Nonnull + public static String capitalizeEachWord(@Nonnull final String text, + @Nonnull final int[] sortedSeparators, @Nonnull final Locale locale) { + final StringBuilder builder = new StringBuilder(); + boolean needsCapsNext = true; + final int len = text.length(); + for (int i = 0; i < len; i = text.offsetByCodePoints(i, 1)) { + final String nextChar = text.substring(i, text.offsetByCodePoints(i, 1)); + if (needsCapsNext) { + builder.append(nextChar.toUpperCase(locale)); + } else { + builder.append(nextChar.toLowerCase(locale)); + } + // We need a capital letter next if this is a separator. + needsCapsNext = (Arrays.binarySearch(sortedSeparators, nextChar.codePointAt(0)) >= 0); + } + return builder.toString(); + } + + /** + * Approximates whether the text before the cursor looks like a URL. + * + * This is not foolproof, but it should work well in the practice. + * Essentially it walks backward from the cursor until it finds something that's not a letter, + * digit, or common URL symbol like underscore. If it hasn't found a period yet, then it + * does not look like a URL. + * If the text: + * - starts with www and contains a period + * - starts with a slash preceded by either a slash, whitespace, or start-of-string + * Then it looks like a URL and we return true. Otherwise, we return false. + * + * Note: this method is called quite often, and should be fast. + * + * TODO: This will return that "abc./def" and ".abc/def" look like URLs to keep down the + * code complexity, but ideally it should not. It's acceptable for now. + */ + public static boolean lastPartLooksLikeURL(@Nonnull final CharSequence text) { + int i = text.length(); + if (0 == i) { + return false; + } + int wCount = 0; + int slashCount = 0; + boolean hasSlash = false; + boolean hasPeriod = false; + int codePoint = 0; + while (i > 0) { + codePoint = Character.codePointBefore(text, i); + if (codePoint < Constants.CODE_PERIOD || codePoint > 'z') { + // Handwavy heuristic to see if that's a URL character. Anything between period + // and z. This includes all lower- and upper-case ascii letters, period, + // underscore, arrobase, question mark, equal sign. It excludes spaces, exclamation + // marks, double quotes... + // Anything that's not a URL-like character causes us to break from here and + // evaluate normally. + break; + } + if (Constants.CODE_PERIOD == codePoint) { + hasPeriod = true; + } + if (Constants.CODE_SLASH == codePoint) { + hasSlash = true; + if (2 == ++slashCount) { + return true; + } + } else { + slashCount = 0; + } + if ('w' == codePoint) { + ++wCount; + } else { + wCount = 0; + } + i = Character.offsetByCodePoints(text, i, -1); + } + // End of the text run. + // If it starts with www and includes a period, then it looks like a URL. + if (wCount >= 3 && hasPeriod) { + return true; + } + // If it starts with a slash, and the code point before is whitespace, it looks like an URL. + if (1 == slashCount && (0 == i || Character.isWhitespace(codePoint))) { + return true; + } + // If it has both a period and a slash, it looks like an URL. + if (hasPeriod && hasSlash) { + return true; + } + // Otherwise, it doesn't look like an URL. + return false; + } + + /** + * Examines the string and returns whether we're inside a double quote. + * + * This is used to decide whether we should put an automatic space before or after a double + * quote character. If we're inside a quotation, then we want to close it, so we want a space + * after and not before. Otherwise, we want to open the quotation, so we want a space before + * and not after. Exception: after a digit, we never want a space because the "inch" or + * "minutes" use cases is dominant after digits. + * In the practice, we determine whether we are in a quotation or not by finding the previous + * double quote character, and looking at whether it's followed by whitespace. If so, that + * was a closing quotation mark, so we're not inside a double quote. If it's not followed + * by whitespace, then it was an opening quotation mark, and we're inside a quotation. + * + * @param text the text to examine. + * @return whether we're inside a double quote. + */ + public static boolean isInsideDoubleQuoteOrAfterDigit(@Nonnull final CharSequence text) { + int i = text.length(); + if (0 == i) { + return false; + } + int codePoint = Character.codePointBefore(text, i); + if (Character.isDigit(codePoint)) { + return true; + } + int prevCodePoint = 0; + while (i > 0) { + codePoint = Character.codePointBefore(text, i); + if (Constants.CODE_DOUBLE_QUOTE == codePoint) { + // If we see a double quote followed by whitespace, then that + // was a closing quote. + if (Character.isWhitespace(prevCodePoint)) { + return false; + } + } + if (Character.isWhitespace(codePoint) && Constants.CODE_DOUBLE_QUOTE == prevCodePoint) { + // If we see a double quote preceded by whitespace, then that + // was an opening quote. No need to continue seeking. + return true; + } + i -= Character.charCount(codePoint); + prevCodePoint = codePoint; + } + // We reached the start of text. If the first char is a double quote, then we're inside + // a double quote. Otherwise we're not. + return Constants.CODE_DOUBLE_QUOTE == codePoint; + } + + public static boolean isEmptyStringOrWhiteSpaces(@Nonnull final String s) { + final int N = codePointCount(s); + for (int i = 0; i < N; ++i) { + if (!Character.isWhitespace(s.codePointAt(i))) { + return false; + } + } + return true; + } + + @UsedForTesting + @Nonnull + public static String byteArrayToHexString(@Nullable final byte[] bytes) { + if (bytes == null || bytes.length == 0) { + return EMPTY_STRING; + } + final StringBuilder sb = new StringBuilder(); + for (final byte b : bytes) { + sb.append(String.format("%02x", b & 0xff)); + } + return sb.toString(); + } + + /** + * Convert hex string to byte array. The string length must be an even number. + */ + @UsedForTesting + @Nullable + public static byte[] hexStringToByteArray(@Nullable final String hexString) { + if (isEmpty(hexString)) { + return null; + } + final int N = hexString.length(); + if (N % 2 != 0) { + throw new NumberFormatException("Input hex string length must be an even number." + + " Length = " + N); + } + final byte[] bytes = new byte[N / 2]; + for (int i = 0; i < N; i += 2) { + bytes[i / 2] = (byte) ((Character.digit(hexString.charAt(i), 16) << 4) + + Character.digit(hexString.charAt(i + 1), 16)); + } + return bytes; + } + + private static final String LANGUAGE_GREEK = "el"; + + @Nonnull + private static Locale getLocaleUsedForToTitleCase(@Nonnull final Locale locale) { + // In Greek locale {@link String#toUpperCase(Locale)} eliminates accents from its result. + // In order to get accented upper case letter, {@link Locale#ROOT} should be used. + if (LANGUAGE_GREEK.equals(locale.getLanguage())) { + return Locale.ROOT; + } + return locale; + } + + @Nullable + public static String toTitleCaseOfKeyLabel(@Nullable final String label, + @Nonnull final Locale locale) { + if (label == null) { + return label; + } + return label.toUpperCase(getLocaleUsedForToTitleCase(locale)); + } + + public static int toTitleCaseOfKeyCode(final int code, @Nonnull final Locale locale) { + if (!Constants.isLetterCode(code)) { + return code; + } + final String label = newSingleCodePointString(code); + final String titleCaseLabel = toTitleCaseOfKeyLabel(label, locale); + return codePointCount(titleCaseLabel) == 1 + ? titleCaseLabel.codePointAt(0) : Constants.CODE_UNSPECIFIED; + } + + public static int getTrailingSingleQuotesCount(@Nonnull final CharSequence charSequence) { + final int lastIndex = charSequence.length() - 1; + int i = lastIndex; + while (i >= 0 && charSequence.charAt(i) == Constants.CODE_SINGLE_QUOTE) { + --i; + } + return lastIndex - i; + } + + @UsedForTesting + public static class Stringizer<E> { + @Nonnull + private static final String[] EMPTY_STRING_ARRAY = new String[0]; + + @UsedForTesting + @Nonnull + public String stringize(@Nullable final E element) { + if (element == null) { + return "null"; + } + return element.toString(); + } + + @UsedForTesting + @Nonnull + public final String join(@Nullable final E[] array) { + return joinStringArray(toStringArray(array), null /* delimiter */); + } + + @UsedForTesting + public final String join(@Nullable final E[] array, @Nullable final String delimiter) { + return joinStringArray(toStringArray(array), delimiter); + } + + @Nonnull + protected String[] toStringArray(@Nullable final E[] array) { + if (array == null) { + return EMPTY_STRING_ARRAY; + } + final String[] stringArray = new String[array.length]; + for (int index = 0; index < array.length; index++) { + stringArray[index] = stringize(array[index]); + } + return stringArray; + } + + @Nonnull + protected String joinStringArray(@Nonnull final String[] stringArray, + @Nullable final String delimiter) { + if (delimiter == null) { + return Arrays.toString(stringArray); + } + final StringBuilder sb = new StringBuilder(); + for (int index = 0; index < stringArray.length; index++) { + sb.append(index == 0 ? "[" : delimiter); + sb.append(stringArray[index]); + } + return sb + "]"; + } + } + + /** + * Returns whether the last composed word contains line-breaking character (e.g. CR or LF). + * @param text the text to be examined. + * @return {@code true} if the last composed word contains line-breaking separator. + */ + public static boolean hasLineBreakCharacter(@Nullable final String text) { + if (isEmpty(text)) { + return false; + } + for (int i = text.length() - 1; i >= 0; --i) { + final char c = text.charAt(i); + switch (c) { + case CHAR_LINE_FEED: + case CHAR_VERTICAL_TAB: + case CHAR_FORM_FEED: + case CHAR_CARRIAGE_RETURN: + case CHAR_NEXT_LINE: + case CHAR_LINE_SEPARATOR: + case CHAR_PARAGRAPH_SEPARATOR: + return true; + } + } + return false; + } +} diff --git a/common/src/org/kelar/inputmethod/latin/common/UnicodeSurrogate.java b/common/src/org/kelar/inputmethod/latin/common/UnicodeSurrogate.java new file mode 100644 index 000000000..3221606b2 --- /dev/null +++ b/common/src/org/kelar/inputmethod/latin/common/UnicodeSurrogate.java @@ -0,0 +1,38 @@ +/* + * Copyright (C) 2015 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 org.kelar.inputmethod.latin.common; + +/** + * Emojis are supplementary characters expressed as a low+high pair. For instance, + * the emoji U+1F625 is encoded as "\uD83D\uDE25" in UTF-16, where '\uD83D' is in + * the range of [0xd800, 0xdbff] and '\uDE25' is in the range of [0xdc00, 0xdfff]. + * {@see http://docs.oracle.com/javase/6/docs/api/java/lang/Character.html#unicode} + */ +public final class UnicodeSurrogate { + private static final char LOW_SURROGATE_MIN = '\uD800'; + private static final char LOW_SURROGATE_MAX = '\uDBFF'; + private static final char HIGH_SURROGATE_MIN = '\uDC00'; + private static final char HIGH_SURROGATE_MAX = '\uDFFF'; + + public static boolean isLowSurrogate(final char c) { + return c >= LOW_SURROGATE_MIN && c <= LOW_SURROGATE_MAX; + } + + public static boolean isHighSurrogate(final char c) { + return c >= HIGH_SURROGATE_MIN && c <= HIGH_SURROGATE_MAX; + } +} |