Chinese Pinyin
This commit is contained in:
sspanak
Dimo Karaivanov
parent
51cd39fe27
commit
c57877ce9a
46 changed files with 497243 additions and 151 deletions
|
|
@ -7,7 +7,7 @@ If you would like to contribute to the project by fixing a bug, adding a new lan
|
|||
- Finally, see what is the actual [contribution process](#contribution-process) and how to get your code merged.
|
||||
|
||||
## Building
|
||||
The recommended way of building is using Android Studio. The project is compatible with Android Studio Ladybug | 2024.2.1.
|
||||
The recommended way of building is using Android Studio. The project is compatible with Android Studio Meerkat | 2024.3.1.
|
||||
|
||||
### Building a Debug .apk
|
||||
If you have not configured Android Studio yet, follow [the official manual](https://developer.android.com/training/basics/firstapp), then follow the simple steps below to get the project running.
|
||||
|
|
@ -57,11 +57,13 @@ To support a new language one needs to:
|
|||
- Keys 2 through 9, just contain the possible letters.
|
||||
- `abcString` _(optional)_. A custom string to display in ABC mode. By default, the first three letters on 2-key are used (e.g. "ABC" or "АБВ"). Set this if the first letters of the alphabet are _not_ on 2-key, like in Hebrew, or if a different string makes more sense.
|
||||
- `currency` _(optional)_. A string representing the currency related to that language, for example: `₪`, `₩`, `﷼`, etc. The character will be displayed in position 3 between the factory currency characters.
|
||||
- `filterBySounds` _(optional; requires `sounds`)_. For transcribed languages (e.g. Korean, Chinese), enables filtering by the Latin transcriptions. When it is set to `no` or missing, the filtering function and virtual key will be disabled. It has no effect for languages with alphabets, such as English, French, Polish, Arabic, Hindi, Bulgarian, etc.
|
||||
- `hasABC` _(optional)_. Disables ABC mode for this language. Useful for languages where the writing system is not an alphabet, such as East Asian languages. Default: `yes`.
|
||||
- `hasSpaceBetweenWords` _(optional)_ set to `no` when the language does not use spaces between words. For example: Thai, Chinese, Japanese, Korean, and so on. The default is `yes`.
|
||||
- `hasUpperCase` _(optional)_ set to `no` when the language has no upper- and lowercase letters. For example: Arabic, Hebrew, East Asian languages, and so on. The default is `yes`.
|
||||
- `name` _(optional)_ is automatically generated and equals the native name of the language (e.g. "English", "Deutsch", "Українська"). However, sometimes, the automatically selected name may be ambiguous. For example, both Portuguese in Portugal and Brazil will default to "Português", so assigning "Português brasileiro" would make it clear it's the language used in Brazil.
|
||||
-`numerals` _(optional)_ can be used to set a custom list of numerals. The list must contain exactly 10 characters equivalent to the digits from 0 to 9. For example, in Arabic you could use: `numerals: [٠,١,٢,٣,٤,٥,٦,٧,٨,٩]`.
|
||||
- `sounds` _(mandatory for non-alphabetic languages)_ is an array of elements in the format: `[sound,digits]`. It is used for East Asian or other languages where there are thousands of different characters, that can not be described in the layout property. `sounds` contains all possible vowel and consonant sounds and their respective digit combinations. There must be no repeating sounds. If a single Latin letter stands for a sound, the letter must be capital. If more than one letter is necessary to represent the sound, the first letter must be capital, while the rest must be small. For example, "A", "P", "Wo", "Ei", "Dd". The sounds are then used in the dictionary format with phonetic transcriptions. See `Korean.yml` and the respective dictionary file for an example.
|
||||
- `numerals` _(optional)_ can be used to set a custom list of numerals. The list must contain exactly 10 characters equivalent to the digits from 0 to 9. For example, in Arabic you could use: `numerals: [٠,١,٢,٣,٤,٥,٦,٧,٨,٩]`.
|
||||
- `sounds` _(mandatory for non-alphabetic languages)_ is an array of elements in the format: `[sound,digits]`. It is used for East Asian or other languages where there are thousands of different characters, that can not be described in the layout property. `sounds` contains all possible vowel and consonant sounds and their respective digit combinations. There must be no repeating sounds. If a single Latin letter stands for a sound, the letter must be capital. If more than one letter is necessary to represent the sound, the first letter must be capital, while the rest must be small. For example, "A", "P", "Wo", "Ei", "Dd". The sounds are then used in the dictionary format with phonetic transcriptions. See `ChinesePinyin.yml` and `Korean.yml` and the respective dictionary files for an example.
|
||||
|
||||
### Dictionary Formats
|
||||
|
||||
|
|
@ -113,6 +115,8 @@ Constraints:
|
|||
- The separator is `TAB`.
|
||||
- The first element is the language character or word
|
||||
- The second element is the phonetic representation with Latin letters. It must be a combination of the `sounds` in the respective YAML definition.
|
||||
- The third element, frequency, is optional. If missing, it is assumed to be 0.
|
||||
- The frequency must be an integer between 0 and 255, when present.
|
||||
|
||||
Example definition:
|
||||
```yaml
|
||||
|
|
@ -125,14 +129,14 @@ Example definition:
|
|||
|
||||
Example dictionary:
|
||||
```csv
|
||||
다 WeWo
|
||||
다 WeWo 45
|
||||
줘 TtWoWe
|
||||
와 Wo
|
||||
와 Wo 3
|
||||
```
|
||||
|
||||
Using the above example, when the user types "221-48-123", it will result in: "줘".
|
||||
|
||||
See `Korean.yml` and `ko-utf8.csv` for more examples.
|
||||
See `Korean.yml` with `ko-utf8.csv`, and `ChinesePinyin.yml` with `zh-pinyin.csv` for more examples.
|
||||
|
||||
## Merging Languages
|
||||
Traditional T9 does not officially support typing in more than one language simultaneously, and it will not provide such a possibility in the future. But if you are interested in a do-it-yourself solution, you can achieve it by merging the dictionaries for two (or more) languages into a new language. Follow the steps in [this discussion](https://github.com/sspanak/tt9/issues/713) about merging Czech and English. Just keep in mind: a) there can be no repeating words; b) it is not recommended to go over 2 million words, especially on low-end devices. Normal performance is not guaranteed beyond that limit.
|
||||
|
|
|
|||
|
|
@ -8,13 +8,13 @@ ext.convertDictionaries = { definitionsInputDir, dictionariesInputDir, dictionar
|
|||
int errorCount = 0
|
||||
|
||||
def errorStream = fileTree(dir: definitionsInputDir).getFiles().parallelStream().map { definition ->
|
||||
def (_, sounds, noSyllables, locale, dictionaryFile, langFileErrorCount, langFileErrorMsg) = parseLanguageDefintion(definition, dictionariesInputDir)
|
||||
def (_, sounds, prependSoundsToWords, noSyllables, locale, dictionaryFile, langFileErrorCount, langFileErrorMsg) = parseLanguageDefintion(definition, dictionariesInputDir)
|
||||
errorCount += langFileErrorCount
|
||||
if (!langFileErrorMsg.isEmpty()) {
|
||||
return langFileErrorMsg
|
||||
}
|
||||
|
||||
def (conversionErrorCount, conversionErrorMessages) = convertDictionary(definition, dictionaryFile, dictionariesOutputDir, dictionariesMetaDir, DICTIONARY_OUTPUT_EXTENSION, sounds, noSyllables, locale, MAX_ERRORS, CSV_DELIMITER)
|
||||
def (conversionErrorCount, conversionErrorMessages) = convertDictionary(definition, dictionaryFile, dictionariesOutputDir, dictionariesMetaDir, DICTIONARY_OUTPUT_EXTENSION, sounds, prependSoundsToWords, noSyllables, locale, MAX_ERRORS, CSV_DELIMITER)
|
||||
errorCount += conversionErrorCount
|
||||
if (!conversionErrorMessages.isEmpty()) {
|
||||
return conversionErrorMessages
|
||||
|
|
@ -31,12 +31,12 @@ ext.convertDictionaries = { definitionsInputDir, dictionariesInputDir, dictionar
|
|||
|
||||
|
||||
// this cannot be static, because DictionaryTools will not be visible
|
||||
def convertDictionary(File definition, File csvDictionary, String dictionariesOutputDir, String dictionariesMetaDir, String outputDictionaryExtension, HashMap<String, String> sounds, boolean noSyllables, Locale locale, int maxErrors, String csvDelimiter) {
|
||||
def convertDictionary(File definition, File csvDictionary, String dictionariesOutputDir, String dictionariesMetaDir, String outputDictionaryExtension, HashMap<String, String> sounds, boolean prependSoundsToWords, boolean noSyllables, Locale locale, int maxErrors, String csvDelimiter) {
|
||||
if (isDictionaryUpToDate(definition, csvDictionary, dictionariesMetaDir)) {
|
||||
return [0, ""]
|
||||
}
|
||||
|
||||
|
||||
final LATIN_ONLY_WORD = "^[A-Za-z]+\$"
|
||||
int errorCount = 0
|
||||
String errorMsg = ''
|
||||
|
||||
|
|
@ -63,7 +63,7 @@ def convertDictionary(File definition, File csvDictionary, String dictionariesOu
|
|||
outputDictionary.put(digitSequence, new ArrayList<>())
|
||||
}
|
||||
// prefix the frequency to sort the words later
|
||||
outputDictionary.get(digitSequence).add("${String.format('%03d', frequency)}${word}")
|
||||
outputDictionary.get(digitSequence).add("${String.format('%03d', frequency)}${prependSoundsToWords && !(word =~ LATIN_ONLY_WORD) ? transcription : ''}${word}")
|
||||
wordCount++
|
||||
}
|
||||
}
|
||||
|
|
|
|||
45
app/languages/definitions/ChinesePinyin.yml
Normal file
45
app/languages/definitions/ChinesePinyin.yml
Normal file
|
|
@ -0,0 +1,45 @@
|
|||
locale: zh-CN
|
||||
dictionaryFile: zh-pinyin.csv
|
||||
name: 中文 / 简体
|
||||
hasABC: no
|
||||
hasSpaceBetweenWords: no
|
||||
hasUpperCase: no
|
||||
layout:
|
||||
- [SPECIAL] # 0
|
||||
- [PUNCTUATION_ZH] # 1
|
||||
- [a, b, c] # 2
|
||||
- [d, e, f] # 3
|
||||
- [g, h, i] # 4
|
||||
- [j, k, l] # 5
|
||||
- [m, n, o] # 6
|
||||
- [p, q, r, s] # 7
|
||||
- [t, u, v] # 8
|
||||
- [w, x, y, z] # 9
|
||||
filterBySounds: yes
|
||||
sounds:
|
||||
- [A,2]
|
||||
- [B,2]
|
||||
- [C,2]
|
||||
- [D,3]
|
||||
- [E,3]
|
||||
- [F,3]
|
||||
- [G,4]
|
||||
- [H,4]
|
||||
- [I,4]
|
||||
- [J,5]
|
||||
- [K,5]
|
||||
- [L,5]
|
||||
- [M,6]
|
||||
- [N,6]
|
||||
- [O,6]
|
||||
- [P,7]
|
||||
- [Q,7]
|
||||
- [R,7]
|
||||
- [S,7]
|
||||
- [T,8]
|
||||
- [U,8]
|
||||
- [V,8]
|
||||
- [W,9]
|
||||
- [X,9]
|
||||
- [Y,9]
|
||||
- [Z,9]
|
||||
|
|
@ -1,6 +1,7 @@
|
|||
locale: ko-KR
|
||||
currency: ₩
|
||||
dictionaryFile: ko-utf8.csv
|
||||
hasABC: no
|
||||
hasUpperCase: no
|
||||
layout: # only used for the virtual key labels
|
||||
- [ㅇ,ㅁ,SPECIAL] # 0
|
||||
|
|
|
|||
496366
app/languages/dictionaries/zh-pinyin.csv
Normal file
496366
app/languages/dictionaries/zh-pinyin.csv
Normal file
File diff suppressed because it is too large
Load diff
|
|
@ -88,7 +88,7 @@ public class DataStore {
|
|||
}
|
||||
|
||||
|
||||
public static void getWords(ConsumerCompat<ArrayList<String>> dataHandler, Language language, String sequence, String filter, int minWords, int maxWords) {
|
||||
public static void getWords(ConsumerCompat<ArrayList<String>> dataHandler, Language language, String sequence, boolean onlyExactSequence, String filter, int minWords, int maxWords) {
|
||||
if (getWordsTask != null && !getWordsTask.isDone()) {
|
||||
dataHandler.accept(new ArrayList<>());
|
||||
getWordsCancellationSignal.cancel();
|
||||
|
|
@ -96,14 +96,14 @@ public class DataStore {
|
|||
}
|
||||
|
||||
getWordsCancellationSignal = new CancellationSignal();
|
||||
getWordsTask = executor.submit(() -> getWordsSync(dataHandler, language, sequence, filter, minWords, maxWords));
|
||||
getWordsTask = executor.submit(() -> getWordsSync(dataHandler, language, sequence, onlyExactSequence, filter, minWords, maxWords));
|
||||
executor.submit(DataStore::setGetWordsTimeout);
|
||||
}
|
||||
|
||||
|
||||
private static void getWordsSync(ConsumerCompat<ArrayList<String>> dataHandler, Language language, String sequence, String filter, int minWords, int maxWords) {
|
||||
private static void getWordsSync(ConsumerCompat<ArrayList<String>> dataHandler, Language language, String sequence, boolean onlyExactSequence, String filter, int minWords, int maxWords) {
|
||||
try {
|
||||
ArrayList<String> data = words.getMany(getWordsCancellationSignal, language, sequence, filter, minWords, maxWords);
|
||||
ArrayList<String> data = words.getMany(getWordsCancellationSignal, language, sequence, onlyExactSequence, filter, minWords, maxWords);
|
||||
asyncReturn.post(() -> dataHandler.accept(data));
|
||||
} catch (Exception e) {
|
||||
Logger.e(LOG_TAG, "Error fetching words: " + e.getMessage());
|
||||
|
|
|
|||
|
|
@ -134,7 +134,7 @@ public class CustomWordsImporter extends AbstractFileProcessor {
|
|||
return false;
|
||||
}
|
||||
|
||||
if (customWord.language == null || customWord.language.isSyllabary() || readOps.exists(sqlite.getDb(), customWord.language, customWord.word)) {
|
||||
if (customWord.language == null || customWord.language.isTranscribed() || readOps.exists(sqlite.getDb(), customWord.language, customWord.word)) {
|
||||
ignoredWords++;
|
||||
} else {
|
||||
InsertOps.insertCustomWord(sqlite.getDb(), customWord.language, customWord.sequence, customWord.word);
|
||||
|
|
|
|||
|
|
@ -42,7 +42,7 @@ public class WordFile extends AssetFile {
|
|||
public WordFile(@NonNull Context context, Language language, AssetManager assets) {
|
||||
super(assets, language != null ? language.getDictionaryFile() : "");
|
||||
this.context = context;
|
||||
hasSyllables = language != null && language.isSyllabary();
|
||||
hasSyllables = language != null && language.isTranscribed();
|
||||
|
||||
lastCharCode = 0;
|
||||
reader = null;
|
||||
|
|
|
|||
|
|
@ -18,7 +18,6 @@ import io.github.sspanak.tt9.db.entities.WordList;
|
|||
import io.github.sspanak.tt9.db.entities.WordPositionsStringBuilder;
|
||||
import io.github.sspanak.tt9.db.wordPairs.WordPair;
|
||||
import io.github.sspanak.tt9.db.words.SlowQueryStats;
|
||||
import io.github.sspanak.tt9.db.words.WordStore;
|
||||
import io.github.sspanak.tt9.languages.EmojiLanguage;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
|
|
@ -131,9 +130,6 @@ public class ReadOps {
|
|||
return new WordList();
|
||||
}
|
||||
|
||||
// EXACT_MATCHES concerns only the positions query
|
||||
filter = filter.equals(WordStore.FILTER_EXACT_MATCHES_ONLY) ? "" : filter;
|
||||
|
||||
String wordsQuery = getWordsQuery(language, positions, filter, maximumWords, fullOutput);
|
||||
if (wordsQuery.isEmpty() || (cancel != null && cancel.isCanceled())) {
|
||||
return new WordList();
|
||||
|
|
@ -157,10 +153,10 @@ public class ReadOps {
|
|||
}
|
||||
|
||||
|
||||
public String getSimilarWordPositions(@NonNull SQLiteDatabase db, @NonNull CancellationSignal cancel, @NonNull Language language, @NonNull String sequence, String wordFilter, int minPositions) {
|
||||
public String getSimilarWordPositions(@NonNull SQLiteDatabase db, @NonNull CancellationSignal cancel, @NonNull Language language, @NonNull String sequence, boolean onlyExactSequenceMatches, String wordFilter, int minPositions) {
|
||||
int generations;
|
||||
|
||||
if (wordFilter.equals(WordStore.FILTER_EXACT_MATCHES_ONLY)) {
|
||||
if (onlyExactSequenceMatches) {
|
||||
generations = 0;
|
||||
} else {
|
||||
generations = switch (sequence.length()) {
|
||||
|
|
@ -176,7 +172,7 @@ public class ReadOps {
|
|||
|
||||
@NonNull
|
||||
public String getWordPositions(@NonNull SQLiteDatabase db, @Nullable CancellationSignal cancel, @NonNull Language language, @NonNull String sequence, int generations, int minPositions, String wordFilter) {
|
||||
if (sequence.length() == 1 || (cancel != null && cancel.isCanceled())) {
|
||||
if ((sequence.length() == 1 && !language.isTranscribed()) || (cancel != null && cancel.isCanceled())) {
|
||||
return sequence;
|
||||
}
|
||||
|
||||
|
|
@ -301,9 +297,13 @@ public class ReadOps {
|
|||
sql.append(" AND word LIKE '").append(filter.replaceAll("'", "''")).append("%'");
|
||||
}
|
||||
|
||||
sql
|
||||
.append(" ORDER BY LENGTH(word), frequency DESC")
|
||||
.append(" LIMIT ").append(maxWords);
|
||||
sql.append(" ORDER BY LENGTH(word), frequency DESC");
|
||||
|
||||
if (maxWords < 0 && maxWordsPerSequence.containsKey(language)) {
|
||||
Integer limit = maxWordsPerSequence.get(language);
|
||||
maxWords = limit != null ? limit : SettingsStore.SUGGESTIONS_POSITIONS_LIMIT;
|
||||
}
|
||||
sql.append(" LIMIT ").append(maxWords);
|
||||
|
||||
String wordsSql = sql.toString();
|
||||
Logger.v(LOG_TAG, "Words SQL: " + wordsSql);
|
||||
|
|
|
|||
|
|
@ -24,13 +24,16 @@ public class WordPair {
|
|||
|
||||
|
||||
boolean isInvalid() {
|
||||
Text w1 = new Text(word1);
|
||||
Text w2 = new Text(word2);
|
||||
|
||||
return
|
||||
language == null
|
||||
|| word1.isEmpty() || word2.isEmpty()
|
||||
|| (word1.length() > SettingsStore.WORD_PAIR_MAX_WORD_LENGTH && word2.length() > SettingsStore.WORD_PAIR_MAX_WORD_LENGTH)
|
||||
|| word1.equals(word2)
|
||||
|| sequence2 == null || word2.length() != sequence2.length() || !(new Text(sequence2).isNumeric())
|
||||
|| !(new Text(word1).isAlphabetic()) || !(new Text(word2).isAlphabetic());
|
||||
|| sequence2 == null || !(new Text(sequence2).isNumeric())
|
||||
|| (w1.codePointLength() > SettingsStore.WORD_PAIR_MAX_WORD_LENGTH && w2.codePointLength() > SettingsStore.WORD_PAIR_MAX_WORD_LENGTH)
|
||||
|| !w1.isAlphabetic() || !w2.isAlphabetic();
|
||||
}
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -137,11 +137,6 @@ public class WordPairStore extends BaseSyncStore {
|
|||
|
||||
int totalPairs = 0;
|
||||
for (Language language : languages) {
|
||||
if (language.isSyllabary()) {
|
||||
Logger.d(LOG_TAG, "Not loading word pairs for syllabary language: " + language.getId());
|
||||
continue;
|
||||
}
|
||||
|
||||
HashMap<WordPair, WordPair> wordPairs = pairs.get(language.getId());
|
||||
if (wordPairs == null) {
|
||||
wordPairs = new HashMap<>();
|
||||
|
|
|
|||
|
|
@ -236,7 +236,7 @@ public class DictionaryLoader {
|
|||
|
||||
|
||||
private int importLetters(Language language) throws InvalidLanguageCharactersException {
|
||||
if (language.isSyllabary()) {
|
||||
if (language.isTranscribed()) {
|
||||
return 0;
|
||||
}
|
||||
|
||||
|
|
|
|||
|
|
@ -26,7 +26,6 @@ import io.github.sspanak.tt9.util.Timer;
|
|||
|
||||
|
||||
public class WordStore extends BaseSyncStore {
|
||||
public static final String FILTER_EXACT_MATCHES_ONLY = "__exact__";
|
||||
private final String LOG_TAG = "sqlite.WordStore";
|
||||
private final ReadOps readOps;
|
||||
|
||||
|
|
@ -60,10 +59,10 @@ public class WordStore extends BaseSyncStore {
|
|||
/**
|
||||
* Loads words matching and similar to a given digit sequence
|
||||
* For example: "7655" -> "roll" (exact match), but also: "rolled", "roller", "rolling", ...
|
||||
* and other similar. When "wordFilter" is set to FILTER_EXACT_MATCHES_ONLY, the word list is
|
||||
* constrained only to the words with length equal to the digit sequence length (exact matches).
|
||||
* and other similar. When "onlyExactSequence" is TRUE, the word list is constrained only to
|
||||
* the words with length equal to the digit sequence length (exact matches).
|
||||
*/
|
||||
public ArrayList<String> getMany(@NonNull CancellationSignal cancel, Language language, String sequence, String wordFilter, int minimumWords, int maximumWords) {
|
||||
public ArrayList<String> getMany(@NonNull CancellationSignal cancel, Language language, String sequence, boolean onlyExactSequence, String wordFilter, int minimumWords, int maximumWords) {
|
||||
if (!checkOrNotify()) {
|
||||
return new ArrayList<>();
|
||||
}
|
||||
|
|
@ -83,11 +82,11 @@ public class WordStore extends BaseSyncStore {
|
|||
long longPositionsTime = Timer.stop("cache_long_positions");
|
||||
|
||||
final int minWords = Math.max(minimumWords, 0);
|
||||
final int maxWords = Math.max(maximumWords, minWords);
|
||||
final int maxWords = maximumWords >= 0 ? Math.max(maximumWords, minWords) : maximumWords;
|
||||
final String filter = wordFilter == null ? "" : wordFilter;
|
||||
|
||||
Timer.start("get_positions");
|
||||
String positions = readOps.getSimilarWordPositions(sqlite.getDb(), cancel, language, sequence, filter, minWords);
|
||||
String positions = readOps.getSimilarWordPositions(sqlite.getDb(), cancel, language, sequence, onlyExactSequence, filter, minWords);
|
||||
long positionsTime = Timer.stop("get_positions");
|
||||
|
||||
Timer.start("get_words");
|
||||
|
|
|
|||
|
|
@ -88,7 +88,7 @@ abstract public class CommandHandler extends TextEditingHandler {
|
|||
return;
|
||||
}
|
||||
|
||||
if (mLanguage.isSyllabary()) {
|
||||
if (mLanguage.isTranscribed()) {
|
||||
UI.toastShortSingle(this, R.string.function_add_word_not_available);
|
||||
return;
|
||||
}
|
||||
|
|
|
|||
|
|
@ -36,7 +36,12 @@ public abstract class HotkeyHandler extends CommandHandler {
|
|||
suggestionOps.cancelDelayedAccept();
|
||||
|
||||
if (!suggestionOps.isEmpty()) {
|
||||
if (mInputMode.shouldReplacePreviousSuggestion()) {
|
||||
mInputMode.onReplaceSuggestion(suggestionOps.getCurrent());
|
||||
} else {
|
||||
onAcceptSuggestionManually(suggestionOps.acceptCurrent(), KeyEvent.KEYCODE_ENTER);
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
|
|
@ -192,7 +197,7 @@ public abstract class HotkeyHandler extends CommandHandler {
|
|||
|
||||
|
||||
public boolean onKeyFilterClear(boolean validateOnly) {
|
||||
if (suggestionOps.isEmpty() || mLanguage.isSyllabary()) {
|
||||
if (suggestionOps.isEmpty() || !mInputMode.supportsFiltering()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
|
|
@ -208,12 +213,13 @@ public abstract class HotkeyHandler extends CommandHandler {
|
|||
// References:
|
||||
// - https://github.com/sspanak/tt9/issues/698#issuecomment-2600441061
|
||||
// - https://github.com/sspanak/tt9/issues/418
|
||||
boolean isFilteringOn = mInputMode.isStemFilterFuzzy() || (mInputMode.getSequenceLength() != mInputMode.getWordStem().length());
|
||||
int stemLength = mInputMode.getWordStem().length();
|
||||
boolean isFilteringOn = mInputMode.isStemFilterFuzzy() || (stemLength > 0 && mInputMode.getSequenceLength() != stemLength);
|
||||
|
||||
if (mInputMode.clearWordStem() && isFilteringOn) {
|
||||
mInputMode
|
||||
.setOnSuggestionsUpdated(this::handleSuggestions)
|
||||
.loadSuggestions(suggestionOps.getCurrent(mInputMode.getSequenceLength()));
|
||||
.loadSuggestions(suggestionOps.getCurrent(mLanguage, mInputMode.getSequenceLength()));
|
||||
return true;
|
||||
}
|
||||
|
||||
|
|
@ -229,7 +235,7 @@ public abstract class HotkeyHandler extends CommandHandler {
|
|||
return false;
|
||||
}
|
||||
|
||||
if (mLanguage.isSyllabary()) {
|
||||
if (!mInputMode.supportsFiltering()) {
|
||||
UI.toastShortSingle(this, R.string.function_filter_suggestions_not_available);
|
||||
return true; // prevent the default key action to acknowledge we have processed the event
|
||||
}
|
||||
|
|
@ -244,7 +250,7 @@ public abstract class HotkeyHandler extends CommandHandler {
|
|||
if (repeat && !suggestionOps.get(1).isEmpty()) {
|
||||
filter = suggestionOps.get(1);
|
||||
} else {
|
||||
filter = suggestionOps.getCurrent(mInputMode.getSequenceLength());
|
||||
filter = suggestionOps.getCurrent(mLanguage, mInputMode.getSequenceLength());
|
||||
}
|
||||
|
||||
if (filter.isEmpty()) {
|
||||
|
|
@ -289,10 +295,11 @@ public abstract class HotkeyHandler extends CommandHandler {
|
|||
detectRTL();
|
||||
|
||||
// for languages that do not have ABC or Predictive, make sure we remain in valid state
|
||||
if (!mInputMode.changeLanguage(mLanguage)) {
|
||||
if (mInputMode.changeLanguage(mLanguage)) {
|
||||
mInputMode.clearWordStem();
|
||||
} else {
|
||||
mInputMode = InputMode.getInstance(settings, mLanguage, inputType, textField, determineInputModeId());
|
||||
}
|
||||
mInputMode.clearWordStem();
|
||||
|
||||
getSuggestions(null);
|
||||
statusBar.setText(mInputMode);
|
||||
|
|
|
|||
|
|
@ -64,11 +64,21 @@ abstract public class MainViewHandler extends HotkeyHandler {
|
|||
}
|
||||
|
||||
|
||||
public boolean isAddingWordsSupported() {
|
||||
return mLanguage == null || !mLanguage.isTranscribed();
|
||||
}
|
||||
|
||||
|
||||
public boolean isDragResizeOn() {
|
||||
return dragResize;
|
||||
}
|
||||
|
||||
|
||||
public boolean isFilteringSupported() {
|
||||
return mInputMode.supportsFiltering();
|
||||
}
|
||||
|
||||
|
||||
public boolean isInputLimited() {
|
||||
return inputType.isLimited();
|
||||
}
|
||||
|
|
@ -114,13 +124,8 @@ abstract public class MainViewHandler extends HotkeyHandler {
|
|||
}
|
||||
|
||||
|
||||
public boolean notLanguageSyllabary() {
|
||||
return mLanguage == null || !mLanguage.isSyllabary();
|
||||
}
|
||||
|
||||
|
||||
public String getABCString() {
|
||||
return mLanguage == null || mLanguage.isSyllabary() ? "ABC" : mLanguage.getAbcString().toUpperCase(mLanguage.getLocale());
|
||||
return mLanguage == null ? "ABC" : mLanguage.getAbcString().toUpperCase(mLanguage.getLocale());
|
||||
}
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -21,7 +21,6 @@ import io.github.sspanak.tt9.ime.modes.InputMode;
|
|||
import io.github.sspanak.tt9.ime.modes.InputModeKind;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.languages.LanguageCollection;
|
||||
import io.github.sspanak.tt9.languages.LanguageKind;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
import io.github.sspanak.tt9.ui.UI;
|
||||
import io.github.sspanak.tt9.util.Text;
|
||||
|
|
@ -294,7 +293,7 @@ public abstract class TypingHandler extends KeyPadHandler {
|
|||
}
|
||||
|
||||
allowedInputModes = new ArrayList<>(inputType.determineInputModes(getApplicationContext()));
|
||||
if (LanguageKind.isKorean(mLanguage)) {
|
||||
if (!mLanguage.hasABC()) {
|
||||
allowedInputModes.remove((Integer) InputMode.MODE_ABC);
|
||||
} else if (!settings.getPredictiveMode()) {
|
||||
allowedInputModes.remove((Integer) InputMode.MODE_PREDICTIVE);
|
||||
|
|
@ -381,7 +380,7 @@ public abstract class TypingHandler extends KeyPadHandler {
|
|||
// last key press makes up a compound word like: (it)'s, (I)'ve, l'(oiseau), or it is
|
||||
// just the end of a sentence, like: "word." or "another?"
|
||||
if (mInputMode.shouldAcceptPreviousSuggestion(suggestionOps.getCurrent())) {
|
||||
String lastWord = suggestionOps.acceptPrevious(mInputMode.getSequenceLength());
|
||||
String lastWord = suggestionOps.acceptPrevious(mLanguage, mInputMode.getSequenceLength());
|
||||
onAcceptSuggestionAutomatically(lastWord);
|
||||
}
|
||||
|
||||
|
|
@ -405,7 +404,7 @@ public abstract class TypingHandler extends KeyPadHandler {
|
|||
// Otherwise, put the first suggestion in the text field,
|
||||
// but cut it off to the length of the sequence (how many keys were pressed),
|
||||
// for a more intuitive experience.
|
||||
String trimmedWord = suggestionOps.getCurrent(mInputMode.getSequenceLength());
|
||||
String trimmedWord = suggestionOps.getCurrent(mLanguage, mInputMode.getSequenceLength());
|
||||
appHacks.setComposingTextWithHighlightedStem(trimmedWord, mInputMode);
|
||||
|
||||
forceShowWindow();
|
||||
|
|
|
|||
|
|
@ -8,10 +8,12 @@ import androidx.annotation.Nullable;
|
|||
|
||||
import java.util.ArrayList;
|
||||
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
import io.github.sspanak.tt9.ui.main.ResizableMainView;
|
||||
import io.github.sspanak.tt9.ui.tray.SuggestionsBar;
|
||||
import io.github.sspanak.tt9.util.ConsumerCompat;
|
||||
import io.github.sspanak.tt9.util.Text;
|
||||
|
||||
public class SuggestionOps {
|
||||
@NonNull private final Handler delayedAcceptHandler;
|
||||
|
|
@ -110,12 +112,12 @@ public class SuggestionOps {
|
|||
}
|
||||
|
||||
|
||||
public String acceptPrevious(int sequenceLength) {
|
||||
public String acceptPrevious(Language language, int sequenceLength) {
|
||||
if (sequenceLength <= 0) {
|
||||
set(null);
|
||||
}
|
||||
|
||||
String lastComposingText = getCurrent(sequenceLength - 1);
|
||||
String lastComposingText = getCurrent(language, sequenceLength - 1);
|
||||
commitCurrent(false);
|
||||
return lastComposingText;
|
||||
}
|
||||
|
|
@ -143,17 +145,17 @@ public class SuggestionOps {
|
|||
}
|
||||
|
||||
|
||||
public String getCurrent(int maxLength) {
|
||||
public String getCurrent(Language language, int maxLength) {
|
||||
if (maxLength == 0 || isEmpty()) {
|
||||
return "";
|
||||
}
|
||||
|
||||
String text = getCurrent();
|
||||
if (maxLength > 0 && !text.isEmpty() && text.length() > maxLength) {
|
||||
text = text.substring(0, maxLength);
|
||||
Text text = new Text(language, getCurrent());
|
||||
if (maxLength > 0 && !text.isEmpty() && text.codePointLength() > maxLength) {
|
||||
return text.substringCodePoints(0, maxLength);
|
||||
}
|
||||
|
||||
return text;
|
||||
return text.toString();
|
||||
}
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -42,6 +42,7 @@ abstract public class InputMode {
|
|||
|
||||
|
||||
protected InputMode(SettingsStore settings, InputType inputType) {
|
||||
allowedTextCases.add(CASE_LOWER);
|
||||
isEmailMode = inputType != null && inputType.isEmail() && !inputType.isDefectiveText();
|
||||
this.settings = settings;
|
||||
}
|
||||
|
|
@ -50,7 +51,9 @@ abstract public class InputMode {
|
|||
public static InputMode getInstance(SettingsStore settings, @Nullable Language language, InputType inputType, TextField textField, int mode) {
|
||||
switch (mode) {
|
||||
case MODE_PREDICTIVE:
|
||||
return (LanguageKind.isKorean(language) ? new ModeCheonjiin(settings, inputType, textField) : new ModeWords(settings, language, inputType, textField));
|
||||
if (LanguageKind.isChinese(language)) return new ModePinyin(settings, language, inputType, textField);
|
||||
if (LanguageKind.isKorean(language)) return new ModeCheonjiin(settings, inputType, textField);
|
||||
return new ModeWords(settings, language, inputType, textField);
|
||||
case MODE_ABC:
|
||||
return new ModeABC(settings, language, inputType);
|
||||
case MODE_PASSTHROUGH:
|
||||
|
|
@ -69,6 +72,7 @@ abstract public class InputMode {
|
|||
// Suggestions
|
||||
public void onAcceptSuggestion(@NonNull String word) { onAcceptSuggestion(word, false); }
|
||||
public void onAcceptSuggestion(@NonNull String word, boolean preserveWordList) {}
|
||||
public void onReplaceSuggestion(@NonNull String word) {}
|
||||
|
||||
/**
|
||||
* loadSuggestions
|
||||
|
|
@ -123,6 +127,7 @@ abstract public class InputMode {
|
|||
// Interaction with the IME. Return "true" if it should perform the respective action.
|
||||
public boolean shouldAcceptPreviousSuggestion(String unacceptedText) { return false; }
|
||||
public boolean shouldAcceptPreviousSuggestion(int nextKey, boolean hold) { return false; }
|
||||
public boolean shouldReplacePreviousSuggestion() { return false; }
|
||||
public boolean shouldAddTrailingSpace(boolean isWordAcceptedManually, int nextKey) { return false; }
|
||||
public boolean shouldAddPrecedingSpace() { return false; }
|
||||
public boolean shouldDeletePrecedingSpace() { return false; }
|
||||
|
|
@ -251,4 +256,5 @@ abstract public class InputMode {
|
|||
public boolean isStemFilterFuzzy() { return false; }
|
||||
public String getWordStem() { return ""; }
|
||||
public boolean setWordStem(String stem, boolean exact) { return false; }
|
||||
public boolean supportsFiltering() { return false; }
|
||||
}
|
||||
|
|
|
|||
|
|
@ -92,7 +92,7 @@ class ModeABC extends InputMode {
|
|||
|
||||
@Override
|
||||
public boolean changeLanguage(@Nullable Language newLanguage) {
|
||||
if (newLanguage != null && newLanguage.isSyllabary()) {
|
||||
if (newLanguage != null && newLanguage.isTranscribed()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
|
|
|
|||
|
|
@ -48,13 +48,15 @@ class ModeCheonjiin extends InputMode {
|
|||
|
||||
SPECIAL_CHAR_SEQUENCE_PREFIX = "11";
|
||||
|
||||
super.setLanguage(LanguageCollection.getLanguage(LanguageKind.KOREAN));
|
||||
|
||||
autoSpace = new AutoSpace(settings).setLanguage(language);
|
||||
digitSequence = "";
|
||||
allowedTextCases.add(CASE_LOWER);
|
||||
this.inputType = inputType;
|
||||
this.textField = textField;
|
||||
|
||||
initPredictions();
|
||||
setLanguage(LanguageCollection.getLanguage(LanguageKind.KOREAN));
|
||||
setSpecialCharacterConstants();
|
||||
|
||||
if (isEmailMode) {
|
||||
|
|
@ -64,8 +66,6 @@ class ModeCheonjiin extends InputMode {
|
|||
} else {
|
||||
setCustomSpecialCharacters();
|
||||
}
|
||||
|
||||
autoSpace = new AutoSpace(settings).setLanguage(language);
|
||||
}
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -0,0 +1,210 @@
|
|||
package io.github.sspanak.tt9.ime.modes;
|
||||
|
||||
import androidx.annotation.NonNull;
|
||||
import androidx.annotation.Nullable;
|
||||
|
||||
import io.github.sspanak.tt9.hacks.InputType;
|
||||
import io.github.sspanak.tt9.ime.helpers.TextField;
|
||||
import io.github.sspanak.tt9.ime.modes.predictions.IdeogramPredictions;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.languages.LanguageKind;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
import io.github.sspanak.tt9.util.Logger;
|
||||
import io.github.sspanak.tt9.util.Text;
|
||||
import io.github.sspanak.tt9.util.TextTools;
|
||||
|
||||
public class ModeIdeograms extends ModeWords {
|
||||
private static final String LOG_TAG = ModeIdeograms.class.getSimpleName();
|
||||
|
||||
private boolean isFiltering = false;
|
||||
|
||||
|
||||
protected ModeIdeograms(SettingsStore settings, Language lang, InputType inputType, TextField textField) {
|
||||
super(settings, lang, inputType, textField);
|
||||
}
|
||||
|
||||
|
||||
@Override public void determineNextWordTextCase() {}
|
||||
@Override protected String adjustSuggestionTextCase(String word, int newTextCase) { return word; }
|
||||
|
||||
|
||||
@Override
|
||||
protected void initPredictions() {
|
||||
predictions = new IdeogramPredictions(settings, textField);
|
||||
predictions.setWordsChangedHandler(this::onPredictions);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean changeLanguage(@Nullable Language newLanguage) {
|
||||
if (newLanguage != null && !newLanguage.isTranscribed() || LanguageKind.isKorean(newLanguage)) {
|
||||
return false;
|
||||
}
|
||||
|
||||
setLanguage(newLanguage);
|
||||
return true;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public void reset() {
|
||||
super.reset();
|
||||
isFiltering = false;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
protected void onPredictions() {
|
||||
if (language.hasTranscriptionsEmbedded()) {
|
||||
if (isFiltering) {
|
||||
((IdeogramPredictions) predictions).stripNativeWords();
|
||||
} else {
|
||||
((IdeogramPredictions) predictions).stripTranscriptions();
|
||||
}
|
||||
}
|
||||
|
||||
if (!isFiltering) {
|
||||
// We can reorder by pairs only after stripping the transcriptions, if any.
|
||||
// Otherwise, the input field words will not match with any pair.
|
||||
((IdeogramPredictions) predictions).orderByPairs();
|
||||
}
|
||||
|
||||
super.onPredictions();
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public void onAcceptSuggestion(@NonNull String currentWord, boolean preserveWords) {
|
||||
if (currentWord.isEmpty() || new Text(currentWord).isNumeric()) {
|
||||
reset();
|
||||
Logger.i(LOG_TAG, "Current word is empty or numeric. Nothing to accept.");
|
||||
return;
|
||||
}
|
||||
|
||||
if (isFiltering) {
|
||||
isFiltering = false;
|
||||
stem = currentWord;
|
||||
loadSuggestions("");
|
||||
return;
|
||||
}
|
||||
|
||||
try {
|
||||
String latinWord = ((IdeogramPredictions) predictions).getTranscription(currentWord);
|
||||
String digits = language.getDigitSequenceForWord(latinWord);
|
||||
((IdeogramPredictions) predictions).onAcceptTranscription(currentWord, latinWord, digits);
|
||||
} catch (Exception e) {
|
||||
Logger.e(LOG_TAG, "Failed incrementing priority of word: '" + currentWord + "'. " + e.getMessage());
|
||||
}
|
||||
|
||||
int len = digitSequence.length();
|
||||
if (preserveWords && len >= 2) {
|
||||
digitSequence = digitSequence.substring(len - 1);
|
||||
loadSuggestions("");
|
||||
} else {
|
||||
reset();
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean shouldAcceptPreviousSuggestion(String s) {
|
||||
return
|
||||
!digitSequence.isEmpty()
|
||||
&& predictions.noDbWords()
|
||||
&& !digitSequence.equals(EMOJI_SEQUENCE)
|
||||
&& !digitSequence.equals(PUNCTUATION_SEQUENCE)
|
||||
&& !digitSequence.equals(SPECIAL_CHAR_SEQUENCE);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean shouldAcceptPreviousSuggestion(int nextKey, boolean hold) {
|
||||
if (digitSequence.isEmpty()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
if (super.shouldAcceptPreviousSuggestion(nextKey, hold)) {
|
||||
return true;
|
||||
}
|
||||
|
||||
String nextSequence = digitSequence + (char)(nextKey + '0');
|
||||
|
||||
return
|
||||
TextTools.containsOtherThan1(nextSequence)
|
||||
&& (
|
||||
nextSequence.endsWith(EMOJI_SEQUENCE) || nextSequence.startsWith(EMOJI_SEQUENCE) ||
|
||||
nextSequence.endsWith(PUNCTUATION_SEQUENCE) || nextSequence.startsWith(PUNCTUATION_SEQUENCE)
|
||||
);
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* When we want to filter by a Latin transcription, we must have discarded it from the text field,
|
||||
* then give it to this method. It will filter the suggestions and show only the ones that match
|
||||
* the given Latin word.
|
||||
*/
|
||||
@Override
|
||||
public void onReplaceSuggestion(@NonNull String word) {
|
||||
if (word.isEmpty() || new Text(word).isNumeric()) {
|
||||
reset();
|
||||
Logger.i(LOG_TAG, "Can not replace an empty or numeric word.");
|
||||
return;
|
||||
}
|
||||
|
||||
isFiltering = false;
|
||||
stem = word;
|
||||
loadSuggestions("");
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* This should be called before accepting a word. It says whether we should discard the current
|
||||
* word. Discarding it means we want to erase it from the text field and instead display a
|
||||
* filtered list of suggestions that matches the word. If we don't discard it, usually we should
|
||||
* accept it.
|
||||
*/
|
||||
@Override
|
||||
public boolean shouldReplacePreviousSuggestion() {
|
||||
return isFiltering;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean setWordStem(String newStem, boolean fromScrolling) {
|
||||
if (!supportsFiltering()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
if (!fromScrolling) {
|
||||
isFiltering = true;
|
||||
} else if (isFiltering) {
|
||||
stem = newStem;
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean clearWordStem() {
|
||||
if (!supportsFiltering()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
isFiltering = false;
|
||||
stem = "";
|
||||
return true;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean supportsFiltering() {
|
||||
return language.hasTranscriptionsEmbedded();
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean isStemFilterFuzzy() {
|
||||
return isFiltering;
|
||||
}
|
||||
}
|
||||
|
|
@ -10,7 +10,6 @@ class ModePassthrough extends InputMode {
|
|||
protected ModePassthrough(SettingsStore settings, InputType inputType) {
|
||||
super(settings, inputType);
|
||||
reset();
|
||||
allowedTextCases.add(CASE_LOWER);
|
||||
}
|
||||
|
||||
@Override public int getId() { return MODE_PASSTHROUGH; }
|
||||
|
|
|
|||
|
|
@ -0,0 +1,57 @@
|
|||
package io.github.sspanak.tt9.ime.modes;
|
||||
|
||||
import io.github.sspanak.tt9.hacks.InputType;
|
||||
import io.github.sspanak.tt9.ime.helpers.TextField;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
import io.github.sspanak.tt9.util.chars.Characters;
|
||||
|
||||
public class ModePinyin extends ModeIdeograms {
|
||||
boolean ignoreNextSpace = false;
|
||||
|
||||
|
||||
protected ModePinyin(SettingsStore settings, Language lang, InputType inputType, TextField textField) {
|
||||
super(settings, lang, inputType, textField);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
protected void onNumberPress(int number) {
|
||||
if (ignoreNextSpace && number == SPECIAL_CHAR_SEQUENCE.charAt(0) - '0') {
|
||||
ignoreNextSpace = false;
|
||||
return;
|
||||
}
|
||||
|
||||
ignoreNextSpace = false;
|
||||
super.onNumberPress(number);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
protected void onNumberHold(int number) {
|
||||
ignoreNextSpace = false;
|
||||
super.onNumberHold(number);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean shouldAcceptPreviousSuggestion(int nextKey, boolean hold) {
|
||||
// In East Asian languages, 0-key must accept the current word, or type a space when there is no word.
|
||||
if (!digitSequence.isEmpty() && !digitSequence.endsWith(SPECIAL_CHAR_SEQUENCE) && nextKey == SPECIAL_CHAR_SEQUENCE.charAt(0) - '0') {
|
||||
ignoreNextSpace = true;
|
||||
}
|
||||
|
||||
return super.shouldAcceptPreviousSuggestion(nextKey, hold);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
protected String getPreferredChar() {
|
||||
final String preferredChar = settings.getDoubleZeroChar();
|
||||
return switch (preferredChar) {
|
||||
case "." -> Characters.ZH_FULL_STOP;
|
||||
case "," -> Characters.ZH_COMMA_LIST;
|
||||
default -> preferredChar;
|
||||
};
|
||||
}
|
||||
}
|
||||
|
|
@ -27,7 +27,7 @@ class ModeWords extends ModeCheonjiin {
|
|||
|
||||
// stem filter
|
||||
private boolean isStemFuzzy = false;
|
||||
private String stem = "";
|
||||
protected String stem = "";
|
||||
|
||||
// text analysis tools
|
||||
private final AutoTextCase autoTextCase;
|
||||
|
|
@ -110,12 +110,18 @@ class ModeWords extends ModeCheonjiin {
|
|||
|
||||
@Override
|
||||
public boolean changeLanguage(@Nullable Language newLanguage) {
|
||||
if (newLanguage != null && newLanguage.isSyllabary()) {
|
||||
if (newLanguage != null && newLanguage.isTranscribed()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
super.setLanguage(newLanguage);
|
||||
setLanguage(newLanguage);
|
||||
return true;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
protected void setLanguage(@Nullable Language newLanguage) {
|
||||
super.setLanguage(newLanguage);
|
||||
autoSpace.setLanguage(language);
|
||||
|
||||
allowedTextCases.clear();
|
||||
|
|
@ -124,14 +130,11 @@ class ModeWords extends ModeCheonjiin {
|
|||
allowedTextCases.add(CASE_CAPITALIZE);
|
||||
allowedTextCases.add(CASE_UPPER);
|
||||
}
|
||||
|
||||
return true;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean recompose(String word) {
|
||||
if (!language.hasSpaceBetweenWords() || language.isSyllabary()) {
|
||||
if (!language.hasSpaceBetweenWords() || language.isTranscribed()) {
|
||||
return false;
|
||||
}
|
||||
|
||||
|
|
@ -258,6 +261,12 @@ class ModeWords extends ModeCheonjiin {
|
|||
}
|
||||
|
||||
|
||||
@Override
|
||||
public boolean supportsFiltering() {
|
||||
return true;
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* loadSuggestions
|
||||
* Loads the possible list of suggestions for the current digitSequence. "currentWord" is used
|
||||
|
|
@ -284,7 +293,7 @@ class ModeWords extends ModeCheonjiin {
|
|||
private boolean loadPreferredChar() {
|
||||
if (digitSequence.startsWith(NaturalLanguage.PREFERRED_CHAR_SEQUENCE)) {
|
||||
suggestions.clear();
|
||||
suggestions.add(settings.getDoubleZeroChar());
|
||||
suggestions.add(getPreferredChar());
|
||||
return true;
|
||||
}
|
||||
|
||||
|
|
@ -292,6 +301,11 @@ class ModeWords extends ModeCheonjiin {
|
|||
}
|
||||
|
||||
|
||||
protected String getPreferredChar() {
|
||||
return settings.getDoubleZeroChar();
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* onAcceptSuggestion
|
||||
* Bring this word up in the suggestions list next time and if necessary preserves the suggestion list
|
||||
|
|
|
|||
|
|
@ -39,7 +39,7 @@ public class AutoSpace {
|
|||
public AutoSpace setLanguage(Language lang) {
|
||||
language = language == null ? new NullLanguage() : lang;
|
||||
isLanguageFrench = LanguageKind.isFrench(lang);
|
||||
isLanguageWithAlphabet = !language.isSyllabary();
|
||||
isLanguageWithAlphabet = !language.isTranscribed();
|
||||
isLanguageWithSpaceBetweenWords = language.hasSpaceBetweenWords();
|
||||
return this;
|
||||
}
|
||||
|
|
|
|||
|
|
@ -0,0 +1,200 @@
|
|||
package io.github.sspanak.tt9.ime.modes.predictions;
|
||||
|
||||
import androidx.annotation.NonNull;
|
||||
|
||||
import java.util.ArrayList;
|
||||
import java.util.Collections;
|
||||
import java.util.HashSet;
|
||||
|
||||
import io.github.sspanak.tt9.db.DataStore;
|
||||
import io.github.sspanak.tt9.ime.helpers.TextField;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
import io.github.sspanak.tt9.util.TextTools;
|
||||
|
||||
public class IdeogramPredictions extends WordPredictions {
|
||||
private boolean isTranscriptionFilterAllowed = false;
|
||||
private String lastTypedWord = "";
|
||||
@NonNull protected ArrayList<String> transcriptions = new ArrayList<>();
|
||||
|
||||
|
||||
public IdeogramPredictions(SettingsStore settings, TextField textField) {
|
||||
super(settings, textField);
|
||||
minWords = 1;
|
||||
maxWords = -1;
|
||||
onlyExactMatches = true;
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public Predictions setLanguage(@NonNull Language language) {
|
||||
isTranscriptionFilterAllowed = language.hasTranscriptionsEmbedded();
|
||||
return super.setLanguage(language);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
public void load() {
|
||||
transcriptions.clear();
|
||||
super.load();
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
protected void onDbWords(ArrayList<String> dbWords, boolean isRetryAllowed) {
|
||||
transcriptions = onlyExactMatches ? reduceFuzzyMatches(dbWords, SettingsStore.SUGGESTIONS_MAX) : dbWords;
|
||||
words = new ArrayList<>(transcriptions);
|
||||
areThereDbWords = !words.isEmpty();
|
||||
onWordsChanged.run();
|
||||
}
|
||||
|
||||
|
||||
public void onAcceptTranscription(String word, String transcription, String sequence) {
|
||||
super.onAccept(transcription + word, sequence);
|
||||
}
|
||||
|
||||
|
||||
@Override
|
||||
@NonNull
|
||||
protected String getWordBeforeCursor(@NonNull String currentWord) {
|
||||
int currentWordLength = currentWord.length();
|
||||
int lastWordLength = lastTypedWord.length();
|
||||
int requiredTextLength = currentWordLength + lastWordLength;
|
||||
String text = textField.getStringBeforeCursor(requiredTextLength);
|
||||
// Logger.d("LOG_TAG", "====+> previous string: " + text);
|
||||
|
||||
return lastWordLength < text.length() ? text.substring(0, lastWordLength) : "";
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Tries to do a best guess what is the previous word and pairs it with the incoming one. Guessing
|
||||
* is because East Asian languages do not have spaces between words, so we try to match the
|
||||
* last typed or just give up.
|
||||
*/
|
||||
@Override
|
||||
protected void pairWithPreviousWord(@NonNull String word, @NonNull String sequence) {
|
||||
if (language.hasSpaceBetweenWords()) {
|
||||
super.pairWithPreviousWord(word, sequence);
|
||||
return;
|
||||
}
|
||||
|
||||
if (!settings.getPredictWordPairs() || sequence.length() != digitSequence.length()) {
|
||||
// Logger.d("LOG_TAG", "====+> sequence length mismatch: " + sequence.length() + " != " + digitSequence.length());
|
||||
return;
|
||||
}
|
||||
|
||||
int latinEnd = TextTools.lastIndexOfLatin(word);
|
||||
String nativeWord = latinEnd < 0 || latinEnd >= word.length() ? word : word.substring(latinEnd + 1);
|
||||
|
||||
if (lastTypedWord.isEmpty() || (!words.isEmpty() && nativeWord.equals(words.get(0)))) {
|
||||
lastTypedWord = nativeWord;
|
||||
// Logger.d("LOG_TAG", "====+> Will not pair the first word. native word: " + nativeWord + " first suggestion: " + words.get(0));
|
||||
// if (lastTypedWord.isEmpty()) {
|
||||
// Logger.d("LOG_TAG", "====+> No previous word to pair with: " + lastTypedWord);
|
||||
// }
|
||||
|
||||
return;
|
||||
}
|
||||
|
||||
String previousWord = getWordBeforeCursor(nativeWord);
|
||||
if (previousWord.equals(lastTypedWord)) {
|
||||
// Logger.d("LOG_TAG", "====+> Pairing words: " + previousWord + " + " + nativeWord);
|
||||
DataStore.addWordPair(language, previousWord, nativeWord, sequence);
|
||||
// } else {
|
||||
// Logger.d("LOG_TAG", "===> Last word mismatch: " + previousWord + " != " + lastTypedWord + ". Not pairing.");
|
||||
}
|
||||
|
||||
lastTypedWord = nativeWord;
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Keeps all exact matches and the first n fuzzy matches. Unlike Latin- or Cyrillic-based languages,
|
||||
* ideograms do not "start with" a sequence of characters, so fuzzy matches have little value.
|
||||
* Just keep some of them, in case there are no exact matches.
|
||||
*/
|
||||
@NonNull
|
||||
public ArrayList<String> reduceFuzzyMatches(ArrayList<String> allWords, int maxWords) {
|
||||
if (allWords.isEmpty()) {
|
||||
return allWords;
|
||||
}
|
||||
|
||||
ArrayList<String> shortWords = new ArrayList<>();
|
||||
final int MAX_LENGTH = Math.max(digitSequence.length() + 1, allWords.get(0).length());
|
||||
|
||||
for (int i = 0, longWords = 0, end = allWords.size(); i < end; i++) {
|
||||
String word = allWords.get(i);
|
||||
int trueLength = isTranscriptionFilterAllowed ? TextTools.lastIndexOfLatin(word) : word.length();
|
||||
|
||||
if (trueLength < MAX_LENGTH) {
|
||||
shortWords.add(word);
|
||||
}
|
||||
|
||||
if (trueLength >= MAX_LENGTH && longWords <= maxWords) {
|
||||
longWords++;
|
||||
shortWords.add(word);
|
||||
}
|
||||
}
|
||||
|
||||
return shortWords;
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Removes the native words and keeps only the unique transcriptions. Directly modifies the words
|
||||
* list, but the original is preserved in this.transcriptions.
|
||||
* Example operation: [SHIWU食物, SHIWU事物, SHIWU事务, SHIZU十足] -> [SHIWU, SHIZU]
|
||||
*/
|
||||
public void stripNativeWords() {
|
||||
HashSet<String> uniqueTranscriptions = new HashSet<>();
|
||||
|
||||
for (int i = 0; i < transcriptions.size(); i++) {
|
||||
String transcription = transcriptions.get(i);
|
||||
int firstNative = TextTools.lastIndexOfLatin(transcription) + 1;
|
||||
uniqueTranscriptions.add(
|
||||
firstNative < 1 || firstNative >= transcription.length() ? transcription : transcription.substring(0, firstNative)
|
||||
);
|
||||
}
|
||||
|
||||
words.clear();
|
||||
words.addAll(uniqueTranscriptions);
|
||||
Collections.sort(words);
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Removes the Latin transcriptions from native words. Directly modifies the words list, but the
|
||||
* original is preserved in this.transcriptions.
|
||||
* Example operation: [SHIWU食物, SHIZU十足] -> [食物, 十足]
|
||||
*/
|
||||
public void stripTranscriptions() {
|
||||
words.clear();
|
||||
for (int i = 0; i < transcriptions.size(); i++) {
|
||||
String transcription = transcriptions.get(i);
|
||||
int firstNative = TextTools.lastIndexOfLatin(transcription) + 1;
|
||||
words.add(firstNative >= transcription.length() ? transcription : transcription.substring(firstNative));
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Similar to "stripNativeWords()", but finds and returns the transcription of the given word.
|
||||
* Returns an empty string if the word is not in the current suggestion list.
|
||||
*/
|
||||
@NonNull
|
||||
public String getTranscription(@NonNull String word) {
|
||||
for (String w : transcriptions) {
|
||||
if (w.endsWith(word)) {
|
||||
return w.replace(word, "");
|
||||
}
|
||||
}
|
||||
|
||||
return "";
|
||||
}
|
||||
|
||||
|
||||
public void orderByPairs() {
|
||||
words = rearrangeByPairFrequency(words);
|
||||
}
|
||||
}
|
||||
|
|
@ -5,7 +5,6 @@ import androidx.annotation.NonNull;
|
|||
import java.util.ArrayList;
|
||||
|
||||
import io.github.sspanak.tt9.db.DataStore;
|
||||
import io.github.sspanak.tt9.db.words.WordStore;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.languages.NullLanguage;
|
||||
import io.github.sspanak.tt9.preferences.settings.SettingsStore;
|
||||
|
|
@ -41,7 +40,7 @@ abstract public class Predictions {
|
|||
}
|
||||
|
||||
|
||||
public Predictions setLanguage(Language language) {
|
||||
public Predictions setLanguage(@NonNull Language language) {
|
||||
this.language = language;
|
||||
return this;
|
||||
}
|
||||
|
|
@ -110,7 +109,8 @@ abstract public class Predictions {
|
|||
(dbWords) -> onDbWords(dbWords, isRetryAllowed()),
|
||||
language,
|
||||
digitSequence,
|
||||
onlyExactMatches ? WordStore.FILTER_EXACT_MATCHES_ONLY : stem,
|
||||
onlyExactMatches,
|
||||
stem,
|
||||
minWords,
|
||||
maxWords
|
||||
);
|
||||
|
|
|
|||
|
|
@ -1,5 +1,7 @@
|
|||
package io.github.sspanak.tt9.ime.modes.predictions;
|
||||
|
||||
import androidx.annotation.NonNull;
|
||||
|
||||
import java.util.ArrayList;
|
||||
|
||||
import io.github.sspanak.tt9.db.DataStore;
|
||||
|
|
@ -11,7 +13,7 @@ import io.github.sspanak.tt9.util.TextTools;
|
|||
import io.github.sspanak.tt9.util.chars.Characters;
|
||||
|
||||
public class WordPredictions extends Predictions {
|
||||
private final TextField textField;
|
||||
protected final TextField textField;
|
||||
private LocaleWordsSorter localeWordsSorter;
|
||||
|
||||
private String inputWord;
|
||||
|
|
@ -30,7 +32,7 @@ public class WordPredictions extends Predictions {
|
|||
|
||||
|
||||
@Override
|
||||
public Predictions setLanguage(Language language) {
|
||||
public Predictions setLanguage(@NonNull Language language) {
|
||||
super.setLanguage(language);
|
||||
localeWordsSorter = new LocaleWordsSorter(language);
|
||||
|
||||
|
|
@ -67,6 +69,7 @@ public class WordPredictions extends Predictions {
|
|||
},
|
||||
language,
|
||||
digitSequence.substring(1),
|
||||
onlyExactMatches,
|
||||
stem.length() > 1 ? stem.substring(1) : "",
|
||||
SettingsStore.SUGGESTIONS_MIN,
|
||||
SettingsStore.SUGGESTIONS_MAX
|
||||
|
|
@ -262,14 +265,16 @@ public class WordPredictions extends Predictions {
|
|||
return;
|
||||
}
|
||||
|
||||
// Second condition note: If the accepted word is longer than the sequence, it is some different word,
|
||||
// not a textonym of the fist suggestion. We don't need to store it.
|
||||
if (settings.getPredictWordPairs() && word.length() == digitSequence.length()) {
|
||||
DataStore.addWordPair(language, textField.getWordBeforeCursor(language, 1, true), word, sequence);
|
||||
pairWithPreviousWord(word, sequence);
|
||||
makeTopWord(word, sequence);
|
||||
}
|
||||
|
||||
// Update the priority only if the user has selected the word, not when we have enforced it
|
||||
// because it is in a popular word pair.
|
||||
|
||||
/**
|
||||
* Update the priority only if the user has selected the word, not when we have enforced it
|
||||
* because it is in a popular word pair.
|
||||
*/
|
||||
protected void makeTopWord(String word, String sequence) {
|
||||
if (!word.equals(lastEnforcedTopWord)) {
|
||||
DataStore.makeTopWord(language, word, sequence);
|
||||
}
|
||||
|
|
@ -283,7 +288,7 @@ public class WordPredictions extends Predictions {
|
|||
* "onAccept()", we have remembered the "am" comes after "I" and "an" comes after "am", we will
|
||||
* not suggest the textonyms "am" or "an" twice (depending on which has the highest frequency).
|
||||
*/
|
||||
private ArrayList<String> rearrangeByPairFrequency(ArrayList<String> words) {
|
||||
protected ArrayList<String> rearrangeByPairFrequency(ArrayList<String> words) {
|
||||
lastEnforcedTopWord = "";
|
||||
|
||||
if (!settings.getPredictWordPairs() || words.size() < 2) {
|
||||
|
|
@ -291,7 +296,7 @@ public class WordPredictions extends Predictions {
|
|||
}
|
||||
|
||||
ArrayList<String> rearrangedWords = new ArrayList<>();
|
||||
String penultimateWord = textField.getWordBeforeCursor(language, 1, true);
|
||||
String penultimateWord = getWordBeforeCursor(words.get(0));
|
||||
|
||||
String pairWord = DataStore.getWord2(language, penultimateWord, digitSequence);
|
||||
int morePopularIndex = TextTools.indexOfIgnoreCase(words, pairWord);
|
||||
|
|
@ -310,4 +315,26 @@ public class WordPredictions extends Predictions {
|
|||
|
||||
return rearrangedWords;
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Pairs the given word and its digit sequence to the last word in the text field.
|
||||
* Second condition note: If the accepted word is longer than the sequence, it is some different word,
|
||||
* not a textonym of the fist suggestion. We don't need to store it.
|
||||
*/
|
||||
protected void pairWithPreviousWord(@NonNull String word, @NonNull String sequence) {
|
||||
if (settings.getPredictWordPairs() && sequence.length() == digitSequence.length()) {
|
||||
DataStore.addWordPair(language, getWordBeforeCursor(word), word, sequence);
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Returns the last word in the text field. The way of finding it depends on the language, so
|
||||
* we have a separate method for that.
|
||||
*/
|
||||
@NonNull
|
||||
protected String getWordBeforeCursor(@NonNull String currentWord) {
|
||||
return textField.getWordBeforeCursor(language, 1, true);
|
||||
}
|
||||
}
|
||||
|
|
|
|||
|
|
@ -13,11 +13,13 @@ abstract public class Language {
|
|||
protected String code;
|
||||
protected String currency;
|
||||
protected String dictionaryFile;
|
||||
protected Locale locale = Locale.ROOT;
|
||||
protected String name;
|
||||
protected boolean hasABC = true;
|
||||
protected boolean hasSpaceBetweenWords = true;
|
||||
protected boolean hasUpperCase = true;
|
||||
protected boolean isSyllabary = false;
|
||||
protected boolean hasTranscriptionsEmbedded = false;
|
||||
protected boolean isTranscribed = false;
|
||||
protected Locale locale = Locale.ROOT;
|
||||
protected String name;
|
||||
|
||||
|
||||
public int getId() {
|
||||
|
|
@ -63,6 +65,10 @@ abstract public class Language {
|
|||
return name;
|
||||
}
|
||||
|
||||
final public boolean hasABC() {
|
||||
return hasABC;
|
||||
}
|
||||
|
||||
final public boolean hasSpaceBetweenWords() {
|
||||
return hasSpaceBetweenWords;
|
||||
}
|
||||
|
|
@ -71,8 +77,12 @@ abstract public class Language {
|
|||
return hasUpperCase;
|
||||
}
|
||||
|
||||
final public boolean isSyllabary() {
|
||||
return isSyllabary;
|
||||
final public boolean hasTranscriptionsEmbedded() {
|
||||
return hasTranscriptionsEmbedded;
|
||||
}
|
||||
|
||||
final public boolean isTranscribed() {
|
||||
return isTranscribed;
|
||||
}
|
||||
|
||||
@NonNull
|
||||
|
|
|
|||
|
|
@ -24,9 +24,11 @@ public class LanguageDefinition {
|
|||
public String abcString = "";
|
||||
public String currency = "";
|
||||
public String dictionaryFile = "";
|
||||
public boolean filterBySounds = false;
|
||||
public boolean hasABC = true;
|
||||
public boolean hasSpaceBetweenWords = true;
|
||||
public boolean hasUpperCase = true;
|
||||
public boolean isSyllabary = false;
|
||||
public boolean isTranscribed = false;
|
||||
public final ArrayList<ArrayList<String>> layout = new ArrayList<>();
|
||||
public String locale = "";
|
||||
public String name = "";
|
||||
|
|
@ -134,6 +136,12 @@ public class LanguageDefinition {
|
|||
case "dictionaryFile":
|
||||
dictionaryFile = value.replaceFirst("\\.\\w+$", "." + BuildConfig.DICTIONARY_EXTENSION);
|
||||
return;
|
||||
case "filterBySounds":
|
||||
filterBySounds = parseYamlBoolean(value);
|
||||
return;
|
||||
case "hasABC":
|
||||
hasABC = parseYamlBoolean(value);
|
||||
return;
|
||||
case "hasSpaceBetweenWords":
|
||||
hasSpaceBetweenWords = parseYamlBoolean(value);
|
||||
return;
|
||||
|
|
@ -141,7 +149,7 @@ public class LanguageDefinition {
|
|||
hasUpperCase = parseYamlBoolean(value);
|
||||
return;
|
||||
case "sounds":
|
||||
isSyllabary = true;
|
||||
isTranscribed = true;
|
||||
return;
|
||||
case "locale":
|
||||
locale = value;
|
||||
|
|
|
|||
|
|
@ -11,6 +11,7 @@ public class LanguageKind {
|
|||
|
||||
public static boolean isArabic(Language language) { return language != null && language.getId() == 502337; }
|
||||
public static boolean isEnglish(Language language) { return language != null && language.getLocale().equals(Locale.ENGLISH); }
|
||||
public static boolean isChinese(Language language) { return language != null && language.getId() == 462106; }
|
||||
public static boolean isFarsi(Language language) { return language != null && language.getId() == 599078; }
|
||||
public static boolean isFrench(Language language) { return language != null && language.getId() == 596550; }
|
||||
public static boolean isGreek(Language language) { return language != null && language.getId() == 597381; }
|
||||
|
|
|
|||
|
|
@ -10,11 +10,10 @@ import java.util.Map;
|
|||
|
||||
import io.github.sspanak.tt9.languages.exceptions.InvalidLanguageCharactersException;
|
||||
import io.github.sspanak.tt9.util.Text;
|
||||
import io.github.sspanak.tt9.util.TextTools;
|
||||
import io.github.sspanak.tt9.util.chars.Characters;
|
||||
|
||||
|
||||
public class NaturalLanguage extends Language implements Comparable<NaturalLanguage> {
|
||||
public class NaturalLanguage extends TranscribedLanguage implements Comparable<NaturalLanguage> {
|
||||
final public static String SPECIAL_CHAR_KEY = "0";
|
||||
final public static String PUNCTUATION_KEY = "1";
|
||||
final public static String PREFERRED_CHAR_SEQUENCE = "00";
|
||||
|
|
@ -34,9 +33,11 @@ public class NaturalLanguage extends Language implements Comparable<NaturalLangu
|
|||
lang.abcString = definition.abcString.isEmpty() ? null : definition.abcString;
|
||||
lang.currency = definition.currency;
|
||||
lang.dictionaryFile = definition.getDictionaryFile();
|
||||
lang.hasABC = definition.hasABC;
|
||||
lang.hasSpaceBetweenWords = definition.hasSpaceBetweenWords;
|
||||
lang.hasUpperCase = definition.hasUpperCase;
|
||||
lang.isSyllabary = definition.isSyllabary;
|
||||
lang.hasTranscriptionsEmbedded = definition.filterBySounds;
|
||||
lang.isTranscribed = definition.isTranscribed;
|
||||
lang.name = definition.name.isEmpty() ? lang.name : definition.name;
|
||||
lang.numerals = definition.numerals;
|
||||
lang.setLocale(definition);
|
||||
|
|
@ -86,6 +87,7 @@ public class NaturalLanguage extends Language implements Comparable<NaturalLangu
|
|||
specialChars.put(SPECIAL_CHARS_PLACEHOLDER, Characters.Special);
|
||||
specialChars.put(PUNCTUATION_PLACEHOLDER, Characters.PunctuationEnglish);
|
||||
specialChars.put(PUNCTUATION_PLACEHOLDER + "_AR", Characters.PunctuationArabic);
|
||||
specialChars.put(PUNCTUATION_PLACEHOLDER + "_ZH", Characters.PunctuationChinese);
|
||||
specialChars.put(PUNCTUATION_PLACEHOLDER + "_FA", Characters.PunctuationFarsi);
|
||||
specialChars.put(PUNCTUATION_PLACEHOLDER + "_FR", Characters.PunctuationFrench);
|
||||
specialChars.put(PUNCTUATION_PLACEHOLDER + "_DE", Characters.PunctuationGerman);
|
||||
|
|
@ -257,7 +259,7 @@ public class NaturalLanguage extends Language implements Comparable<NaturalLangu
|
|||
if (
|
||||
word == null
|
||||
|| word.isEmpty()
|
||||
|| (isSyllabary && LanguageKind.isKorean(this) && TextTools.isHangul(word))
|
||||
|| (super.isValidWord(word))
|
||||
|| (word.length() == 1 && Character.isDigit(word.charAt(0)))
|
||||
) {
|
||||
return true;
|
||||
|
|
|
|||
|
|
@ -0,0 +1,16 @@
|
|||
package io.github.sspanak.tt9.languages;
|
||||
|
||||
import io.github.sspanak.tt9.util.TextTools;
|
||||
|
||||
abstract class TranscribedLanguage extends Language {
|
||||
|
||||
@Override
|
||||
public boolean isValidWord(String word) {
|
||||
if (!isTranscribed) {
|
||||
return false;
|
||||
}
|
||||
|
||||
return (LanguageKind.isKorean(this) && TextTools.isHangul(word))
|
||||
|| (LanguageKind.isChinese(this) && TextTools.isChinese(word));
|
||||
}
|
||||
}
|
||||
|
|
@ -21,8 +21,8 @@ class ItemSelectZeroKeyCharacter extends ItemDropDown {
|
|||
|
||||
public ItemSelectZeroKeyCharacter populate() {
|
||||
LinkedHashMap<String, String> items = new LinkedHashMap<>();
|
||||
items.put(".", context.getString(R.string.char_dot));
|
||||
items.put(",", context.getString(R.string.char_comma));
|
||||
items.put(".", ".");
|
||||
items.put(",", ",");
|
||||
items.put("\\n", context.getString(R.string.char_newline)); // SharedPreferences return a corrupted string when using the real "\n"... :(
|
||||
items.put(" ", context.getString(R.string.char_space));
|
||||
|
||||
|
|
|
|||
|
|
@ -28,7 +28,7 @@ public class SoftKeyAddWord extends BaseSoftKeyWithIcons {
|
|||
@Override
|
||||
public void render() {
|
||||
if (tt9 != null) {
|
||||
setEnabled(!tt9.isVoiceInputActive() && tt9.notLanguageSyllabary() && !tt9.isTextEditingActive());
|
||||
setEnabled(!tt9.isVoiceInputActive() && tt9.isAddingWordsSupported() && !tt9.isTextEditingActive());
|
||||
}
|
||||
super.render();
|
||||
}
|
||||
|
|
|
|||
|
|
@ -60,7 +60,7 @@ public class SoftKeyFilter extends BaseSoftKeyWithIcons {
|
|||
&& !tt9.isVoiceInputActive()
|
||||
&& (
|
||||
LanguageKind.isKorean(tt9.getLanguage())
|
||||
|| (tt9.notLanguageSyllabary() && !tt9.isTextEditingActive())
|
||||
|| (tt9.isFilteringSupported() && !tt9.isTextEditingActive())
|
||||
)
|
||||
);
|
||||
}
|
||||
|
|
|
|||
|
|
@ -36,7 +36,12 @@ public class SoftKeyPunctuation extends SoftKey {
|
|||
@Override
|
||||
protected String getTitle() {
|
||||
String keyChar = getKeyChar();
|
||||
return "*".equals(keyChar) ? "✱" : keyChar;
|
||||
return switch (keyChar) {
|
||||
case "*" -> "✱";
|
||||
case Characters.ZH_QUESTION_MARK -> "?";
|
||||
case Characters.ZH_EXCLAMATION_MARK -> "!";
|
||||
default -> keyChar;
|
||||
};
|
||||
}
|
||||
|
||||
protected String getKeyChar() {
|
||||
|
|
@ -55,6 +60,8 @@ public class SoftKeyPunctuation extends SoftKey {
|
|||
if (tt9.isInputModePhone()) return "*";
|
||||
if (tt9.isInputModeNumeric()) return ",";
|
||||
|
||||
if (LanguageKind.isChinese(tt9.getLanguage())) return Characters.ZH_EXCLAMATION_MARK;
|
||||
|
||||
return "!";
|
||||
}
|
||||
|
||||
|
|
@ -63,8 +70,9 @@ public class SoftKeyPunctuation extends SoftKey {
|
|||
if (tt9.isInputModePhone()) return "#";
|
||||
if (tt9.isInputModeNumeric()) return ".";
|
||||
|
||||
if (LanguageKind.isArabic(tt9.getLanguage())) return "؟";
|
||||
if (LanguageKind.isArabic(tt9.getLanguage())) return Characters.AR_QUESTION_MARK;
|
||||
if (LanguageKind.isGreek(tt9.getLanguage())) return Characters.GR_QUESTION_MARK;
|
||||
if (LanguageKind.isChinese(tt9.getLanguage())) return Characters.ZH_QUESTION_MARK;
|
||||
|
||||
return "?";
|
||||
}
|
||||
|
|
|
|||
|
|
@ -206,7 +206,7 @@ public class SuggestionsBar {
|
|||
return;
|
||||
}
|
||||
|
||||
stem = containsGenerated ? newSuggestions.get(0).substring(0, newSuggestions.get(0).length() - 1) : "";
|
||||
stem = containsGenerated && newSuggestions.get(0).length() > 1 ? newSuggestions.get(0).substring(0, newSuggestions.get(0).length() - 1) : "";
|
||||
|
||||
// Do not modify single letter + punctuation, such as "j'" or "l'". They look better as they are.
|
||||
stem = (stem.length() == 1 && newSuggestions.get(0).length() == 2 && !Character.isAlphabetic(newSuggestions.get(0).charAt(1))) ? "" : stem;
|
||||
|
|
|
|||
|
|
@ -7,6 +7,7 @@ import java.util.Locale;
|
|||
|
||||
import io.github.sspanak.tt9.ime.modes.InputMode;
|
||||
import io.github.sspanak.tt9.languages.Language;
|
||||
import io.github.sspanak.tt9.languages.LanguageKind;
|
||||
import io.github.sspanak.tt9.util.chars.Characters;
|
||||
|
||||
public class Text extends TextTools {
|
||||
|
|
@ -172,11 +173,22 @@ public class Text extends TextTools {
|
|||
}
|
||||
|
||||
|
||||
/**
|
||||
* Returns the number of regular 8-bit chars
|
||||
*/
|
||||
public int length() {
|
||||
return text == null ? 0 : text.length();
|
||||
}
|
||||
|
||||
|
||||
/**
|
||||
* Returns the number of UTF-16 chars
|
||||
*/
|
||||
public int codePointLength() {
|
||||
return text == null ? 0 : text.codePointCount(0, text.length());
|
||||
}
|
||||
|
||||
|
||||
public int lastBoundaryIndex() {
|
||||
if (text == null || text.length() < 2) {
|
||||
return -1;
|
||||
|
|
@ -224,6 +236,28 @@ public class Text extends TextTools {
|
|||
}
|
||||
|
||||
|
||||
/**
|
||||
* A safe substring method that works with code points (UTF-16 chars), instead of 8-bit chars.
|
||||
* Useful for languages with complex characters, like Chinese.
|
||||
*/
|
||||
public String substringCodePoints(int start, int end) {
|
||||
if (text == null) {
|
||||
return "";
|
||||
}
|
||||
|
||||
if (!LanguageKind.isChinese(language)) {
|
||||
return text.substring(start, end);
|
||||
}
|
||||
|
||||
StringBuilder output = new StringBuilder();
|
||||
for (int i = Math.max(start, 0), finish = Math.min(text.length(), end); i < finish; i++) {
|
||||
output.append(text.charAt(i));
|
||||
}
|
||||
|
||||
return output.toString();
|
||||
}
|
||||
|
||||
|
||||
public String subStringEndingWord(boolean keepApostrophe, boolean keepQuote) {
|
||||
if (text == null) {
|
||||
return "";
|
||||
|
|
|
|||
|
|
@ -11,22 +11,23 @@ import java.util.regex.Pattern;
|
|||
import io.github.sspanak.tt9.util.chars.Characters;
|
||||
|
||||
public class TextTools {
|
||||
private static final Pattern containsOtherThan1 = Pattern.compile("[02-9]");
|
||||
private static final Pattern combiningString = Pattern.compile("^\\p{M}+$");
|
||||
private static final Pattern nextIsPunctuation = Pattern.compile("^\\p{Punct}");
|
||||
private static final Pattern isHangul = Pattern.compile("[\u1100-\u11FF\u302E-\u302F\u3131-\u318F\u3200-\u321F\u3260-\u327E\uA960-\uA97F\uAC00-\uD7FB\uFFA0-\uFFDF]+");
|
||||
private static final Pattern nextToWord = Pattern.compile("\\b$");
|
||||
private static final Pattern previousIsLetter = Pattern.compile("[\\p{L}\\p{M}]$");
|
||||
private static final Pattern startOfSentence = Pattern.compile("(?<!\\.)(^|[.?!؟¿¡])\\s+$");
|
||||
private static final Pattern CONTAINS_OTHER_THAN_1 = Pattern.compile("[02-9]");
|
||||
private static final Pattern COMBINING_STRING = Pattern.compile("^\\p{M}+$");
|
||||
private static final Pattern NEXT_IS_PUNCTUATION = Pattern.compile("^\\p{Punct}");
|
||||
private static final Pattern IS_CHINESE = Pattern.compile("\\p{script=Han}+");
|
||||
private static final Pattern IS_HANGUL = Pattern.compile("[\u1100-\u11FF\u302E-\u302F\u3131-\u318F\u3200-\u321F\u3260-\u327E\uA960-\uA97F\uAC00-\uD7FB\uFFA0-\uFFDF]+");
|
||||
private static final Pattern NEXT_TO_WORD = Pattern.compile("\\b$");
|
||||
private static final Pattern PREVIOUS_IS_LETTER = Pattern.compile("[\\p{L}\\p{M}]$");
|
||||
private static final Pattern START_OF_SENTENCE = Pattern.compile("(?<!\\.)(^|[.?!؟¿¡])\\s+$");
|
||||
|
||||
|
||||
public static boolean containsOtherThan1(String str) {
|
||||
return str != null && containsOtherThan1.matcher(str).find();
|
||||
return str != null && CONTAINS_OTHER_THAN_1.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
public static boolean isCombining(String str) {
|
||||
return str != null && combiningString.matcher(str).find();
|
||||
return str != null && COMBINING_STRING.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
|
|
@ -45,8 +46,13 @@ public class TextTools {
|
|||
}
|
||||
|
||||
|
||||
public static boolean isChinese(String str) {
|
||||
return str != null && IS_CHINESE.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
public static boolean isHangul(String str) {
|
||||
return str != null && isHangul.matcher(str).find();
|
||||
return str != null && IS_HANGUL.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
|
|
@ -61,23 +67,35 @@ public class TextTools {
|
|||
}
|
||||
|
||||
|
||||
public static int lastIndexOfLatin(String str) {
|
||||
for (int i = str != null ? str.length() - 1 : -1; i >= 0; i--) {
|
||||
char ch = str.charAt(i);
|
||||
if ((ch >= 'A' && ch <= 'Z') || (ch >= 'a' && ch <= 'z')) {
|
||||
return i;
|
||||
}
|
||||
}
|
||||
|
||||
return -1;
|
||||
}
|
||||
|
||||
|
||||
public static boolean isStartOfSentence(String str) {
|
||||
return str != null && startOfSentence.matcher(str).find();
|
||||
return str != null && START_OF_SENTENCE.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
public static boolean isNextToWord(String str) {
|
||||
return str != null && nextToWord.matcher(str).find();
|
||||
return str != null && NEXT_TO_WORD.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
public static boolean nextIsPunctuation(String str) {
|
||||
return str != null && !str.isEmpty() && nextIsPunctuation.matcher(str).find();
|
||||
return str != null && !str.isEmpty() && NEXT_IS_PUNCTUATION.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
public static boolean previousIsLetter(String str) {
|
||||
return str != null && previousIsLetter.matcher(str).find();
|
||||
return str != null && PREVIOUS_IS_LETTER.matcher(str).find();
|
||||
}
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -10,7 +10,13 @@ import io.github.sspanak.tt9.languages.Language;
|
|||
import io.github.sspanak.tt9.languages.LanguageKind;
|
||||
|
||||
class Punctuation {
|
||||
public static final String AR_QUESTION_MARK = "؟";
|
||||
public static final String GR_QUESTION_MARK = ";";
|
||||
public static final String ZH_COMMA_LIST = "、";
|
||||
public static final String ZH_FULL_STOP = "。";
|
||||
public static final String ZH_QUESTION_MARK = "？";
|
||||
public static final String ZH_EXCLAMATION_MARK = "！";
|
||||
|
||||
public static final String NEW_LINE = DeviceInfo.AT_LEAST_ANDROID_6 && new Paint().hasGlyph("⏎") ? "⏎" : "\\n";
|
||||
public static final String ZWJ = "\u200D";
|
||||
public static final String ZWJ_GRAPHIC = "ZWJ";
|
||||
|
|
@ -22,7 +28,7 @@ class Punctuation {
|
|||
));
|
||||
|
||||
final public static ArrayList<Character> CombiningPunctuationFarsi = new ArrayList<>(Arrays.asList(
|
||||
'،', ZWNJ.charAt(0), '-', '\'', ':', '؛', '!', '؟', '.'
|
||||
'،', ZWNJ.charAt(0), '-', '\'', ':', AR_QUESTION_MARK.charAt(0), '!', '؛', '.'
|
||||
));
|
||||
|
||||
final private static ArrayList<Character> CombiningPunctuationGujarati = new ArrayList<>(Arrays.asList(
|
||||
|
|
@ -38,7 +44,11 @@ class Punctuation {
|
|||
));
|
||||
|
||||
final public static ArrayList<String> PunctuationArabic = new ArrayList<>(Arrays.asList(
|
||||
"،", ".", "-", "(", ")", "&", "~", "`", "'", "\"", "؛", ":", "!", "؟"
|
||||
"،", ".", "-", "(", ")", "&", "~", "`", "'", "\"", "؛", ":", "!", AR_QUESTION_MARK
|
||||
));
|
||||
|
||||
final public static ArrayList<String> PunctuationChinese = new ArrayList<>(Arrays.asList(
|
||||
"，", ZH_COMMA_LIST, ZH_FULL_STOP, "—", "～", "（", "）", ".", "「", "」", "『", "』", "•", "《", "》", "〈", "〉", "'", "“", "”", "；", "：", ZH_EXCLAMATION_MARK, ZH_QUESTION_MARK
|
||||
));
|
||||
|
||||
final public static ArrayList<String> PunctuationEnglish = new ArrayList<>(Arrays.asList(
|
||||
|
|
|
|||
|
|
@ -224,8 +224,6 @@
|
|||
<string name="key_yellow">Yellow Button</string>
|
||||
<string name="key_blue">Blue Button</string>
|
||||
|
||||
<string name="char_comma" translatable="false">,</string>
|
||||
<string name="char_dot" translatable="false">.</string>
|
||||
<string name="char_newline">New Line</string>
|
||||
<string name="char_space">Space</string>
|
||||
|
||||
|
|
|
|||
|
|
@ -9,7 +9,7 @@ ext.validateLanguageFiles = { definitionsDir, dictionariesDir, validationDir ->
|
|||
return "Too many errors! Skipping: ${definition}\n"
|
||||
}
|
||||
|
||||
def (alphabet, sounds, isAlphabeticLanguage, locale, dictionaryFile, langFileErrorCount, langFileErrorMsg) = parseLanguageDefintion(definition, dictionariesDir)
|
||||
def (alphabet, sounds, _, isAlphabeticLanguage, locale, dictionaryFile, langFileErrorCount, langFileErrorMsg) = parseLanguageDefintion(definition, dictionariesDir)
|
||||
|
||||
def languageHash = DictionaryTools.getLanguageHash(definition, dictionaryFile)
|
||||
def validationFile = new File("${validationDir}/${definition.name.replace(".yml", "")}.txt")
|
||||
|
|
@ -52,8 +52,11 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
int errorCount = 0
|
||||
String errorMsg = ""
|
||||
|
||||
String abcString = ""
|
||||
boolean hasABC = true
|
||||
boolean hasLayout = false
|
||||
boolean hasSounds = false
|
||||
boolean filterBySounds = false
|
||||
boolean areNumeralsValid = true
|
||||
String localeString = ""
|
||||
String dictionaryFileName = ""
|
||||
|
|
@ -64,6 +67,8 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
&& !rawLine.startsWith("abcString")
|
||||
&& !rawLine.startsWith("currency")
|
||||
&& !rawLine.startsWith("dictionaryFile")
|
||||
&& !rawLine.startsWith("filterBySounds")
|
||||
&& !rawLine.startsWith("hasABC")
|
||||
&& !rawLine.startsWith("hasSpaceBetweenWords")
|
||||
&& !rawLine.startsWith("hasUpperCase")
|
||||
&& !rawLine.startsWith("layout")
|
||||
|
|
@ -71,6 +76,7 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
&& !rawLine.startsWith("name")
|
||||
&& !rawLine.startsWith("numerals")
|
||||
&& !rawLine.startsWith("sounds")
|
||||
|
||||
) {
|
||||
def parts = rawLine.split(":")
|
||||
def property = parts.length > 0 ? parts[0] : rawLine
|
||||
|
|
@ -81,14 +87,29 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
|
||||
String line = rawLine.replaceFirst("#[\\s\\S]+\$", "")
|
||||
|
||||
if (
|
||||
(line.startsWith("hasUpperCase") || line.startsWith("hasSpaceBetweenWords"))
|
||||
&& !line.endsWith("yes") && !line.endsWith("no")
|
||||
) {
|
||||
def property = line.replaceAll(":.*\$", "")
|
||||
def invalidVal = line.replace("hasUpperCase:", "").trim()
|
||||
def booleanProperties = ["hasUpperCase", "hasSpaceBetweenWords", "filterBySounds", "hasABC"]
|
||||
for (String property : booleanProperties) {
|
||||
String booleanError = validateBooleanProperty(line, property, languageFile.name)
|
||||
if (booleanError) {
|
||||
errorCount++
|
||||
errorMsg += "Language '${languageFile.name}' is invalid. Unrecognized '${property}' value: '${invalidVal}'. Only 'yes' and 'no' are allowed.\n"
|
||||
errorMsg += booleanError
|
||||
}
|
||||
}
|
||||
|
||||
if (line.startsWith("abcString")) {
|
||||
abcString = line.replace("abcString:", "").trim()
|
||||
}
|
||||
|
||||
if (line.startsWith("dictionaryFile")) {
|
||||
dictionaryFileName = line.replace("dictionaryFile:", "").trim()
|
||||
}
|
||||
|
||||
if (line.startsWith("filterBySounds")) {
|
||||
filterBySounds = line.endsWith("yes")
|
||||
}
|
||||
|
||||
if (line.startsWith("hasABC")) {
|
||||
hasABC = line.endsWith("yes")
|
||||
}
|
||||
|
||||
if (line.startsWith("numerals")) {
|
||||
|
|
@ -107,10 +128,6 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
localeString = line.replace("locale:", "").trim()
|
||||
}
|
||||
|
||||
if (line.startsWith("dictionaryFile")) {
|
||||
dictionaryFileName = line.replace("dictionaryFile:", "").trim()
|
||||
}
|
||||
|
||||
// alphabet string
|
||||
def lineCharacters = extractAlphabetCharsFromLine(line)
|
||||
lineCharacters = lineCharacters.isEmpty() ? extractAlphabetExtraCharsFromLine(languageFile.name, line) : lineCharacters
|
||||
|
|
@ -135,6 +152,11 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
}
|
||||
}
|
||||
|
||||
if (!hasABC && !abcString.isEmpty()) {
|
||||
errorCount++
|
||||
errorMsg += "Language '${languageFile.name}' is invalid. hasABC must be 'true' when abcString is provided.\n"
|
||||
}
|
||||
|
||||
if (!hasLayout) {
|
||||
errorCount++
|
||||
errorMsg += "Language '${languageFile.name}' is invalid. Missing 'layout' property.\n"
|
||||
|
|
@ -150,6 +172,11 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
errorMsg += "Language '${languageFile.name}' is invalid. 'sounds' property must contain series of phonetic transcriptions per digit sequence in the format: ' - [Yae,1221]' and so on.\n"
|
||||
}
|
||||
|
||||
if (filterBySounds && !hasSounds) {
|
||||
errorCount++
|
||||
errorMsg += "Language '${languageFile.name}' is invalid. 'filterBySounds' property can only be used with 'sounds' property.\n"
|
||||
}
|
||||
|
||||
if (!localeString.matches("^[a-z]{2,3}(?:-[A-Z]{2})?\$")) {
|
||||
errorCount++
|
||||
def msg = localeString.isEmpty() ? "Missing 'locale' property." : "Unrecognized locale format: '${localeString}'"
|
||||
|
|
@ -176,7 +203,7 @@ ext.parseLanguageDefintion = { File languageFile, String dictionariesDir ->
|
|||
}
|
||||
}
|
||||
|
||||
return [alphabet, sounds, !hasSounds, locale, dictionaryFile, errorCount, errorMsg]
|
||||
return [alphabet, sounds, filterBySounds, !hasSounds, locale, dictionaryFile, errorCount, errorMsg]
|
||||
}
|
||||
|
||||
|
||||
|
|
@ -298,6 +325,16 @@ static def isLayoutLine(String line) {
|
|||
|
||||
//////////////////// VALIDATION ////////////////////
|
||||
|
||||
static def validateBooleanProperty(String line, String propertyName, String languageFileName) {
|
||||
if (line.startsWith(propertyName) && !line.endsWith("yes") && !line.endsWith("no")) {
|
||||
def property = line.replaceAll(":.*\$", "")
|
||||
def invalidVal = line.replace("hasUpperCase:", "").trim()
|
||||
return "Language '${languageFileName}' is invalid. Unrecognized '${property}' value: '${invalidVal}'. Only 'yes' and 'no' are allowed.\n"
|
||||
}
|
||||
|
||||
return ''
|
||||
}
|
||||
|
||||
static def validateNoWhitespace(String line, int lineNumber) {
|
||||
if (line == "") {
|
||||
return "There is no word on line ${lineNumber}. Remove all empty lines.\n"
|
||||
|
|
|
|||
11
docs/dictionaries/pinyinWordlistReadme.txt
Normal file
11
docs/dictionaries/pinyinWordlistReadme.txt
Normal file
|
|
@ -0,0 +1,11 @@
|
|||
Chinese Pinyin word list by Vkiki2233:
|
||||
Source: https://github.com/sspanak/tt9/issues/260
|
||||
|
||||
Chinese frequency list 1 by Cai & Brysbaert
|
||||
Source: http://crr.ugent.be/programs-data/subtitle-frequencies/subtlex-ch
|
||||
Version: 20231108150220
|
||||
Reference: Cai, Q., & Brysbaert, M. (2010). SUBTLEX-CH: Chinese Word and Character Frequencies Based on Film Subtitles. Plos ONE, 5(6), e10729
|
||||
|
||||
Chinese frequency list 2 by HanziCraft
|
||||
Source: https://hanzicraft.com/lists/frequency
|
||||
Version 2025-03-19
|
||||
BIN
downloads/zh-pinyin.zip
Normal file
BIN
downloads/zh-pinyin.zip
Normal file
Binary file not shown.
Loading…
Add table
Add a link
Reference in a new issue