Improve emoji search results.
This commit is contained in:
Greyson Parrelli
parent
18eac51576
commit
907abf72d3
3 changed files with 109 additions and 85 deletions
|
|
@ -1,84 +0,0 @@
|
|||
package org.thoughtcrime.securesms.database;
|
||||
|
||||
import android.content.ContentValues;
|
||||
import android.content.Context;
|
||||
import android.database.Cursor;
|
||||
import android.text.TextUtils;
|
||||
|
||||
import androidx.annotation.NonNull;
|
||||
|
||||
import org.signal.core.util.CursorUtil;
|
||||
import org.signal.core.util.SqlUtil;
|
||||
import org.thoughtcrime.securesms.database.model.EmojiSearchData;
|
||||
import org.thoughtcrime.securesms.util.FtsUtil;
|
||||
|
||||
import java.util.LinkedList;
|
||||
import java.util.List;
|
||||
|
||||
/**
|
||||
* Contains all info necessary for full-text search of emoji tags.
|
||||
*/
|
||||
public class EmojiSearchDatabase extends Database {
|
||||
|
||||
public static final String TABLE_NAME = "emoji_search";
|
||||
|
||||
public static final String LABEL = "label";
|
||||
public static final String EMOJI = "emoji";
|
||||
|
||||
public static final String CREATE_TABLE = "CREATE VIRTUAL TABLE " + TABLE_NAME + " USING fts5(" + LABEL + ", " + EMOJI + " UNINDEXED)";
|
||||
|
||||
public EmojiSearchDatabase(@NonNull Context context, @NonNull SignalDatabase databaseHelper) {
|
||||
super(context, databaseHelper);
|
||||
}
|
||||
|
||||
/**
|
||||
* @param query A search query. Doesn't need any special formatted -- it'll be sanitized.
|
||||
* @return A list of emoji that are related to the search term, ordered by relevance.
|
||||
*/
|
||||
public @NonNull List<String> query(@NonNull String query, int limit) {
|
||||
SQLiteDatabase db = databaseHelper.getSignalReadableDatabase();
|
||||
String matchString = FtsUtil.createPrefixMatchString(query);
|
||||
List<String> results = new LinkedList<>();
|
||||
|
||||
if (TextUtils.isEmpty(matchString)) {
|
||||
return results;
|
||||
}
|
||||
|
||||
String[] projection = new String[] { EMOJI };
|
||||
String selection = LABEL + " MATCH (?)";
|
||||
String[] args = SqlUtil.buildArgs(matchString);
|
||||
|
||||
try (Cursor cursor = db.query(true, TABLE_NAME, projection, selection, args, null, null, "rank", String.valueOf(limit))) {
|
||||
while (cursor.moveToNext()) {
|
||||
results.add(CursorUtil.requireString(cursor, EMOJI));
|
||||
}
|
||||
}
|
||||
|
||||
return results;
|
||||
}
|
||||
|
||||
/**
|
||||
* Deletes the content of the current search index and replaces it with the new one.
|
||||
*/
|
||||
public void setSearchIndex(@NonNull List<EmojiSearchData> searchIndex) {
|
||||
SQLiteDatabase db = databaseHelper.getSignalReadableDatabase();
|
||||
|
||||
db.beginTransaction();
|
||||
try {
|
||||
db.delete(TABLE_NAME, null, null);
|
||||
|
||||
for (EmojiSearchData searchData : searchIndex) {
|
||||
for (String label : searchData.getTags()) {
|
||||
ContentValues values = new ContentValues(2);
|
||||
values.put(LABEL, label);
|
||||
values.put(EMOJI, searchData.getEmoji());
|
||||
db.insert(TABLE_NAME, null, values);
|
||||
}
|
||||
}
|
||||
db.setTransactionSuccessful();
|
||||
} finally {
|
||||
db.endTransaction();
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -0,0 +1,108 @@
|
|||
package org.thoughtcrime.securesms.database
|
||||
|
||||
import android.content.Context
|
||||
import android.text.TextUtils
|
||||
import androidx.core.content.contentValuesOf
|
||||
import org.signal.core.util.requireNonNullString
|
||||
import org.signal.core.util.select
|
||||
import org.signal.core.util.withinTransaction
|
||||
import org.thoughtcrime.securesms.database.model.EmojiSearchData
|
||||
import kotlin.math.max
|
||||
import kotlin.math.roundToInt
|
||||
|
||||
/**
|
||||
* Contains all info necessary for full-text search of emoji tags.
|
||||
*/
|
||||
class EmojiSearchDatabase(context: Context, databaseHelper: SignalDatabase) : Database(context, databaseHelper) {
|
||||
|
||||
companion object {
|
||||
const val TABLE_NAME = "emoji_search"
|
||||
const val LABEL = "label"
|
||||
const val EMOJI = "emoji"
|
||||
const val CREATE_TABLE = "CREATE VIRTUAL TABLE $TABLE_NAME USING fts5($LABEL, $EMOJI UNINDEXED)"
|
||||
}
|
||||
|
||||
/**
|
||||
* @param query A search query. Doesn't need any special formatted -- it'll be sanitized.
|
||||
* @return A list of emoji that are related to the search term, ordered by relevance.
|
||||
*/
|
||||
fun query(originalQuery: String, originalLimit: Int): List<String> {
|
||||
val query: String = originalQuery.trim()
|
||||
|
||||
if (TextUtils.isEmpty(query)) {
|
||||
return emptyList()
|
||||
}
|
||||
|
||||
val limit: Int = max(originalLimit, 100)
|
||||
val entries = mutableListOf<Entry>()
|
||||
|
||||
readableDatabase
|
||||
.select(LABEL, EMOJI)
|
||||
.from(TABLE_NAME)
|
||||
.where("$LABEL LIKE ?", "%$query%")
|
||||
.limit(limit)
|
||||
.run()
|
||||
.use { cursor ->
|
||||
while (cursor.moveToNext()) {
|
||||
entries += Entry(
|
||||
label = cursor.requireNonNullString(LABEL),
|
||||
emoji = cursor.requireNonNullString(EMOJI)
|
||||
)
|
||||
}
|
||||
}
|
||||
|
||||
return entries
|
||||
.sortedWith { lhs, rhs ->
|
||||
similarityScore(query, lhs.label) - similarityScore(query, rhs.label)
|
||||
}
|
||||
.take(originalLimit)
|
||||
.map { it.emoji }
|
||||
}
|
||||
|
||||
/**
|
||||
* Deletes the content of the current search index and replaces it with the new one.
|
||||
*/
|
||||
fun setSearchIndex(searchIndex: List<EmojiSearchData>) {
|
||||
val db = databaseHelper.signalReadableDatabase
|
||||
|
||||
db.withinTransaction {
|
||||
db.delete(TABLE_NAME, null, null)
|
||||
|
||||
for (searchData in searchIndex) {
|
||||
for (label in searchData.tags) {
|
||||
val values = contentValuesOf(
|
||||
LABEL to label,
|
||||
EMOJI to searchData.emoji
|
||||
)
|
||||
db.insert(TABLE_NAME, null, values)
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Ranks how "similar" a match is to the original search term.
|
||||
* A lower score means more similar, with 0 being a perfect match.
|
||||
*
|
||||
* We know that the `searchTerm` must be a substring of the `match`.
|
||||
* We determine similarity by how many letters appear before or after the `searchTerm` in the `match`.
|
||||
* We give letters that come before the term a bigger weight than those that come after as a way to prefer matches that are prefixed by the `searchTerm`.
|
||||
*/
|
||||
private fun similarityScore(searchTerm: String, match: String): Int {
|
||||
if (searchTerm == match) {
|
||||
return 0
|
||||
}
|
||||
|
||||
val startIndex = match.indexOf(searchTerm)
|
||||
|
||||
val prefixCount = startIndex
|
||||
val suffixCount = match.length - (startIndex + searchTerm.length)
|
||||
|
||||
val prefixRankWeight = 1.5f
|
||||
val suffixRankWeight = 1f
|
||||
|
||||
return ((prefixCount * prefixRankWeight) + (suffixCount * suffixRankWeight)).roundToInt()
|
||||
}
|
||||
|
||||
private data class Entry(val label: String, val emoji: String)
|
||||
}
|
||||
|
|
@ -26,7 +26,7 @@ class EmojiSearchRepository(private val context: Context) {
|
|||
|
||||
fun submitQuery(query: String, limit: Int = EMOJI_SEARCH_LIMIT): Single<List<String>> {
|
||||
val result = if (query.length >= MINIMUM_INLINE_QUERY_THRESHOLD && NOT_PUNCTUATION.matches(query.substring(query.lastIndex))) {
|
||||
Single.fromCallable<List<String>> { emojiSearchDatabase.query(query, limit) }
|
||||
Single.fromCallable { emojiSearchDatabase.query(query, limit) }
|
||||
} else {
|
||||
Single.just(emptyList())
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Reference in a new issue