parent
feac8a91f9
commit
d2579c4832
|
@ -535,10 +535,13 @@ public final class BinaryDictIOUtils {
|
||||||
// TODO: Consolidate this code with the code that computes the size of the bigram list
|
// TODO: Consolidate this code with the code that computes the size of the bigram list
|
||||||
// in BinaryDictionaryInputOutput#computeActualNodeSize
|
// in BinaryDictionaryInputOutput#computeActualNodeSize
|
||||||
for (int i = 0; i < info.mBigrams.size(); ++i) {
|
for (int i = 0; i < info.mBigrams.size(); ++i) {
|
||||||
final int bigramOffset = info.mBigrams.get(i).mAddress - info.mOriginalAddress;
|
|
||||||
final int bigramFrequency = info.mBigrams.get(i).mFrequency;
|
final int bigramFrequency = info.mBigrams.get(i).mFrequency;
|
||||||
int bigramFlags = (i < info.mBigrams.size() - 1)
|
int bigramFlags = (i < info.mBigrams.size() - 1)
|
||||||
? FormatSpec.FLAG_ATTRIBUTE_HAS_NEXT : 0;
|
? FormatSpec.FLAG_ATTRIBUTE_HAS_NEXT : 0;
|
||||||
|
size++;
|
||||||
|
final int bigramOffset = info.mBigrams.get(i).mAddress - (info.mOriginalAddress
|
||||||
|
+ size);
|
||||||
bigramFlags |= (bigramOffset < 0) ? FormatSpec.FLAG_ATTRIBUTE_OFFSET_NEGATIVE : 0;
|
bigramFlags |= (bigramOffset < 0) ? FormatSpec.FLAG_ATTRIBUTE_OFFSET_NEGATIVE : 0;
|
||||||
switch (BinaryDictInputOutput.getByteSize(bigramOffset)) {
|
switch (BinaryDictInputOutput.getByteSize(bigramOffset)) {
|
||||||
case 1:
|
case 1:
|
||||||
|
@ -553,7 +556,6 @@ public final class BinaryDictIOUtils {
|
||||||
}
|
}
|
||||||
bigramFlags |= bigramFrequency & FormatSpec.FLAG_ATTRIBUTE_FREQUENCY;
|
bigramFlags |= bigramFrequency & FormatSpec.FLAG_ATTRIBUTE_FREQUENCY;
|
||||||
destination.write((byte)bigramFlags);
|
destination.write((byte)bigramFlags);
|
||||||
size++;
|
|
||||||
size += writeVariableAddress(destination, Math.abs(bigramOffset));
|
size += writeVariableAddress(destination, Math.abs(bigramOffset));
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
@ -717,7 +719,7 @@ public final class BinaryDictIOUtils {
|
||||||
if (position == FormatSpec.NOT_VALID_WORD) {
|
if (position == FormatSpec.NOT_VALID_WORD) {
|
||||||
// TODO: figure out what is the correct thing to do here.
|
// TODO: figure out what is the correct thing to do here.
|
||||||
} else {
|
} else {
|
||||||
bigrams.add(new PendingAttribute(position, bigram.mFrequency));
|
bigrams.add(new PendingAttribute(bigram.mFrequency, position));
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
@ -947,4 +949,25 @@ public final class BinaryDictIOUtils {
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
/**
|
||||||
|
* Find a word from the buffer.
|
||||||
|
*
|
||||||
|
* @param buffer the buffer representing the body of the dictionary file.
|
||||||
|
* @param word the word searched
|
||||||
|
* @return the found group
|
||||||
|
* @throws IOException
|
||||||
|
* @throws UnsupportedFormatException
|
||||||
|
*/
|
||||||
|
public static CharGroupInfo findWordFromBuffer(final FusionDictionaryBufferInterface buffer,
|
||||||
|
final String word) throws IOException, UnsupportedFormatException {
|
||||||
|
int position = getTerminalPosition(buffer, word);
|
||||||
|
if (position != FormatSpec.NOT_VALID_WORD) {
|
||||||
|
buffer.position(0);
|
||||||
|
final FileHeader header = BinaryDictInputOutput.readHeader(buffer);
|
||||||
|
buffer.position(position);
|
||||||
|
return BinaryDictInputOutput.readCharGroup(buffer, position, header.mFormatOptions);
|
||||||
|
}
|
||||||
|
return null;
|
||||||
|
}
|
||||||
}
|
}
|
||||||
|
|
|
@ -28,6 +28,7 @@ import android.test.AndroidTestCase;
|
||||||
import android.test.MoreAsserts;
|
import android.test.MoreAsserts;
|
||||||
import android.util.Log;
|
import android.util.Log;
|
||||||
|
|
||||||
|
import java.io.BufferedOutputStream;
|
||||||
import java.io.BufferedWriter;
|
import java.io.BufferedWriter;
|
||||||
import java.io.File;
|
import java.io.File;
|
||||||
import java.io.FileInputStream;
|
import java.io.FileInputStream;
|
||||||
|
@ -41,7 +42,7 @@ import java.util.Arrays;
|
||||||
import java.util.HashMap;
|
import java.util.HashMap;
|
||||||
import java.util.Random;
|
import java.util.Random;
|
||||||
|
|
||||||
public class BinaryDictIOUtilsTests extends AndroidTestCase{
|
public class BinaryDictIOUtilsTests extends AndroidTestCase {
|
||||||
private static final String TAG = BinaryDictIOUtilsTests.class.getSimpleName();
|
private static final String TAG = BinaryDictIOUtilsTests.class.getSimpleName();
|
||||||
private static final FormatSpec.FormatOptions FORMAT_OPTIONS =
|
private static final FormatSpec.FormatOptions FORMAT_OPTIONS =
|
||||||
new FormatSpec.FormatOptions(3, true);
|
new FormatSpec.FormatOptions(3, true);
|
||||||
|
@ -150,26 +151,50 @@ public class BinaryDictIOUtilsTests extends AndroidTestCase{
|
||||||
return position;
|
return position;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
private CharGroupInfo findWordFromFile(final File file, final String word) {
|
||||||
|
FileInputStream inStream = null;
|
||||||
|
CharGroupInfo info = null;
|
||||||
|
try {
|
||||||
|
inStream = new FileInputStream(file);
|
||||||
|
final FusionDictionaryBufferInterface buffer = new ByteBufferWrapper(
|
||||||
|
inStream.getChannel().map(FileChannel.MapMode.READ_ONLY, 0, file.length()));
|
||||||
|
info = BinaryDictIOUtils.findWordFromBuffer(buffer, word);
|
||||||
|
} catch (IOException e) {
|
||||||
|
} catch (UnsupportedFormatException e) {
|
||||||
|
} finally {
|
||||||
|
if (inStream != null) {
|
||||||
|
try {
|
||||||
|
inStream.close();
|
||||||
|
} catch (IOException e) {
|
||||||
|
// do nothing
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
return info;
|
||||||
|
}
|
||||||
|
|
||||||
// return amount of time to insert a word
|
// return amount of time to insert a word
|
||||||
private long insertAndCheckWord(final File file, final String word, final int frequency,
|
private long insertAndCheckWord(final File file, final String word, final int frequency,
|
||||||
final boolean exist) {
|
final boolean exist, final ArrayList<WeightedString> bigrams,
|
||||||
|
final ArrayList<WeightedString> shortcuts) {
|
||||||
RandomAccessFile raFile = null;
|
RandomAccessFile raFile = null;
|
||||||
FileOutputStream outStream = null;
|
BufferedOutputStream outStream = null;
|
||||||
FusionDictionaryBufferInterface buffer = null;
|
FusionDictionaryBufferInterface buffer = null;
|
||||||
long amountOfTime = -1;
|
long amountOfTime = -1;
|
||||||
try {
|
try {
|
||||||
raFile = new RandomAccessFile(file, "rw");
|
raFile = new RandomAccessFile(file, "rw");
|
||||||
buffer = new ByteBufferWrapper(raFile.getChannel().map(
|
buffer = new ByteBufferWrapper(raFile.getChannel().map(
|
||||||
FileChannel.MapMode.READ_WRITE, 0, file.length()));
|
FileChannel.MapMode.READ_WRITE, 0, file.length()));
|
||||||
outStream = new FileOutputStream(file, true);
|
outStream = new BufferedOutputStream(new FileOutputStream(file, true));
|
||||||
|
|
||||||
if (!exist) {
|
if (!exist) {
|
||||||
assertEquals(FormatSpec.NOT_VALID_WORD, getWordPosition(file, word));
|
assertEquals(FormatSpec.NOT_VALID_WORD, getWordPosition(file, word));
|
||||||
}
|
}
|
||||||
final long now = System.nanoTime();
|
final long now = System.nanoTime();
|
||||||
BinaryDictIOUtils.insertWord(buffer, outStream, word, frequency, null, null, false,
|
BinaryDictIOUtils.insertWord(buffer, outStream, word, frequency, bigrams, shortcuts,
|
||||||
false);
|
false, false);
|
||||||
amountOfTime = System.nanoTime() - now;
|
amountOfTime = System.nanoTime() - now;
|
||||||
|
outStream.flush();
|
||||||
MoreAsserts.assertNotEqual(FormatSpec.NOT_VALID_WORD, getWordPosition(file, word));
|
MoreAsserts.assertNotEqual(FormatSpec.NOT_VALID_WORD, getWordPosition(file, word));
|
||||||
outStream.close();
|
outStream.close();
|
||||||
raFile.close();
|
raFile.close();
|
||||||
|
@ -215,8 +240,6 @@ public class BinaryDictIOUtilsTests extends AndroidTestCase{
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
|
||||||
|
|
||||||
private void checkReverseLookup(final File file, final String word, final int position) {
|
private void checkReverseLookup(final File file, final String word, final int position) {
|
||||||
FileInputStream inStream = null;
|
FileInputStream inStream = null;
|
||||||
try {
|
try {
|
||||||
|
@ -242,7 +265,7 @@ public class BinaryDictIOUtilsTests extends AndroidTestCase{
|
||||||
public void testInsertWord() {
|
public void testInsertWord() {
|
||||||
File file = null;
|
File file = null;
|
||||||
try {
|
try {
|
||||||
file = File.createTempFile("testInsertWord", ".dict");
|
file = File.createTempFile("testInsertWord", ".dict", getContext().getCacheDir());
|
||||||
} catch (IOException e) {
|
} catch (IOException e) {
|
||||||
fail("IOException while creating temporary file: " + e);
|
fail("IOException while creating temporary file: " + e);
|
||||||
}
|
}
|
||||||
|
@ -263,36 +286,74 @@ public class BinaryDictIOUtilsTests extends AndroidTestCase{
|
||||||
}
|
}
|
||||||
|
|
||||||
MoreAsserts.assertNotEqual(FormatSpec.NOT_VALID_WORD, getWordPosition(file, "abcd"));
|
MoreAsserts.assertNotEqual(FormatSpec.NOT_VALID_WORD, getWordPosition(file, "abcd"));
|
||||||
insertAndCheckWord(file, "abcde", 10, false);
|
insertAndCheckWord(file, "abcde", 10, false, null, null);
|
||||||
|
|
||||||
insertAndCheckWord(file, "abcdefghijklmn", 10, false);
|
insertAndCheckWord(file, "abcdefghijklmn", 10, false, null, null);
|
||||||
checkReverseLookup(file, "abcdefghijklmn", getWordPosition(file, "abcdefghijklmn"));
|
checkReverseLookup(file, "abcdefghijklmn", getWordPosition(file, "abcdefghijklmn"));
|
||||||
|
|
||||||
insertAndCheckWord(file, "abcdabcd", 10, false);
|
insertAndCheckWord(file, "abcdabcd", 10, false, null, null);
|
||||||
checkReverseLookup(file, "abcdabcd", getWordPosition(file, "abcdabcd"));
|
checkReverseLookup(file, "abcdabcd", getWordPosition(file, "abcdabcd"));
|
||||||
|
|
||||||
// update the existing word.
|
// update the existing word.
|
||||||
insertAndCheckWord(file, "abcdabcd", 15, true);
|
insertAndCheckWord(file, "abcdabcd", 15, true, null, null);
|
||||||
|
|
||||||
// split 1
|
// split 1
|
||||||
insertAndCheckWord(file, "ab", 20, false);
|
insertAndCheckWord(file, "ab", 20, false, null, null);
|
||||||
|
|
||||||
// split 2
|
// split 2
|
||||||
insertAndCheckWord(file, "ami", 30, false);
|
insertAndCheckWord(file, "ami", 30, false, null, null);
|
||||||
|
|
||||||
deleteWord(file, "ami");
|
deleteWord(file, "ami");
|
||||||
assertEquals(FormatSpec.NOT_VALID_WORD, getWordPosition(file, "ami"));
|
assertEquals(FormatSpec.NOT_VALID_WORD, getWordPosition(file, "ami"));
|
||||||
|
|
||||||
insertAndCheckWord(file, "abcdabfg", 30, false);
|
insertAndCheckWord(file, "abcdabfg", 30, false, null, null);
|
||||||
|
|
||||||
deleteWord(file, "abcd");
|
deleteWord(file, "abcd");
|
||||||
assertEquals(FormatSpec.NOT_VALID_WORD, getWordPosition(file, "abcd"));
|
assertEquals(FormatSpec.NOT_VALID_WORD, getWordPosition(file, "abcd"));
|
||||||
}
|
}
|
||||||
|
|
||||||
|
public void testInsertWordWithBigrams() {
|
||||||
|
File file = null;
|
||||||
|
try {
|
||||||
|
file = File.createTempFile("testInsertWordWithBigrams", ".dict",
|
||||||
|
getContext().getCacheDir());
|
||||||
|
} catch (IOException e) {
|
||||||
|
fail("IOException while creating temporary file: " + e);
|
||||||
|
}
|
||||||
|
|
||||||
|
// set an initial dictionary.
|
||||||
|
final FusionDictionary dict = new FusionDictionary(new Node(),
|
||||||
|
new FusionDictionary.DictionaryOptions(new HashMap<String,String>(), false, false));
|
||||||
|
dict.add("abcd", 10, null, false);
|
||||||
|
dict.add("efgh", 15, null, false);
|
||||||
|
|
||||||
|
try {
|
||||||
|
final FileOutputStream out = new FileOutputStream(file);
|
||||||
|
BinaryDictInputOutput.writeDictionaryBinary(out, dict, FORMAT_OPTIONS);
|
||||||
|
out.close();
|
||||||
|
} catch (IOException e) {
|
||||||
|
fail("IOException while writing an initial dictionary : " + e);
|
||||||
|
} catch (UnsupportedFormatException e) {
|
||||||
|
fail("UnsupportedFormatException while writing an initial dictionary : " + e);
|
||||||
|
}
|
||||||
|
|
||||||
|
final ArrayList<WeightedString> banana = new ArrayList<WeightedString>();
|
||||||
|
banana.add(new WeightedString("banana", 10));
|
||||||
|
|
||||||
|
insertAndCheckWord(file, "banana", 0, false, null, null);
|
||||||
|
insertAndCheckWord(file, "recursive", 60, true, banana, null);
|
||||||
|
|
||||||
|
final CharGroupInfo info = findWordFromFile(file, "recursive");
|
||||||
|
int bananaPos = getWordPosition(file, "banana");
|
||||||
|
assertNotNull(info.mBigrams);
|
||||||
|
assertEquals(info.mBigrams.size(), 1);
|
||||||
|
assertEquals(info.mBigrams.get(0).mAddress, bananaPos);
|
||||||
|
}
|
||||||
|
|
||||||
public void testRandomWords() {
|
public void testRandomWords() {
|
||||||
File file = null;
|
File file = null;
|
||||||
try {
|
try {
|
||||||
file = File.createTempFile("testRandomWord", ".dict");
|
file = File.createTempFile("testRandomWord", ".dict", getContext().getCacheDir());
|
||||||
} catch (IOException e) {
|
} catch (IOException e) {
|
||||||
}
|
}
|
||||||
assertNotNull(file);
|
assertNotNull(file);
|
||||||
|
@ -317,7 +378,8 @@ public class BinaryDictIOUtilsTests extends AndroidTestCase{
|
||||||
long minTimeToInsert = 100000000; // 1000000000 is an upper bound for minTimeToInsert.
|
long minTimeToInsert = 100000000; // 1000000000 is an upper bound for minTimeToInsert.
|
||||||
int cnt = 0;
|
int cnt = 0;
|
||||||
for (final String word : sWords) {
|
for (final String word : sWords) {
|
||||||
final long diff = insertAndCheckWord(file, word, cnt%255, false);
|
final long diff = insertAndCheckWord(file, word,
|
||||||
|
cnt % FormatSpec.MAX_TERMINAL_FREQUENCY, false, null, null);
|
||||||
maxTimeToInsert = Math.max(maxTimeToInsert, diff);
|
maxTimeToInsert = Math.max(maxTimeToInsert, diff);
|
||||||
minTimeToInsert = Math.min(minTimeToInsert, diff);
|
minTimeToInsert = Math.min(minTimeToInsert, diff);
|
||||||
sum += diff;
|
sum += diff;
|
||||||
|
|
Loading…
Reference in New Issue