nyancrimew/LatinIME
1
0
Fork 0
Code Issues Pull requests Projects Releases Wiki Activity

Finish up the "info" command in dicttool. (A6)

Browse source 
Bug: 7388857
Change-Id: I704f12a6be76ce1644ec5e8dd3b667f112e9c04a
This commit is contained in:
Jean Chalard 2012-10-25 18:15:42 +09:00
parent f1d35ac5dc
commit 47cac57e45
4 changed files with 75 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
java/src/com/android/inputmethod/latin/makedict/FormatSpec.java
View file

@ -224,6 +224,8 @@ public final class FormatSpec {
static final int MAX_TERMINAL_FREQUENCY = 255;
static final int MAX_BIGRAM_FREQUENCY = 15;
public static final int SHORTCUT_WHITELIST_FREQUENCY = 15;
// This option needs to be the same numeric value as the one in binary_format.h.
static final int NOT_VALID_WORD = -99;
static final int SIGNED_CHILDREN_ADDRESS_SIZE = 3;

36
java/src/com/android/inputmethod/latin/makedict/FusionDictionary.java
View file

@ -22,6 +22,7 @@ import com.android.inputmethod.latin.Constants;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.Date;
import java.util.HashMap;
import java.util.Iterator;
import java.util.LinkedList;
@ -255,8 +256,6 @@ public final class FusionDictionary implements Iterable<Word> {
/**
* Options global to the dictionary.
*
* There are no options at the moment, so this class is empty.
*/
public static final class DictionaryOptions {
public final boolean mGermanUmlautProcessing;
@ -268,6 +267,39 @@ public final class FusionDictionary implements Iterable<Word> {
mGermanUmlautProcessing = germanUmlautProcessing;
mFrenchLigatureProcessing = frenchLigatureProcessing;
}
@Override
public String toString() { // Convenience method
return toString(0);
}
public String toString(final int indentCount) {
final StringBuilder indent = new StringBuilder();
for (int i = 0; i < indentCount; ++i) {
indent.append(" ");
}
final StringBuilder s = new StringBuilder();
for (final String optionKey : mAttributes.keySet()) {
s.append(indent);
s.append(optionKey);
s.append(" = ");
if ("date".equals(optionKey)) {
// Date needs a number of milliseconds, but the dictionary contains seconds
s.append(new Date(
1000 * Long.parseLong(mAttributes.get(optionKey))).toString());
} else {
s.append(mAttributes.get(optionKey));
}
s.append("\n");
}
if (mGermanUmlautProcessing) {
s.append(indent);
s.append("Needs German umlaut processing\n");
}
if (mFrenchLigatureProcessing) {
s.append(indent);
s.append("Needs French ligature processing\n");
}
return s.toString();
}
}
public final DictionaryOptions mOptions;

6
tools/dicttool/src/com/android/inputmethod/latin/dicttool/CombinedInputOutput.java
View file

@ -16,6 +16,7 @@
package com.android.inputmethod.latin.dicttool;
import com.android.inputmethod.latin.makedict.FormatSpec;
import com.android.inputmethod.latin.makedict.FusionDictionary;
import com.android.inputmethod.latin.makedict.FusionDictionary.DictionaryOptions;
import com.android.inputmethod.latin.makedict.FusionDictionary.Node;
@ -150,8 +151,9 @@ public class CombinedInputOutput {
if (SHORTCUT_TAG.equals(params[0])) {
shortcut = params[1];
} else if (FREQUENCY_TAG.equals(params[0])) {
shortcutFreq =
WHITELIST_TAG.equals(params[1]) ? 15 : Integer.parseInt(params[1]);
shortcutFreq = WHITELIST_TAG.equals(params[1])
? FormatSpec.SHORTCUT_WHITELIST_FREQUENCY
: Integer.parseInt(params[1]);
}
}
if (null != shortcut) {

37
tools/dicttool/src/com/android/inputmethod/latin/dicttool/Info.java
View file

@ -18,8 +18,11 @@ package com.android.inputmethod.latin.dicttool;
import com.android.inputmethod.latin.dicttool.BinaryDictOffdeviceUtils.DecoderChainSpec;
import com.android.inputmethod.latin.makedict.BinaryDictInputOutput;
import com.android.inputmethod.latin.makedict.FormatSpec;
import com.android.inputmethod.latin.makedict.FusionDictionary;
import com.android.inputmethod.latin.makedict.FusionDictionary.WeightedString;
import com.android.inputmethod.latin.makedict.UnsupportedFormatException;
import com.android.inputmethod.latin.makedict.Word;
import org.xml.sax.SAXException;
@ -38,8 +41,9 @@ public class Info extends Dicttool.Command {
public Info() {
}
@Override
public String getHelp() {
return "info <filename>: prints various information about a dictionary file";
return COMMAND + "<filename>: prints various information about a dictionary file";
}
private static void crash(final String filename, final Exception e) {
@ -72,6 +76,7 @@ public class Info extends Dicttool.Command {
FileChannel.MapMode.READ_ONLY, 0, decodedSpec.mFile.length());
System.out.println("Format : Binary dictionary format");
System.out.println("Packaging : " + decodedSpec.describeChain());
System.out.println("Uncompressed size : " + decodedSpec.mFile.length());
return BinaryDictInputOutput.readDictionaryBinary(
new BinaryDictInputOutput.ByteBufferWrapper(buffer), null);
}
@ -87,12 +92,40 @@ public class Info extends Dicttool.Command {
return null;
}
private static void showInfo(final FusionDictionary dict) {
System.out.println("Header attributes :");
System.out.print(dict.mOptions.toString(2));
int wordCount = 0;
int bigramCount = 0;
int shortcutCount = 0;
int whitelistCount = 0;
for (final Word w : dict) {
++wordCount;
if (null != w.mBigrams) {
bigramCount += w.mBigrams.size();
}
if (null != w.mShortcutTargets) {
shortcutCount += w.mShortcutTargets.size();
for (WeightedString shortcutTarget : w.mShortcutTargets) {
if (FormatSpec.SHORTCUT_WHITELIST_FREQUENCY == shortcutTarget.mFrequency) {
++whitelistCount;
}
}
}
}
System.out.println("Words in the dictionary : " + wordCount);
System.out.println("Bigram count : " + bigramCount);
System.out.println("Shortcuts : " + shortcutCount + " (out of which " + whitelistCount
+ " whitelist entries)");
}
@Override
public void run() {
// TODO: implement this
if (mArgs.length < 1) {
throw new RuntimeException("Not enough arguments for command " + COMMAND);
}
final String filename = mArgs[0];
final FusionDictionary dict = getDictionary(filename);
showInfo(dict);
}
}