[Refactor] Move some helper methods to BinaryDictIOUtils.

Change-Id: Ib817a975dc1f82241f732b236c44b042fda25b3c
main
Yuichiro Hanada 2013-08-14 19:18:28 +09:00
parent be4d38e61f
commit 1db93c9c04
3 changed files with 78 additions and 80 deletions

View File

@ -87,7 +87,7 @@ public final class BinaryDictIOUtils {
if (p.mNumOfCharGroup == Position.NOT_READ_GROUPCOUNT) { if (p.mNumOfCharGroup == Position.NOT_READ_GROUPCOUNT) {
p.mNumOfCharGroup = BinaryDictInputOutput.readCharGroupCount(buffer); p.mNumOfCharGroup = BinaryDictInputOutput.readCharGroupCount(buffer);
p.mAddress += BinaryDictInputOutput.getGroupCountSize(p.mNumOfCharGroup); p.mAddress += getGroupCountSize(p.mNumOfCharGroup);
p.mPosition = 0; p.mPosition = 0;
} }
if (p.mNumOfCharGroup == 0) { if (p.mNumOfCharGroup == 0) {
@ -101,9 +101,9 @@ public final class BinaryDictIOUtils {
} }
p.mPosition++; p.mPosition++;
final boolean isMovedGroup = BinaryDictInputOutput.isMovedGroup(info.mFlags, final boolean isMovedGroup = isMovedGroup(info.mFlags,
formatOptions); formatOptions);
final boolean isDeletedGroup = BinaryDictInputOutput.isDeletedGroup(info.mFlags, final boolean isDeletedGroup = isDeletedGroup(info.mFlags,
formatOptions); formatOptions);
if (!isMovedGroup && !isDeletedGroup if (!isMovedGroup && !isDeletedGroup
&& info.mFrequency != FusionDictionary.CharGroup.NOT_A_TERMINAL) {// found word && info.mFrequency != FusionDictionary.CharGroup.NOT_A_TERMINAL) {// found word
@ -130,7 +130,7 @@ public final class BinaryDictIOUtils {
p.mAddress = buffer.position(); p.mAddress = buffer.position();
} }
if (!isMovedGroup && BinaryDictInputOutput.hasChildrenAddress(info.mChildrenAddress)) { if (!isMovedGroup && hasChildrenAddress(info.mChildrenAddress)) {
Position childrenPos = new Position(info.mChildrenAddress + headerSize, index); Position childrenPos = new Position(info.mChildrenAddress + headerSize, index);
stack.push(childrenPos); stack.push(childrenPos);
} }
@ -187,12 +187,10 @@ public final class BinaryDictIOUtils {
final int charGroupPos = buffer.position(); final int charGroupPos = buffer.position();
final CharGroupInfo currentInfo = BinaryDictInputOutput.readCharGroup(buffer, final CharGroupInfo currentInfo = BinaryDictInputOutput.readCharGroup(buffer,
buffer.position(), header.mFormatOptions); buffer.position(), header.mFormatOptions);
final boolean isMovedGroup = final boolean isMovedGroup = isMovedGroup(currentInfo.mFlags,
BinaryDictInputOutput.isMovedGroup(currentInfo.mFlags, header.mFormatOptions);
header.mFormatOptions); final boolean isDeletedGroup = isDeletedGroup(currentInfo.mFlags,
final boolean isDeletedGroup = header.mFormatOptions);
BinaryDictInputOutput.isDeletedGroup(currentInfo.mFlags,
header.mFormatOptions);
if (isMovedGroup) continue; if (isMovedGroup) continue;
boolean same = true; boolean same = true;
for (int p = 0, j = word.offsetByCodePoints(0, wordPos); for (int p = 0, j = word.offsetByCodePoints(0, wordPos);
@ -490,8 +488,8 @@ public final class BinaryDictIOUtils {
*/ */
static int writeNode(final OutputStream destination, final CharGroupInfo[] infos) static int writeNode(final OutputStream destination, final CharGroupInfo[] infos)
throws IOException { throws IOException {
int size = BinaryDictInputOutput.getGroupCountSize(infos.length); int size = getGroupCountSize(infos.length);
switch (BinaryDictInputOutput.getGroupCountSize(infos.length)) { switch (getGroupCountSize(infos.length)) {
case 1: case 1:
destination.write((byte)infos.length); destination.write((byte)infos.length);
break; break;
@ -566,4 +564,52 @@ public final class BinaryDictIOUtils {
return null; return null;
} }
} }
/**
* Helper method to hide the actual value of the no children address.
*/
public static boolean hasChildrenAddress(final int address) {
return FormatSpec.NO_CHILDREN_ADDRESS != address;
}
/**
* Helper method to check whether the group is moved.
*/
public static boolean isMovedGroup(final int flags, final FormatOptions options) {
return options.mSupportsDynamicUpdate
&& ((flags & FormatSpec.MASK_GROUP_ADDRESS_TYPE) == FormatSpec.FLAG_IS_MOVED);
}
/**
* Helper method to check whether the dictionary can be updated dynamically.
*/
public static boolean supportsDynamicUpdate(final FormatOptions options) {
return options.mVersion >= FormatSpec.FIRST_VERSION_WITH_DYNAMIC_UPDATE
&& options.mSupportsDynamicUpdate;
}
/**
* Helper method to check whether the group is deleted.
*/
public static boolean isDeletedGroup(final int flags, final FormatOptions formatOptions) {
return formatOptions.mSupportsDynamicUpdate
&& ((flags & FormatSpec.MASK_GROUP_ADDRESS_TYPE) == FormatSpec.FLAG_IS_DELETED);
}
/**
* Compute the binary size of the group count
* @param count the group count
* @return the size of the group count, either 1 or 2 bytes.
*/
public static int getGroupCountSize(final int count) {
if (FormatSpec.MAX_CHARGROUPS_FOR_ONE_BYTE_CHARGROUP_COUNT >= count) {
return 1;
} else if (FormatSpec.MAX_CHARGROUPS_IN_A_NODE >= count) {
return 2;
} else {
throw new RuntimeException("Can't have more than "
+ FormatSpec.MAX_CHARGROUPS_IN_A_NODE + " groups in a node (found " + count
+ ")");
}
}
} }

View File

@ -306,30 +306,13 @@ public final class BinaryDictInputOutput {
return getGroupCharactersSize(group.mChars); return getGroupCharactersSize(group.mChars);
} }
/**
* Compute the binary size of the group count
* @param count the group count
* @return the size of the group count, either 1 or 2 bytes.
*/
public static int getGroupCountSize(final int count) {
if (FormatSpec.MAX_CHARGROUPS_FOR_ONE_BYTE_CHARGROUP_COUNT >= count) {
return 1;
} else if (FormatSpec.MAX_CHARGROUPS_IN_A_NODE >= count) {
return 2;
} else {
throw new RuntimeException("Can't have more than "
+ FormatSpec.MAX_CHARGROUPS_IN_A_NODE + " groups in a node (found " + count
+ ")");
}
}
/** /**
* Compute the binary size of the group count for a node * Compute the binary size of the group count for a node
* @param node the node * @param node the node
* @return the size of the group count, either 1 or 2 bytes. * @return the size of the group count, either 1 or 2 bytes.
*/ */
private static int getGroupCountSize(final Node node) { private static int getGroupCountSize(final Node node) {
return getGroupCountSize(node.mData.size()); return BinaryDictIOUtils.getGroupCountSize(node.mData.size());
} }
/** /**
@ -403,37 +386,6 @@ public final class BinaryDictInputOutput {
node.mCachedSize = size; node.mCachedSize = size;
} }
/**
* Helper method to hide the actual value of the no children address.
*/
public static boolean hasChildrenAddress(final int address) {
return FormatSpec.NO_CHILDREN_ADDRESS != address;
}
/**
* Helper method to check whether the group is moved.
*/
public static boolean isMovedGroup(final int flags, final FormatOptions options) {
return options.mSupportsDynamicUpdate
&& ((flags & FormatSpec.MASK_GROUP_ADDRESS_TYPE) == FormatSpec.FLAG_IS_MOVED);
}
/**
* Helper method to check whether the group is deleted.
*/
public static boolean isDeletedGroup(final int flags, final FormatOptions formatOptions) {
return formatOptions.mSupportsDynamicUpdate
&& ((flags & FormatSpec.MASK_GROUP_ADDRESS_TYPE) == FormatSpec.FLAG_IS_DELETED);
}
/**
* Helper method to check whether the dictionary can be updated dynamically.
*/
public static boolean supportsDynamicUpdate(final FormatOptions options) {
return options.mVersion >= FormatSpec.FIRST_VERSION_WITH_DYNAMIC_UPDATE
&& options.mSupportsDynamicUpdate;
}
/** /**
* Compute the size of the header (flag + [parent address] + characters size) of a CharGroup. * Compute the size of the header (flag + [parent address] + characters size) of a CharGroup.
* *
@ -441,7 +393,7 @@ public final class BinaryDictInputOutput {
* @param options file format options. * @param options file format options.
*/ */
private static int getGroupHeaderSize(final CharGroup group, final FormatOptions options) { private static int getGroupHeaderSize(final CharGroup group, final FormatOptions options) {
if (supportsDynamicUpdate(options)) { if (BinaryDictIOUtils.supportsDynamicUpdate(options)) {
return FormatSpec.GROUP_FLAGS_SIZE + FormatSpec.PARENT_ADDRESS_SIZE return FormatSpec.GROUP_FLAGS_SIZE + FormatSpec.PARENT_ADDRESS_SIZE
+ getGroupCharactersSize(group); + getGroupCharactersSize(group);
} else { } else {
@ -461,7 +413,7 @@ public final class BinaryDictInputOutput {
*/ */
static int getByteSize(final int address) { static int getByteSize(final int address) {
assert(address <= FormatSpec.UINT24_MAX); assert(address <= FormatSpec.UINT24_MAX);
if (!hasChildrenAddress(address)) { if (!BinaryDictIOUtils.hasChildrenAddress(address)) {
return 0; return 0;
} else if (Math.abs(address) <= FormatSpec.UINT8_MAX) { } else if (Math.abs(address) <= FormatSpec.UINT8_MAX) {
return 1; return 1;
@ -802,7 +754,7 @@ public final class BinaryDictInputOutput {
*/ */
private static int writeVariableSignedAddress(final byte[] buffer, int index, private static int writeVariableSignedAddress(final byte[] buffer, int index,
final int address) { final int address) {
if (!hasChildrenAddress(address)) { if (!BinaryDictIOUtils.hasChildrenAddress(address)) {
buffer[index] = buffer[index + 1] = buffer[index + 2] = 0; buffer[index] = buffer[index + 1] = buffer[index + 2] = 0;
} else { } else {
final int absAddress = Math.abs(address); final int absAddress = Math.abs(address);
@ -966,7 +918,7 @@ public final class BinaryDictInputOutput {
private static final int writeParentAddress(final byte[] buffer, final int index, private static final int writeParentAddress(final byte[] buffer, final int index,
final int address, final FormatOptions formatOptions) { final int address, final FormatOptions formatOptions) {
if (supportsDynamicUpdate(formatOptions)) { if (BinaryDictIOUtils.supportsDynamicUpdate(formatOptions)) {
if (address == FormatSpec.NO_PARENT_ADDRESS) { if (address == FormatSpec.NO_PARENT_ADDRESS) {
buffer[index] = buffer[index + 1] = buffer[index + 2] = 0; buffer[index] = buffer[index + 1] = buffer[index + 2] = 0;
} else { } else {
@ -1315,7 +1267,7 @@ public final class BinaryDictInputOutput {
static int readParentAddress(final FusionDictionaryBufferInterface buffer, static int readParentAddress(final FusionDictionaryBufferInterface buffer,
final FormatOptions formatOptions) { final FormatOptions formatOptions) {
if (supportsDynamicUpdate(formatOptions)) { if (BinaryDictIOUtils.supportsDynamicUpdate(formatOptions)) {
final int parentAddress = buffer.readUnsignedInt24(); final int parentAddress = buffer.readUnsignedInt24();
final int sign = ((parentAddress & FormatSpec.MSB24) != 0) ? -1 : 1; final int sign = ((parentAddress & FormatSpec.MSB24) != 0) ? -1 : 1;
return sign * (parentAddress & FormatSpec.SINT24_MAX); return sign * (parentAddress & FormatSpec.SINT24_MAX);
@ -1332,7 +1284,7 @@ public final class BinaryDictInputOutput {
++addressPointer; ++addressPointer;
final int parentAddress = readParentAddress(buffer, options); final int parentAddress = readParentAddress(buffer, options);
if (supportsDynamicUpdate(options)) { if (BinaryDictIOUtils.supportsDynamicUpdate(options)) {
addressPointer += 3; addressPointer += 3;
} }
@ -1459,7 +1411,7 @@ public final class BinaryDictInputOutput {
final int originalPointer = buffer.position(); final int originalPointer = buffer.position();
buffer.position(address); buffer.position(address);
if (supportsDynamicUpdate(formatOptions)) { if (BinaryDictIOUtils.supportsDynamicUpdate(formatOptions)) {
result = getWordAtAddressWithParentAddress(buffer, headerSize, address, formatOptions); result = getWordAtAddressWithParentAddress(buffer, headerSize, address, formatOptions);
} else { } else {
result = getWordAtAddressWithoutParentAddress(buffer, headerSize, address, result = getWordAtAddressWithoutParentAddress(buffer, headerSize, address,
@ -1488,13 +1440,13 @@ public final class BinaryDictInputOutput {
do { do {
buffer.position(currentAddress + headerSize); buffer.position(currentAddress + headerSize);
currentInfo = readCharGroup(buffer, currentAddress, options); currentInfo = readCharGroup(buffer, currentAddress, options);
if (isMovedGroup(currentInfo.mFlags, options)) { if (BinaryDictIOUtils.isMovedGroup(currentInfo.mFlags, options)) {
currentAddress = currentInfo.mParentAddress + currentInfo.mOriginalAddress; currentAddress = currentInfo.mParentAddress + currentInfo.mOriginalAddress;
} }
if (DBG && loopCounter++ > MAX_JUMPS) { if (DBG && loopCounter++ > MAX_JUMPS) {
MakedictLog.d("Too many jumps - probably a bug"); MakedictLog.d("Too many jumps - probably a bug");
} }
} while (isMovedGroup(currentInfo.mFlags, options)); } while (BinaryDictIOUtils.isMovedGroup(currentInfo.mFlags, options));
if (Integer.MIN_VALUE == frequency) frequency = currentInfo.mFrequency; if (Integer.MIN_VALUE == frequency) frequency = currentInfo.mFrequency;
for (int i = 0; i < currentInfo.mCharacters.length; ++i) { for (int i = 0; i < currentInfo.mCharacters.length; ++i) {
sGetWordBuffer[index--] = sGetWordBuffer[index--] =
@ -1514,7 +1466,7 @@ public final class BinaryDictInputOutput {
final FormatOptions options) { final FormatOptions options) {
buffer.position(headerSize); buffer.position(headerSize);
final int count = readCharGroupCount(buffer); final int count = readCharGroupCount(buffer);
int groupOffset = getGroupCountSize(count); int groupOffset = BinaryDictIOUtils.getGroupCountSize(count);
final StringBuilder builder = new StringBuilder(); final StringBuilder builder = new StringBuilder();
WeightedString result = null; WeightedString result = null;
@ -1527,23 +1479,23 @@ public final class BinaryDictInputOutput {
result = new WeightedString(builder.toString(), info.mFrequency); result = new WeightedString(builder.toString(), info.mFrequency);
break; // and return break; // and return
} }
if (hasChildrenAddress(info.mChildrenAddress)) { if (BinaryDictIOUtils.hasChildrenAddress(info.mChildrenAddress)) {
if (info.mChildrenAddress > address) { if (info.mChildrenAddress > address) {
if (null == last) continue; if (null == last) continue;
builder.append(new String(last.mCharacters, 0, last.mCharacters.length)); builder.append(new String(last.mCharacters, 0, last.mCharacters.length));
buffer.position(last.mChildrenAddress + headerSize); buffer.position(last.mChildrenAddress + headerSize);
i = readCharGroupCount(buffer); i = readCharGroupCount(buffer);
groupOffset = last.mChildrenAddress + getGroupCountSize(i); groupOffset = last.mChildrenAddress + BinaryDictIOUtils.getGroupCountSize(i);
last = null; last = null;
continue; continue;
} }
last = info; last = info;
} }
if (0 == i && hasChildrenAddress(last.mChildrenAddress)) { if (0 == i && BinaryDictIOUtils.hasChildrenAddress(last.mChildrenAddress)) {
builder.append(new String(last.mCharacters, 0, last.mCharacters.length)); builder.append(new String(last.mCharacters, 0, last.mCharacters.length));
buffer.position(last.mChildrenAddress + headerSize); buffer.position(last.mChildrenAddress + headerSize);
i = readCharGroupCount(buffer); i = readCharGroupCount(buffer);
groupOffset = last.mChildrenAddress + getGroupCountSize(i); groupOffset = last.mChildrenAddress + BinaryDictIOUtils.getGroupCountSize(i);
last = null; last = null;
continue; continue;
} }
@ -1576,10 +1528,10 @@ public final class BinaryDictInputOutput {
do { // Scan the linked-list node. do { // Scan the linked-list node.
final int nodeHeadPosition = buffer.position() - headerSize; final int nodeHeadPosition = buffer.position() - headerSize;
final int count = readCharGroupCount(buffer); final int count = readCharGroupCount(buffer);
int groupOffset = nodeHeadPosition + getGroupCountSize(count); int groupOffset = nodeHeadPosition + BinaryDictIOUtils.getGroupCountSize(count);
for (int i = count; i > 0; --i) { // Scan the array of CharGroup. for (int i = count; i > 0; --i) { // Scan the array of CharGroup.
CharGroupInfo info = readCharGroup(buffer, groupOffset, options); CharGroupInfo info = readCharGroup(buffer, groupOffset, options);
if (isMovedGroup(info.mFlags, options)) continue; if (BinaryDictIOUtils.isMovedGroup(info.mFlags, options)) continue;
ArrayList<WeightedString> shortcutTargets = info.mShortcutTargets; ArrayList<WeightedString> shortcutTargets = info.mShortcutTargets;
ArrayList<WeightedString> bigrams = null; ArrayList<WeightedString> bigrams = null;
if (null != info.mBigrams) { if (null != info.mBigrams) {
@ -1592,7 +1544,7 @@ public final class BinaryDictInputOutput {
bigrams.add(new WeightedString(word.mWord, reconstructedFrequency)); bigrams.add(new WeightedString(word.mWord, reconstructedFrequency));
} }
} }
if (hasChildrenAddress(info.mChildrenAddress)) { if (BinaryDictIOUtils.hasChildrenAddress(info.mChildrenAddress)) {
Node children = reverseNodeMap.get(info.mChildrenAddress); Node children = reverseNodeMap.get(info.mChildrenAddress);
if (null == children) { if (null == children) {
final int currentPosition = buffer.position(); final int currentPosition = buffer.position();

View File

@ -85,7 +85,7 @@ public final class DynamicBinaryDictIOUtils {
throw new RuntimeException("this file format does not support parent addresses"); throw new RuntimeException("this file format does not support parent addresses");
} }
final int flags = buffer.readUnsignedByte(); final int flags = buffer.readUnsignedByte();
if (BinaryDictInputOutput.isMovedGroup(flags, formatOptions)) { if (BinaryDictIOUtils.isMovedGroup(flags, formatOptions)) {
// if the group is moved, the parent address is stored in the destination group. // if the group is moved, the parent address is stored in the destination group.
// We are guaranteed to process the destination group later, so there is no need to // We are guaranteed to process the destination group later, so there is no need to
// update anything here. // update anything here.
@ -286,7 +286,7 @@ public final class DynamicBinaryDictIOUtils {
address = buffer.position(); address = buffer.position();
final CharGroupInfo currentInfo = BinaryDictInputOutput.readCharGroup(buffer, final CharGroupInfo currentInfo = BinaryDictInputOutput.readCharGroup(buffer,
buffer.position(), header.mFormatOptions); buffer.position(), header.mFormatOptions);
final boolean isMovedGroup = BinaryDictInputOutput.isMovedGroup(currentInfo.mFlags, final boolean isMovedGroup = BinaryDictIOUtils.isMovedGroup(currentInfo.mFlags,
header.mFormatOptions); header.mFormatOptions);
if (isMovedGroup) continue; if (isMovedGroup) continue;
nodeParentAddress = (currentInfo.mParentAddress == FormatSpec.NO_PARENT_ADDRESS) nodeParentAddress = (currentInfo.mParentAddress == FormatSpec.NO_PARENT_ADDRESS)