org.apache.lucene.analysis.compound.hyphenation.TernaryTree Maven / Gradle / Ivy
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.lucene.analysis.compound.hyphenation;
import java.io.PrintStream;
import java.util.Enumeration;
import java.util.Stack;
/**
* Ternary Search Tree.
*
*
* A ternary search tree is a hybrid between a binary tree and a digital search
* tree (trie). Keys are limited to strings. A data value of type char is stored
* in each leaf node. It can be used as an index (or pointer) to the data.
* Branches that only contain one key are compressed to one node by storing a
* pointer to the trailer substring of the key. This class is intended to serve
* as base class or helper class to implement Dictionary collections or the
* like. Ternary trees have some nice properties as the following: the tree can
* be traversed in sorted order, partial matches (wildcard) can be implemented,
* retrieval of all keys within a given distance from the target, etc. The
* storage requirements are higher than a binary tree but a lot less than a
* trie. Performance is comparable with a hash table, sometimes it outperforms a
* hash function (most of the time can determine a miss faster than a hash).
*
*
*
* The main purpose of this java port is to serve as a base for implementing
* TeX's hyphenation algorithm (see The TeXBook, appendix H). Each language
* requires from 5000 to 15000 hyphenation patterns which will be keys in this
* tree. The strings patterns are usually small (from 2 to 5 characters), but
* each char in the tree is stored in a node. Thus memory usage is the main
* concern. We will sacrifice 'elegance' to keep memory requirements to the
* minimum. Using java's char type as pointer (yes, I know pointer it is a
* forbidden word in java) we can keep the size of the node to be just 8 bytes
* (3 pointers and the data char). This gives room for about 65000 nodes. In my
* tests the english patterns took 7694 nodes and the german patterns 10055
* nodes, so I think we are safe.
*
*
*
* All said, this is a map with strings as keys and char as value. Pretty
* limited!. It can be extended to a general map by using the string
* representation of an object and using the char value as an index to an array
* that contains the object values.
*
*
* This class has been taken from the Apache FOP project (http://xmlgraphics.apache.org/fop/). They have been slightly modified.
*/
public class TernaryTree implements Cloneable {
/**
* We use 4 arrays to represent a node. I guess I should have created a proper
* node class, but somehow Knuth's pascal code made me forget we now have a
* portable language with virtual memory management and automatic garbage
* collection! And now is kind of late, furthermore, if it ain't broken, don't
* fix it.
*/
/**
* Pointer to low branch and to rest of the key when it is stored directly in
* this node, we don't have unions in java!
*/
protected char[] lo;
/**
* Pointer to high branch.
*/
protected char[] hi;
/**
* Pointer to equal branch and to data when this node is a string terminator.
*/
protected char[] eq;
/**
*
* The character stored in this node: splitchar. Two special values are
* reserved:
*
*
* - 0x0000 as string terminator
* - 0xFFFF to indicate that the branch starting at this node is compressed
*
*
* This shouldn't be a problem if we give the usual semantics to strings since
* 0xFFFF is guaranteed not to be an Unicode character.
*
*/
protected char[] sc;
/**
* This vector holds the trailing of the keys when the branch is compressed.
*/
protected CharVector kv;
protected char root;
protected char freenode;
protected int length; // number of items in tree
protected static final int BLOCK_SIZE = 2048; // allocation size for arrays
TernaryTree() {
init();
}
protected void init() {
root = 0;
freenode = 1;
length = 0;
lo = new char[BLOCK_SIZE];
hi = new char[BLOCK_SIZE];
eq = new char[BLOCK_SIZE];
sc = new char[BLOCK_SIZE];
kv = new CharVector();
}
/**
* Branches are initially compressed, needing one node per key plus the size
* of the string key. They are decompressed as needed when another key with
* same prefix is inserted. This saves a lot of space, specially for long
* keys.
*/
public void insert(String key, char val) {
// make sure we have enough room in the arrays
int len = key.length() + 1; // maximum number of nodes that may be generated
if (freenode + len > eq.length) {
redimNodeArrays(eq.length + BLOCK_SIZE);
}
char strkey[] = new char[len--];
key.getChars(0, len, strkey, 0);
strkey[len] = 0;
root = insert(root, strkey, 0, val);
}
public void insert(char[] key, int start, char val) {
int len = strlen(key) + 1;
if (freenode + len > eq.length) {
redimNodeArrays(eq.length + BLOCK_SIZE);
}
root = insert(root, key, start, val);
}
/**
* The actual insertion function, recursive version.
*/
private char insert(char p, char[] key, int start, char val) {
int len = strlen(key, start);
if (p == 0) {
// this means there is no branch, this node will start a new branch.
// Instead of doing that, we store the key somewhere else and create
// only one node with a pointer to the key
p = freenode++;
eq[p] = val; // holds data
length++;
hi[p] = 0;
if (len > 0) {
sc[p] = 0xFFFF; // indicates branch is compressed
lo[p] = (char) kv.alloc(len + 1); // use 'lo' to hold pointer to key
strcpy(kv.getArray(), lo[p], key, start);
} else {
sc[p] = 0;
lo[p] = 0;
}
return p;
}
if (sc[p] == 0xFFFF) {
// branch is compressed: need to decompress
// this will generate garbage in the external key array
// but we can do some garbage collection later
char pp = freenode++;
lo[pp] = lo[p]; // previous pointer to key
eq[pp] = eq[p]; // previous pointer to data
lo[p] = 0;
if (len > 0) {
sc[p] = kv.get(lo[pp]);
eq[p] = pp;
lo[pp]++;
if (kv.get(lo[pp]) == 0) {
// key completly decompressed leaving garbage in key array
lo[pp] = 0;
sc[pp] = 0;
hi[pp] = 0;
} else {
// we only got first char of key, rest is still there
sc[pp] = 0xFFFF;
}
} else {
// In this case we can save a node by swapping the new node
// with the compressed node
sc[pp] = 0xFFFF;
hi[p] = pp;
sc[p] = 0;
eq[p] = val;
length++;
return p;
}
}
char s = key[start];
if (s < sc[p]) {
lo[p] = insert(lo[p], key, start, val);
} else if (s == sc[p]) {
if (s != 0) {
eq[p] = insert(eq[p], key, start + 1, val);
} else {
// key already in tree, overwrite data
eq[p] = val;
}
} else {
hi[p] = insert(hi[p], key, start, val);
}
return p;
}
/**
* Compares 2 null terminated char arrays
*/
public static int strcmp(char[] a, int startA, char[] b, int startB) {
for (; a[startA] == b[startB]; startA++, startB++) {
if (a[startA] == 0) {
return 0;
}
}
return a[startA] - b[startB];
}
/**
* Compares a string with null terminated char array
*/
public static int strcmp(String str, char[] a, int start) {
int i, d, len = str.length();
for (i = 0; i < len; i++) {
d = (int) str.charAt(i) - a[start + i];
if (d != 0) {
return d;
}
if (a[start + i] == 0) {
return d;
}
}
if (a[start + i] != 0) {
return -a[start + i];
}
return 0;
}
public static void strcpy(char[] dst, int di, char[] src, int si) {
while (src[si] != 0) {
dst[di++] = src[si++];
}
dst[di] = 0;
}
public static int strlen(char[] a, int start) {
int len = 0;
for (int i = start; i < a.length && a[i] != 0; i++) {
len++;
}
return len;
}
public static int strlen(char[] a) {
return strlen(a, 0);
}
public int find(String key) {
int len = key.length();
char strkey[] = new char[len + 1];
key.getChars(0, len, strkey, 0);
strkey[len] = 0;
return find(strkey, 0);
}
public int find(char[] key, int start) {
int d;
char p = root;
int i = start;
char c;
while (p != 0) {
if (sc[p] == 0xFFFF) {
if (strcmp(key, i, kv.getArray(), lo[p]) == 0) {
return eq[p];
} else {
return -1;
}
}
c = key[i];
d = c - sc[p];
if (d == 0) {
if (c == 0) {
return eq[p];
}
i++;
p = eq[p];
} else if (d < 0) {
p = lo[p];
} else {
p = hi[p];
}
}
return -1;
}
public boolean knows(String key) {
return (find(key) >= 0);
}
// redimension the arrays
private void redimNodeArrays(int newsize) {
int len = newsize < lo.length ? newsize : lo.length;
char[] na = new char[newsize];
System.arraycopy(lo, 0, na, 0, len);
lo = na;
na = new char[newsize];
System.arraycopy(hi, 0, na, 0, len);
hi = na;
na = new char[newsize];
System.arraycopy(eq, 0, na, 0, len);
eq = na;
na = new char[newsize];
System.arraycopy(sc, 0, na, 0, len);
sc = na;
}
public int size() {
return length;
}
@Override
public TernaryTree clone() {
TernaryTree t = new TernaryTree();
t.lo = this.lo.clone();
t.hi = this.hi.clone();
t.eq = this.eq.clone();
t.sc = this.sc.clone();
t.kv = this.kv.clone();
t.root = this.root;
t.freenode = this.freenode;
t.length = this.length;
return t;
}
/**
* Recursively insert the median first and then the median of the lower and
* upper halves, and so on in order to get a balanced tree. The array of keys
* is assumed to be sorted in ascending order.
*/
protected void insertBalanced(String[] k, char[] v, int offset, int n) {
int m;
if (n < 1) {
return;
}
m = n >> 1;
insert(k[m + offset], v[m + offset]);
insertBalanced(k, v, offset, m);
insertBalanced(k, v, offset + m + 1, n - m - 1);
}
/**
* Balance the tree for best search performance
*/
public void balance() {
// System.out.print("Before root splitchar = ");
// System.out.println(sc[root]);
int i = 0, n = length;
String[] k = new String[n];
char[] v = new char[n];
Iterator iter = new Iterator();
while (iter.hasMoreElements()) {
v[i] = iter.getValue();
k[i++] = iter.nextElement();
}
init();
insertBalanced(k, v, 0, n);
// With uniform letter distribution sc[root] should be around 'm'
// System.out.print("After root splitchar = ");
// System.out.println(sc[root]);
}
/**
* Each node stores a character (splitchar) which is part of some key(s). In a
* compressed branch (one that only contain a single string key) the trailer
* of the key which is not already in nodes is stored externally in the kv
* array. As items are inserted, key substrings decrease. Some substrings may
* completely disappear when the whole branch is totally decompressed. The
* tree is traversed to find the key substrings actually used. In addition,
* duplicate substrings are removed using a map (implemented with a
* TernaryTree!).
*
*/
public void trimToSize() {
// first balance the tree for best performance
balance();
// redimension the node arrays
redimNodeArrays(freenode);
// ok, compact kv array
CharVector kx = new CharVector();
kx.alloc(1);
TernaryTree map = new TernaryTree();
compact(kx, map, root);
kv = kx;
kv.trimToSize();
}
private void compact(CharVector kx, TernaryTree map, char p) {
int k;
if (p == 0) {
return;
}
if (sc[p] == 0xFFFF) {
k = map.find(kv.getArray(), lo[p]);
if (k < 0) {
k = kx.alloc(strlen(kv.getArray(), lo[p]) + 1);
strcpy(kx.getArray(), k, kv.getArray(), lo[p]);
map.insert(kx.getArray(), k, (char) k);
}
lo[p] = (char) k;
} else {
compact(kx, map, lo[p]);
if (sc[p] != 0) {
compact(kx, map, eq[p]);
}
compact(kx, map, hi[p]);
}
}
public Enumeration keys() {
return new Iterator();
}
public class Iterator implements Enumeration {
/**
* current node index
*/
int cur;
/**
* current key
*/
String curkey;
private class Item implements Cloneable {
char parent;
char child;
public Item() {
parent = 0;
child = 0;
}
public Item(char p, char c) {
parent = p;
child = c;
}
@Override
public Item clone() {
return new Item(parent, child);
}
}
/**
* Node stack
*/
Stack- ns;
/**
* key stack implemented with a StringBuilder
*/
StringBuilder ks;
public Iterator() {
cur = -1;
ns = new Stack<>();
ks = new StringBuilder();
rewind();
}
public void rewind() {
ns.removeAllElements();
ks.setLength(0);
cur = root;
run();
}
@Override
public String nextElement() {
String res = curkey;
cur = up();
run();
return res;
}
public char getValue() {
if (cur >= 0) {
return eq[cur];
}
return 0;
}
@Override
public boolean hasMoreElements() {
return (cur != -1);
}
/**
* traverse upwards
*/
private int up() {
Item i = new Item();
int res = 0;
if (ns.empty()) {
return -1;
}
if (cur != 0 && sc[cur] == 0) {
return lo[cur];
}
boolean climb = true;
while (climb) {
i = ns.pop();
i.child++;
switch (i.child) {
case 1:
if (sc[i.parent] != 0) {
res = eq[i.parent];
ns.push(i.clone());
ks.append(sc[i.parent]);
} else {
i.child++;
ns.push(i.clone());
res = hi[i.parent];
}
climb = false;
break;
case 2:
res = hi[i.parent];
ns.push(i.clone());
if (ks.length() > 0) {
ks.setLength(ks.length() - 1); // pop
}
climb = false;
break;
default:
if (ns.empty()) {
return -1;
}
climb = true;
break;
}
}
return res;
}
/**
* traverse the tree to find next key
*/
private int run() {
if (cur == -1) {
return -1;
}
boolean leaf = false;
while (true) {
// first go down on low branch until leaf or compressed branch
while (cur != 0) {
if (sc[cur] == 0xFFFF) {
leaf = true;
break;
}
ns.push(new Item((char) cur, '\u0000'));
if (sc[cur] == 0) {
leaf = true;
break;
}
cur = lo[cur];
}
if (leaf) {
break;
}
// nothing found, go up one node and try again
cur = up();
if (cur == -1) {
return -1;
}
}
// The current node should be a data node and
// the key should be in the key stack (at least partially)
StringBuilder buf = new StringBuilder(ks.toString());
if (sc[cur] == 0xFFFF) {
int p = lo[cur];
while (kv.get(p) != 0) {
buf.append(kv.get(p++));
}
}
curkey = buf.toString();
return 0;
}
}
public void printStats(PrintStream out) {
out.println("Number of keys = " + Integer.toString(length));
out.println("Node count = " + Integer.toString(freenode));
// System.out.println("Array length = " + Integer.toString(eq.length));
out.println("Key Array length = " + Integer.toString(kv.length()));
/*
* for(int i=0; i