All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.apache.xerces.util.SymbolTable Maven / Gradle / Ivy

Go to download

Xerces2 provides high performance, fully compliant XML parsers in the Apache Xerces family. This new version of Xerces continues to build upon the Xerces Native Interface (XNI), a complete framework for building parser components and configurations that is extremely modular and easy to program. The Apache Xerces2 parser is the reference implementation of XNI but other parser components, configurations, and parsers can be written using the Xerces Native Interface. For complete design and implementation documents, refer to the XNI Manual. Xerces2 provides fully conforming XML Schema 1.0 and 1.1 processors. An experimental implementation of the "XML Schema Definition Language (XSD): Component Designators (SCD) Candidate Recommendation (January 2010)" is also provided for evaluation. For more information, refer to the XML Schema page. Xerces2 also provides a complete implementation of the Document Object Model Level 3 Core and Load/Save W3C Recommendations and provides a complete implementation of the XML Inclusions (XInclude) W3C Recommendation. It also provides support for OASIS XML Catalogs v1.1. Xerces2 is able to parse documents written according to the XML 1.1 Recommendation, except that it does not yet provide an option to enable normalization checking as described in section 2.13 of this specification. It also handles namespaces according to the XML Namespaces 1.1 Recommendation, and will correctly serialize XML 1.1 documents if the DOM level 3 load/save APIs are in use.

The newest version!
/*
 * 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.xerces.util;

/**
 * This class is a symbol table implementation that guarantees that
 * strings used as identifiers are unique references. Multiple calls
 * to addSymbol will always return the same string
 * reference.
 * 

* The symbol table performs the same task as String.intern() * with the following differences: *

    *
  • * A new string object does not need to be created in order to * retrieve a unique reference. Symbols can be added by using * a series of characters in a character array. *
  • *
  • * Users of the symbol table can provide their own symbol hashing * implementation. For example, a simple string hashing algorithm * may fail to produce a balanced set of hashcodes for symbols * that are mostly unique. Strings with similar leading * characters are especially prone to this poor hashing behavior. *
  • *
* * An instance of SymbolTable has two parameters that affect its * performance: initial capacity and load factor. The * capacity is the number of buckets in the SymbolTable, and the * initial capacity is simply the capacity at the time the SymbolTable * is created. Note that the SymbolTable is open: in the case of a "hash * collision", a single bucket stores multiple entries, which must be searched * sequentially. The load factor is a measure of how full the SymbolTable * is allowed to get before its capacity is automatically increased. * When the number of entries in the SymbolTable exceeds the product of the load * factor and the current capacity, the capacity is increased by calling the * rehash method.

* * Generally, the default load factor (.75) offers a good tradeoff between * time and space costs. Higher values decrease the space overhead but * increase the time cost to look up an entry (which is reflected in most * SymbolTable operations, including addSymbol and containsSymbol).

* * The initial capacity controls a tradeoff between wasted space and the * need for rehash operations, which are time-consuming. * No rehash operations will ever occur if the initial * capacity is greater than the maximum number of entries the * Hashtable will contain divided by its load factor. However, * setting the initial capacity too high can waste space.

* * If many entries are to be made into a SymbolTable, * creating it with a sufficiently large capacity may allow the * entries to be inserted more efficiently than letting it perform * automatic rehashing as needed to grow the table.

* @see SymbolHash * * @author Andy Clark * @author John Kim, IBM * * @version $Id: SymbolTable.java 1358350 2012-07-06 19:04:35Z mrglavas $ */ public class SymbolTable { // // Constants // /** Default table size. */ protected static final int TABLE_SIZE = 101; /** Maximum hash collisions per bucket for a table with load factor == 1. */ protected static final int MAX_HASH_COLLISIONS = 40; protected static final int MULTIPLIERS_SIZE = 1 << 5; protected static final int MULTIPLIERS_MASK = MULTIPLIERS_SIZE - 1; // // Data // /** Buckets. */ protected Entry[] fBuckets = null; /** actual table size **/ protected int fTableSize; /** The total number of entries in the hash table. */ protected transient int fCount; /** The table is rehashed when its size exceeds this threshold. (The * value of this field is (int)(capacity * loadFactor).) */ protected int fThreshold; /** The load factor for the SymbolTable. */ protected float fLoadFactor; /** * A new hash function is selected and the table is rehashed when * the number of keys in the bucket exceeds this threshold. */ protected final int fCollisionThreshold; /** * Array of randomly selected hash function multipliers or null * if the default String.hashCode() function should be used. */ protected int[] fHashMultipliers; // // Constructors // /** * Constructs a new, empty SymbolTable with the specified initial * capacity and the specified load factor. * * @param initialCapacity the initial capacity of the SymbolTable. * @param loadFactor the load factor of the SymbolTable. * @throws IllegalArgumentException if the initial capacity is less * than zero, or if the load factor is nonpositive. */ public SymbolTable(int initialCapacity, float loadFactor) { if (initialCapacity < 0) { throw new IllegalArgumentException("Illegal Capacity: " + initialCapacity); } if (loadFactor <= 0 || Float.isNaN(loadFactor)) { throw new IllegalArgumentException("Illegal Load: " + loadFactor); } if (initialCapacity == 0) { initialCapacity = 1; } fLoadFactor = loadFactor; fTableSize = initialCapacity; fBuckets = new Entry[fTableSize]; fThreshold = (int)(fTableSize * loadFactor); fCollisionThreshold = (int)(MAX_HASH_COLLISIONS * loadFactor); fCount = 0; } /** * Constructs a new, empty SymbolTable with the specified initial capacity * and default load factor, which is 0.75. * * @param initialCapacity the initial capacity of the hashtable. * @throws IllegalArgumentException if the initial capacity is less * than zero. */ public SymbolTable(int initialCapacity) { this(initialCapacity, 0.75f); } /** * Constructs a new, empty SymbolTable with a default initial capacity (101) * and load factor, which is 0.75. */ public SymbolTable() { this(TABLE_SIZE, 0.75f); } // // Public methods // /** * Adds the specified symbol to the symbol table and returns a * reference to the unique symbol. If the symbol already exists, * the previous symbol reference is returned instead, in order * guarantee that symbol references remain unique. * * @param symbol The new symbol. */ public String addSymbol(String symbol) { // search for identical symbol int collisionCount = 0; int bucket = hash(symbol) % fTableSize; for (Entry entry = fBuckets[bucket]; entry != null; entry = entry.next) { if (entry.symbol.equals(symbol)) { return entry.symbol; } ++collisionCount; } return addSymbol0(symbol, bucket, collisionCount); } // addSymbol(String):String private String addSymbol0(String symbol, int bucket, int collisionCount) { if (fCount >= fThreshold) { // Rehash the table if the threshold is exceeded rehash(); bucket = hash(symbol) % fTableSize; } else if (collisionCount >= fCollisionThreshold) { // Select a new hash function and rehash the table if // the collision threshold is exceeded. rebalance(); bucket = hash(symbol) % fTableSize; } // create new entry Entry entry = new Entry(symbol, fBuckets[bucket]); fBuckets[bucket] = entry; ++fCount; return entry.symbol; } // addSymbol0(String,int,int):String /** * Adds the specified symbol to the symbol table and returns a * reference to the unique symbol. If the symbol already exists, * the previous symbol reference is returned instead, in order * guarantee that symbol references remain unique. * * @param buffer The buffer containing the new symbol. * @param offset The offset into the buffer of the new symbol. * @param length The length of the new symbol in the buffer. */ public String addSymbol(char[] buffer, int offset, int length) { // search for identical symbol int collisionCount = 0; int bucket = hash(buffer, offset, length) % fTableSize; OUTER: for (Entry entry = fBuckets[bucket]; entry != null; entry = entry.next) { if (length == entry.characters.length) { for (int i = 0; i < length; i++) { if (buffer[offset + i] != entry.characters[i]) { ++collisionCount; continue OUTER; } } return entry.symbol; } ++collisionCount; } return addSymbol0(buffer, offset, length, bucket, collisionCount); } // addSymbol(char[],int,int):String private String addSymbol0(char[] buffer, int offset, int length, int bucket, int collisionCount) { if (fCount >= fThreshold) { // Rehash the table if the threshold is exceeded rehash(); bucket = hash(buffer, offset, length) % fTableSize; } else if (collisionCount >= fCollisionThreshold) { // Select a new hash function and rehash the table if // the collision threshold is exceeded. rebalance(); bucket = hash(buffer, offset, length) % fTableSize; } // add new entry Entry entry = new Entry(buffer, offset, length, fBuckets[bucket]); fBuckets[bucket] = entry; ++fCount; return entry.symbol; } // addSymbol0(char[],int,int,int,int):String /** * Returns a hashcode value for the specified symbol. The value * returned by this method must be identical to the value returned * by the hash(char[],int,int) method when called * with the character array that comprises the symbol string. * * @param symbol The symbol to hash. */ public int hash(String symbol) { if (fHashMultipliers == null) { return symbol.hashCode() & 0x7FFFFFFF; } return hash0(symbol); } // hash(String):int private int hash0(String symbol) { int code = 0; final int length = symbol.length(); final int[] multipliers = fHashMultipliers; for (int i = 0; i < length; ++i) { code = code * multipliers[i & MULTIPLIERS_MASK] + symbol.charAt(i); } return code & 0x7FFFFFFF; } // hash0(String):int /** * Returns a hashcode value for the specified symbol information. * The value returned by this method must be identical to the value * returned by the hash(String) method when called * with the string object created from the symbol information. * * @param buffer The character buffer containing the symbol. * @param offset The offset into the character buffer of the start * of the symbol. * @param length The length of the symbol. */ public int hash(char[] buffer, int offset, int length) { if (fHashMultipliers == null) { int code = 0; for (int i = 0; i < length; ++i) { code = code * 31 + buffer[offset + i]; } return code & 0x7FFFFFFF; } return hash0(buffer, offset, length); } // hash(char[],int,int):int private int hash0(char[] buffer, int offset, int length) { int code = 0; final int[] multipliers = fHashMultipliers; for (int i = 0; i < length; ++i) { code = code * multipliers[i & MULTIPLIERS_MASK] + buffer[offset + i]; } return code & 0x7FFFFFFF; } // hash0(char[],int,int):int /** * Increases the capacity of and internally reorganizes this * SymbolTable, in order to accommodate and access its entries more * efficiently. This method is called automatically when the * number of keys in the SymbolTable exceeds this hashtable's capacity * and load factor. */ protected void rehash() { rehashCommon(fBuckets.length * 2 + 1); } /** * Randomly selects a new hash function and reorganizes this SymbolTable * in order to more evenly distribute its entries across the table. This * method is called automatically when the number keys in one of the * SymbolTable's buckets exceeds the given collision threshold. */ protected void rebalance() { if (fHashMultipliers == null) { fHashMultipliers = new int[MULTIPLIERS_SIZE]; } PrimeNumberSequenceGenerator.generateSequence(fHashMultipliers); rehashCommon(fBuckets.length); } private void rehashCommon(final int newCapacity) { int oldCapacity = fBuckets.length; Entry[] oldTable = fBuckets; Entry[] newTable = new Entry[newCapacity]; fThreshold = (int)(newCapacity * fLoadFactor); fBuckets = newTable; fTableSize = fBuckets.length; for (int i = oldCapacity ; i-- > 0 ;) { for (Entry old = oldTable[i] ; old != null ; ) { Entry e = old; old = old.next; int index = hash(e.symbol) % newCapacity; e.next = newTable[index]; newTable[index] = e; } } } /** * Returns true if the symbol table already contains the specified * symbol. * * @param symbol The symbol to look for. */ public boolean containsSymbol(String symbol) { // search for identical symbol int bucket = hash(symbol) % fTableSize; int length = symbol.length(); OUTER: for (Entry entry = fBuckets[bucket]; entry != null; entry = entry.next) { if (length == entry.characters.length) { for (int i = 0; i < length; i++) { if (symbol.charAt(i) != entry.characters[i]) { continue OUTER; } } return true; } } return false; } // containsSymbol(String):boolean /** * Returns true if the symbol table already contains the specified * symbol. * * @param buffer The buffer containing the symbol to look for. * @param offset The offset into the buffer. * @param length The length of the symbol in the buffer. */ public boolean containsSymbol(char[] buffer, int offset, int length) { // search for identical symbol int bucket = hash(buffer, offset, length) % fTableSize; OUTER: for (Entry entry = fBuckets[bucket]; entry != null; entry = entry.next) { if (length == entry.characters.length) { for (int i = 0; i < length; i++) { if (buffer[offset + i] != entry.characters[i]) { continue OUTER; } } return true; } } return false; } // containsSymbol(char[],int,int):boolean // // Classes // /** * This class is a symbol table entry. Each entry acts as a node * in a linked list. */ protected static final class Entry { // // Data // /** Symbol. */ public final String symbol; /** * Symbol characters. This information is duplicated here for * comparison performance. */ public final char[] characters; /** The next entry. */ public Entry next; // // Constructors // /** * Constructs a new entry from the specified symbol and next entry * reference. */ public Entry(String symbol, Entry next) { this.symbol = symbol.intern(); characters = new char[symbol.length()]; symbol.getChars(0, characters.length, characters, 0); this.next = next; } /** * Constructs a new entry from the specified symbol information and * next entry reference. */ public Entry(char[] ch, int offset, int length, Entry next) { characters = new char[length]; System.arraycopy(ch, offset, characters, 0, length); symbol = new String(characters).intern(); this.next = next; } } // class Entry } // class SymbolTable





© 2015 - 2024 Weber Informatics LLC | Privacy Policy