src.java.com.ctc.wstx.dtd.ElementIdMap Maven / Gradle / Ivy
/* Woodstox XML processor
*
* Copyright (c) 2004- Tatu Saloranta, [email protected]
*
* Licensed under the License specified in the file LICENSE which is
* included with the source code.
* You may not use this file except in compliance with the License.
*
* 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 com.ctc.wstx.dtd;
import javax.xml.stream.Location;
import com.ctc.wstx.util.ExceptionUtil;
/**
* This class is a specialized type-safe linked hash map used for
* storing {@link ElementId} instances. {@link ElementId} instances
* represent both id definitions (values of element attributes that
* have type ID in DTD), and references (values of element attributes
* of type IDREF and IDREFS). These definitions and references are
* stored for the purpose of verifying
* that all referenced id values are defined, and that none are defined
* more than once.
*/
final class ElementIdMap
{
/**
* Default initial table size; set so that usually it need not
* be expanded.
*/
protected static final int DEFAULT_SIZE = 128;
protected static final int MIN_SIZE = 16;
/**
* Let's use 80% fill factor...
*/
protected static final int FILL_PCT = 80;
/*
////////////////////////////////////////
// Actual hash table structure
////////////////////////////////////////
*/
/**
* Actual hash table area
*/
protected ElementId[] mTable;
/**
* Current size (number of entries); needed to know if and when
* rehash.
*/
protected int mSize;
/**
* Limit that indicates maximum size this instance can hold before
* it needs to be expanded and rehashed. Calculated using fill
* factor passed in to constructor.
*/
protected int mSizeThreshold;
/**
* Mask used to get index from hash values; equal to
* mBuckets.length - 1, when mBuckets.length is
* a power of two.
*/
protected int mIndexMask;
/*
////////////////////////////////////////
// Linked list info
////////////////////////////////////////
*/
protected ElementId mHead;
protected ElementId mTail;
/*
////////////////////////////////////////
// Life-cycle:
////////////////////////////////////////
*/
public ElementIdMap()
{
this(DEFAULT_SIZE);
}
/**
* This constructor is mainly used for testing, as it can be sized
* appropriately to test rehashing etc.
*/
public ElementIdMap(int initialSize)
{
int actual = MIN_SIZE;
while (actual < initialSize) {
actual += actual;
}
mTable = new ElementId[actual];
// Mask is easy to calc for powers of two.
mIndexMask = actual - 1;
mSize = 0;
mSizeThreshold = (actual * FILL_PCT) / 100;
mHead = mTail = null;
}
/*
////////////////////////////////////////////////////
// Public API
////////////////////////////////////////////////////
*/
public ElementId getFirstUndefined()
{
/* Since the linked list is pruned to always start with
* the first (in doc order) undefined id, we can just
* return head:
*/
return mHead;
}
/**
* Method called when a reference to id is encountered. If so, need
* to check if specified id entry (ref or definiton) exists; and if not,
* to add a reference marker.
*/
public ElementId addReferenced(char[] buffer, int start, int len, int hash,
Location loc, NameKey elemName, NameKey attrName)
{
int index = (hash & mIndexMask);
ElementId id = mTable[index];
while (id != null) {
if (id.idMatches(buffer, start, len)) { // found existing one
return id;
}
id = id.mNextColl;
}
// Not found, need to create a placeholder...
// But first, do we need more room?
if (mSize >= mSizeThreshold) {
rehash();
// Index changes, for the new entr:
index = (hash & mIndexMask);
}
++mSize;
// Ok, then, let's create the entry
String idStr = new String(buffer, start, len);
id = new ElementId(idStr, loc, false, elemName, attrName);
// First, let's link it to Map; all ids have to be connected
id.mNextColl = mTable[index];
mTable[index] = id;
// And then add the undefined entry at the end of list
if (mHead == null) {
mHead = mTail = id;
} else {
mTail.mNextUndefd = id;
mTail = id;
}
return id;
}
/**
* Method called when an id definition is encountered. If so, need
* to check if specified id entry (ref or definiton) exists. If not,
* need to add the definition marker. If it does exist, need to
* 'upgrade it', if it was a reference marker; otherwise need to
* just return the old entry, and expect caller to check for dups
* and report the error.
*/
public ElementId addDefined(char[] buffer, int start, int len, int hash,
Location loc, NameKey elemName, NameKey attrName)
{
int index = (hash & mIndexMask);
ElementId id = mTable[index];
while (id != null) {
if (id.idMatches(buffer, start, len)) {
break;
}
id = id.mNextColl;
}
/* Not found, can just add it to the Map; no need to add to the
* linked list as it's not undefined
*/
if (id == null) {
// First, do we need more room?
if (mSize >= mSizeThreshold) {
rehash();
index = (hash & mIndexMask);
}
++mSize;
String idStr = new String(buffer, start, len);
id = new ElementId(idStr, loc, true, elemName, attrName);
id.mNextColl = mTable[index];
mTable[index] = id;
} else {
/* If already defined, nothing additional to do (we could
* signal an error here, though... for now, we'll let caller
* do that
*/
if (id.mDefined) {
;
} else {
/* Not defined, just need to upgrade, and possibly remove from
* the linked list.
*/
id.markDefined(loc);
/* Ok; if it was the first undefined, need to unlink it, as
* well as potentially next items.
*/
if (id == mHead) {
do {
mHead = mHead.mNextUndefd;
} while (mHead != null && mHead.mDefined);
// Did we clear up all undefined ids?
if (mHead == null) {
mTail = null;
}
}
}
}
return id;
}
/**
* Implementation of a hashing method for variable length
* Strings. Most of the time intention is that this calculation
* is done by caller during parsing, not here; however, sometimes
* it needs to be done for parsed "String" too.
*
* Note: identical to {@link com.ctc.wstx.util.SymbolTable#calcHash},
* although not required to be.
*
* @param len Length of String; has to be at least 1 (caller guarantees
* this pre-condition)
*/
public static int calcHash(char[] buffer, int start, int len)
{
int hash = (int) buffer[0];
for (int i = 1; i < len; ++i) {
hash = (hash * 31) + (int) buffer[i];
}
return hash;
}
public static int calcHash(String key) {
int hash = (int) key.charAt(0);
for (int i = 1, len = key.length(); i < len; ++i) {
hash = (hash * 31) + (int) key.charAt(i);
}
return hash;
}
/*
//////////////////////////////////////////////////////////
// Internal methods
//////////////////////////////////////////////////////////
*/
/**
* Method called when size (number of entries) of symbol table grows
* so big that load factor is exceeded. Since size has to remain
* power of two, arrays will then always be doubled. Main work
* is really redistributing old entries into new String/Bucket
* entries.
*/
private void rehash()
{
int size = mTable.length;
/* Let's grow aggressively; this should minimize number of
* resizes, while adding to mem usage. But since these Maps
* are never long-lived (only during parsing and validation of
* a single doc), that shouldn't greatly matter.
*/
int newSize = (size << 2);
ElementId[] oldSyms = mTable;
mTable = new ElementId[newSize];
// Let's update index mask, threshold, now (needed for rehashing)
mIndexMask = newSize - 1;
mSizeThreshold <<= 2;
int count = 0; // let's do sanity check
for (int i = 0; i < size; ++i) {
for (ElementId id = oldSyms[i]; id != null; ) {
++count;
int index = calcHash(id.mIdValue) & mIndexMask;
ElementId nextIn = id.mNextColl;
id.mNextColl = mTable[index];
mTable[index] = id;
id = nextIn;
}
}
if (count != mSize) {
ExceptionUtil.throwInternal("on rehash(): had "+mSize+" entries; now have "+count+".");
}
}
}