Kevin Roast f651abe34a Merged BRANCHES/DEV/V4.1-BUG-FIX to HEAD
43598: Merged HEAD to BRANCHES/DEV/V4.1-BUG-FIX *RECORD ONLY*
             41906: ALF-11378: REST API has been modified to return extra information about a user whether s/he belongs to a group or not.
   44003: Merged BRANCHES/DEV/BELARUS/V4.1-BUG-FIX-2012_11_22 to BRANCHES/DEV/V4.1-BUG-FIX:
            ALF-15210: Inconsistency in the '?' icon over the dashlets projects\slingshot\source\web\js\share.js DashletTitleBarActions_onReady() function was updated. Fix initialize the style for actionsNode elements into the DOM for IE.
   44004: Merged BRANCHES/DEV/BELARUS/V4.1-BUG-FIX-2012_11_22 to V4.1-BUG-FIX:
            ALF-15793: edit offline hides version history projects\slingshot\source\web\components\document-details\document-actions.js onActionUploadNewVersion function was updated. Fix sets version variable correctly now using asset.workingCopy property instead asset.custom property, which is undefined.
   44018: ALF-16540 : CMIS: createDocument with VersioningState.CHECKEDOUT causes NodeLockedException for types with mandatory versionable aspect
            The ckeck for lock was disabled when beforeCreateVersion policy is handled.
            Version could be created for a locked node. 
   44054: Fix for ALF-16337. Datalist assignee not searchable by full name.
   44056: Trivial change. Fixing some compiler warnings under org.alfresco.repo.content.metadata including a noisy Tika one.
   44143: Merged BRANCHES/DEV/BELARUS/V4.1-BUG-FIX-2012_10_19 to BRANCHES/DEV/V4.1-BUG-FIX:
            42989: ALF-16331: Wrong user for "completed by" information provided in Group Review And Approve workflow
   44147: Merged BRANCHES/DEV/V3.4-BUG-FIX to BRANCHES/DEV/V4.1-BUG-FIX:
   44146: Merged BRANCHES/DEV/BELARUS/V3.4-BUG-FIX-2012_05_22 to BRANCHES/DEV/V3.4-BUG-FIX:
            37733: ALF-12051: Webdav - Cannot open files containing "?" character in the filename in WinXP
   44152: ALF-17009 : Merged V3.4-BUG-FIX (3.4.12) to V4.1-BUG-FIX (4.1.3)
            44151: ALF-14035 Tiny HTML file that causes Jodconverter to launch a 100% CPU soffice instance
               - HTML to PDF is now done via ODT as the direct transform hangs if there are <sub> tags in the HTML.
               - Added in 'unsupportedTransformations' to stop a bare transformer.JodConverter from doing HTML to PDF 
               - TransformerDebug test file debugTransformers.txt no longer needs to be 18 bytes, as it made it too fiddly.
               - Modified debug from RuntimeExec so less editing is required to running from the command line
            - Removed tabs that had been added to enterprise/content-services-context.xml in 4.1-BUG-FIX
   44192: ALF-16560 - CIFS: Word document version history lost after saving content in Word:mac 2011 on Mac Mountain Lion
   44224: ALF-16896 Exception with TIKA meta data extractor.
            - Patch POI to handle parsing of Unicode properties that starts on a 4 byte boundary
              rather than the specified offset. Example file was created using http://www.aspose.com/
   44241: Merged DEV to V4.1-BUG-FIX
            44208: ALF-14591 : Ordering not supported for IMAP properties defining IMAP sort fields in Share
                   Make properties from imap:imapContent aspect indexable for SOLR. 
   44253: Merged BRANCHES/DEV/AMILLER/CLOUD1 to BRANCHES/DEV/V4.1-BUG-FIX:
            38927: CLOUD-128 - Update rules works incorrectly
          This is a partial fix for ALF-14568. The rest is coming in a separate check-in.
          I made some minor adjustments to this change - trivial spelling fix and whitespace changes.
   44257: ALF-16563 - CIFS: Image document version history lost after saving content in Preview on Mac Mountain Lion
   44260: Fix for ALF-16430 - List of values shown in alphabetical order in Share Forms. Values now only sorted if the Forms config 'field' element has the sorted='true' attribute.
   44269: Completion of fix for ALF-14568 - Update rule works incorrectly.
   44318: Fix for ALF-17055 - remoteadm webscript set a Last-Modified HTTP header whose date format does not conform to RFC 2616 hence breaking proxy caching
   44320: Fix for ALF-16463 - documentLibrary RSS feed does not pass the w3c validator, in particular pubDate breaks RFC-822, date not displayed when using non English locale
   44352: Merged BRANCHES/DEV/BELARUS/V4.1-BUG-FIX-2012_11_12 to BRANCHES/DEV/V4.1-BUG-FIX
            43860: ALF-16263: Search using a "Stop Word" not displaying any result

git-svn-id: https://svn.alfresco.com/repos/alfresco-enterprise/alfresco/HEAD/root@44459 c4b6b30b-aa2e-2d43-bbcb-ca4b014f7261
2012-12-07 14:04:23 +00:00

301 lines
10 KiB
Java

/*
* Copyright (C) 2005-2012 Jesper Steen Møller
*
* This file is part of Alfresco
*
* Alfresco is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* Alfresco is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with Alfresco. If not, see <http://www.gnu.org/licenses/>.
*/
package org.alfresco.repo.content.metadata;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.locks.Lock;
import java.util.concurrent.locks.ReadWriteLock;
import java.util.concurrent.locks.ReentrantReadWriteLock;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
/**
* Holds and provides the most appropriate metadate extracter for a particular
* mimetype.
* <p>
* The extracters themselves know how well they are able to extract metadata.
*
* @see org.alfresco.repo.content.metadata.MetadataExtracter
* @author Jesper Steen Møller
*/
public class MetadataExtracterRegistry
{
private static final Log logger = LogFactory.getLog(MetadataExtracterRegistry.class);
private List<MetadataExtracter> extracters;
private Map<String, List<MetadataExtracter>> extracterCache;
private Map<String, List<MetadataEmbedder>> embedderCache;
/** Controls read access to the cache */
private Lock extracterCacheReadLock;
/** controls write access to the cache */
private Lock extracterCacheWriteLock;
public MetadataExtracterRegistry()
{
// initialise lists
extracters = new ArrayList<MetadataExtracter>(10);
extracterCache = new HashMap<String, List<MetadataExtracter>>(17);
embedderCache = new HashMap<String, List<MetadataEmbedder>>(17);
// create lock objects for access to the cache
ReadWriteLock extractionCacheLock = new ReentrantReadWriteLock();
extracterCacheReadLock = extractionCacheLock.readLock();
extracterCacheWriteLock = extractionCacheLock.writeLock();
}
/**
* Force the registry to drop its cache of extractors. This is useful for the case where an extractor
* becomes available only after the registry has initialized the cache.
*/
public void resetCache()
{
extracterCacheWriteLock.lock();
try
{
extracterCache.clear();
embedderCache.clear();
}
finally
{
extracterCacheWriteLock.unlock();
}
}
/**
* Register an instance of an extracter for use
*
* @param extracter an extracter
*/
public void register(MetadataExtracter extracter)
{
if (logger.isDebugEnabled())
{
logger.debug("Registering metadata extracter: " + extracter);
}
extracterCacheWriteLock.lock();
try
{
extracters.add(extracter);
extracterCache.clear();
embedderCache.clear();
}
finally
{
extracterCacheWriteLock.unlock();
}
}
/**
* Gets the best metadata extracter. This is a combination of the most
* reliable and the most performant extracter.
* <p>
* The result is cached for quicker access next time.
*
* @param mimetype the source MIME of the extraction
* @return Returns a metadata extracter that can extract metadata from the
* chosen MIME type.
*/
public MetadataExtracter getExtracter(String sourceMimetype)
{
logger.debug("Get extractors for " + sourceMimetype);
List<MetadataExtracter> extractors = null;
extracterCacheReadLock.lock();
try
{
if (extracterCache.containsKey(sourceMimetype))
{
// the translation has been requested before
// it might have been null
extractors = extracterCache.get(sourceMimetype);
}
}
finally
{
extracterCacheReadLock.unlock();
}
if (extractors == null)
{
// No request has been made before
// Get a write lock on the cache
// No double check done as it is not an expensive task
extracterCacheWriteLock.lock();
try
{
// find the most suitable transformer - may be empty list
extractors = findBestExtracters(sourceMimetype);
// store the result even if it is null
extracterCache.put(sourceMimetype, extractors);
}
finally
{
extracterCacheWriteLock.unlock();
}
}
// We have the list of extractors that supposedly work (as registered).
// Take the last one that still claims to work
MetadataExtracter liveExtractor = null;
for (MetadataExtracter extractor : extractors)
{
// An extractor may dynamically become unavailable
if (!extractor.isSupported(sourceMimetype))
{
logger.debug("Get unsupported: "+getName(extractor));
continue;
}
logger.debug("Get supported: "+getName(extractor));
liveExtractor = extractor;
}
logger.debug("Get returning: "+getName(liveExtractor));
return liveExtractor;
}
@SuppressWarnings("deprecation")
private String getName(MetadataExtracter extractor)
{
return extractor == null
? null
: extractor instanceof AbstractMetadataExtracter
? ((AbstractMetadataExtracter)extractor).getBeanName()
: extractor instanceof AbstractMappingMetadataExtracter
? ((AbstractMappingMetadataExtracter)extractor).getBeanName()
: extractor.getClass().getSimpleName();
}
/**
* @param sourceMimetype The MIME type under examination
* @return Returns a set of extractors that will work for the given mimetype
*/
private List<MetadataExtracter> findBestExtracters(String sourceMimetype)
{
logger.debug("Finding extractors for " + sourceMimetype);
List<MetadataExtracter> extractors = new ArrayList<MetadataExtracter>(1);
for (MetadataExtracter extractor : extracters)
{
if (!extractor.isSupported(sourceMimetype))
{
// extraction not achievable
logger.debug("Find unsupported: "+getName(extractor));
continue;
}
logger.debug("Find supported: "+getName(extractor));
extractors.add(extractor);
}
logger.debug("Find returning: "+extractors);
return extractors;
}
/**
* Gets the best metadata embedder. This is a combination of the most
* reliable and the most performant embedder.
* <p>
* The result is cached for quicker access next time.
*
* @param mimetype the source MIME of the extraction
* @return Returns a metadata embedder that can embed metadata in the
* chosen MIME type.
*/
public MetadataEmbedder getEmbedder(String sourceMimetype)
{
List<MetadataEmbedder> embedders = null;
extracterCacheReadLock.lock();
try
{
if (embedderCache.containsKey(sourceMimetype))
{
// the translation has been requested before
// it might have been null
embedders = embedderCache.get(sourceMimetype);
}
}
finally
{
extracterCacheReadLock.unlock();
}
if (embedders == null)
{
// No request has been made before
// Get a write lock on the cache
// No double check done as it is not an expensive task
extracterCacheWriteLock.lock();
try
{
// find the most suitable transformer - may be empty list
embedders = findBestEmbedders(sourceMimetype);
// store the result even if it is null
embedderCache.put(sourceMimetype, embedders);
}
finally
{
extracterCacheWriteLock.unlock();
}
}
// We have the list of embedders that supposedly work (as registered).
// Take the last one that still claims to work
MetadataEmbedder liveEmbedder = null;
for (MetadataEmbedder embedder : embedders)
{
// An extractor may dynamically become unavailable
if (!embedder.isEmbeddingSupported(sourceMimetype))
{
continue;
}
liveEmbedder = embedder;
}
return liveEmbedder;
}
/**
* @param sourceMimetype The MIME type under examination
* @return Returns a set of embedders that will work for the given mimetype
*/
private List<MetadataEmbedder> findBestEmbedders(String sourceMimetype)
{
logger.debug("Finding embedders for " + sourceMimetype);
List<MetadataEmbedder> embedders = new ArrayList<MetadataEmbedder>(1);
for (MetadataExtracter extractor : extracters)
{
if (!(extractor instanceof MetadataEmbedder))
{
continue;
}
if (!((MetadataEmbedder) extractor).isEmbeddingSupported(sourceMimetype))
{
// extraction not achievable
continue;
}
embedders.add((MetadataEmbedder)extractor);
}
return embedders;
}
}