writerpoolprocessor.java

Heritrix是一个开源,可扩展的web爬虫项目。Heritrix设计成严格按照robots.txt文件的排除指示和META robots标签。

/* WriterPoolProcessor
 *
 * $Id: WriterPoolProcessor.java 5029 2007-03-29 23:53:50Z gojomo $
 *
 * Created on July 19th, 2006
 *
 * Copyright (C) 2006 Internet Archive.
 *
 * This file is part of the Heritrix web crawler (crawler.archive.org).
 *
 * Heritrix is free software; you can redistribute it and/or modify
 * it under the terms of the GNU Lesser Public License as published by
 * the Free Software Foundation; either version 2.1 of the License, or
 * any later version.
 *
 * Heritrix 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 Public License for more details.
 *
 * You should have received a copy of the GNU Lesser Public License
 * along with Heritrix; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 */
package org.archive.crawler.framework;

import java.io.DataInputStream;
import java.io.DataOutputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.FileOutputStream;
import java.io.IOException;
import java.io.ObjectInputStream;
import java.io.StringWriter;
import java.net.InetAddress;
import java.net.UnknownHostException;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Iterator;
import java.util.List;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.logging.Logger;

import javax.management.AttributeNotFoundException;
import javax.management.MBeanException;
import javax.management.ReflectionException;
import javax.xml.transform.SourceLocator;
import javax.xml.transform.Templates;
import javax.xml.transform.Transformer;
import javax.xml.transform.TransformerConfigurationException;
import javax.xml.transform.TransformerException;
import javax.xml.transform.TransformerFactory;
import javax.xml.transform.stream.StreamResult;
import javax.xml.transform.stream.StreamSource;

import org.archive.crawler.Heritrix;
import org.archive.crawler.datamodel.CoreAttributeConstants;
import org.archive.crawler.datamodel.CrawlHost;
import org.archive.crawler.datamodel.CrawlOrder;
import org.archive.crawler.datamodel.CrawlURI;
import org.archive.crawler.datamodel.FetchStatusCodes;
import org.archive.crawler.deciderules.recrawl.IdenticalDigestDecideRule;
import org.archive.crawler.event.CrawlStatusListener;
import org.archive.crawler.settings.SimpleType;
import org.archive.crawler.settings.StringList;
import org.archive.crawler.settings.Type;
import org.archive.crawler.settings.XMLSettingsHandler;
import org.archive.io.ObjectPlusFilesInputStream;
import org.archive.io.WriterPool;
import org.archive.io.WriterPoolMember;

/**
 * Abstract implementation of a file pool processor.
 * Subclass to implement for a particular {@link WriterPoolMember} instance.
 * @author Parker Thompson
 * @author stack
 */
public abstract class WriterPoolProcessor extends Processor
implements CoreAttributeConstants, CrawlStatusListener, FetchStatusCodes {
    private final Logger logger = Logger.getLogger(this.getClass().getName());

    /**
     * Key to use asking settings for file compression value.
     */
    public static final String ATTR_COMPRESS = "compress";

    /**
     * Default as to whether we do compression of files.
     */
    public static final boolean DEFAULT_COMPRESS = true;

    /**
     * Key to use asking settings for file prefix value.
     */
    public static final String ATTR_PREFIX = "prefix";    

    /**
     * Key to use asking settings for arc path value.
     */
    public static final String ATTR_PATH ="path";

    /**
     * Key to use asking settings for file suffix value.
     */
    public static final String ATTR_SUFFIX = "suffix";

    /**
     * Key to use asking settings for file max size value.
     */
    public static final String ATTR_MAX_SIZE_BYTES = "max-size-bytes";
    
    /**
     * Key to get maximum pool size.
     *
     * This key is for maximum files active in the pool.
     */
    public static final String ATTR_POOL_MAX_ACTIVE = "pool-max-active";

    /**
     * Key to get maximum wait on pool object before we give up and
     * throw IOException.
     */
    public static final String ATTR_POOL_MAX_WAIT = "pool-max-wait";

    /**
     * Key for the maximum bytes to write attribute.
     */
    public static final String ATTR_MAX_BYTES_WRITTEN =
    	"total-bytes-to-write";
    
    /**
     * Key for whether to skip writing records of content-digest repeats 
     */
    public static final String ATTR_SKIP_IDENTICAL_DIGESTS =
        "skip-identical-digests";
    
    /**
     * CrawlURI annotation indicating no record was written
     */
    protected static final String ANNOTATION_UNWRITTEN = "unwritten";
    
    /**
     * Default maximum file size.
     * TODO: Check that subclasses can set a different MAX_FILE_SIZE and
     * it will be used in the constructor as default.
     */
    private static final int DEFAULT_MAX_FILE_SIZE = 100000000;
    
    /**
     * Default path list.
     * 
     * TODO: Confirm this one gets picked up.
     */
    private static final String [] DEFAULT_PATH = {"crawl-store"};

    /**
     * Reference to pool.
     */
    transient private WriterPool pool = null;
    
    /**
     * Total number of bytes written to disc.
     */
    private long totalBytesWritten = 0;
    
    /**
     * Calculate metadata once only.
     */
    transient private List<String> cachedMetadata = null;


    /**
     * @param name Name of this processor.
     */
    public WriterPoolProcessor(String name) {
    	this(name, "Pool of files processor");
    }
    	
    /**
     * @param name Name of this processor.
     * @param description Description for this processor.
     */
    public WriterPoolProcessor(final String name,
        		final String description) {
        super(name, description);
        Type e = addElementToDefinition(
            new SimpleType(ATTR_COMPRESS, "Compress files when " +
            	"writing to disk.", new Boolean(DEFAULT_COMPRESS)));
        e.setOverrideable(false);
        e = addElementToDefinition(
            new SimpleType(ATTR_PREFIX, 
                "File prefix. " +
                "The text supplied here will be used as a prefix naming " +
                "writer files.  For example if the prefix is 'IAH', " +
                "then file names will look like " +
                "IAH-20040808101010-0001-HOSTNAME.arc.gz " +
                "...if writing ARCs (The prefix will be " +
                "separated from the date by a hyphen).",
                WriterPoolMember.DEFAULT_PREFIX));
        e = addElementToDefinition(
            new SimpleType(ATTR_SUFFIX, "Suffix to tag onto " +
                "files. If value is '${HOSTNAME}', will use hostname for " +
                "suffix. If empty, no suffix will be added.",
                WriterPoolMember.DEFAULT_SUFFIX));
        e.setOverrideable(false);
        e = addElementToDefinition(
            new SimpleType(ATTR_MAX_SIZE_BYTES, "Max size of each file",
                new Long(DEFAULT_MAX_FILE_SIZE)));
        e.setOverrideable(false);
        e = addElementToDefinition(
            new StringList(ATTR_PATH, "Where to files. " +
                "Supply absolute or relative path.  If relative, files " +
                "will be written relative to " +
                "the " + CrawlOrder.ATTR_DISK_PATH + "setting." +
                " If more than one path specified, we'll round-robin" +
                " dropping files to each.  This setting is safe" +
                " to change midcrawl (You can remove and add new dirs" +
                " as the crawler progresses).", getDefaultPath()));
        e.setOverrideable(false);
        e = addElementToDefinition(new SimpleType(ATTR_POOL_MAX_ACTIVE,
            "Maximum active files in pool. " +
            "This setting cannot be varied over the life of a crawl.",
            new Integer(WriterPool.DEFAULT_MAX_ACTIVE)));
        e.setOverrideable(false);
        e = addElementToDefinition(new SimpleType(ATTR_POOL_MAX_WAIT,
            "Maximum time to wait on pool element" +
            " (milliseconds). This setting cannot be varied over the life" +
            " of a crawl.",
            new Integer(WriterPool.DEFAULT_MAXIMUM_WAIT)));
        e.setOverrideable(false);
        e = addElementToDefinition(new SimpleType(ATTR_MAX_BYTES_WRITTEN,
            "Total file bytes to write to disk." +
            " Once the size of all files on disk has exceeded this " +
            "limit, this processor will stop the crawler. " +
            "A value of zero means no upper limit.", new Long(0)));
        e.setOverrideable(false);
        e.setExpertSetting(true);
        e = addElementToDefinition(new SimpleType(ATTR_SKIP_IDENTICAL_DIGESTS,
                "Whether to skip the writing of a record when URI " +
                "history information is available and indicates the " +
                "prior fetch had an identical content digest. " +
                "Default is false.", new Boolean(false)));
        e.setOverrideable(true);
        e.setExpertSetting(true);
    }
    
    protected String [] getDefaultPath() {
    	return DEFAULT_PATH;
	}

    public synchronized void initialTasks() {
        // Add this class to crawl state listeners and setup pool.
        getSettingsHandler().getOrder().getController().
            addCrawlStatusListener(this);
        setupPool(new AtomicInteger());
        // Run checkpoint recovery code.
        if (getSettingsHandler().getOrder().getController().
        		isCheckpointRecover()) {
        	checkpointRecover();
        }
    }
    
    protected AtomicInteger getSerialNo() {
        return ((WriterPool)getPool()).getSerialNo();
    }

    /**
     * Set up pool of files.
     */
    protected abstract void setupPool(final AtomicInteger serialNo);

    /**
     * Writes a CrawlURI and its associated data to store file.
     *
     * Currently this method understands the following uri types: dns, http, 
     * and https.
     *
     * @param curi CrawlURI to process.
     */
    protected abstract void innerProcess(CrawlURI curi);
    
    protected void checkBytesWritten() {
        long max = getMaxToWrite();
        if (max <= 0) {
            return;
        }
        if (max <= this.totalBytesWritten) {
            getController().requestCrawlStop("Finished - Maximum bytes (" +
                Long.toString(max) + ") written");
        }
    }
    
    /**
     * Whether the given CrawlURI should be written to archive files. 
     * Annotates CrawlURI with a reason for any negative answer. 
     * 
     * @param curi CrawlURI
     * @return true if URI should be written; false otherwise
     */
    protected boolean shouldWrite(CrawlURI curi) {
        // check for duplicate content write suppression
        if(((Boolean)getUncheckedAttribute(curi, ATTR_SKIP_IDENTICAL_DIGESTS)) 
            && IdenticalDigestDecideRule.hasIdenticalDigest(curi)) {
            curi.addAnnotation(ANNOTATION_UNWRITTEN + ":identicalDigest");
            return false; 
        }
        String scheme = curi.getUURI().getScheme().toLowerCase();
        // TODO: possibly move this sort of isSuccess() test into CrawlURI
        boolean retVal; 
        if (scheme.equals("dns")) {
            retVal = curi.getFetchStatus() == S_DNS_SUCCESS;
        } else if (scheme.equals("http") || scheme.equals("https")) {
            retVal = curi.getFetchStatus() > 0 && curi.isHttpTransaction();
        } else if (scheme.equals("ftp")) {
            retVal = curi.getFetchStatus() == 200;
        } else {
            // unsupported scheme
            curi.addAnnotation(ANNOTATION_UNWRITTEN + ":scheme");
            return false; 
        }
        if (retVal == false) {
            // status not deserving writing
            curi.addAnnotation(ANNOTATION_UNWRITTEN + ":status");
            return false; 
        }
        return true; 
    }
    
    /**
     * Return IP address of given URI suitable for recording (as in a
     * classic ARC 5-field header line).
     * 
     * @param curi CrawlURI
     * @return String of IP address
     */
    protected String getHostAddress(CrawlURI curi) {
        // special handling for DNS URIs: want address of DNS server
        if(curi.getUURI().getScheme().toLowerCase().equals("dns")) {
            return curi.getString(A_DNS_SERVER_IP_LABEL);
        }
        // otherwise, host referenced in URI
        CrawlHost h = getController().getServerCache().getHostFor(curi);
        if (h == null) {
            throw new NullPointerException("Crawlhost is null for " +
                curi + " " + curi.getVia());
        }
        InetAddress a = h.getIP();
        if (a == null) {
            throw new NullPointerException("Address is null for " +
                curi + " " + curi.getVia() + ". Address " +
                ((h.getIpFetched() == CrawlHost.IP_NEVER_LOOKED_UP)?
                     "was never looked up.":
                     (System.currentTimeMillis() - h.getIpFetched()) +
                         " ms ago."));
        }
        return h.getIP().getHostAddress();
    }