experimentalwarcwriterprocessor.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。

/* $Id: ExperimentalWARCWriterProcessor.java 4935 2007-02-23 00:27:24Z gojomo $
 *
 * Created on August 1st, 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.writer;

import java.io.ByteArrayInputStream;
import java.io.IOException;
import java.net.URI;
import java.net.URISyntaxException;
import java.util.Collection;
import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.logging.Level;
import java.util.logging.Logger;

import org.apache.commons.httpclient.Header;
import org.apache.commons.httpclient.HttpMethodBase;
import org.apache.commons.httpclient.HttpStatus;
import org.archive.crawler.datamodel.CoreAttributeConstants;
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.extractor.Link;
import org.archive.crawler.framework.WriterPoolProcessor;
import org.archive.crawler.settings.SimpleType;
import org.archive.crawler.settings.Type;
import org.archive.io.ReplayInputStream;
import org.archive.io.WriterPoolMember;
import org.archive.io.WriterPoolSettings;
import org.archive.io.warc.ExperimentalWARCWriter;
import org.archive.io.warc.WARCConstants;
import org.archive.io.warc.WARCWriterPool;
import org.archive.uid.GeneratorFactory;
import org.archive.util.ArchiveUtils;
import org.archive.util.anvl.ANVLRecord;

/**
 * Experimental WARCWriterProcessor.
 * Goes against the pending release of 0.12 of the WARC specification, the
 * "Marcel Marceau" release. See <a href="https://archive-access.svn.sourceforge.net/svnroot/archive-access/branches/gjm_warc_0_12/warc/warc_file_format.html">latest revision</a>
 * for current state.  The 0.10 WARC implemenation has been moved to
 * {@link ExperimentalV10WARCWriterProcessor}.
 * 
 * <p>TODO: Remove ANVLRecord. Rename NameValue or use RFC822
 * (commons-httpclient?) or find something else.
 * 
 * @author stack
 */
public class ExperimentalWARCWriterProcessor extends WriterPoolProcessor
implements CoreAttributeConstants, CrawlStatusListener,
WriterPoolSettings, FetchStatusCodes, WARCConstants {
    private static final long serialVersionUID = 6182850087635847443L;

    private final Logger logger = Logger.getLogger(this.getClass().getName());
    
    /**
     * Key for whether to write 'request' type records where possible
     */
    public static final String ATTR_WRITE_REQUESTS =
        "write-requests";
    
    /**
     * Key for whether to write 'metadata' type records where possible
     */
    public static final String ATTR_WRITE_METADATA =
        "write-metadata";
    
    /**
     * Key for whether to write 'revisit' type records when
     * consecutive identical digest
     */
    public static final String ATTR_WRITE_REVISIT_FOR_IDENTICAL_DIGESTS =
        "write-revisit-for-identical-digests";
    
    /**
     * Key for whether to write 'revisit' type records for server
     * "304 not modified" responses
     */
    public static final String ATTR_WRITE_REVISIT_FOR_NOT_MODIFIED =
        "write-revisit-for-not-modified";
    
    /**
     * Default path list.
     */
    private static final String [] DEFAULT_PATH = {"warcs"};

    protected String [] getDefaultPath() {
        return DEFAULT_PATH;
    }
    
    /**
     * @param name Name of this writer.
     */
    public ExperimentalWARCWriterProcessor(final String name) {
        super(name, "Experimental WARCWriter processor (Version 0.12)");
        Type e = addElementToDefinition(
                new SimpleType(ATTR_WRITE_REQUESTS,
                "Whether to write 'request' type records. " +
                "Default is true.", new Boolean(true)));
        e.setOverrideable(true);
        e.setExpertSetting(true);
        e = addElementToDefinition(
                new SimpleType(ATTR_WRITE_METADATA,
                "Whether to write 'metadata' type records. " +
                "Default is true.", new Boolean(true)));
        e.setOverrideable(true);
        e.setExpertSetting(true);
        e = addElementToDefinition(
                new SimpleType(ATTR_WRITE_REVISIT_FOR_IDENTICAL_DIGESTS,
                "Whether to write 'revisit' type records when a URI's " +
                "history indicates the previous fetch had an identical " +
                "content digest. " +
                "Default is true.", new Boolean(true)));
        e.setOverrideable(true);
        e.setExpertSetting(true);
        e = addElementToDefinition(
                new SimpleType(ATTR_WRITE_REVISIT_FOR_NOT_MODIFIED,
                "Whether to write 'revisit' type records when a " +
                "304-Not Modified response is received. " +
                "Default is true.", new Boolean(true)));
        e.setOverrideable(true);
        e.setExpertSetting(true);
    }

    protected void setupPool(final AtomicInteger serialNo) {
		setPool(new WARCWriterPool(serialNo, this, getPoolMaximumActive(),
            getPoolMaximumWait()));
    }
    
    /**
     * 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 void innerProcess(CrawlURI curi) {
        // If failure, or we haven't fetched the resource yet, return
        if (curi.getFetchStatus() <= 0) {
            return;
        }
        
        // If no recorded content at all, don't write record.
        long recordLength = curi.getContentSize();
        if (recordLength <= 0) {
            // getContentSize() should be > 0 if any material (even just
            // HTTP headers with zero-length body) is available. 
        	return;
        }
        
        String scheme = curi.getUURI().getScheme().toLowerCase();
        try {
            if (shouldWrite(curi)) {
                write(scheme, curi);
            } else {
                logger.info("This writer does not write out scheme " +
                        scheme + " content");
            }
        } catch (IOException e) {
            curi.addLocalizedError(this.getName(), e, "WriteRecord: " +
                curi.toString());
            logger.log(Level.SEVERE, "Failed write of Record: " +
                curi.toString(), e);
        }
    }
    
    protected void write(final String lowerCaseScheme, final CrawlURI curi)
    throws IOException {
        WriterPoolMember writer = getPool().borrowFile();
        long position = writer.getPosition();
        // See if we need to open a new file because we've exceeed maxBytes.
        // Call to checkFileSize will open new file if we're at maximum for
        // current file.
        writer.checkSize();
        if (writer.getPosition() != position) {
            // We just closed the file because it was larger than maxBytes.
            // Add to the totalBytesWritten the size of the first record
            // in the file, if any.
            setTotalBytesWritten(getTotalBytesWritten() +
            	(writer.getPosition() - position));
            position = writer.getPosition();
        }
        
        ExperimentalWARCWriter w = (ExperimentalWARCWriter)writer;
        try {
            // Write a request, response, and metadata all in the one
            // 'transaction'.
            final URI baseid = getRecordID();
            final String timestamp =
                ArchiveUtils.getLog14Date(curi.getLong(A_FETCH_BEGAN_TIME));
            if (lowerCaseScheme.startsWith("http")) {
                // Add named fields for ip, checksum, and relate the metadata
                // and request to the resource field.
                // TODO: Use other than ANVL (or rename ANVL as NameValue or
                // use RFC822 (commons-httpclient?).
                ANVLRecord headers = new ANVLRecord(5);
                if (curi.getContentDigest() != null) {
                    headers.addLabelValue(HEADER_KEY_CHECKSUM,
                        curi.getContentDigestSchemeString());
                }
                headers.addLabelValue(HEADER_KEY_IP, getHostAddress(curi));
                URI rid;
                
                if (IdenticalDigestDecideRule.hasIdenticalDigest(curi) && 
                        ((Boolean)getUncheckedAttribute(curi, 
                                ATTR_WRITE_REVISIT_FOR_IDENTICAL_DIGESTS))) {
                    rid = writeRevisitDigest(w, timestamp, HTTP_RESPONSE_MIMETYPE,
                            baseid, curi, headers);
                } else if (curi.getFetchStatus() == HttpStatus.SC_NOT_MODIFIED && 
                        ((Boolean)getUncheckedAttribute(curi, 
                                ATTR_WRITE_REVISIT_FOR_NOT_MODIFIED))) {
                    rid = writeRevisitNotModified(w, timestamp,
                            baseid, curi, headers);
                } else {
                    if (curi.isTruncatedFetch()) {
                        String value = curi.isTimeTruncatedFetch()?
                            NAMED_FIELD_TRUNCATED_VALUE_TIME:
                            curi.isLengthTruncatedFetch()?