You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
309 lines
12 KiB
309 lines
12 KiB
// SurrogateReader.java
|
|
// (C) 2009 by Michael Peter Christen; mc@yacy.net, Frankfurt a. M., Germany
|
|
// first published 15.04.2009 on http://yacy.net
|
|
//
|
|
// $LastChangedDate$
|
|
// $LastChangedRevision$
|
|
// $LastChangedBy$
|
|
//
|
|
// LICENSE
|
|
//
|
|
// This program is free software; you can redistribute it and/or modify
|
|
// it under the terms of the GNU General Public License as published by
|
|
// the Free Software Foundation; either version 2 of the License, or
|
|
// (at your option) any later version.
|
|
//
|
|
// This program 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 General Public License for more details.
|
|
//
|
|
// You should have received a copy of the GNU General Public License
|
|
// along with this program; if not, write to the Free Software
|
|
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
|
|
|
|
package net.yacy.document.content;
|
|
|
|
import java.io.BufferedReader;
|
|
import java.io.IOException;
|
|
import java.io.InputStream;
|
|
import java.io.InputStreamReader;
|
|
import java.io.PushbackInputStream;
|
|
import java.io.Reader;
|
|
import java.io.StringReader;
|
|
import java.net.MalformedURLException;
|
|
import java.nio.charset.Charset;
|
|
import java.nio.charset.StandardCharsets;
|
|
import java.util.Map;
|
|
import java.util.concurrent.ArrayBlockingQueue;
|
|
import java.util.concurrent.BlockingQueue;
|
|
|
|
import javax.xml.parsers.ParserConfigurationException;
|
|
import javax.xml.parsers.SAXParser;
|
|
import javax.xml.parsers.SAXParserFactory;
|
|
|
|
import org.apache.solr.client.solrj.impl.XMLResponseParser;
|
|
import org.apache.solr.common.SolrDocument;
|
|
import org.apache.solr.common.SolrInputDocument;
|
|
import org.apache.solr.common.util.NamedList;
|
|
import org.xml.sax.Attributes;
|
|
import org.xml.sax.InputSource;
|
|
import org.xml.sax.SAXException;
|
|
import org.xml.sax.SAXParseException;
|
|
import org.xml.sax.helpers.DefaultHandler;
|
|
|
|
import net.yacy.cora.document.id.DigestURL;
|
|
import net.yacy.cora.util.ConcurrentLog;
|
|
import net.yacy.crawler.CrawlStacker;
|
|
import net.yacy.search.schema.CollectionConfiguration;
|
|
|
|
|
|
public class SurrogateReader extends DefaultHandler implements Runnable {
|
|
|
|
// definition of the surrogate main element
|
|
public final static String SURROGATES_MAIN_ELEMENT_NAME =
|
|
"surrogates";
|
|
public final static String SURROGATES_MAIN_ELEMENT_OPEN =
|
|
"<" + SURROGATES_MAIN_ELEMENT_NAME +
|
|
" xmlns:dc=\"http://purl.org/dc/elements/1.1/\"" +
|
|
" xmlns:yacy=\"http://yacy.net/\"" +
|
|
" xmlns:geo=\"http://www.w3.org/2003/01/geo/wgs84_pos#\">";
|
|
public final static String SURROGATES_MAIN_ELEMENT_CLOSE =
|
|
"</" + SURROGATES_MAIN_ELEMENT_NAME + ">";
|
|
public final static SolrInputDocument POISON_DOCUMENT = new SolrInputDocument();
|
|
|
|
// class variables
|
|
private final StringBuilder buffer;
|
|
private boolean parsingValue;
|
|
private DCEntry dcEntry;
|
|
private String elementName;
|
|
/** Surrogates are either SolrInputDocument or DCEntry instances*/
|
|
private final BlockingQueue<Object> surrogates;
|
|
private SAXParser saxParser;
|
|
private final InputSource inputSource;
|
|
private final PushbackInputStream inputStream;
|
|
private final CrawlStacker crawlStacker;
|
|
private final CollectionConfiguration configuration;
|
|
private final int concurrency;
|
|
private Charset charset = StandardCharsets.UTF_8;
|
|
|
|
private static final ThreadLocal<SAXParser> tlSax = new ThreadLocal<SAXParser>();
|
|
private static SAXParser getParser() throws SAXException {
|
|
SAXParser parser = tlSax.get();
|
|
if (parser == null) {
|
|
try {
|
|
parser = SAXParserFactory.newInstance().newSAXParser();
|
|
} catch (final ParserConfigurationException e) {
|
|
throw new SAXException(e.getMessage(), e);
|
|
}
|
|
tlSax.set(parser);
|
|
}
|
|
return parser;
|
|
}
|
|
|
|
public SurrogateReader(final InputStream stream, int queueSize, CrawlStacker crawlStacker, CollectionConfiguration configuration, int concurrency) throws IOException {
|
|
this(new PushbackInputStream(stream, 200), queueSize, crawlStacker, configuration, concurrency);
|
|
}
|
|
|
|
public SurrogateReader(final PushbackInputStream stream, int queueSize, CrawlStacker crawlStacker, CollectionConfiguration configuration, int concurrency) throws IOException {
|
|
this.crawlStacker = crawlStacker;
|
|
this.configuration = configuration;
|
|
this.concurrency = concurrency;
|
|
this.buffer = new StringBuilder(300);
|
|
this.parsingValue = false;
|
|
this.dcEntry = null;
|
|
this.elementName = null;
|
|
this.surrogates = new ArrayBlockingQueue<>(queueSize);
|
|
|
|
Reader reader = new BufferedReader(new InputStreamReader(stream, this.charset));
|
|
this.inputSource = new InputSource(reader);
|
|
this.inputSource.setEncoding(this.charset.name());
|
|
this.inputStream = stream;
|
|
|
|
try {
|
|
this.saxParser = getParser();
|
|
} catch (final SAXException e) {
|
|
ConcurrentLog.logException(e);
|
|
throw new IOException(e.getMessage());
|
|
}
|
|
}
|
|
|
|
@Override
|
|
public void run() {
|
|
// test the syntax of the stream by reading parts of the beginning
|
|
try {
|
|
if (isSolrDump()) {
|
|
BufferedReader br = new BufferedReader(new InputStreamReader(this.inputStream, this.charset));
|
|
String line;
|
|
while ((line = br.readLine()) != null) {
|
|
if (!line.startsWith("<doc>")) continue;
|
|
try {
|
|
NamedList<Object> nl = new XMLResponseParser().processResponse(new StringReader("<result>" + line + "</result>")); //
|
|
SolrDocument doc = (SolrDocument) nl.iterator().next().getValue();
|
|
|
|
// check if url is in accepted domain
|
|
String u = (String) doc.getFieldValue("sku");
|
|
if (u != null) {
|
|
try {
|
|
DigestURL url = new DigestURL(u);
|
|
final String urlRejectReason = this.crawlStacker.urlInAcceptedDomain(url);
|
|
if ( urlRejectReason == null ) {
|
|
// convert SolrDocument to SolrInputDocument
|
|
this.surrogates.put(this.configuration.toSolrInputDocument(doc));
|
|
}
|
|
} catch (MalformedURLException e) {
|
|
}
|
|
}
|
|
} catch (Throwable ee) {
|
|
// bad line
|
|
}
|
|
}
|
|
} else {
|
|
this.saxParser.parse(this.inputSource, this);
|
|
}
|
|
} catch (final SAXParseException e) {
|
|
ConcurrentLog.logException(e);
|
|
} catch (final SAXException e) {
|
|
ConcurrentLog.logException(e);
|
|
} catch (final IOException e) {
|
|
ConcurrentLog.logException(e);
|
|
} finally {
|
|
for (int i = 0; i < this.concurrency; i++) {
|
|
try {
|
|
this.surrogates.put(POISON_DOCUMENT);
|
|
} catch (final InterruptedException e1) {
|
|
ConcurrentLog.logException(e1);
|
|
}
|
|
}
|
|
try {
|
|
this.inputStream.close();
|
|
} catch (final IOException e) {
|
|
ConcurrentLog.logException(e);
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* @return true when inputStream is likely to contain a rich and full-text Solr xml data dump (see IndexExport_p.html)
|
|
*/
|
|
private boolean isSolrDump() {
|
|
boolean res = false;
|
|
byte[] b = new byte[100];
|
|
int nbRead = -1;
|
|
try {
|
|
nbRead = this.inputStream.read(b);
|
|
if(nbRead > 0) {
|
|
String s = new String(b, 0, nbRead, this.charset);
|
|
if ((s.contains("<response>") && s.contains("<result>")) || s.startsWith("<doc>")) {
|
|
res = true;
|
|
}
|
|
}
|
|
} catch (IOException e) {
|
|
ConcurrentLog.logException(e);
|
|
} finally {
|
|
if (nbRead > 0) {
|
|
try {
|
|
this.inputStream.unread(b, 0, nbRead);
|
|
} catch (IOException e2) {
|
|
ConcurrentLog.logException(e2);
|
|
}
|
|
}
|
|
}
|
|
return res;
|
|
}
|
|
|
|
@Override
|
|
public void startElement(final String uri, final String name, String tag, final Attributes atts) throws SAXException {
|
|
if (tag == null) return;
|
|
tag = tag.toLowerCase();
|
|
if ("record".equals(tag) || "document".equals(tag) || "doc".equals(tag)) {
|
|
this.dcEntry = new DCEntry();
|
|
} else if ("element".equals(tag) || "str".equals(tag) || "int".equals(tag) || "bool".equals(tag) || "long".equals(tag)) {
|
|
this.elementName = atts.getValue("name");
|
|
this.parsingValue = true;
|
|
} else if ("value".equals(tag)) {
|
|
this.buffer.setLength(0);
|
|
this.parsingValue = true;
|
|
} else if (tag.startsWith("dc:") || tag.startsWith("geo:") || tag.startsWith("md:")) {
|
|
// parse dublin core attribute
|
|
this.elementName = tag;
|
|
this.parsingValue = true;
|
|
}
|
|
}
|
|
|
|
@Override
|
|
public void endElement(final String uri, final String name, String tag) {
|
|
if (tag == null) return;
|
|
tag = tag.toLowerCase();
|
|
if ("record".equals(tag) || "document".equals(tag) || "doc".equals(tag)) {
|
|
//System.out.println("A Title: " + this.surrogate.title());
|
|
try {
|
|
// check if url is in accepted domain
|
|
final String urlRejectReason = this.crawlStacker.urlInAcceptedDomain(this.dcEntry.getIdentifier(true));
|
|
if ( urlRejectReason == null ) {
|
|
// DCEntry can not be converted to SolrInputDocument as DC schema has nothing to do with Solr collection schema
|
|
this.surrogates.put(this.dcEntry);
|
|
}
|
|
} catch (final InterruptedException e) {
|
|
ConcurrentLog.logException(e);
|
|
} finally {
|
|
//System.out.println("B Title: " + this.surrogate.title());
|
|
this.dcEntry = null;
|
|
this.buffer.setLength(0);
|
|
this.parsingValue = false;
|
|
}
|
|
} else if ("element".equals(tag)) {
|
|
this.buffer.setLength(0);
|
|
this.parsingValue = false;
|
|
} else if ("str".equals(tag) || "int".equals(tag) || "bool".equals(tag) || "long".equals(tag)){
|
|
final String value = buffer.toString().trim();
|
|
if (this.elementName != null) {
|
|
this.dcEntry.getMap().put(this.elementName, new String[]{value});
|
|
}
|
|
this.buffer.setLength(0);
|
|
this.parsingValue = false;
|
|
} else if ("value".equals(tag)) {
|
|
//System.out.println("BUFFER-SIZE=" + buffer.length());
|
|
final String value = buffer.toString().trim();
|
|
if (this.elementName != null) {
|
|
this.dcEntry.getMap().put(this.elementName, new String[]{value});
|
|
}
|
|
this.buffer.setLength(0);
|
|
this.parsingValue = false;
|
|
} else if (tag.startsWith("dc:") || tag.startsWith("geo:") || tag.startsWith("md:")) {
|
|
final String value = buffer.toString().trim();
|
|
if (this.elementName != null && tag.equals(this.elementName)) {
|
|
Map<String,String[]> map = this.dcEntry.getMap();
|
|
String[] oldcontent = map.get(this.elementName);
|
|
if (oldcontent == null || oldcontent.length == 0) {
|
|
map.put(this.elementName, new String[]{value});
|
|
} else {
|
|
String[] newcontent = new String[oldcontent.length + 1];
|
|
System.arraycopy(oldcontent, 0, newcontent, 0, oldcontent.length);
|
|
newcontent[oldcontent.length] = value;
|
|
map.put(this.elementName, newcontent);
|
|
}
|
|
}
|
|
this.buffer.setLength(0);
|
|
this.parsingValue = false;
|
|
}
|
|
}
|
|
|
|
@Override
|
|
public void characters(final char ch[], final int start, final int length) {
|
|
if (parsingValue) {
|
|
buffer.append(ch, start, length);
|
|
}
|
|
}
|
|
|
|
public Object take() {
|
|
try {
|
|
return this.surrogates.take();
|
|
} catch (final InterruptedException e) {
|
|
ConcurrentLog.logException(e);
|
|
return null;
|
|
}
|
|
}
|
|
|
|
}
|