yacy_search_server/source/de/anomic/plasma/plasmaWordIndexEntry.java

310 lines
18 KiB
Java
Raw Normal View History

// plasmaIndexEntry.java
// -----------------------
// part of YaCy
// (C) by Michael Peter Christen; mc@anomic.de
// first published on http://www.anomic.de
// Frankfurt, Germany, 2004
//
// $LastChangedDate$
// $LastChangedRevision$
// $LastChangedBy$
//
// 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
//
// Using this software in any meaning (reading, learning, copying, compiling,
// running) means that you agree that the Author(s) is (are) not responsible
// for cost, loss of data or any harm that may be caused directly or indirectly
// by usage of this softare or this documentation. The usage of this software
// is on your own risk. The installation and usage (starting/running) of this
// software may allow other people or application to access your computer and
// any attached devices and is highly dependent on the configuration of the
// software which must be done by the user of the software; the author(s) is
// (are) also not responsible for proper configuration and usage of the
// software, even if provoked by documentation provided together with
// the software.
//
// Any changes to this file according to the GPL as documented in the file
// gpl.txt aside this file in the shipment you received can be done to the
// lines that follows this copyright notice here, but changes must not be
// done inside the copyright notive above. A re-distribution must contain
// the intact and unchanged copyright notice.
// Contributions and changes to the program code must be marked as such.
/*
This class defines the structures of an index entry
*/
package de.anomic.plasma;
import java.util.Properties;
import de.anomic.index.indexEntryAttribute;
import de.anomic.kelondro.kelondroBase64Order;
public final class plasmaWordIndexEntry implements Cloneable {
// an wordEntry can be filled in either of two ways:
// by the discrete values of the entry
// or by the encoded entry-string
// the size of the index entry attributes
public static final int attrSpace = 24;
// the associated hash
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
private final String urlHash;
// discrete values
private int hitcount; // number of this words in file
private int wordcount; // number of all words in the file
private int phrasecount; // number of all phrases in the file
private int posintext; // first position of the word in text as number of word; 0=unknown or irrelevant position
private int posinphrase; // position within a phrase of the word
private int posofphrase; // position of the phrase in the text as count of sentences; 0=unknown; 1=path; 2=keywords; 3=headline; >4: in text
private int worddistance;// distance between the words, only used if the index is artificial (from a conjunction)
private long lastModified;// calculated by using last-modified
private int quality; // result of a heuristic on the source file
private byte[] language; // essentially the country code (the TLD as heuristic), two letters lowercase only
private char doctype; // type of source
private char localflag; // indicates if the index was created locally
// the class instantiation can only be done by a plasmaStore method
// therefore they are all public
public plasmaWordIndexEntry(String urlHash,
int urlLength, // byte-length of complete URL
int urlComps, // number of path components
int titleLength, // length of description/length (longer are better?)
int hitcount, //*how often appears this word in the text
int wordcount, //*total number of words
int phrasecount, //*total number of phrases
int posintext, //*position of word in all words
int posinphrase, //*position of word in its phrase
int posofphrase, //*number of the phrase where word appears
int distance, //*word distance; this is 0 by default, and set to the difference of posintext from two indexes if these are combined (simultanous search). If stored, this shows that the result was obtained by remote search
int sizeOfPage, // # of bytes of the page
long lastmodified, //*last-modified time of the document where word appears
long updatetime, // update time; this is needed to compute a TTL for the word, so it can be removed easily if the TTL is short
int quality, //*the entropy value
String language, //*(guessed) language of document
char doctype, //*type of document
int outlinksSame, // outlinks to same domain
int outlinksOther,// outlinks to other domain
boolean local //*flag shows that this index was generated locally; othervise its from a remote peer
) {
// more needed attributes:
// - boolean: appearance attributes: title, appears in header, anchor-descr, image-tag etc
// - boolean: URL attributes
if ((language == null) || (language.length() != plasmaURL.urlLanguageLength)) language = "uk";
this.urlHash = urlHash;
this.hitcount = hitcount;
this.wordcount = wordcount;
this.phrasecount = phrasecount;
this.posintext = posintext;
this.posinphrase = posinphrase;
this.posofphrase = posofphrase;
this.worddistance = distance;
this.lastModified = lastmodified;
this.quality = quality;
this.language = language.getBytes();
this.doctype = doctype;
this.localflag = (local) ? indexEntryAttribute.LT_LOCAL : indexEntryAttribute.LT_GLOBAL;
}
public plasmaWordIndexEntry(String urlHash, String code) {
// the code is not parsed but used later on
this.urlHash = urlHash;
this.hitcount = (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(6, 8));
this.lastModified = plasmaWordIndex.reverseMicroDateDays((int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(3, 6)));
this.quality = (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(0, 3));
this.language = code.substring(8, 10).getBytes();
this.doctype = code.charAt(10);
this.localflag = code.charAt(11);
this.posintext = (code.length() >= 14) ? (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(12, 14)) : 0;
this.posinphrase = (code.length() >= 15) ? (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(14, 16)) : 0;
this.posofphrase = (code.length() >= 17) ? (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(16, 18)) : 0;
this.worddistance = (code.length() >= 19) ? (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(18, 20)) : 0;
this.wordcount = (code.length() >= 21) ? (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(20, 22)) : 0;
this.phrasecount = (code.length() >= 23) ? (int) kelondroBase64Order.enhancedCoder.decodeLong(code.substring(22, 24)) : 0;
if (hitcount == 0) hitcount = 1;
if (wordcount == 0) wordcount = 1000;
if (phrasecount == 0) phrasecount = 100;
}
public plasmaWordIndexEntry(String external) {
// parse external form
String[] elts = external.substring(1, external.length() - 1).split(",");
Properties pr = new Properties();
int p;
for (int i = 0; i < elts.length; i++) {
pr.put(elts[i].substring(0, (p = elts[i].indexOf("="))), elts[i].substring(p + 1));
}
// set values
this.urlHash = pr.getProperty("h", "");
this.hitcount = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("c", "A"));
this.wordcount = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("w", "__"));
this.phrasecount = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("p", "__"));
this.posintext = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("t", "__"));
this.posinphrase = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("r", "__"));
this.posofphrase = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("o", "__"));
this.worddistance = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("i", "__"));
this.lastModified = plasmaWordIndex.reverseMicroDateDays((int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("a", "A")));
this.quality = (int) kelondroBase64Order.enhancedCoder.decodeLong(pr.getProperty("q", "__"));
this.language = pr.getProperty("l", "uk").getBytes();
this.doctype = pr.getProperty("d", "u").charAt(0);
this.localflag = pr.getProperty("f", ""+indexEntryAttribute.LT_LOCAL).charAt(0);
}
public Object clone() {
return new plasmaWordIndexEntry(this.toExternalForm());
}
public String toEncodedForm() {
// attention: this integrates NOT the URL hash into the encoding
// if you need a complete dump, use toExternalForm()
StringBuffer buf = new StringBuffer(attrSpace);
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
buf.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.quality, plasmaURL.urlQualityLength))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(plasmaWordIndex.microDateDays(this.lastModified), 3))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.hitcount, 2))
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
.append(new String(this.language))
.append(this.doctype)
.append(this.localflag)
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.posintext, 2))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.posinphrase, 2))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.posofphrase, 2))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.worddistance, 2))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.wordcount, 2))
.append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.phrasecount, 2)); // 3+3+2+2+1+1+2+2+2+2+2+2= 24 bytes
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
return buf.toString();
}
public String toExternalForm() {
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
StringBuffer str = new StringBuffer(61);
str.append("{")
.append( "h=").append(this.urlHash)
.append(",q=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.quality, plasmaURL.urlQualityLength))
.append(",a=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(plasmaWordIndex.microDateDays(this.lastModified), 3))
.append(",c=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.hitcount, 2))
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
.append(",l=").append(new String(this.language))
.append(",d=").append(this.doctype)
.append(",f=").append(this.localflag)
.append(",t=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.posintext, 2))
.append(",r=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.posinphrase, 2))
.append(",o=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.posofphrase, 2))
.append(",i=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.worddistance, 2))
.append(",w=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.wordcount, 2))
.append(",p=").append(kelondroBase64Order.enhancedCoder.encodeLongSmart(this.phrasecount, 2))
*) Asynchronous queuing of crawl job URLs (stackCrawl) various checks like the blacklist check or the robots.txt disallow check are now done by a separate thread to unburden the indexer thread(s) TODO: maybe we have to introduce a threadpool here if it turn out that this single thread is a bottleneck because of the time consuming robots.txt downloads *) improved index transfer The index selection and transmission is done in parallel now to improve index transfer performance. TODO: maybe we could speed up performance by unsing multiple transmission threads in parallel instead of only a single one. *) gzip encoded post requests it is now configureable if a gzip encoded post request should be send on intex transfer/distribution *) storage Peer (very experimentell and not optimized yet) Now it's possible to send the result of the yacy indexer thread to a remote peer istead of storing the indexed words locally. This could be done by setting the property "storagePeerHash" in the yacy config file - Please note that if the index transfer fails, the index ist stored locally. - TODO: currently this index transfer is done by the indexer thread. To seedup the indexer a) this transmission should be done in parallel and b) multiple chunks should be bundled and transfered together *) general performance improvements - better memory cleanup after http request processing has finished - replacing some string concatenations with stringBuffers - replacing BufferedInputStreams with serverByteBuffer - replacing vectors with arraylists wherever possible - replacing hashtables with hashmaps wherever possible This was done because function calls to verctor or hashtable functions take 3 time longer than calls to functions of arraylists or hashmaps. TODO: we should take a look on the class serverObject which is inherited from hashmap Do we realy need a synchronization for this class? TODO: replace arraylists with linkedLists if random access to the list elements is not needed *) Robots Parser supports if-modified-since downloads now If the downloaded robots.txt file is older than 7 days the robots parser tries to download the robots.txt with the if-modified-since header to avoid unnecessary downloads if the file was not changed. Additionally the ETag header is used to detect changes. *) Crawler: better handling of unsupported mimeTypes + FileExtension *) Bugfix: plasmaWordIndexEntity was not closed correctly in - query.java - plasmaswitchboard.java *) function minimizeUrlDB added to yacy.java this function tests the current urlHashDB for unused urls ATTENTION: please don't use this function at the moment because it causes the wordIndexDB to flush all words into the word directory! git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@853 6c8d7289-2bf4-0310-a012-ef5d649a1542
2005-10-05 12:45:33 +02:00
.append("}");
return str.toString();
}
public void combineDistance(plasmaWordIndexEntry oe) {
this.worddistance = this.worddistance + oe.worddistance + Math.abs(this.posintext - oe.posintext);
this.posintext = Math.min(this.posintext, oe.posintext);
if (this.posofphrase != oe.posofphrase) this.posinphrase = 0; // (unknown)
this.posofphrase = Math.min(this.posofphrase, oe.posofphrase);
this.wordcount = (this.wordcount + oe.wordcount) / 2;
}
public void min(plasmaWordIndexEntry other) {
if (this.hitcount > other.hitcount) this.hitcount = other.hitcount;
if (this.wordcount > other.wordcount) this.wordcount = other.wordcount;
if (this.phrasecount > other.phrasecount) this.phrasecount = other.phrasecount;
if (this.posintext > other.posintext) this.posintext = other.posintext;
if (this.posinphrase > other.posinphrase) this.posinphrase = other.posinphrase;
if (this.posofphrase > other.posofphrase) this.posofphrase = other.posofphrase;
if (this.worddistance > other.worddistance) this.worddistance = other.worddistance;
if (this.lastModified > other.lastModified) this.lastModified = other.lastModified;
if (this.quality > other.quality) this.quality = other.quality;
}
public void max(plasmaWordIndexEntry other) {
if (this.hitcount < other.hitcount) this.hitcount = other.hitcount;
if (this.wordcount < other.wordcount) this.wordcount = other.wordcount;
if (this.phrasecount < other.phrasecount) this.phrasecount = other.phrasecount;
if (this.posintext < other.posintext) this.posintext = other.posintext;
if (this.posinphrase < other.posinphrase) this.posinphrase = other.posinphrase;
if (this.posofphrase < other.posofphrase) this.posofphrase = other.posofphrase;
if (this.worddistance < other.worddistance) this.worddistance = other.worddistance;
if (this.lastModified < other.lastModified) this.lastModified = other.lastModified;
if (this.quality < other.quality) this.quality = other.quality;
}
public void normalize(plasmaWordIndexEntry min, plasmaWordIndexEntry max) {
this.hitcount = (this.hitcount == 0) ? 0 : 1 + 255 * (this.hitcount - min.hitcount ) / (1 + max.hitcount - min.hitcount);
this.wordcount = (this.wordcount == 0) ? 0 : 1 + 255 * (this.wordcount - min.wordcount ) / (1 + max.wordcount - min.wordcount);
this.phrasecount = (this.phrasecount == 0) ? 0 : 1 + 255 * (this.phrasecount - min.phrasecount ) / (1 + max.phrasecount - min.phrasecount);
this.posintext = (this.posintext == 0) ? 0 : 1 + 255 * (this.posintext - min.posintext ) / (1 + max.posintext - min.posintext);
this.posinphrase = (this.posinphrase == 0) ? 0 : 1 + 255 * (this.posinphrase - min.posinphrase ) / (1 + max.posinphrase - min.posinphrase);
this.posofphrase = (this.posofphrase == 0) ? 0 : 1 + 255 * (this.posofphrase - min.posofphrase ) / (1 + max.posofphrase - min.posofphrase);
this.worddistance = (this.worddistance == 0) ? 0 : 1 + 255 * (this.worddistance - min.worddistance) / (1 + max.worddistance - min.worddistance);
this.lastModified = (this.lastModified == 0) ? 0 : 1 + 255 * (this.lastModified - min.lastModified) / (1 + max.lastModified - min.lastModified);
this.quality = (this.quality == 0) ? 0 : 1 + 255 * (this.quality - min.quality ) / (1 + max.quality - min.quality);
}
public plasmaWordIndexEntry generateNormalized(plasmaWordIndexEntry min, plasmaWordIndexEntry max) {
plasmaWordIndexEntry e = (plasmaWordIndexEntry) this.clone();
e.normalize(min, max);
return e;
}
public String getUrlHash() { return urlHash; }
public int getQuality() { return quality; }
public int getVirtualAge() { return plasmaWordIndex.microDateDays(lastModified); }
public long getLastModified() { return lastModified; }
public int hitcount() { return hitcount; }
public int posintext() { return posintext; }
public int posinphrase() { return posinphrase; }
public int posofphrase() { return posofphrase; }
public int worddistance() { return worddistance; }
public int wordcount() { return wordcount; }
public int phrasecount() { return phrasecount; }
public String getLanguage() { return new String(language); }
public char getType() { return doctype; }
public boolean isLocal() { return localflag == indexEntryAttribute.LT_LOCAL; }
public boolean isNewer(plasmaWordIndexEntry other) {
if (other == null) return true;
if (this.lastModified > other.lastModified) return true;
if (this.lastModified == other.getLastModified()) {
if (this.quality > other.quality) return true;
}
return false;
}
public boolean isOlder(plasmaWordIndexEntry other) {
if (other == null) return false;
if (this.lastModified < other.getLastModified()) return true;
if (this.lastModified == other.getLastModified()) {
if (this.quality < other.quality) return true;
}
return false;
}
public int domlengthNormalized() {
return 255 * plasmaURL.domLengthEstimation(this.urlHash) / 30;
}
public static void main(String[] args) {
// outputs the word hash to a given word
if (args.length != 1) System.exit(0);
System.out.println("WORDHASH: " + indexEntryAttribute.word2hash(args[0]));
}
}