yacy_search_server/source/net/yacy/search/query/RankingProcess.java

525 lines
23 KiB
Java
Raw Normal View History

// RankingProcess.java
// (C) 2007 by Michael Peter Christen; mc@yacy.net, Frankfurt a. M., Germany
// first published 07.11.2007 on http://yacy.net
//
// This is a part of YaCy, a peer-to-peer based web search engine
//
// $LastChangedDate$
// $LastChangedRevision$
// $LastChangedBy$
//
// 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.search.query;
import java.util.HashMap;
import java.util.Iterator;
import java.util.Map;
import java.util.SortedMap;
import java.util.concurrent.BlockingQueue;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.TimeUnit;
import java.util.concurrent.atomic.AtomicInteger;
import java.util.regex.Pattern;
import net.yacy.cora.document.Classification.ContentDomain;
import net.yacy.cora.document.MultiProtocolURI;
import net.yacy.cora.document.UTF8;
import net.yacy.cora.federate.yacy.CacheStrategy;
import net.yacy.cora.lod.JenaTripleStore;
import net.yacy.cora.lod.vocabulary.Tagging;
import net.yacy.cora.lod.vocabulary.YaCyMetadata;
import net.yacy.cora.sorting.ConcurrentScoreMap;
import net.yacy.cora.sorting.ScoreMap;
import net.yacy.cora.sorting.WeakPriorityBlockingQueue;
import net.yacy.cora.sorting.WeakPriorityBlockingQueue.ReverseElement;
import net.yacy.cora.storage.HandleSet;
import net.yacy.cora.util.SpaceExceededException;
import net.yacy.document.Condenser;
import net.yacy.document.LibraryProvider;
import net.yacy.kelondro.data.meta.DigestURI;
import net.yacy.kelondro.data.meta.URIMetadataNode;
import net.yacy.kelondro.data.meta.URIMetadataRow;
import net.yacy.kelondro.data.word.Word;
import net.yacy.kelondro.data.word.WordReference;
import net.yacy.kelondro.data.word.WordReferenceVars;
import net.yacy.kelondro.index.RowHandleSet;
import net.yacy.kelondro.logging.Log;
import net.yacy.kelondro.rwi.ReferenceContainer;
import net.yacy.kelondro.rwi.TermSearch;
import net.yacy.kelondro.util.Bitfield;
import net.yacy.peers.graphics.ProfilingGraph;
import net.yacy.search.EventTracker;
import net.yacy.search.Switchboard;
import net.yacy.search.index.Segment;
import net.yacy.search.ranking.ReferenceOrder;
import net.yacy.search.snippet.ResultEntry;
import com.hp.hpl.jena.rdf.model.RDFNode;
import com.hp.hpl.jena.rdf.model.Resource;
public final class RankingProcess extends Thread {
protected static final int max_results_preparation = 3000, max_results_preparation_special = -1; // -1 means 'no limit'
//private final SearchEvent searchEvent;
private final QueryParams query;
private SortedMap<byte[], ReferenceContainer<WordReference>> localSearchInclusion;
private final ScoreMap<String> ref; // reference score computation for the commonSense heuristic
private final long maxtime;
protected final WeakPriorityBlockingQueue<WordReferenceVars> rwiStack;
protected final ConcurrentHashMap<String, WeakPriorityBlockingQueue<WordReferenceVars>> doubleDomCache; // key = domhash (6 bytes); value = like stack
private final int[] flagcount; // flag counter
private final AtomicInteger feedersAlive, feedersTerminated;
private boolean addRunning;
protected final AtomicInteger receivedRemoteReferences;
protected final ReferenceOrder order;
protected final HandleSet urlhashes; // map for double-check; String/Long relation, addresses ranking number (backreference for deletion)
protected final ScoreMap<String> hostNavigator; // a counter for the appearance of the host hash
protected final Map<String, byte[]> hostResolver; // a mapping from a host hash (6 bytes) to the full url hash of one of these urls that have the host hash
protected final Map<String, String> taggingPredicates; // a map from tagging vocabulary names to tagging predicate uris
protected final Map<String, ScoreMap<String>> vocabularyNavigator; // counters for Vocabularies; key is metatag.getVocabularyName()
private boolean remote;
protected RankingProcess(final QueryParams query, boolean remote) {
// we collect the urlhashes and construct a list with urlEntry objects
// attention: if minEntries is too high, this method will not terminate within the maxTime
// sortorder: 0 = hash, 1 = url, 2 = ranking
this.query = query;
this.remote = remote;
this.localSearchInclusion = null;
this.ref = new ConcurrentScoreMap<String>();
this.maxtime = query.maxtime;
int stackMaxsize = query.snippetCacheStrategy == null || query.snippetCacheStrategy == CacheStrategy.CACHEONLY ? max_results_preparation_special : max_results_preparation;
this.rwiStack = new WeakPriorityBlockingQueue<WordReferenceVars>(stackMaxsize, false);
this.doubleDomCache = new ConcurrentHashMap<String, WeakPriorityBlockingQueue<WordReferenceVars>>();
this.flagcount = new int[32];
for ( int i = 0; i < 32; i++ ) {
this.flagcount[i] = 0;
}
this.feedersAlive = new AtomicInteger(0);
this.feedersTerminated = new AtomicInteger(0);
this.addRunning = true;
this.receivedRemoteReferences = new AtomicInteger(0);
this.order = new ReferenceOrder(this.query.ranking, UTF8.getBytes(this.query.targetlang));
this.urlhashes = new RowHandleSet(URIMetadataRow.rowdef.primaryKeyLength, URIMetadataRow.rowdef.objectOrder, 100);
this.hostNavigator = new ConcurrentScoreMap<String>();
this.hostResolver = new ConcurrentHashMap<String, byte[]>();
this.vocabularyNavigator = new ConcurrentHashMap<String, ScoreMap<String>>();
this.taggingPredicates = new HashMap<String, String>();
for (Tagging t: LibraryProvider.autotagging.getVocabularies()) {
this.taggingPredicates.put(t.getName(), t.getPredicate());
}
}
public ReferenceOrder getOrder() {
return this.order;
}
protected boolean feedingIsFinished() {
return
this.feedersTerminated.intValue() > (this.remote ? 1 : 0) &&
this.feedersAlive.get() == 0;// &&
//(!this.remote || this.remote_indexCount > 0);
}
/**
* method to signal the incoming stack that one feeder has terminated
*/
public void oneFeederTerminated() {
this.feedersTerminated.incrementAndGet();
final int c = this.feedersAlive.decrementAndGet();
assert c >= 0 : "feeders = " + c;
}
public void oneFeederStarted() {
this.feedersAlive.addAndGet(1);
}
public QueryParams getQuery() {
return this.query;
}
public int[] flagCount() {
return this.flagcount;
}
protected void addBegin() {
this.addRunning = true;
}
public void addFinalize() {
this.addRunning = false;
}
protected boolean addRunning() {
return this.addRunning;
}
public boolean rwiIsEmpty() {
if ( !this.rwiStack.isEmpty() ) {
return false;
}
for ( final WeakPriorityBlockingQueue<WordReferenceVars> s : this.doubleDomCache.values() ) {
if ( !s.isEmpty() ) {
return false;
}
}
return true;
}
protected int rwiQueueSize() {
int c = this.rwiStack.sizeQueue();
for ( final WeakPriorityBlockingQueue<WordReferenceVars> s : this.doubleDomCache.values() ) {
c += s.sizeQueue();
}
return c;
}
protected boolean testFlags(final Bitfield flags) {
if (this.query.constraint == null) return true;
// test if ientry matches with filter
// if all = true: let only entries pass that has all matching bits
// if all = false: let all entries pass that has at least one matching bit
if (this.query.allofconstraint) {
for ( int i = 0; i < 32; i++ ) {
if ((this.query.constraint.get(i)) && (!flags.get(i))) return false;
}
return true;
}
for (int i = 0; i < 32; i++) {
if ((this.query.constraint.get(i)) && (flags.get(i))) return true;
}
return false;
}
@Override
public void run() {
// do a search
oneFeederStarted();
// sort the local containers and truncate it to a limited count,
// so following sortings together with the global results will be fast
try {
final long timer = System.currentTimeMillis();
final TermSearch<WordReference> search =
this.query
.getSegment()
.termIndex()
.query(
this.query.query_include_hashes,
this.query.query_exclude_hashes,
null,
Segment.wordReferenceFactory,
this.query.maxDistance);
this.localSearchInclusion = search.inclusion();
final ReferenceContainer<WordReference> index = search.joined();
EventTracker.update(
EventTracker.EClass.SEARCH,
new ProfilingGraph.EventSearch(
this.query.id(true),
SearchEventType.JOIN,
this.query.queryString,
index.size(),
System.currentTimeMillis() - timer),
false);
if ( !index.isEmpty() ) {
add(index, true, "local index: " + this.query.getSegment().getLocation(), -1, this.maxtime);
this.addFinalize();
}
} catch ( final Exception e ) {
Log.logException(e);
} finally {
oneFeederTerminated();
}
}
public void add(
final ReferenceContainer<WordReference> index,
final boolean local,
final String resourceName,
final int fullResource,
final long maxtime) {
// we collect the urlhashes and construct a list with urlEntry objects
// attention: if minEntries is too high, this method will not terminate within the maxTime
//Log.logInfo("RWIProcess", "added a container, size = " + index.size());
this.addRunning = true;
assert (index != null);
if (index.isEmpty()) return;
if (local) {
this.query.local_rwi_stored.addAndGet(fullResource);
} else {
assert fullResource >= 0 : "fullResource = " + fullResource;
this.query.remote_stored.addAndGet(fullResource);
this.query.remote_peerCount.incrementAndGet();
}
long timer = System.currentTimeMillis();
// normalize entries
final BlockingQueue<WordReferenceVars> decodedEntries = this.order.normalizeWith(index, maxtime);
int is = index.size();
EventTracker.update(EventTracker.EClass.SEARCH, new ProfilingGraph.EventSearch(
this.query.id(true),
SearchEventType.NORMALIZING,
resourceName,
is,
System.currentTimeMillis() - timer), false);
if (!local) this.receivedRemoteReferences.addAndGet(is);
// iterate over normalized entries and select some that are better than currently stored
timer = System.currentTimeMillis();
final boolean nav_hosts = this.query.navigators.equals("all") || this.query.navigators.indexOf("hosts", 0) >= 0;
// apply all constraints
long timeout = System.currentTimeMillis() + maxtime;
try {
WordReferenceVars iEntry;
final String pattern = this.query.urlMask.pattern();
final boolean httpPattern = pattern.equals("http://.*");
final boolean noHttpButProtocolPattern = pattern.equals("https://.*") || pattern.equals("ftp://.*") || pattern.equals("smb://.*") || pattern.equals("file://.*");
long remaining;
pollloop: while ( true ) {
remaining = timeout - System.currentTimeMillis();
if (remaining <= 0) {
Log.logWarning("RWIProcess", "terminated 'add' loop before poll time-out = " + remaining + ", decodedEntries.size = " + decodedEntries.size());
break;
}
iEntry = decodedEntries.poll(remaining, TimeUnit.MILLISECONDS);
if (iEntry == null) {
Log.logWarning("RWIProcess", "terminated 'add' loop after poll time-out = " + remaining + ", decodedEntries.size = " + decodedEntries.size());
break pollloop;
}
if (iEntry == WordReferenceVars.poison) {
break pollloop;
}
assert (iEntry.urlhash().length == index.row().primaryKeyLength);
// doublecheck for urls
if (this.urlhashes.has(iEntry.urlhash())) continue pollloop;
// increase flag counts
Bitfield flags = iEntry.flags();
for (int j = 0; j < 32; j++) {
if (flags.get(j)) this.flagcount[j]++;
}
// check constraints
if (!this.testFlags(flags)) continue pollloop;
// check document domain
if (this.query.contentdom.getCode() > 0 &&
((this.query.contentdom == ContentDomain.AUDIO && !(flags.get(Condenser.flag_cat_hasaudio))) ||
(this.query.contentdom == ContentDomain.VIDEO && !(flags.get(Condenser.flag_cat_hasvideo))) ||
(this.query.contentdom == ContentDomain.IMAGE && !(flags.get(Condenser.flag_cat_hasimage))) ||
(this.query.contentdom == ContentDomain.APP && !(flags.get(Condenser.flag_cat_hasapp))))) {
continue pollloop;
}
// count domZones
//this.domZones[DigestURI.domDomain(iEntry.metadataHash())]++;
// check site constraints
final String hosthash = iEntry.hosthash();
if ( this.query.sitehash == null ) {
if (this.query.siteexcludes != null && this.query.siteexcludes.contains(hosthash)) continue pollloop;
} else {
// filter out all domains that do not match with the site constraint
if (!hosthash.equals(this.query.sitehash)) continue pollloop;
}
// collect host navigation information (even if we have only one; this is to provide a switch-off button)
if (this.query.navigators.isEmpty() && (nav_hosts || this.query.urlMask_isCatchall)) {
this.hostNavigator.inc(hosthash);
this.hostResolver.put(hosthash, iEntry.urlhash());
}
// check protocol
if (!this.query.urlMask_isCatchall) {
final boolean httpFlagSet = DigestURI.flag4HTTPset(iEntry.urlHash);
if (httpPattern && !httpFlagSet) continue pollloop;
if (noHttpButProtocolPattern && httpFlagSet) continue pollloop;
}
// check vocabulary constraint
String subject = YaCyMetadata.hashURI(iEntry.urlhash());
Resource resource = JenaTripleStore.getResource(subject);
if (this.query.metatags != null && !this.query.metatags.isEmpty()) {
// all metatags must appear in the tags list
for (Tagging.Metatag metatag: this.query.metatags) {
Iterator<RDFNode> ni = JenaTripleStore.getObjects(resource, metatag.getPredicate());
if (!ni.hasNext()) continue pollloop;
String tags = ni.next().toString();
if (tags.indexOf(metatag.getObject()) < 0) continue pollloop;
}
}
// add navigators using the triplestore
for (Map.Entry<String, String> v: this.taggingPredicates.entrySet()) {
Iterator<RDFNode> ni = JenaTripleStore.getObjects(resource, v.getValue());
while (ni.hasNext()) {
String[] tags = ni.next().toString().split(",");
for (String tag: tags) {
ScoreMap<String> voc = this.vocabularyNavigator.get(v.getKey());
if (voc == null) {
voc = new ConcurrentScoreMap<String>();
this.vocabularyNavigator.put(v.getKey(), voc);
}
voc.inc(tag);
}
}
}
// finally extend the double-check and insert result to stack
this.urlhashes.putUnique(iEntry.urlhash());
rankingtryloop: while (true) {
try {
this.rwiStack.put(new ReverseElement<WordReferenceVars>(iEntry, this.order.cardinal(iEntry))); // inserts the element and removes the worst (which is smallest)
break rankingtryloop;
} catch ( final ArithmeticException e ) {
// this may happen if the concurrent normalizer changes values during cardinal computation
continue rankingtryloop;
}
}
// increase counter for statistics
if (local) this.query.local_rwi_available.incrementAndGet(); else this.query.remote_available.incrementAndGet();
}
if (System.currentTimeMillis() >= timeout) Log.logWarning("RWIProcess", "rwi normalization ended with timeout = " + maxtime);
} catch ( final InterruptedException e ) {
} catch ( final SpaceExceededException e ) {
}
//if ((query.neededResults() > 0) && (container.size() > query.neededResults())) remove(true, true);
EventTracker.update(EventTracker.EClass.SEARCH, new ProfilingGraph.EventSearch(
this.query.id(true),
SearchEventType.PRESORT,
resourceName,
index.size(),
System.currentTimeMillis() - timer), false);
}
protected Map<byte[], ReferenceContainer<WordReference>> searchContainerMap() {
// direct access to the result maps is needed for abstract generation
// this is only available if execQuery() was called before
return this.localSearchInclusion;
}
public ScoreMap<String> getHostNavigator() {
final ScoreMap<String> result = new ConcurrentScoreMap<String>();
if ( !this.query.navigators.equals("all") && this.query.navigators.indexOf("hosts", 0) < 0 ) {
return result;
}
final Iterator<String> domhashs = this.hostNavigator.keys(false);
URIMetadataNode row;
byte[] urlhash;
String hosthash, hostname;
if ( this.hostResolver != null ) {
while ( domhashs.hasNext() && result.sizeSmaller(30) ) {
hosthash = domhashs.next();
if ( hosthash == null ) {
continue;
}
urlhash = this.hostResolver.get(hosthash);
row = urlhash == null ? null : this.query.getSegment().fulltext().getMetadata(urlhash);
hostname = row == null ? null : row.url().getHost();
if ( hostname != null ) {
result.set(hostname, this.hostNavigator.get(hosthash));
}
}
}
return result;
}
public Map<String,ScoreMap<String>> getVocabularyNavigators() {
return this.vocabularyNavigator;
}
protected ScoreMap<String> getTopicNavigator(final int count) {
// create a list of words that had been computed by statistics over all
// words that appeared in the url or the description of all urls
final ScoreMap<String> result = new ConcurrentScoreMap<String>();
if ( !this.query.navigators.equals("all") && this.query.navigators.indexOf("topics", 0) < 0 ) {
return result;
}
if ( this.ref.sizeSmaller(2) ) {
this.ref.clear(); // navigators with one entry are not useful
}
final Map<String, Float> counts = new HashMap<String, Float>();
final Iterator<String> i = this.ref.keys(false);
String word;
int c;
float q, min = Float.MAX_VALUE, max = Float.MIN_VALUE;
int ic = count;
while ( ic-- > 0 && i.hasNext() ) {
word = i.next();
if ( word == null ) {
continue;
}
c = this.query.getSegment().getQueryCount(word);
if ( c > 0 ) {
q = ((float) this.ref.get(word)) / ((float) c);
min = Math.min(min, q);
max = Math.max(max, q);
counts.put(word, q);
}
}
if ( max > min ) {
for ( final Map.Entry<String, Float> ce : counts.entrySet() ) {
result.set(ce.getKey(), (int) (((double) count) * (ce.getValue() - min) / (max - min)));
}
}
return this.ref;
}
private final static Pattern lettermatch = Pattern.compile("[a-z]+");
public void addTopic(final String[] words) {
String word;
for ( final String w : words ) {
word = w.toLowerCase();
if ( word.length() > 2
&& "http_html_php_ftp_www_com_org_net_gov_edu_index_home_page_for_usage_the_and_zum_der_die_das_und_the_zur_bzw_mit_blog_wiki_aus_bei_off"
.indexOf(word) < 0
&& !this.query.query_include_hashes.has(Word.word2hash(word))
&& lettermatch.matcher(word).matches()
&& !Switchboard.badwords.contains(word)
&& !Switchboard.stopwords.contains(word) ) {
this.ref.inc(word);
}
}
}
protected void addTopics(final ResultEntry resultEntry) {
// take out relevant information for reference computation
if ( (resultEntry.url() == null) || (resultEntry.title() == null) ) {
return;
}
//final String[] urlcomps = htmlFilterContentScraper.urlComps(resultEntry.url().toNormalform(true, true)); // word components of the url
final String[] descrcomps = MultiProtocolURI.splitpattern.split(resultEntry.title().toLowerCase()); // words in the description
// add references
//addTopic(urlcomps);
addTopic(descrcomps);
}
}