Package net.yacy.kelondro.data.meta

Examples of net.yacy.kelondro.data.meta.URIMetadataRow.metadata()

617
618
619
620
621
622
623
624
625
626
627
        statsDump = new ArrayList<HostStat>();
        TreeSet<String> set = new TreeSet<String>();
        for (URLHashCounter hs: domainSamples.values()) {
            if (hs == null) continue;
            urlref = this.load(hs.urlhashb);
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            set.add(urlref.metadata().url().getHost());
            count--;
            if (count == 0) break;
        }
        return set;
View Full Code Here
617
618
619
620
621
622
623
624
625
626
627
        statsDump = new ArrayList<HostStat>();
        TreeSet<String> set = new TreeSet<String>();
        for (URLHashCounter hs: domainSamples.values()) {
            if (hs == null) continue;
            urlref = this.load(hs.urlhashb);
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            set.add(urlref.metadata().url().getHost());
            count--;
            if (count == 0) break;
        }
        return set;
View Full Code Here
617
618
619
620
621
622
623
624
625
626
627
        statsDump = new ArrayList<HostStat>();
        TreeSet<String> set = new TreeSet<String>();
        for (URLHashCounter hs: domainSamples.values()) {
            if (hs == null) continue;
            urlref = this.load(hs.urlhashb);
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            set.add(urlref.metadata().url().getHost());
            count--;
            if (count == 0) break;
        }
        return set;
View Full Code Here
618
619
620
621
622
623
624
625
626
627
628
        TreeSet<String> set = new TreeSet<String>();
        for (URLHashCounter hs: domainSamples.values()) {
            if (hs == null) continue;
            urlref = this.load(hs.urlhashb);
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            set.add(urlref.metadata().url().getHost());
            count--;
            if (count == 0) break;
        }
        return set;
    }
View Full Code Here
657
658
659
660
661
662
663
664
665
666
667
        }
        URIMetadataRow.Components comps;
        DigestURI url;
        for (Map.Entry<String, URLHashCounter> e: domainSamples.entrySet()) {
            urlref = this.load(e.getValue().urlhashb);
            comps = urlref.metadata();
            url = comps.url();
            hostMap.put(e.getKey(), new HostStat(url.getHost(), url.getPort(), e.getKey(), hosthashScore.get(e.getKey())));
        }
        return hostMap;
    }
View Full Code Here
681
682
683
684
685
686
687
688
689
690
691
        DigestURI url;
        while (j.hasNext()) {
            urlhash = j.next();
            if (urlhash == null) continue;
            urlref = this.load(ASCII.getBytes(urlhash));
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            if (statsDump == null) return new ArrayList<HostStat>().iterator(); // some other operation has destroyed the object
            comps = urlref.metadata();
            url = comps.url();
            statsDump.add(new HostStat(url.getHost(), url.getPort(), urlhash.substring(6), domainScore.get(urlhash)));
            count--;
View Full Code Here
681
682
683
684
685
686
687
688
689
690
691
        DigestURI url;
        while (j.hasNext()) {
            urlhash = j.next();
            if (urlhash == null) continue;
            urlref = this.load(ASCII.getBytes(urlhash));
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            if (statsDump == null) return new ArrayList<HostStat>().iterator(); // some other operation has destroyed the object
            comps = urlref.metadata();
            url = comps.url();
            statsDump.add(new HostStat(url.getHost(), url.getPort(), urlhash.substring(6), domainScore.get(urlhash)));
            count--;
View Full Code Here
681
682
683
684
685
686
687
688
689
690
691
        DigestURI url;
        while (j.hasNext()) {
            urlhash = j.next();
            if (urlhash == null) continue;
            urlref = this.load(ASCII.getBytes(urlhash));
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            if (statsDump == null) return new ArrayList<HostStat>().iterator(); // some other operation has destroyed the object
            comps = urlref.metadata();
            url = comps.url();
            statsDump.add(new HostStat(url.getHost(), url.getPort(), urlhash.substring(6), domainScore.get(urlhash)));
            count--;
View Full Code Here
683
684
685
686
687
688
689
690
691
692
693
            urlhash = j.next();
            if (urlhash == null) continue;
            urlref = this.load(ASCII.getBytes(urlhash));
            if (urlref == null || urlref.metadata() == null || urlref.metadata().url() == null || urlref.metadata().url().getHost() == null) continue;
            if (statsDump == null) return new ArrayList<HostStat>().iterator(); // some other operation has destroyed the object
            comps = urlref.metadata();
            url = comps.url();
            statsDump.add(new HostStat(url.getHost(), url.getPort(), urlhash.substring(6), domainScore.get(urlhash)));
            count--;
            if (count == 0) break;
        }
View Full Code Here
376
377
378
379
380
381
382
383
384
385
386
                    if (entry == null) {
                        if (Log.isFine("URLDBCLEANER")) Log.logFine("URLDBCLEANER", "entry == null");
                    } else if (entry.hash() == null) {
                        if (Log.isFine("URLDBCLEANER")) Log.logFine("URLDBCLEANER", ++blacklistedUrls + " blacklisted (" + ((double) blacklistedUrls / totalSearchedUrls) * 100 + "%): " + "hash == null");
                    } else {
                        final URIMetadataRow.Components metadata = entry.metadata();
                        totalSearchedUrls++;
                        if (metadata == null) {
                            if (Log.isFine("URLDBCLEANER")) Log.logFine("URLDBCLEANER", "corrupted entry for hash = " + ASCII.String(entry.hash()));
                            remove(entry.hash());
                            continue;
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.