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.
yacy_search_server/source/de/anomic/crawler/RSSLoader.java

153 lines
6.5 KiB

/**
* RSSLoader
* Copyright 2010 by Michael Peter Christen, mc@yacy.net, Frankfurt a. M., Germany
* First released 27.8.2010 at http://yacy.net
*
* $LastChangedDate$
* $LastChangedRevision$
* $LastChangedBy$
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library 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 General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program in the file lgpl21.txt
* If not, see <http://www.gnu.org/licenses/>.
*/
package de.anomic.crawler;
import java.io.IOException;
import java.net.MalformedURLException;
import java.util.Date;
import net.yacy.cora.document.RSSFeed;
import net.yacy.cora.document.RSSMessage;
import net.yacy.cora.document.RSSReader;
import net.yacy.cora.document.UTF8;
import net.yacy.cora.storage.ARC;
import net.yacy.cora.storage.ComparableARC;
import net.yacy.document.Parser.Failure;
import net.yacy.kelondro.blob.Tables;
import net.yacy.kelondro.data.meta.DigestURI;
import net.yacy.kelondro.index.RowSpaceExceededException;
import net.yacy.kelondro.logging.Log;
import net.yacy.kelondro.order.Base64Order;
import de.anomic.crawler.retrieval.Response;
import de.anomic.data.WorkTables;
import de.anomic.search.Segments;
import de.anomic.search.Switchboard;
import de.anomic.server.serverObjects;
public class RSSLoader extends Thread {
public static final ARC<byte[], Date> indexTriggered = new ComparableARC<byte[], Date>(1000, Base64Order.enhancedCoder);
DigestURI urlf;
Switchboard sb;
public RSSLoader(Switchboard sb, DigestURI urlf) {
this.sb = sb;
this.urlf = urlf;
}
public void run() {
RSSReader rss = null;
try {
Response response = sb.loader.load(sb.loader.request(urlf, true, false), CrawlProfile.CacheStrategy.NOCACHE, Long.MAX_VALUE, true);
byte[] resource = response == null ? null : response.getContent();
rss = resource == null ? null : RSSReader.parse(RSSFeed.DEFAULT_MAXSIZE, resource);
} catch (MalformedURLException e) {
Log.logWarning("Load_RSS", "rss loading for url '" + this.getName().substring(9) + "' failed: " + e.getMessage());
return;
} catch (IOException e) {
Log.logWarning("Load_RSS", "rss loading for url '" + urlf.toNormalform(true, false) + "' failed: " + e.getMessage());
return;
}
if (rss == null) {
Log.logWarning("Load_RSS", "no rss for url " + urlf.toNormalform(true, false));
return;
}
RSSFeed feed = rss.getFeed();
indexAllRssFeed(sb, urlf, feed);
// add the feed also to the scheduler
recordAPI(sb, null, urlf, feed, 7, "seldays");
}
public static void indexAllRssFeed(Switchboard sb, DigestURI url, RSSFeed feed) {
int loadCount = 0;
loop: for (RSSMessage message: feed) {
try {
DigestURI messageurl = new DigestURI(message.getLink());
if (indexTriggered.containsKey(messageurl.hash())) continue loop;
if (sb.urlExists(Segments.Process.LOCALCRAWLING, messageurl.hash()) != null) continue loop;
sb.addToIndex(messageurl, null, null);
indexTriggered.put(messageurl.hash(), new Date());
loadCount++;
} catch (IOException e) {
Log.logException(e);
} catch (Failure e) {
Log.logException(e);
}
}
// update info for loading
try {
Tables.Data rssRow = sb.tables.select("rss", url.hash());
if (rssRow == null) rssRow = new Tables.Data();
Date lastLoadDate = rssRow.get("last_load_date", new Date(0));
long deltaTime = Math.min(System.currentTimeMillis() - lastLoadDate.getTime(), 1000 * 60 * 60 * 24);
int allLoadCount = rssRow.get("all_load_count", 0);
int lastAvg = rssRow.get("avg_upd_per_day", 0);
long thisAvg = 1000 * 60 * 60 * 24 / deltaTime * loadCount;
long nextAvg = lastAvg == 0 ? thisAvg : (thisAvg + lastAvg * 2) / 3;
rssRow.put("url", UTF8.getBytes(url.toNormalform(true, false)));
rssRow.put("title", feed.getChannel().getTitle());
rssRow.put("last_load_date", new Date());
rssRow.put("last_load_count", loadCount);
rssRow.put("all_load_count", allLoadCount + loadCount);
rssRow.put("avg_upd_per_day", nextAvg);
sb.tables.update("rss", url.hash(), rssRow);
} catch (IOException e) {
Log.logException(e);
} catch (RowSpaceExceededException e) {
Log.logException(e);
}
}
public static void recordAPI(Switchboard sb, String apicall_pk, DigestURI url, RSSFeed feed, int repeat_time, String repeat_unit) {
// record API action
byte[] pk = null;
serverObjects post = new serverObjects();
post.put("url", url.toNormalform(true, false));
post.put("indexAllItemContent", "");
if (apicall_pk != null) post.put(WorkTables.TABLE_API_COL_APICALL_PK, apicall_pk);
if (repeat_time > 0) {
// store as scheduled api call
pk = sb.tables.recordAPICall(post, "Load_RSS_p.html", WorkTables.TABLE_API_TYPE_CRAWLER, "import feed " + url.toNormalform(true, false), repeat_time, repeat_unit.substring(3));
} else {
// store just a protocol
pk = sb.tables.recordAPICall(post, "Load_RSS_p.html", WorkTables.TABLE_API_TYPE_CRAWLER, "import feed " + url.toNormalform(true, false));
}
// store pk of api table into rss table to show that the entry has been recorded
assert pk != null;
Tables.Data rssRow = new Tables.Data();
rssRow.put("url", UTF8.getBytes(url.toNormalform(true, false)));
rssRow.put("title", feed.getChannel().getTitle());
rssRow.put("api_pk", pk);
try {
sb.tables.update("rss", url.hash(), rssRow);
} catch (IOException e) {
Log.logException(e);
}
}
}