|
|
|
@ -750,16 +750,21 @@ public class MediawikiImporter extends Thread implements Importer {
|
|
|
|
|
|
|
|
|
|
public static void main(final String[] s) {
|
|
|
|
|
if (s.length == 0) {
|
|
|
|
|
ConcurrentLog.info("WIKITRANSLATION", "usage:");
|
|
|
|
|
ConcurrentLog.info("WIKITRANSLATION", " -index <wikipedia-dump>");
|
|
|
|
|
ConcurrentLog.info("WIKITRANSLATION", " -read <start> <len> <idx-file>");
|
|
|
|
|
ConcurrentLog.info("WIKITRANSLATION", " -find <title> <wikipedia-dump>");
|
|
|
|
|
ConcurrentLog.info("WIKITRANSLATION", " -convert <wikipedia-dump-xml.bz2> <convert-target-dir> <url-stub>");
|
|
|
|
|
System.exit(0);
|
|
|
|
|
System.out.println("usage:");
|
|
|
|
|
System.out.println(" -index <wikipedia-dump>");
|
|
|
|
|
System.out.println(" -read <start> <len> <idx-file>");
|
|
|
|
|
System.out.println(" -find <title> <wikipedia-dump>");
|
|
|
|
|
System.out.println(" -convert <wikipedia-dump-xml.bz2> <convert-target-dir> <url-stub>");
|
|
|
|
|
ConcurrentLog.shutdown();
|
|
|
|
|
return;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
try {
|
|
|
|
|
// example:
|
|
|
|
|
// java -Xmx2000m -cp classes:lib/bzip2.jar de.anomic.tools.mediawikiIndex -convert DATA/HTCACHE/dewiki-20090311-pages-articles.xml.bz2 DATA/SURROGATES/in/ http://de.wikipedia.org/wiki/
|
|
|
|
|
// java -Xmx2000m -cp classes:lib/bzip2.jar
|
|
|
|
|
// de.anomic.tools.mediawikiIndex -convert
|
|
|
|
|
// DATA/HTCACHE/dewiki-20090311-pages-articles.xml.bz2
|
|
|
|
|
// DATA/SURROGATES/in/ http://de.wikipedia.org/wiki/
|
|
|
|
|
|
|
|
|
|
if (s[0].equals("-convert") && s.length > 2) {
|
|
|
|
|
final File sourcefile = new File(s[1]);
|
|
|
|
@ -802,7 +807,9 @@ public class MediawikiImporter extends Thread implements Importer {
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
}
|
|
|
|
|
System.exit(0);
|
|
|
|
|
} finally {
|
|
|
|
|
ConcurrentLog.shutdown();
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
}
|
|
|
|
|