/**
* This method checks to see if symlinks are to be create for the localized cache files in the
* current working directory
*
* @param conf the jobconf
* @return true if symlinks are to be created- else return false
*/
public static boolean getSymlink(Configuration conf) {
String result = conf.get("mapred.create.symlink");
if ("yes".equals(result)) {
return true;
}
return false;
}
/**
* This method checks to see if symlinks are to be create for the localized cache files in the
* current working directory Used by internal DistributedCache code.
*
* @param conf the jobconf
* @return true if symlinks are to be created- else return false
*/
public static boolean getSymlink(Configuration conf) {
String result = conf.get(CACHE_SYMLINK);
if ("yes".equals(result)) {
return true;
}
return false;
}
public int run(String[] args) throws Exception {
Configuration argConf = getConf();
// JobConf conf = new JobConf(diffdb.class);
Configuration config = HBaseConfiguration.create();
HBaseAdmin hbAdmin = new HBaseAdmin(config);
dbutil db_util = new dbutil(config);
HTable runTable = new HTable(config, "gestore_runs");
Get runGet = new Get(argConf.get("id").getBytes());
Result pipeline = runTable.get(runGet);
NavigableMap<byte[], byte[]> pipeMap = pipeline.getFamilyMap("d".getBytes());
Map.Entry<byte[], byte[]> results = pipeMap.pollFirstEntry();
HashMap<String, HashMap<String, String>> resultMap =
new HashMap<String, HashMap<String, String>>();
while (results != null) {
String resultKey = new String(results.getKey());
String resultValue = new String(results.getValue());
String field = "type";
HashMap<String, String> tempMap = new HashMap<String, String>();
String entry = resultKey;
if (resultKey.endsWith("_db_timestamp")) {
field = "db_timestamp";
entry = resultKey.substring(0, resultKey.lastIndexOf("_db_timestamp"));
} else if (resultKey.endsWith("_filename")) {
field = "filename";
entry = resultKey.substring(0, resultKey.lastIndexOf("_filename"));
} else if (resultKey.endsWith("_regex")) {
field = "regex";
entry = resultKey.substring(0, resultKey.lastIndexOf("_regex"));
}
if (resultMap.containsKey(entry)) {
tempMap = resultMap.get(entry);
}
tempMap.put(field, resultValue);
resultMap.put(entry, tempMap);
// System.out.println("Key: " + resultKey + " Value: " + resultValue);
results = pipeMap.pollFirstEntry();
}
for (String key : resultMap.keySet()) {
System.out.println("File ID: " + key);
for (String subKey : resultMap.get(key).keySet()) {
// System.out.println("\t " + subKey + "\t\t" + resultMap.get(key).get(subKey));
System.out.format(" %1$-20s %2$s\n", subKey, resultMap.get(key).get(subKey));
}
}
return 0;
}
public static InetSocketAddress getAddress(Configuration conf) {
String jobTrackerStr = conf.get("mapred.job.tracker", "localhost:8012");
int colon = jobTrackerStr.indexOf(":");
if (colon < 0) {
throw new RuntimeException("Bad mapred.job.tracker: " + jobTrackerStr);
}
String jobTrackerName = jobTrackerStr.substring(0, colon);
int jobTrackerPort = Integer.parseInt(jobTrackerStr.substring(colon + 1));
return new InetSocketAddress(jobTrackerName, jobTrackerPort);
}
/**
* Add a file path to the current set of classpath entries. It adds the file to cache as well.
* Intended to be used by user code.
*
* @param file Path of the file to be added
* @param conf Configuration that contains the classpath setting
* @param fs FileSystem with respect to which {@code archivefile} should be interpreted.
*/
public static void addFileToClassPath(Path file, Configuration conf, FileSystem fs)
throws IOException {
String filepath = file.toUri().getPath();
String classpath = conf.get("mapred.job.classpath.files");
conf.set(
"mapred.job.classpath.files",
classpath == null ? filepath : classpath + System.getProperty("path.separator") + filepath);
URI uri = fs.makeQualified(file).toUri();
addCacheFile(uri, conf);
}
/**
* Get the archive entries in classpath as an array of Path
*
* @param conf Configuration that contains the classpath setting
*/
public static Path[] getArchiveClassPaths(Configuration conf) {
String classpath = conf.get("mapred.job.classpath.archives");
if (classpath == null) return null;
ArrayList list =
Collections.list(new StringTokenizer(classpath, System.getProperty("path.separator")));
Path[] paths = new Path[list.size()];
for (int i = 0; i < list.size(); i++) {
paths[i] = new Path((String) list.get(i));
}
return paths;
}
/**
* Add an archive path to the current set of classpath entries. It adds the archive to cache as
* well. Intended to be used by user code.
*
* @param archive Path of the archive to be added
* @param conf Configuration that contains the classpath setting
* @param fs FileSystem with respect to which {@code archive} should be interpreted.
*/
public static void addArchiveToClassPath(Path archive, Configuration conf, FileSystem fs)
throws IOException {
String archivepath = archive.toUri().getPath();
String classpath = conf.get("mapred.job.classpath.archives");
conf.set(
"mapred.job.classpath.archives",
classpath == null
? archivepath
: classpath + System.getProperty("path.separator") + archivepath);
URI uri = fs.makeQualified(archive).toUri();
addCacheArchive(uri, conf);
}
private static URI addArchiveToClassPathHelper(Path archive, Configuration conf)
throws IOException {
String classpath = conf.get("mapred.job.classpath.archives");
// the scheme/authority use ':' as separator. put the unqualified path in classpath
String archivePath = archive.toUri().getPath();
conf.set(
"mapred.job.classpath.archives",
classpath == null
? archivePath
: classpath + System.getProperty("path.separator") + archivePath);
return archive.makeQualified(archive.getFileSystem(conf)).toUri();
}
/**
* Construct in a named directory.
*
* @param conf
* @param dir
* @throws IOException
*/
public NutchBean(Configuration conf, Path dir) throws IOException {
this.conf = conf;
this.fs = FileSystem.get(this.conf);
if (dir == null) {
dir = new Path(this.conf.get("searcher.dir", "crawl"));
}
final Path luceneConfig = new Path(dir, "search-servers.txt");
final Path solrConfig = new Path(dir, "solr-servers.txt");
if (fs.exists(luceneConfig) || fs.exists(solrConfig)) {
searchBean = new DistributedSearchBean(conf, luceneConfig, solrConfig);
} else {
final Path indexDir = new Path(dir, "index");
final Path indexesDir = new Path(dir, "indexes");
searchBean = new LuceneSearchBean(conf, indexDir, indexesDir);
}
String tableName = conf.get("table.name", "webtable");
table = new HTable(tableName);
summarizerFactory = new SummarizerFactory(conf);
}
/**
* Add a file to be localized to the conf
*
* @param uri The uri of the cache to be localized
* @param conf Configuration to add the cache to
*/
public static void addCacheFile(URI uri, Configuration conf) {
String files = conf.get("mapred.cache.files");
conf.set("mapred.cache.files", files == null ? uri.toString() : files + "," + uri.toString());
}
/**
* Add a archives to be localized to the conf
*
* @param uri The uri of the cache to be localized
* @param conf Configuration to add the cache to
*/
public static void addSharedCacheArchive(URI uri, Configuration conf) {
String archives = conf.get("mapred.cache.shared.archives");
conf.set(
"mapred.cache.shared.archives",
archives == null ? uri.toString() : archives + "," + uri.toString());
}
/** Sets up configuration based on params */
private static boolean setup(Hashtable<String, String> curConf, Configuration argConf) {
if (argConf.get("file") == null) {
logger.fatal("Missing file parameter");
System.exit(1);
}
if (argConf.get("hdfs_base_path") == null) {
logger.fatal("Missing HDFS base path, check gestore-conf.xml");
System.exit(1);
}
if (argConf.get("hdfs_temp_path") == null) {
logger.fatal("Missing HDFS temp path, check gestore-conf.xml");
System.exit(1);
}
if (argConf.get("local_temp_path") == null) {
logger.fatal("Missing local temp path, check gestore-conf.xml");
System.exit(1);
}
// Input paramaters
curConf.put("run_id", argConf.get("run", ""));
curConf.put("task_id", argConf.get("task", ""));
curConf.put("file_id", argConf.get("file"));
curConf.put("local_path", argConf.get("path", ""));
curConf.put("type", argConf.get("type", "l2r"));
curConf.put("timestamp_start", argConf.get("timestamp_start", "1"));
curConf.put(
"timestamp_stop", argConf.get("timestamp_stop", Integer.toString(Integer.MAX_VALUE)));
curConf.put("delimiter", argConf.get("regex", "ID=.*"));
curConf.put("taxon", argConf.get("taxon", "all"));
curConf.put("intermediate", argConf.get("full_run", "false"));
curConf.put("quick_add", argConf.get("quick_add", "false"));
Boolean full_run = curConf.get("intermediate").matches("(?i).*true.*");
curConf.put("format", argConf.get("format", "unknown"));
curConf.put("split", argConf.get("split", "1"));
curConf.put("copy", argConf.get("copy", "true"));
// Constants
curConf.put("base_path", argConf.get("hdfs_base_path"));
curConf.put("temp_path", argConf.get("hdfs_temp_path"));
curConf.put("local_temp_path", argConf.get("local_temp_path"));
curConf.put("db_name_files", argConf.get("hbase_file_table"));
curConf.put("db_name_runs", argConf.get("hbase_run_table"));
curConf.put("db_name_updates", argConf.get("hbase_db_update_table"));
// Timestamps
Date currentTime = new Date();
Date endDate = new Date(new Long(curConf.get("timestamp_stop")));
curConf.put("timestamp_real", Long.toString(currentTime.getTime()));
return true;
}
/**
* Add a file to be localized to the conf. Intended to be used by user code.
*
* @param uri The uri of the cache to be localized
* @param conf Configuration to add the cache to
*/
public static void addCacheFile(URI uri, Configuration conf) {
String files = conf.get(CACHE_FILES);
conf.set(CACHE_FILES, files == null ? uri.toString() : files + "," + uri.toString());
}
/**
* Add a archives to be localized to the conf. Intended to be used by user code.
*
* @param uri The uri of the cache to be localized
* @param conf Configuration to add the cache to
*/
public static void addCacheArchive(URI uri, Configuration conf) {
String archives = conf.get(CACHE_ARCHIVES);
conf.set(CACHE_ARCHIVES, archives == null ? uri.toString() : archives + "," + uri.toString());
}
/**
* Add a file that has been localized to the conf.. Used by internal DistributedCache code.
*
* @param conf The conf to modify to contain the localized caches
* @param str a comma separated list of local files
*/
public static void addLocalFiles(Configuration conf, String str) {
String files = conf.get(CACHE_LOCALFILES);
conf.set(CACHE_LOCALFILES, files == null ? str : files + "," + str);
}
/**
* Add a archive that has been localized to the conf. Used by internal DistributedCache code.
*
* @param conf The conf to modify to contain the localized caches
* @param str a comma separated list of local archives
*/
public static void addLocalArchives(Configuration conf, String str) {
String archives = conf.get(CACHE_LOCALARCHIVES);
conf.set(CACHE_LOCALARCHIVES, archives == null ? str : archives + "," + str);
}