Package org.apache.pig.backend.datastorage

Examples of org.apache.pig.backend.datastorage.DataStorage

175
176
177
178
179
180
181
182
183
184
185
186
        return getSize(fileName, ConfigurationUtil.toProperties(conf));
    }
   
    public static long getSize(String fileName, Properties properties) throws IOException {
      DataStorage dds = new HDataStorage(properties);
        ElementDescriptor elem = dds.asElement(fileName);
      
        // recursively get all the files under this path
        ElementDescriptor[] allElems = getFileElementDescriptors(elem);
       
        long size = 0;
View Full Code Here
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
     * @return an array of Element descriptors for files present (found by traversing all levels of dirs)
     *  in the input element descriptor
     * @throws DataStorageException
     */
    private static ElementDescriptor[] getFileElementDescriptors(ElementDescriptor elem) throws DataStorageException {
        DataStorage store = elem.getDataStorage();
        ElementDescriptor[] elems = store.asCollection(elem.toString());
        // elems could have directories in it, if so
        // get the files out so that it contains only files
        List<ElementDescriptor> paths = new ArrayList<ElementDescriptor>();
        List<ElementDescriptor> filePaths = new ArrayList<ElementDescriptor>();
        for (int m = 0; m < elems.length; m++) {
            paths.add(elems[m]);
        }
        for (int j = 0; j < paths.size(); j++) {
            ElementDescriptor fullPath = store.asElement(store
                    .getActiveContainer(), paths.get(j));
            // Skip hadoop's private/meta files ...
            if (fullPath.systemElement()) {
                continue;
            }
View Full Code Here
984
985
986
987
988
989
990
991
992
993
994
995
996
    public long capacity() throws IOException {
        if (pigContext.getExecType() == ExecType.LOCAL) {
            throw new IOException("capacity only supported for non-local execution");
        }
        else {
            DataStorage dds = pigContext.getDfs();
           
            Map<String, Object> stats = dds.getStatistics();

            String rawCapacityStr = (String) stats.get(DataStorage.RAW_CAPACITY_KEY);
            String rawUsedStr = (String) stats.get(DataStorage.RAW_USED_KEY);
           
            if ((rawCapacityStr == null) || (rawUsedStr == null)) {
View Full Code Here
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
     * @param filename
     * @return length of the file in bytes
     * @throws IOException
     */
    public long fileSize(String filename) throws IOException {
        DataStorage dfs = pigContext.getDfs();
        ElementDescriptor elem = dfs.asElement(filename);
        Map<String, Object> stats = elem.getStatistics();
        long length = (Long) stats.get(ElementDescriptor.LENGTH_KEY);
        int replication = (Short) stats
                .get(ElementDescriptor.BLOCK_REPLICATION_KEY);

View Full Code Here
469
470
471
472
473
474
475
476
477
478
479
480
481
    public long capacity() throws IOException {
        if (pigContext.getExecType() == ExecType.LOCAL) {
            throw new IOException("capacity only supported for non-local execution");
        }
        else {
            DataStorage dds = pigContext.getDfs();
           
            Map<String, Object> stats = dds.getStatistics();

            String rawCapacityStr = (String) stats.get(DataStorage.RAW_CAPACITY_KEY);
            String rawUsedStr = (String) stats.get(DataStorage.RAW_USED_KEY);
           
            if ((rawCapacityStr == null) || (rawUsedStr == null)) {
View Full Code Here
493
494
495
496
497
498
499
500
501
502
503
504
     * Returns the length of a file in bytes which exists in the HDFS (accounts for replication).
     * @param filename
     * @throws IOException
     */
    public long fileSize(String filename) throws IOException {
        DataStorage dfs = pigContext.getDfs();
        ElementDescriptor elem = dfs.asElement(filename);
        Map<String, Object> stats = elem.getStatistics();
        long length = (Long) stats.get(ElementDescriptor.LENGTH_KEY);
        int replication = (Short) stats
                .get(ElementDescriptor.BLOCK_REPLICATION_KEY);

View Full Code Here
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
        src.rename(dst);

    }

    public void copy(String src, String dst, boolean localDst) throws IOException {
        DataStorage dstStorage = dfs;
       
        if (localDst) {
            dstStorage = lfs;
        }
       
        ElementDescriptor srcElement = null;
        ElementDescriptor dstElement = null;

        try {
            srcElement = dfs.asElement(src);
            dstElement = dstStorage.asElement(dst);
        }
        catch (DataStorageException e) {
            StringBuilder sb = new StringBuilder();
            sb.append("Unable to copy ");
            sb.append(src);
View Full Code Here
153
154
155
156
157
158
159
160
161
        return openDFSFile(fileName, ConfigurationUtil.toProperties(wrapper.getJobConf()));
       
    }

    public static InputStream openDFSFile(String fileName, Properties properties) throws IOException{
        DataStorage dds = new HDataStorage(properties);
        ElementDescriptor elem = dds.asElement(fileName);
        return openDFSFile(elem);
    }
View Full Code Here
105
106
107
108
109
110
111
112
113
114
115
116
        return lastConf;
    }

    public RecordReader<Text, Tuple> makeReader(JobConf job) throws IOException {
        lastConf = job;       
        DataStorage store = new HDataStorage(ConfigurationUtil.toProperties(job));
        store.setActiveContainer(store.asContainer("/user/" + job.getUser()));
        wrapped.init(store);
       
        // Mimic org.apache.hadoop.mapred.FileSplit if feasible...
        String[] locations = wrapped.getLocations();
        if (locations.length > 0) {
View Full Code Here
86
87
88
89
90
91
92
93
94
95
96
        }

        FileSystem fs = FileSystem.get(job);
        List<SliceWrapper> splits = new ArrayList<SliceWrapper>();
        for (int i = 0; i < inputs.size(); i++) {
            DataStorage store = new HDataStorage(ConfigurationUtil.toProperties(job));
            ValidatingInputFileSpec spec;
            if (inputs.get(i) instanceof ValidatingInputFileSpec) {
                spec = (ValidatingInputFileSpec) inputs.get(i);
            } else {
                spec = new ValidatingInputFileSpec(inputs.get(i), store);
View Full Code Here
TOP

Related Classes of org.apache.pig.backend.datastorage.DataStorage

  • com.packetloop.packetpig.storage.JsonMetadata
  • org.apache.pig.backend.hadoop.executionengine.mapreduceExec.PigInputFormat
  • org.apache.pig.backend.hadoop.executionengine.mapreduceExec.SliceWrapper
  • org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigInputFormat
  • org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.SliceWrapper
  • org.apache.pig.builtin.JsonMetadata
  • org.apache.pig.impl.io.FileLocalizer
  • org.apache.pig.impl.PigContext
  • org.apache.pig.parser.QueryParserUtils
  • org.apache.pig.piggybank.storage.JsonMetadata
    • Copyright © 2018 www.massapicom. 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.