Package org.apache.flink.api.common.io.statistics

Examples of org.apache.flink.api.common.io.statistics.BaseStatistics

40
41
42
43
44
45
46
47
48
49
50
    try {
      final DummyFileInputFormat format = new DummyFileInputFormat();
      format.setFilePath("file:///some/none/existing/directory/");
      format.configure(new Configuration());
     
      BaseStatistics stats = format.getStatistics(null);
      Assert.assertNull("The file statistics should be null.", stats);
    } catch (Exception ex) {
      ex.printStackTrace();
      Assert.fail(ex.getMessage());
    }
View Full Code Here
58
59
60
61
62
63
64
65
66
67
68
69
     
      final DummyFileInputFormat format = new DummyFileInputFormat();
      format.setFilePath(tempFile);
      format.configure(new Configuration());
     
      BaseStatistics stats = format.getStatistics(null);
      Assert.assertEquals("The file size from the statistics is wrong.", SIZE, stats.getTotalInputSize());
    } catch (Exception ex) {
      ex.printStackTrace();
      Assert.fail(ex.getMessage());
    }
  }
View Full Code Here
80
81
82
83
84
85
86
87
88
89
90
91
     
      final DummyFileInputFormat format = new DummyFileInputFormat();
      format.setFilePath(tempDir);
      format.configure(new Configuration());
     
      BaseStatistics stats = format.getStatistics(null);
      Assert.assertEquals("The file size from the statistics is wrong.", TOTAL, stats.getTotalInputSize());
    } catch (Exception ex) {
      ex.printStackTrace();
      Assert.fail(ex.getMessage());
    }
  }
View Full Code Here
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
      format = new DummyFileInputFormat();
      format.setFilePath(tempFile);
      format.configure(new Configuration());
     
      FileBaseStatistics fakeStats = new FileBaseStatistics(stats.getLastModificationTime(), FAKE_SIZE, BaseStatistics.AVG_RECORD_BYTES_UNKNOWN);
      BaseStatistics latest = format.getStatistics(fakeStats);
      Assert.assertEquals("The file size from the statistics is wrong.", FAKE_SIZE, latest.getTotalInputSize());
     
      // insert fake stats with the expired modification time. the call should return new accurate stats
      format = new DummyFileInputFormat();
      format.setFilePath(tempFile);
      format.configure(new Configuration());
     
      FileBaseStatistics outDatedFakeStats = new FileBaseStatistics(stats.getLastModificationTime()-1, FAKE_SIZE, BaseStatistics.AVG_RECORD_BYTES_UNKNOWN);
      BaseStatistics reGathered = format.getStatistics(outDatedFakeStats);
      Assert.assertEquals("The file size from the statistics is wrong.", SIZE, reGathered.getTotalInputSize());
     
    } catch (Exception ex) {
      ex.printStackTrace();
      Assert.fail(ex.getMessage());
    }
View Full Code Here
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
      format = new DummyFileInputFormat();
      format.setFilePath(tempDir);
      format.configure(new Configuration());
     
      FileBaseStatistics fakeStats = new FileBaseStatistics(stats.getLastModificationTime(), FAKE_SIZE, BaseStatistics.AVG_RECORD_BYTES_UNKNOWN);
      BaseStatistics latest = format.getStatistics(fakeStats);
      Assert.assertEquals("The file size from the statistics is wrong.", FAKE_SIZE, latest.getTotalInputSize());
     
      // insert fake stats with the correct modification time. the call should return the fake stats
      format = new DummyFileInputFormat();
      format.setFilePath(tempDir);
      format.configure(new Configuration());
     
      FileBaseStatistics outDatedFakeStats = new FileBaseStatistics(stats.getLastModificationTime()-1, FAKE_SIZE, BaseStatistics.AVG_RECORD_BYTES_UNKNOWN);
      BaseStatistics reGathered = format.getStatistics(outDatedFakeStats);
      Assert.assertEquals("The file size from the statistics is wrong.", TOTAL, reGathered.getTotalInputSize());
     
    } catch (Exception ex) {
      ex.printStackTrace();
      Assert.fail(ex.getMessage());
    }
View Full Code Here
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
      final Configuration conf = new Configuration();
     
      final TestDelimitedInputFormat format = new TestDelimitedInputFormat();
      format.setFilePath(tempFile);
      format.configure(conf);
      BaseStatistics stats = format.getStatistics(null);
     
      final int numLines = TEST_DATA_1_LINES;
      final float avgWidth = ((float) TEST_DATA1.length()) / TEST_DATA_1_LINES;
      Assert.assertTrue("Wrong record count.", stats.getNumberOfRecords() < numLines + 1 & stats.getNumberOfRecords() > numLines - 1);
      Assert.assertTrue("Wrong avg record size.", stats.getAverageRecordWidth() < avgWidth + 1 & stats.getAverageRecordWidth() > avgWidth - 1);
    } catch (Exception e) {
      e.printStackTrace();
      Assert.fail(e.getMessage());
    }
  }
View Full Code Here
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
      final Configuration conf = new Configuration();
     
      final TestDelimitedInputFormat format = new TestDelimitedInputFormat();
      format.setFilePath(tempFile);
      format.configure(conf);
      BaseStatistics stats = format.getStatistics(null);
     
      final int maxNumLines = (int) Math.ceil(TOTAL_SIZE / ((double) Math.min(TEST_DATA_1_LINEWIDTH, TEST_DATA_2_LINEWIDTH)));
      final int minNumLines = (int) (TOTAL_SIZE / ((double) Math.max(TEST_DATA_1_LINEWIDTH, TEST_DATA_2_LINEWIDTH)));
      final float maxAvgWidth = ((float) (TOTAL_SIZE)) / minNumLines;
      final float minAvgWidth = ((float) (TOTAL_SIZE)) / maxNumLines;
     
      if (!(stats.getNumberOfRecords() <= maxNumLines  & stats.getNumberOfRecords() >= minNumLines)) {
        System.err.println("Records: " + stats.getNumberOfRecords() + " out of (" + minNumLines + ", " + maxNumLines + ").");
        Assert.fail("Wrong record count.");
      }
      if (!(stats.getAverageRecordWidth() <= maxAvgWidth & stats.getAverageRecordWidth() >= minAvgWidth)) {
        Assert.fail("Wrong avg record size.");
      }
    } catch (Exception e) {
      e.printStackTrace();
      Assert.fail(e.getMessage());
View Full Code Here
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
      final TestDelimitedInputFormat format = new TestDelimitedInputFormat();
      format.setFilePath(tempFile);
      format.setDelimiter(DELIMITER);
      format.configure(conf);
     
      BaseStatistics stats = format.getStatistics(null);
      final int numLines = TEST_DATA_1_LINES;
      final float avgWidth = ((float) testData.length()) / TEST_DATA_1_LINES;
     
      Assert.assertTrue("Wrong record count.", stats.getNumberOfRecords() < numLines + 1 & stats.getNumberOfRecords() > numLines - 1);
      Assert.assertTrue("Wrong avg record size.", stats.getAverageRecordWidth() < avgWidth + 1 & stats.getAverageRecordWidth() > avgWidth - 1);
    } catch (Exception e) {
      e.printStackTrace();
      Assert.fail(e.getMessage());
    }
  }
View Full Code Here
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
      final TestDelimitedInputFormat format = new TestDelimitedInputFormat();
      format.setFilePath("test://" + tempFile);
      format.configure(conf);
     
      TestFileSystem.resetStreamOpenCounter();
      BaseStatistics stats = format.getStatistics(null);
      Assert.assertEquals("Wrong number of samples taken.", DEFAULT_NUM_SAMPLES, TestFileSystem.getNumtimeStreamOpened());
     
      final TestDelimitedInputFormat format2 = new TestDelimitedInputFormat();
      format2.setFilePath("test://" + tempFile);
      format2.configure(conf);
     
      TestFileSystem.resetStreamOpenCounter();
      BaseStatistics stats2 = format2.getStatistics(stats);
      Assert.assertTrue("Using cached statistics should cicumvent sampling.", 0 == TestFileSystem.getNumtimeStreamOpened());
      Assert.assertTrue("Using cached statistics should cicumvent sampling.", stats == stats2);
     
    } catch (Exception e) {
      e.printStackTrace();
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
      }
      catch (Throwable t) {}
     
      // first of all, get the statistics from the cache
      final String statisticsKey = getPactContract().getStatisticsKey();
      final BaseStatistics cachedStatistics = statistics.getBaseStatistics(statisticsKey);
     
      BaseStatistics bs = null;
      try {
        bs = format.getStatistics(cachedStatistics);
      }
      catch (Throwable t) {
        if (PactCompiler.LOG.isWarnEnabled()) {
          PactCompiler.LOG.warn("Error obtaining statistics from input format: " + t.getMessage(), t);
        }
      }
     
      if (bs != null) {
        final long len = bs.getTotalInputSize();
        if (len == BaseStatistics.SIZE_UNKNOWN) {
          if (PactCompiler.LOG.isInfoEnabled()) {
            PactCompiler.LOG.info("Compiler could not determine the size of input '" + inFormatDescription + "'. Using default estimates.");
          }
        }
        else if (len >= 0) {
          this.estimatedOutputSize = len;
        }
       
        final long card = bs.getNumberOfRecords();
        if (card != BaseStatistics.NUM_RECORDS_UNKNOWN) {
          this.estimatedNumRecords = card;
        }
      }
    }
View Full Code Here
TOP

Related Classes of org.apache.flink.api.common.io.statistics.BaseStatistics

  • org.apache.flink.api.common.io.DelimitedInputFormatSamplingTest
  • org.apache.flink.api.common.io.FileInputFormatTest
  • org.apache.flink.api.common.io.SequentialFormatTest
  • org.apache.flink.compiler.dag.DataSourceNode
    • 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.