Package com.datasalt.pangool.io.Schema

Examples of com.datasalt.pangool.io.Schema.Field

85
86
87
88
89
90
91
92
93
94
95
96
  }

  void write(Schema destinationSchema, ITuple tuple, int[] translationTable, Serializer[] customSerializers)
      throws IOException {
    for(int i = 0; i < destinationSchema.getFields().size(); i++) {
      Field field = destinationSchema.getField(i);
      Type fieldType = field.getType();
      Object element;
      if(translationTable == null) {
        element = tuple.get(i);
      } else {
        element = tuple.get(translationTable[i]);
View Full Code Here
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
            }
            String currentValue = "";
            try {
              currentValue = readLine.get(index);
              if(currentValue != null) {
                Field field = schema.getFields().get(i);
                switch(field.getType()) {
                case INT:
                case LONG:
                case FLOAT:
                case DOUBLE:
                  processNumber(field.getType(), tuple, i, currentValue);
                  break;
                case ENUM:
                  Class clazz = field.getObjectClass();
                  tuple.set(i, Enum.valueOf(clazz, currentValue.trim()));
                  break;
                case STRING:
                  if (type == InputType.CSV) {
                    tuple.set(i, currentValue);
View Full Code Here
123
124
125
126
127
128
129
130
131
      }
    }
  }

  private static Schema getPangoolTweetSchema() {
    Field tweetIdField = Field.create("tweet_id", Schema.Field.Type.INT);
    Field tweetHashTags = Fields.createAvroField("tweet_hashtags", getAvroStringArraySchema(), false);
    return new Schema("tweet", Arrays.asList(tweetIdField, tweetHashTags));
  }
View Full Code Here
129
130
131
132
133
134
135
136
137
    Field tweetHashTags = Fields.createAvroField("tweet_hashtags", getAvroStringArraySchema(), false);
    return new Schema("tweet", Arrays.asList(tweetIdField, tweetHashTags));
  }

  private static Schema getPangoolRetweetSchema() {
    Field userId = Field.create("username", Schema.Field.Type.STRING);
    Field tweetId = Field.create("tweet_id", Schema.Field.Type.INT);
    return new Schema("retweet", Arrays.asList(userId, tweetId));
  }
View Full Code Here
124
125
126
127
128
129
130
131
  public AvroTopicalWordCount() {
    super("Usage: AvroTopicalWordCount [input_path] [output_path]");
  }

  static Schema getSchema() {
    Field avroField = Fields.createAvroField("my_avro",getAvroSchema(),false);
    return new Schema("schema",Arrays.asList(avroField));
  }
View Full Code Here
124
125
126
127
128
129
130
131
132
    super("Usage: AvroCustomSerializationJob [input_path] [output_path]");
  }

  static Schema getSchema() {
    org.apache.avro.Schema avroSchema = getAvroSchema();
    Field avroField = Field.createObject("my_avro",AvroFieldSerializer.class,AvroFieldDeserializer.class);
    avroField.addProp("avro.schema",avroSchema.toString());
    return new Schema("schema",Arrays.asList(avroField));
  }
View Full Code Here
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
            }
            String currentValue = "";
            try {
              currentValue = readLine.get(index);
              if(currentValue != null) {
                Field field = schema.getFields().get(i);
                switch(field.getType()) {
                case INT:
                case LONG:
                case FLOAT:
                case DOUBLE:
                  processNumber(field.getType(), tuple, i, currentValue);
                  break;
                case ENUM:
                  Class clazz = field.getObjectClass();
                  tuple.set(i, Enum.valueOf(clazz, currentValue.trim()));
                  break;
                case STRING:
                  if (type == InputType.CSV) {
                    tuple.set(i, currentValue);
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
121
122
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
    }

    // Field by field deseralization
    for(int index = 0; index < schema.getFields().size(); index++) {
      Deserializer customDeser = customDeserializers[index];
      Field field = schema.getField(index);

      // Nulls control
      if (field.isNullable() && nullsAbsolute.flags[index]) {
        // Null field. Nothing to deserialize.
        continue;
      }

      switch(field.getType()) {
      case INT:
        tuple.set(index, WritableUtils.readVInt(input));
        break;
      case LONG:
        tuple.set(index, WritableUtils.readVLong(input));
        break;
      case DOUBLE:
        tuple.set(index, input.readDouble());
        break;
      case FLOAT:
        tuple.set(index, input.readFloat());
        break;
      case STRING:
        readUtf8(input, tuple, index);
        break;
      case BOOLEAN:
        byte b = input.readByte();
        tuple.set(index, (b != 0));
        break;
      case ENUM:
        readEnum(input, tuple, field.getObjectClass(), index);
        break;
      case BYTES:
        readBytes(input, tuple, index);
        break;
      case OBJECT:
        readCustomObject(input, tuple, field.getObjectClass(), index, customDeser);
        break;
      default:
        throw new IOException("Not supported type:" + field.getType());
      }
    }
  }
View Full Code Here
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
      }
      nulls.ser(out);
    }

    for (int i = 0; i < destinationSchema.getFields().size(); i++) {
      Field field = destinationSchema.getField(i);
      Type fieldType = field.getType();
      Object element = valueAt(i, tuple, translationTable);
      if (element == null) {
        if (field.isNullable()) {
          // Nullable null fields don't need serialization.
          continue;
        } else {
          raiseUnexpectedNullException(field, element);
        }
View Full Code Here
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
  public static String toString(ITuple tuple) {
    Schema schema = tuple.getSchema();
    StringBuilder b = new StringBuilder();
    b.append("{");
    for (int i = 0; i < schema.getFields().size(); i++) {
      Field f = schema.getField(i);
      if (i != 0) {
        b.append(",");
      }
      b.append("\"").append(f.getName()).append("\"").append(":");
      if (tuple.get(i) == null) {
        b.append("null");
        continue;
      }
      switch (f.getType()) {
        case INT:
        case LONG:
        case FLOAT:
        case DOUBLE:
        case BOOLEAN:
          b.append(tuple.get(i));
          break;
        case STRING:
        case ENUM:

          b.append("\"").append(tuple.get(i)).append("\"");
          break;
        case OBJECT:
          b.append("{").append(tuple.get(i).toString()).append("}");
          break;
        case BYTES:
          Object o = tuple.get(i);
          b.append("{\"bytes\": \"");
          byte[] bytes;
          int offset, length;
          if (o instanceof ByteBuffer) {
            ByteBuffer byteBuffer = (ByteBuffer) o;
            bytes = byteBuffer.array();
            offset = byteBuffer.arrayOffset() + byteBuffer.position();
            length = byteBuffer.limit() - byteBuffer.position();
          } else {
            //byte[]
            bytes = (byte[]) o;
            offset = 0;
            length = bytes.length;
          }
          for (int p = offset; p < offset + length; p++) {
            b.append((char) bytes[p]);
          }
          b.append("\"}");
          break;
        default:
          throw new PangoolRuntimeException("Not stringifiable type :" + f.getType());
      }
    }
    b.append("}");
    return b.toString();
  }
View Full Code Here
TOP

Related Classes of com.datasalt.pangool.io.Schema.Field

  • com.datasalt.pangool.BaseTest
  • com.datasalt.pangool.examples.avro.AvroCustomSerializationJob
  • com.datasalt.pangool.examples.avro.AvroTopicalWordCount
  • com.datasalt.pangool.examples.avro.AvroTweetsJoin
  • com.datasalt.pangool.io.Fields
  • com.datasalt.pangool.io.Tuple
  • com.datasalt.pangool.tuplemr.mapred.lib.input.TupleTextInputFormat$TupleTextInputReader
  • com.datasalt.pangool.tuplemr.mapred.SortComparator
  • com.datasalt.pangool.tuplemr.mapred.TestComparators
  • com.datasalt.pangool.tuplemr.serialization.SimpleTupleDeserializer
    • 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.