Package org.apache.pig.newplan.logical.relational

Examples of org.apache.pig.newplan.logical.relational.LOJoin

99
100
101
102
103
104
105
106
107
108
109
110
            "B = LOAD 'data2' using "+ DummyIndexableLoader.class.getName() +"() as (id, name, grade);" +
            "C = join A by id left, B by id using 'merge';" +
            "store C into 'out';";
            LogicalPlan lp = Util.buildLp(pigServer, query);
            LOStore store = (LOStore)lp.getSinks().get(0);
            LOJoin join = (LOJoin)lp.getPredecessors(store).get(0);
            assertEquals(LOJoin.JOINTYPE.MERGE, join.getJoinType());

            PigContext pc = new PigContext(ExecType.MAPREDUCE,cluster.getProperties());
            pc.connect();
            PhysicalPlan phyP = Util.buildPp(pigServer, query);
            PhysicalOperator phyOp = phyP.getLeaves().get(0);
View Full Code Here
141
142
143
144
145
146
147
148
149
150
151
152
        "B = LOAD 'data2' using "+ DummyIndexableLoader.class.getName() +"() as (id, name, grade);" +
        "C = join E by A.id, B by id using 'merge';" +
        "store C into 'output';";
        LogicalPlan lp = Util.buildLp(pigServer, query);
        Operator op = lp.getSinks().get(0);
        LOJoin join = (LOJoin)lp.getPredecessors(op).get(0);
        assertEquals(LOJoin.JOINTYPE.MERGE, join.getJoinType());

        PigContext pc = new PigContext(ExecType.MAPREDUCE,cluster.getProperties());
        pc.connect();
        boolean exceptionCaught = false;
        try{
View Full Code Here
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
        cschema.getField(0).uid = 1;
        cschema.getField(1).uid = 2;

        MultiMap<Integer, LogicalExpressionPlan> mm =
            new MultiMap<Integer, LogicalExpressionPlan>();
        LOJoin C = new LOJoin(lp, mm, JOINTYPE.HASH, new boolean[] {true, true});
       
        LogicalExpressionPlan aprojplan = new LogicalExpressionPlan();
        ProjectExpression x = new ProjectExpression(aprojplan, 0, 0, C);
        x.neverUseForRealSetFieldSchema(new LogicalFieldSchema(null, null,
                DataType.INTEGER, 1));
        LogicalExpressionPlan bprojplan = new LogicalExpressionPlan();
        ProjectExpression y = new ProjectExpression(bprojplan, 1, 0, C);
        y.neverUseForRealSetFieldSchema(new LogicalFieldSchema(null, null,
                DataType.INTEGER, 2));
        mm.put(0, aprojplan);
        mm.put(1, bprojplan);
       
        C.neverUseForRealSetSchema(cschema);
        // Don't add it to the plan quite yet
       
        // D = filter
        LogicalExpressionPlan filterPlan = new LogicalExpressionPlan();
        LOFilter D = new LOFilter(lp, filterPlan);
View Full Code Here
303
304
305
306
307
308
309
        }
        return alias;
    }
   
    LOJoin createJoinOp() {
        return new LOJoin( plan );
    }
View Full Code Here
109
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
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
          cschema.addField(new LogicalSchema.LogicalFieldSchema(
              "b", null, DataType.BYTEARRAY));
          MultiMap<Integer, LogicalExpressionPlan> mm =
                new MultiMap<Integer, LogicalExpressionPlan>();
          LogicalExpressionPlan aprojplan = new LogicalExpressionPlan();
          LOJoin C = new LOJoin(lp, mm, JOINTYPE.HASH, new boolean[] {true, true});
          new ProjectExpression(aprojplan, 0, 0, C);
          LogicalExpressionPlan bprojplan = new LogicalExpressionPlan();
          new ProjectExpression(bprojplan, 1, 0, C);
          mm.put(0, aprojplan);
          mm.put(1, bprojplan);
         
          C.setAlias("C");
          lp.add(C);
          lp.connect(A, C);
          lp.connect(B, C);
       
          // D = filter
          LogicalExpressionPlan filterPlan = new LogicalExpressionPlan();
          LOFilter D = new LOFilter(lp, filterPlan);
          ProjectExpression fx = new ProjectExpression(filterPlan, 0, 0, D);
          ConstantExpression fc0 = new ConstantExpression(filterPlan, new Integer(0));
          EqualExpression eq1 = new EqualExpression(filterPlan, fx, fc0);
          ProjectExpression fanotherx = new ProjectExpression(filterPlan, 0, 0, D);
          ProjectExpression fa = new ProjectExpression(filterPlan, 0, 2, D);
          EqualExpression eq2 = new EqualExpression(filterPlan, fanotherx, fa);
          AndExpression and1 = new AndExpression(filterPlan, eq1, eq2);
          ProjectExpression fb = new ProjectExpression(filterPlan, 0, 3, D);
          ConstantExpression fc1 = new ConstantExpression(filterPlan, new Integer(1));
          EqualExpression eq3 = new EqualExpression(filterPlan, fb, fc1);
          AndExpression and2 = new AndExpression(filterPlan, and1, eq3);
          ProjectExpression fanotherb = new ProjectExpression(filterPlan, 0, 3, D);
          ProjectExpression fy = new ProjectExpression(filterPlan, 0, 1, D);
          EqualExpression eq4 = new EqualExpression(filterPlan, fy, fanotherb);
          new AndExpression(filterPlan, and2, eq4);
         
          D.setAlias("D");
          // Connect D to B, since the transform has happened.
          lp.add(D);
          lp.connect(C, D);
        }
       
        System.out.println(lp);
        LogicalPlanOptimizer optimizer = new LogicalPlanOptimizer(lp, 500, null);
        optimizer.optimize();
       
        LogicalPlan expected = new LogicalPlan();
        {
            // A = load
          LogicalSchema aschema = new LogicalSchema();
          aschema.addField(new LogicalSchema.LogicalFieldSchema(
              "x", null, DataType.BYTEARRAY));
          aschema.addField(new LogicalSchema.LogicalFieldSchema(
              "y", null, DataType.BYTEARRAY));
          LOLoad A = new LOLoad(new FileSpec("bla", new FuncSpec("PigStorage", "\t")), aschema, expected, conf);
          expected.add(A);
         
          // DA = filter
          LogicalExpressionPlan DAfilterPlan = new LogicalExpressionPlan();
          LOFilter DA = new LOFilter(expected, DAfilterPlan);
          ProjectExpression fx = new ProjectExpression(DAfilterPlan, 0, 0, DA);
          fx.neverUseForRealSetFieldSchema(new LogicalFieldSchema(null, null, DataType.BYTEARRAY));
          ConstantExpression fc0 = new ConstantExpression(DAfilterPlan, new Integer(0));
          new EqualExpression(DAfilterPlan, fx, fc0);
         
          DA.neverUseForRealSetSchema(aschema);
          expected.add(DA);
          expected.connect(A, DA);
         
          // A = foreach
            LOForEach foreachA = org.apache.pig.newplan.logical.Util.addForEachAfter(expected, DA, 0, new HashSet<Integer>());
            foreachA.setAlias("A");
            foreachA.neverUseForRealSetSchema(aschema);
         
          // B = load
          LogicalSchema bschema = new LogicalSchema();
          bschema.addField(new LogicalSchema.LogicalFieldSchema(
              "a", null, DataType.BYTEARRAY));
          bschema.addField(new LogicalSchema.LogicalFieldSchema(
              "b", null, DataType.BYTEARRAY));
          LOLoad B = new LOLoad(new FileSpec("morebla", new FuncSpec("PigStorage", "\t")), bschema, expected, conf);
          expected.add(B);
         
          // DB = filter
          LogicalExpressionPlan DBfilterPlan = new LogicalExpressionPlan();
            LOFilter DB = new LOFilter(expected, DBfilterPlan);
          ProjectExpression fb = new ProjectExpression(DBfilterPlan, 0, 1, DB);
          fb.neverUseForRealSetFieldSchema(new LogicalFieldSchema(null, null, DataType.BYTEARRAY));
          ConstantExpression fc1 = new ConstantExpression(DBfilterPlan, new Integer(1));
          new EqualExpression(DBfilterPlan, fb, fc1);

          DB.neverUseForRealSetSchema(bschema);
          expected.add(DB);
          expected.connect(B, DB);
         
            // B = foreach
            LOForEach foreachB = org.apache.pig.newplan.logical.Util.addForEachAfter(expected, DB, 0, new HashSet<Integer>());
            foreachB.setAlias("B");
            foreachB.neverUseForRealSetSchema(bschema);
           
          // C = join
          LogicalSchema cschema = new LogicalSchema();
          cschema.addField(new LogicalSchema.LogicalFieldSchema(
              "D::x", null, DataType.BYTEARRAY));
          cschema.addField(new LogicalSchema.LogicalFieldSchema(
              "D::y", null, DataType.BYTEARRAY));
          cschema.addField(new LogicalSchema.LogicalFieldSchema(
              "null::a", null, DataType.BYTEARRAY));
          cschema.addField(new LogicalSchema.LogicalFieldSchema(
              "null::b", null, DataType.BYTEARRAY));
          cschema.getField(0).uid = 1;
          cschema.getField(1).uid = 2;
          cschema.getField(2).uid = 3;
          cschema.getField(3).uid = 4;
          LogicalExpressionPlan aprojplan = new LogicalExpressionPlan();
            MultiMap<Integer, LogicalExpressionPlan> mm =
                new MultiMap<Integer, LogicalExpressionPlan>();
            LOJoin C = new LOJoin(expected, mm, JOINTYPE.HASH, new boolean[] {true, true});

          ProjectExpression x = new ProjectExpression(aprojplan, 0, 0, C);
          x.neverUseForRealSetFieldSchema(new LogicalFieldSchema(null, null, DataType.BYTEARRAY));
          LogicalExpressionPlan bprojplan = new LogicalExpressionPlan();
          ProjectExpression y = new ProjectExpression(bprojplan, 1, 0, C);
          y.neverUseForRealSetFieldSchema(new LogicalFieldSchema(null, null, DataType.BYTEARRAY));
          mm.put(0, aprojplan);
          mm.put(1, bprojplan);
          C.neverUseForRealSetSchema(cschema);
          expected.add(C);
          expected.connect(foreachA, C);
          expected.connect(foreachB, C);
         
          // D = filter
View Full Code Here
44
45
46
47
48
49
50
51
52
53
54
55
56
            pigServer.setBatchOn();
            pigServer.registerQuery("a = load 'tmp1' as (foo, bar);");
            pigServer.registerQuery("b = load 'tmp1' as (foo, bar);");
            pigServer.registerQuery("c = join a by foo, b by foo using '"+joinTypes[i]+"';");
            pigServer.registerQuery("store c into 'tmp2';");
            LOJoin op = (LOJoin)TestPigStats.getLogicalPlan(pigServer).findByAlias("c");
            assertTrue("did "+joinTypes[i]+" join get pinned? ",
                    op.isPinnedOption(LOJoin.OPTION_JOIN));
            assertEquals("did the right join type get set? ",
                    ((LOJoin) op).getJoinType().toString(), expectedJoinTypes[i]);
            pigServer.discardBatch();
        }
    }
View Full Code Here
60
61
62
63
64
65
66
67
68
69
70
        pigServer.setBatchOn();
        pigServer.registerQuery("a = load 'tmp1' as (foo, bar);");
        pigServer.registerQuery("b = load 'tmp1' as (foo, bar);");
        pigServer.registerQuery("c = join a by foo, b by foo;");
        pigServer.registerQuery("store c into 'tmp2';");
        LOJoin op = (LOJoin)TestPigStats.getLogicalPlan(pigServer).findByAlias("c");
        assertEquals("default join should be hash",
                ((LOJoin) op).getJoinType().toString(), "HASH");
        assertFalse(op.isPinnedOption(LOJoin.OPTION_JOIN));
        pigServer.discardBatch();
    }
View Full Code Here
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
            new ProjectExpression(bprojplan, 1, 0, null);
            MultiMap<Integer, LogicalExpressionPlan> mm =
                new MultiMap<Integer, LogicalExpressionPlan>();
            mm.put(0, aprojplan);
            mm.put(1, bprojplan);
            LOJoin C = new LOJoin(lp, mm, JOINTYPE.HASH, new boolean[] {true, true});
            C.neverUseForRealSetSchema(cschema);
            lp.add(C);
            lp.connect(A, C);
            lp.connect(B, C);
           
            // D = filter
            LogicalExpressionPlan filterPlan = new LogicalExpressionPlan();
            ProjectExpression fy = new ProjectExpression(filterPlan, 0, 1, null);
            ConstantExpression fc = new ConstantExpression(filterPlan, new Integer(0));
            new EqualExpression(filterPlan, fy, fc);
            LOFilter D = new LOFilter(lp, filterPlan);
            D.neverUseForRealSetSchema(cschema);
            lp.add(D);
            lp.connect(C, D);
        }
       
        // Build a second similar plan to test equality
        // A = load
        LogicalPlan lp1 = new LogicalPlan();
        {
            LogicalSchema aschema = new LogicalSchema();
            aschema.addField(new LogicalSchema.LogicalFieldSchema(
                "x", null, DataType.INTEGER));
            LOLoad A = new LOLoad(new FileSpec("/abc",
                new FuncSpec(DummyLoad.class.getName(), new String[] {"x", "y"})), aschema, lp1, conf);
            lp1.add(A);
           
            // B = load
            LogicalSchema bschema = new LogicalSchema();
            bschema.addField(new LogicalSchema.LogicalFieldSchema(
                "y", null, DataType.INTEGER));
            LOLoad B = new LOLoad(new FileSpec("/def",
                new FuncSpec("PigStorage", "\t")), bschema, lp1, conf);
            lp1.add(B);
           
            // C = join
            LogicalSchema cschema = new LogicalSchema();
            cschema.addField(new LogicalSchema.LogicalFieldSchema(
                "x", null, DataType.INTEGER));
            cschema.addField(new LogicalSchema.LogicalFieldSchema(
                "y", null, DataType.INTEGER));
            LogicalExpressionPlan aprojplan = new LogicalExpressionPlan();
            new ProjectExpression(aprojplan, 0, 0, null);
            LogicalExpressionPlan bprojplan = new LogicalExpressionPlan();
            new ProjectExpression(bprojplan, 1, 0, null);
            MultiMap<Integer, LogicalExpressionPlan> mm =
                new MultiMap<Integer, LogicalExpressionPlan>();
            mm.put(0, aprojplan);
            mm.put(1, bprojplan);
            LOJoin C = new LOJoin(lp1, mm, JOINTYPE.HASH, new boolean[] {true, true});
            C.neverUseForRealSetSchema(cschema);
            lp1.add(C);
            lp1.connect(A, C);
            lp1.connect(B, C);
               
           
View Full Code Here
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
        new ProjectExpression(bprojplan1, 1, 0, null);
        MultiMap<Integer, LogicalExpressionPlan> mm1 =
            new MultiMap<Integer, LogicalExpressionPlan>();
        mm1.put(0, aprojplan1);
        mm1.put(1, bprojplan1);
        LOJoin C1 = new LOJoin(lp, mm1, JOINTYPE.HASH, new boolean[] {true, true});
        C1.neverUseForRealSetSchema(jcschema1);
        lp.add(C1);
        lp.connect(A1, C1);
        lp.connect(B1, C1);
       
        // A = load
        LogicalSchema jaschema2 = new LogicalSchema();
        jaschema2.addField(new LogicalSchema.LogicalFieldSchema(
           "x", null, DataType.INTEGER));
        LOLoad A2 = new LOLoad(new FileSpec("/abc",
           new FuncSpec("/fooload", new String[] {"x", "y"})), jaschema2, lp, conf);
        lp.add(A2);
       
        // B = load
        LogicalSchema jbschema2 = new LogicalSchema();
        jbschema2.addField(new LogicalSchema.LogicalFieldSchema(
            "y", null, DataType.INTEGER));
        LOLoad B2 = new LOLoad(new FileSpec("/def",
            new FuncSpec("PigStorage", "\t")), jbschema2, lp, conf);
        lp.add(B2);
       
        // C = join
        LogicalSchema jcschema2 = new LogicalSchema();
        jcschema2.addField(new LogicalSchema.LogicalFieldSchema(
            "x", null, DataType.INTEGER));
        jcschema2.addField(new LogicalSchema.LogicalFieldSchema(
            "y", null, DataType.INTEGER));
        LogicalExpressionPlan aprojplan2 = new LogicalExpressionPlan();
        new ProjectExpression(aprojplan2, 0, 0, null);
        LogicalExpressionPlan bprojplan2 = new LogicalExpressionPlan();
        new ProjectExpression(bprojplan2, 1, 0, null);
        MultiMap<Integer, LogicalExpressionPlan> mm2 =
            new MultiMap<Integer, LogicalExpressionPlan>();
        mm2.put(0, aprojplan2);
        mm2.put(1, bprojplan2);
        LOJoin C2 = new LOJoin(lp, mm2, JOINTYPE.SKEWED, new boolean[] {true, true});
        C2.neverUseForRealSetSchema(jcschema2);
        lp.add(C2);
        lp.connect(A2, C2);
        lp.connect(B2, C2);
       
        assertFalse(C1.isEqual(C2));
View Full Code Here
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
        new ProjectExpression(bprojplan1, 1, 0, null);
        MultiMap<Integer, LogicalExpressionPlan> mm1 =
            new MultiMap<Integer, LogicalExpressionPlan>();
        mm1.put(0, aprojplan1);
        mm1.put(1, bprojplan1);
        LOJoin C1 = new LOJoin(lp, mm1, JOINTYPE.HASH, new boolean[] {true, true});
        C1.neverUseForRealSetSchema(jcschema1);
        lp.add(C1);
        lp.connect(A1, C1);
        lp.connect(B1, C1);
       
        // Test different inner status
        // A = load
        LogicalSchema jaschema3 = new LogicalSchema();
        jaschema3.addField(new LogicalSchema.LogicalFieldSchema(
           "x", null, DataType.INTEGER));
        LOLoad A3 = new LOLoad(new FileSpec("/abc",
           new FuncSpec("/fooload", new String[] {"x", "y"})), jaschema3, lp, conf);
        lp.add(A3);
       
        // B = load
        LogicalSchema jbschema3 = new LogicalSchema();
        jbschema3.addField(new LogicalSchema.LogicalFieldSchema(
            "y", null, DataType.INTEGER));
        LOLoad B3 = new LOLoad(new FileSpec("/def",
            new FuncSpec("PigStorage", "\t")), jbschema3, lp, conf);
        lp.add(B3);
       
        // C = join
        LogicalSchema jcschema3 = new LogicalSchema();
        jcschema3.addField(new LogicalSchema.LogicalFieldSchema(
            "x", null, DataType.INTEGER));
        jcschema3.addField(new LogicalSchema.LogicalFieldSchema(
            "y", null, DataType.INTEGER));
        LogicalExpressionPlan aprojplan3 = new LogicalExpressionPlan();
        new ProjectExpression(aprojplan3, 0, 0, null);
        LogicalExpressionPlan bprojplan3 = new LogicalExpressionPlan();
        new ProjectExpression(bprojplan3, 1, 0, null);
        MultiMap<Integer, LogicalExpressionPlan> mm3 =
            new MultiMap<Integer, LogicalExpressionPlan>();
        mm3.put(0, aprojplan3);
        mm3.put(1, bprojplan3);
        LOJoin C3 = new LOJoin(lp, mm3, JOINTYPE.HASH, new boolean[] {true, false});
        C3.neverUseForRealSetSchema(jcschema3);
        lp.add(C3);
        lp.connect(A3, C3);
        lp.connect(B3, C3);
       
       
View Full Code Here
TOP

Related Classes of org.apache.pig.newplan.logical.relational.LOJoin

  • org.apache.pig.newplan.logical.rules.PushDownForEachFlatten$PushDownForEachFlattenTransformer
  • org.apache.pig.parser.LogicalPlanBuilder
  • org.apache.pig.test.TestJoin
  • org.apache.pig.test.TestMergeForEachOptimization
  • org.apache.pig.test.TestMergeJoinOuter
  • org.apache.pig.test.TestNewPlanFilterRule
  • org.apache.pig.test.TestNewPlanListener
  • org.apache.pig.test.TestNewPlanLogicalOptimizer
  • org.apache.pig.test.TestNewPlanOperatorPlan
  • org.apache.pig.test.TestPinOptions
    • 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.