Search in sources :

Example 16 with UnionObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector in project hive by apache.

the class SerDeUtils method buildJSONString.

static void buildJSONString(StringBuilder sb, Object o, ObjectInspector oi, String nullStr) {
    switch(oi.getCategory()) {
        case PRIMITIVE:
            {
                PrimitiveObjectInspector poi = (PrimitiveObjectInspector) oi;
                if (o == null) {
                    sb.append(nullStr);
                } else {
                    switch(poi.getPrimitiveCategory()) {
                        case BOOLEAN:
                            {
                                boolean b = ((BooleanObjectInspector) poi).get(o);
                                sb.append(b ? "true" : "false");
                                break;
                            }
                        case BYTE:
                            {
                                sb.append(((ByteObjectInspector) poi).get(o));
                                break;
                            }
                        case SHORT:
                            {
                                sb.append(((ShortObjectInspector) poi).get(o));
                                break;
                            }
                        case INT:
                            {
                                sb.append(((IntObjectInspector) poi).get(o));
                                break;
                            }
                        case LONG:
                            {
                                sb.append(((LongObjectInspector) poi).get(o));
                                break;
                            }
                        case FLOAT:
                            {
                                sb.append(((FloatObjectInspector) poi).get(o));
                                break;
                            }
                        case DOUBLE:
                            {
                                sb.append(((DoubleObjectInspector) poi).get(o));
                                break;
                            }
                        case STRING:
                            {
                                sb.append('"');
                                sb.append(escapeString(((StringObjectInspector) poi).getPrimitiveJavaObject(o)));
                                sb.append('"');
                                break;
                            }
                        case CHAR:
                            {
                                sb.append('"');
                                sb.append(escapeString(((HiveCharObjectInspector) poi).getPrimitiveJavaObject(o).toString()));
                                sb.append('"');
                                break;
                            }
                        case VARCHAR:
                            {
                                sb.append('"');
                                sb.append(escapeString(((HiveVarcharObjectInspector) poi).getPrimitiveJavaObject(o).toString()));
                                sb.append('"');
                                break;
                            }
                        case DATE:
                            {
                                sb.append('"');
                                sb.append(((DateObjectInspector) poi).getPrimitiveWritableObject(o));
                                sb.append('"');
                                break;
                            }
                        case TIMESTAMP:
                            {
                                sb.append('"');
                                sb.append(((TimestampObjectInspector) poi).getPrimitiveWritableObject(o));
                                sb.append('"');
                                break;
                            }
                        case TIMESTAMPLOCALTZ:
                            {
                                sb.append('"');
                                sb.append(((TimestampLocalTZObjectInspector) poi).getPrimitiveWritableObject(o));
                                sb.append('"');
                                break;
                            }
                        case BINARY:
                            {
                                BytesWritable bw = ((BinaryObjectInspector) oi).getPrimitiveWritableObject(o);
                                Text txt = new Text();
                                txt.set(bw.getBytes(), 0, bw.getLength());
                                sb.append(txt.toString());
                                break;
                            }
                        case DECIMAL:
                            {
                                sb.append(((HiveDecimalObjectInspector) oi).getPrimitiveJavaObject(o));
                                break;
                            }
                        case INTERVAL_YEAR_MONTH:
                            {
                                sb.append(((HiveIntervalYearMonthObjectInspector) oi).getPrimitiveJavaObject(o));
                                break;
                            }
                        case INTERVAL_DAY_TIME:
                            {
                                sb.append(((HiveIntervalDayTimeObjectInspector) oi).getPrimitiveJavaObject(o));
                                break;
                            }
                        default:
                            throw new RuntimeException("Unknown primitive type: " + poi.getPrimitiveCategory());
                    }
                }
                break;
            }
        case LIST:
            {
                ListObjectInspector loi = (ListObjectInspector) oi;
                ObjectInspector listElementObjectInspector = loi.getListElementObjectInspector();
                List<?> olist = loi.getList(o);
                if (olist == null) {
                    sb.append(nullStr);
                } else {
                    sb.append(LBRACKET);
                    for (int i = 0; i < olist.size(); i++) {
                        if (i > 0) {
                            sb.append(COMMA);
                        }
                        buildJSONString(sb, olist.get(i), listElementObjectInspector, JSON_NULL);
                    }
                    sb.append(RBRACKET);
                }
                break;
            }
        case MAP:
            {
                MapObjectInspector moi = (MapObjectInspector) oi;
                ObjectInspector mapKeyObjectInspector = moi.getMapKeyObjectInspector();
                ObjectInspector mapValueObjectInspector = moi.getMapValueObjectInspector();
                Map<?, ?> omap = moi.getMap(o);
                if (omap == null) {
                    sb.append(nullStr);
                } else {
                    sb.append(LBRACE);
                    boolean first = true;
                    for (Object entry : omap.entrySet()) {
                        if (first) {
                            first = false;
                        } else {
                            sb.append(COMMA);
                        }
                        Map.Entry<?, ?> e = (Map.Entry<?, ?>) entry;
                        buildJSONString(sb, e.getKey(), mapKeyObjectInspector, JSON_NULL);
                        sb.append(COLON);
                        buildJSONString(sb, e.getValue(), mapValueObjectInspector, JSON_NULL);
                    }
                    sb.append(RBRACE);
                }
                break;
            }
        case STRUCT:
            {
                StructObjectInspector soi = (StructObjectInspector) oi;
                List<? extends StructField> structFields = soi.getAllStructFieldRefs();
                if (o == null) {
                    sb.append(nullStr);
                } else {
                    sb.append(LBRACE);
                    for (int i = 0; i < structFields.size(); i++) {
                        if (i > 0) {
                            sb.append(COMMA);
                        }
                        sb.append(QUOTE);
                        sb.append(structFields.get(i).getFieldName());
                        sb.append(QUOTE);
                        sb.append(COLON);
                        buildJSONString(sb, soi.getStructFieldData(o, structFields.get(i)), structFields.get(i).getFieldObjectInspector(), JSON_NULL);
                    }
                    sb.append(RBRACE);
                }
                break;
            }
        case UNION:
            {
                UnionObjectInspector uoi = (UnionObjectInspector) oi;
                if (o == null) {
                    sb.append(nullStr);
                } else {
                    sb.append(LBRACE);
                    sb.append(uoi.getTag(o));
                    sb.append(COLON);
                    buildJSONString(sb, uoi.getField(o), uoi.getObjectInspectors().get(uoi.getTag(o)), JSON_NULL);
                    sb.append(RBRACE);
                }
                break;
            }
        default:
            throw new RuntimeException("Unknown type in ObjectInspector!");
    }
}
Also used : ListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector) TimestampLocalTZObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.TimestampLocalTZObjectInspector) HiveVarcharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveVarcharObjectInspector) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) HiveCharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveCharObjectInspector) IntObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.IntObjectInspector) HiveDecimalObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveDecimalObjectInspector) BooleanObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BooleanObjectInspector) PrimitiveObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector) ShortObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.ShortObjectInspector) HiveIntervalYearMonthObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveIntervalYearMonthObjectInspector) ObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector) LongObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.LongObjectInspector) BinaryObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BinaryObjectInspector) MapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector) FloatObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.FloatObjectInspector) StringObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.StringObjectInspector) HiveIntervalDayTimeObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveIntervalDayTimeObjectInspector) ByteObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.ByteObjectInspector) DoubleObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.DoubleObjectInspector) TimestampObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.TimestampObjectInspector) DateObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.DateObjectInspector) BinaryObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BinaryObjectInspector) BytesWritable(org.apache.hadoop.io.BytesWritable) Text(org.apache.hadoop.io.Text) StringObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.StringObjectInspector) StructField(org.apache.hadoop.hive.serde2.objectinspector.StructField) MapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector) ListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector) PrimitiveObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector) List(java.util.List) BooleanObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BooleanObjectInspector) Map(java.util.Map) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector)

Example 17 with UnionObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector in project hive by apache.

the class JsonSerDe method buildJSONString.

// TODO : code section copied over from SerDeUtils because of non-standard json production there
// should use quotes for all field names. We should fix this there, and then remove this copy.
// See http://jackson.codehaus.org/1.7.3/javadoc/org/codehaus/jackson/JsonParser.Feature.html#ALLOW_UNQUOTED_FIELD_NAMES
// for details - trying to enable Jackson to ignore that doesn't seem to work(compilation failure
// when attempting to use that feature, so having to change the production itself.
// Also, throws IOException when Binary is detected.
private static void buildJSONString(StringBuilder sb, Object o, ObjectInspector oi) throws IOException {
    switch(oi.getCategory()) {
        case PRIMITIVE:
            {
                PrimitiveObjectInspector poi = (PrimitiveObjectInspector) oi;
                if (o == null) {
                    sb.append("null");
                } else {
                    switch(poi.getPrimitiveCategory()) {
                        case BOOLEAN:
                            {
                                boolean b = ((BooleanObjectInspector) poi).get(o);
                                sb.append(b ? "true" : "false");
                                break;
                            }
                        case BYTE:
                            {
                                sb.append(((ByteObjectInspector) poi).get(o));
                                break;
                            }
                        case SHORT:
                            {
                                sb.append(((ShortObjectInspector) poi).get(o));
                                break;
                            }
                        case INT:
                            {
                                sb.append(((IntObjectInspector) poi).get(o));
                                break;
                            }
                        case LONG:
                            {
                                sb.append(((LongObjectInspector) poi).get(o));
                                break;
                            }
                        case FLOAT:
                            {
                                sb.append(((FloatObjectInspector) poi).get(o));
                                break;
                            }
                        case DOUBLE:
                            {
                                sb.append(((DoubleObjectInspector) poi).get(o));
                                break;
                            }
                        case STRING:
                            {
                                String s = SerDeUtils.escapeString(((StringObjectInspector) poi).getPrimitiveJavaObject(o));
                                appendWithQuotes(sb, s);
                                break;
                            }
                        case BINARY:
                            byte[] b = ((BinaryObjectInspector) oi).getPrimitiveJavaObject(o);
                            Text txt = new Text();
                            txt.set(b, 0, b.length);
                            appendWithQuotes(sb, SerDeUtils.escapeString(txt.toString()));
                            break;
                        case DATE:
                            Date d = ((DateObjectInspector) poi).getPrimitiveJavaObject(o);
                            appendWithQuotes(sb, d.toString());
                            break;
                        case TIMESTAMP:
                            {
                                Timestamp t = ((TimestampObjectInspector) poi).getPrimitiveJavaObject(o);
                                appendWithQuotes(sb, t.toString());
                                break;
                            }
                        case DECIMAL:
                            sb.append(((HiveDecimalObjectInspector) poi).getPrimitiveJavaObject(o));
                            break;
                        case VARCHAR:
                            {
                                String s = SerDeUtils.escapeString(((HiveVarcharObjectInspector) poi).getPrimitiveJavaObject(o).toString());
                                appendWithQuotes(sb, s);
                                break;
                            }
                        case CHAR:
                            {
                                // this should use HiveChar.getPaddedValue() but it's protected; currently (v0.13)
                                // HiveChar.toString() returns getPaddedValue()
                                String s = SerDeUtils.escapeString(((HiveCharObjectInspector) poi).getPrimitiveJavaObject(o).toString());
                                appendWithQuotes(sb, s);
                                break;
                            }
                        default:
                            throw new RuntimeException("Unknown primitive type: " + poi.getPrimitiveCategory());
                    }
                }
                break;
            }
        case LIST:
            {
                ListObjectInspector loi = (ListObjectInspector) oi;
                ObjectInspector listElementObjectInspector = loi.getListElementObjectInspector();
                List<?> olist = loi.getList(o);
                if (olist == null) {
                    sb.append("null");
                } else {
                    sb.append(SerDeUtils.LBRACKET);
                    for (int i = 0; i < olist.size(); i++) {
                        if (i > 0) {
                            sb.append(SerDeUtils.COMMA);
                        }
                        buildJSONString(sb, olist.get(i), listElementObjectInspector);
                    }
                    sb.append(SerDeUtils.RBRACKET);
                }
                break;
            }
        case MAP:
            {
                MapObjectInspector moi = (MapObjectInspector) oi;
                ObjectInspector mapKeyObjectInspector = moi.getMapKeyObjectInspector();
                ObjectInspector mapValueObjectInspector = moi.getMapValueObjectInspector();
                Map<?, ?> omap = moi.getMap(o);
                if (omap == null) {
                    sb.append("null");
                } else {
                    sb.append(SerDeUtils.LBRACE);
                    boolean first = true;
                    for (Object entry : omap.entrySet()) {
                        if (first) {
                            first = false;
                        } else {
                            sb.append(SerDeUtils.COMMA);
                        }
                        Map.Entry<?, ?> e = (Map.Entry<?, ?>) entry;
                        StringBuilder keyBuilder = new StringBuilder();
                        buildJSONString(keyBuilder, e.getKey(), mapKeyObjectInspector);
                        String keyString = keyBuilder.toString().trim();
                        if ((!keyString.isEmpty()) && (keyString.charAt(0) != SerDeUtils.QUOTE)) {
                            appendWithQuotes(sb, keyString);
                        } else {
                            sb.append(keyString);
                        }
                        sb.append(SerDeUtils.COLON);
                        buildJSONString(sb, e.getValue(), mapValueObjectInspector);
                    }
                    sb.append(SerDeUtils.RBRACE);
                }
                break;
            }
        case STRUCT:
            {
                StructObjectInspector soi = (StructObjectInspector) oi;
                List<? extends StructField> structFields = soi.getAllStructFieldRefs();
                if (o == null) {
                    sb.append("null");
                } else {
                    sb.append(SerDeUtils.LBRACE);
                    for (int i = 0; i < structFields.size(); i++) {
                        if (i > 0) {
                            sb.append(SerDeUtils.COMMA);
                        }
                        appendWithQuotes(sb, structFields.get(i).getFieldName());
                        sb.append(SerDeUtils.COLON);
                        buildJSONString(sb, soi.getStructFieldData(o, structFields.get(i)), structFields.get(i).getFieldObjectInspector());
                    }
                    sb.append(SerDeUtils.RBRACE);
                }
                break;
            }
        case UNION:
            {
                UnionObjectInspector uoi = (UnionObjectInspector) oi;
                if (o == null) {
                    sb.append("null");
                } else {
                    sb.append(SerDeUtils.LBRACE);
                    sb.append(uoi.getTag(o));
                    sb.append(SerDeUtils.COLON);
                    buildJSONString(sb, uoi.getField(o), uoi.getObjectInspectors().get(uoi.getTag(o)));
                    sb.append(SerDeUtils.RBRACE);
                }
                break;
            }
        default:
            throw new RuntimeException("Unknown type in ObjectInspector!");
    }
}
Also used : UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) HiveDecimalObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveDecimalObjectInspector) BooleanObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BooleanObjectInspector) ShortObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.ShortObjectInspector) ObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector) MapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector) FloatObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.FloatObjectInspector) StringObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.StringObjectInspector) DateObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.DateObjectInspector) ListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector) HiveVarcharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveVarcharObjectInspector) HiveCharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveCharObjectInspector) IntObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.IntObjectInspector) PrimitiveObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector) LongObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.LongObjectInspector) BinaryObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BinaryObjectInspector) ByteObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.ByteObjectInspector) DoubleObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.DoubleObjectInspector) TimestampObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.TimestampObjectInspector) Text(org.apache.hadoop.io.Text) Timestamp(java.sql.Timestamp) Date(java.sql.Date) TimestampObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.TimestampObjectInspector) StructField(org.apache.hadoop.hive.serde2.objectinspector.StructField) HiveVarcharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveVarcharObjectInspector) MapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector) ListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector) PrimitiveObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector) HiveDecimalObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveDecimalObjectInspector) List(java.util.List) ArrayList(java.util.ArrayList) BooleanObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BooleanObjectInspector) HiveCharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveCharObjectInspector) Map(java.util.Map) LinkedHashMap(java.util.LinkedHashMap) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector)

Example 18 with UnionObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector in project hive by apache.

the class VectorRandomRowSource method getObjectInspector.

private ObjectInspector getObjectInspector(TypeInfo typeInfo) {
    final ObjectInspector objectInspector;
    switch(typeInfo.getCategory()) {
        case PRIMITIVE:
            {
                final PrimitiveTypeInfo primitiveType = (PrimitiveTypeInfo) typeInfo;
                objectInspector = PrimitiveObjectInspectorFactory.getPrimitiveWritableObjectInspector(primitiveType);
            }
            break;
        case MAP:
            {
                final MapTypeInfo mapType = (MapTypeInfo) typeInfo;
                final MapObjectInspector mapInspector = ObjectInspectorFactory.getStandardMapObjectInspector(getObjectInspector(mapType.getMapKeyTypeInfo()), getObjectInspector(mapType.getMapValueTypeInfo()));
                objectInspector = mapInspector;
            }
            break;
        case LIST:
            {
                final ListTypeInfo listType = (ListTypeInfo) typeInfo;
                final ListObjectInspector listInspector = ObjectInspectorFactory.getStandardListObjectInspector(getObjectInspector(listType.getListElementTypeInfo()));
                objectInspector = listInspector;
            }
            break;
        case STRUCT:
            {
                final StructTypeInfo structType = (StructTypeInfo) typeInfo;
                final List<TypeInfo> fieldTypes = structType.getAllStructFieldTypeInfos();
                final List<ObjectInspector> fieldInspectors = new ArrayList<ObjectInspector>();
                for (TypeInfo fieldType : fieldTypes) {
                    fieldInspectors.add(getObjectInspector(fieldType));
                }
                final StructObjectInspector structInspector = ObjectInspectorFactory.getStandardStructObjectInspector(structType.getAllStructFieldNames(), fieldInspectors);
                objectInspector = structInspector;
            }
            break;
        case UNION:
            {
                final UnionTypeInfo unionType = (UnionTypeInfo) typeInfo;
                final List<TypeInfo> fieldTypes = unionType.getAllUnionObjectTypeInfos();
                final List<ObjectInspector> fieldInspectors = new ArrayList<ObjectInspector>();
                for (TypeInfo fieldType : fieldTypes) {
                    fieldInspectors.add(getObjectInspector(fieldType));
                }
                final UnionObjectInspector unionInspector = ObjectInspectorFactory.getStandardUnionObjectInspector(fieldInspectors);
                objectInspector = unionInspector;
            }
            break;
        default:
            throw new RuntimeException("Unexpected category " + typeInfo.getCategory());
    }
    Preconditions.checkState(objectInspector != null);
    return objectInspector;
}
Also used : WritableIntObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableIntObjectInspector) WritableByteObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableByteObjectInspector) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) StandardStructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardStructObjectInspector) WritableDoubleObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableDoubleObjectInspector) ObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector) MapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector) StandardListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardListObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector) WritableStringObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableStringObjectInspector) ListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector) WritableHiveCharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableHiveCharObjectInspector) WritableHiveVarcharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableHiveVarcharObjectInspector) WritableBooleanObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableBooleanObjectInspector) WritableTimestampObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableTimestampObjectInspector) WritableHiveIntervalDayTimeObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableHiveIntervalDayTimeObjectInspector) WritableShortObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableShortObjectInspector) StandardMapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardMapObjectInspector) WritableHiveIntervalYearMonthObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableHiveIntervalYearMonthObjectInspector) WritableFloatObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableFloatObjectInspector) WritableLongObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableLongObjectInspector) StandardUnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardUnionObjectInspector) WritableDateObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableDateObjectInspector) WritableHiveDecimalObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableHiveDecimalObjectInspector) StructTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.StructTypeInfo) MapTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.MapTypeInfo) StructTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.StructTypeInfo) PrimitiveTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.PrimitiveTypeInfo) VarcharTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.VarcharTypeInfo) ListTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.ListTypeInfo) DecimalTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.DecimalTypeInfo) TypeInfo(org.apache.hadoop.hive.serde2.typeinfo.TypeInfo) UnionTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.UnionTypeInfo) CharTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.CharTypeInfo) PrimitiveTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.PrimitiveTypeInfo) MapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector) StandardMapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardMapObjectInspector) ListTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.ListTypeInfo) StandardListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardListObjectInspector) ListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector) MapTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.MapTypeInfo) List(java.util.List) ArrayList(java.util.ArrayList) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) StandardUnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardUnionObjectInspector) StandardStructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardStructObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector) UnionTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.UnionTypeInfo)

Example 19 with UnionObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector in project hive by apache.

the class StatsUtils method getSizeOfComplexTypes.

/**
 * Get the size of complex data types
 * @return raw data size
 */
public static long getSizeOfComplexTypes(HiveConf conf, ObjectInspector oi) {
    long result = 0;
    int length = 0;
    int listEntries = HiveConf.getIntVar(conf, HiveConf.ConfVars.HIVE_STATS_LIST_NUM_ENTRIES);
    int mapEntries = HiveConf.getIntVar(conf, HiveConf.ConfVars.HIVE_STATS_MAP_NUM_ENTRIES);
    switch(oi.getCategory()) {
        case PRIMITIVE:
            String colTypeLowerCase = oi.getTypeName().toLowerCase();
            if (colTypeLowerCase.equals(serdeConstants.STRING_TYPE_NAME) || colTypeLowerCase.startsWith(serdeConstants.VARCHAR_TYPE_NAME) || colTypeLowerCase.startsWith(serdeConstants.CHAR_TYPE_NAME)) {
                int avgColLen = (int) getAvgColLenOf(conf, oi, colTypeLowerCase);
                result += JavaDataModel.get().lengthForStringOfLength(avgColLen);
            } else if (colTypeLowerCase.equals(serdeConstants.BINARY_TYPE_NAME)) {
                int avgColLen = (int) getAvgColLenOf(conf, oi, colTypeLowerCase);
                result += JavaDataModel.get().lengthForByteArrayOfSize(avgColLen);
            } else {
                result += getAvgColLenOfFixedLengthTypes(colTypeLowerCase);
            }
            break;
        case LIST:
            if (oi instanceof StandardConstantListObjectInspector) {
                // constant list projection of known length
                StandardConstantListObjectInspector scloi = (StandardConstantListObjectInspector) oi;
                length = scloi.getWritableConstantValue().size();
                // check if list elements are primitive or Objects
                ObjectInspector leoi = scloi.getListElementObjectInspector();
                if (leoi.getCategory().equals(ObjectInspector.Category.PRIMITIVE)) {
                    result += getSizeOfPrimitiveTypeArraysFromType(leoi.getTypeName(), length, conf);
                } else {
                    result += JavaDataModel.get().lengthForObjectArrayOfSize(length);
                }
            } else {
                StandardListObjectInspector sloi = (StandardListObjectInspector) oi;
                // list overhead + (configured number of element in list * size of element)
                long elemSize = getSizeOfComplexTypes(conf, sloi.getListElementObjectInspector());
                result += JavaDataModel.get().arrayList() + (listEntries * elemSize);
            }
            break;
        case MAP:
            if (oi instanceof StandardConstantMapObjectInspector) {
                // constant map projection of known length
                StandardConstantMapObjectInspector scmoi = (StandardConstantMapObjectInspector) oi;
                result += getSizeOfMap(scmoi);
            } else {
                StandardMapObjectInspector smoi = (StandardMapObjectInspector) oi;
                result += getSizeOfComplexTypes(conf, smoi.getMapKeyObjectInspector());
                result += getSizeOfComplexTypes(conf, smoi.getMapValueObjectInspector());
                // hash map overhead
                result += JavaDataModel.get().hashMap(mapEntries);
            }
            break;
        case STRUCT:
            if (oi instanceof StandardConstantStructObjectInspector) {
                // constant map projection of known length
                StandardConstantStructObjectInspector scsoi = (StandardConstantStructObjectInspector) oi;
                result += getSizeOfStruct(scsoi);
            } else {
                StructObjectInspector soi = (StructObjectInspector) oi;
                // add constant object overhead for struct
                result += JavaDataModel.get().object();
                // add constant struct field names references overhead
                result += soi.getAllStructFieldRefs().size() * JavaDataModel.get().ref();
                for (StructField field : soi.getAllStructFieldRefs()) {
                    result += getSizeOfComplexTypes(conf, field.getFieldObjectInspector());
                }
            }
            break;
        case UNION:
            UnionObjectInspector uoi = (UnionObjectInspector) oi;
            // add constant object overhead for union
            result += JavaDataModel.get().object();
            // add constant size for unions tags
            result += uoi.getObjectInspectors().size() * JavaDataModel.get().primitive1();
            for (ObjectInspector foi : uoi.getObjectInspectors()) {
                result += getSizeOfComplexTypes(conf, foi);
            }
            break;
        default:
            break;
    }
    return result;
}
Also used : WritableIntObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableIntObjectInspector) WritableByteObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableByteObjectInspector) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) ObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector) WritableTimestampLocalTZObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableTimestampLocalTZObjectInspector) StandardConstantListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantListObjectInspector) StandardConstantMapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantMapObjectInspector) HiveCharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveCharObjectInspector) WritableBinaryObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableBinaryObjectInspector) WritableTimestampObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableTimestampObjectInspector) WritableShortObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableShortObjectInspector) WritableLongObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableLongObjectInspector) WritableDateObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableDateObjectInspector) ConstantObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.ConstantObjectInspector) WritableHiveDecimalObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableHiveDecimalObjectInspector) WritableDoubleObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableDoubleObjectInspector) StandardListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardListObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector) StringObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.StringObjectInspector) WritableStringObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableStringObjectInspector) HiveVarcharObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveVarcharObjectInspector) WritableBooleanObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableBooleanObjectInspector) StandardConstantStructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantStructObjectInspector) StandardMapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardMapObjectInspector) BinaryObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.BinaryObjectInspector) WritableFloatObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableFloatObjectInspector) StandardConstantMapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantMapObjectInspector) StandardMapObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardMapObjectInspector) StructField(org.apache.hadoop.hive.serde2.objectinspector.StructField) StandardConstantListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantListObjectInspector) StandardListObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardListObjectInspector) StandardConstantStructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantStructObjectInspector) UnionObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector) StructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector) StandardConstantStructObjectInspector(org.apache.hadoop.hive.serde2.objectinspector.StandardConstantStructObjectInspector)

Example 20 with UnionObjectInspector

use of org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector in project hive by apache.

the class TestStandardObjectInspectors method testStandardUnionObjectInspector.

@SuppressWarnings("unchecked")
public void testStandardUnionObjectInspector() throws Throwable {
    try {
        ArrayList<ObjectInspector> objectInspectors = new ArrayList<ObjectInspector>();
        // add primitive types
        objectInspectors.add(PrimitiveObjectInspectorFactory.javaIntObjectInspector);
        objectInspectors.add(PrimitiveObjectInspectorFactory.javaStringObjectInspector);
        objectInspectors.add(PrimitiveObjectInspectorFactory.javaBooleanObjectInspector);
        // add a list
        objectInspectors.add(ObjectInspectorFactory.getStandardListObjectInspector(PrimitiveObjectInspectorFactory.javaIntObjectInspector));
        // add a map
        objectInspectors.add(ObjectInspectorFactory.getStandardMapObjectInspector(PrimitiveObjectInspectorFactory.javaIntObjectInspector, PrimitiveObjectInspectorFactory.javaStringObjectInspector));
        // add a struct
        List<String> fieldNames = new ArrayList<String>();
        fieldNames.add("myDouble");
        fieldNames.add("myLong");
        ArrayList<ObjectInspector> fieldObjectInspectors = new ArrayList<ObjectInspector>();
        fieldObjectInspectors.add(PrimitiveObjectInspectorFactory.javaDoubleObjectInspector);
        fieldObjectInspectors.add(PrimitiveObjectInspectorFactory.javaLongObjectInspector);
        objectInspectors.add(ObjectInspectorFactory.getStandardStructObjectInspector(fieldNames, fieldObjectInspectors));
        StandardUnionObjectInspector uoi1 = ObjectInspectorFactory.getStandardUnionObjectInspector(objectInspectors);
        StandardUnionObjectInspector uoi2 = ObjectInspectorFactory.getStandardUnionObjectInspector((ArrayList<ObjectInspector>) objectInspectors.clone());
        assertEquals(uoi1, uoi2);
        assertEquals(ObjectInspectorUtils.getObjectInspectorName(uoi1), ObjectInspectorUtils.getObjectInspectorName(uoi2));
        assertTrue(ObjectInspectorUtils.compareTypes(uoi1, uoi2));
        // compareSupported returns false because Union can contain
        // an object of Map
        assertFalse(ObjectInspectorUtils.compareSupported(uoi1));
        // construct unionObjectInspector without Map field.
        ArrayList<ObjectInspector> ois = (ArrayList<ObjectInspector>) objectInspectors.clone();
        ois.set(4, PrimitiveObjectInspectorFactory.javaIntObjectInspector);
        assertTrue(ObjectInspectorUtils.compareSupported(ObjectInspectorFactory.getStandardUnionObjectInspector(ois)));
        // metadata
        assertEquals(Category.UNION, uoi1.getCategory());
        List<? extends ObjectInspector> uois = uoi1.getObjectInspectors();
        assertEquals(6, uois.size());
        for (int i = 0; i < 6; i++) {
            assertEquals(objectInspectors.get(i), uois.get(i));
        }
        StringBuilder unionTypeName = new StringBuilder();
        unionTypeName.append("uniontype<");
        for (int i = 0; i < uois.size(); i++) {
            if (i > 0) {
                unionTypeName.append(",");
            }
            unionTypeName.append(uois.get(i).getTypeName());
        }
        unionTypeName.append(">");
        assertEquals(unionTypeName.toString(), uoi1.getTypeName());
        // TypeInfo
        TypeInfo typeInfo1 = TypeInfoUtils.getTypeInfoFromObjectInspector(uoi1);
        assertEquals(Category.UNION, typeInfo1.getCategory());
        assertEquals(UnionTypeInfo.class.getName(), typeInfo1.getClass().getName());
        assertEquals(typeInfo1.getTypeName(), uoi1.getTypeName());
        assertEquals(typeInfo1, TypeInfoUtils.getTypeInfoFromTypeString(uoi1.getTypeName()));
        TypeInfo typeInfo2 = TypeInfoUtils.getTypeInfoFromObjectInspector(uoi2);
        assertEquals(typeInfo1, typeInfo2);
        assertEquals(TypeInfoUtils.getStandardJavaObjectInspectorFromTypeInfo(typeInfo1), TypeInfoUtils.getStandardJavaObjectInspectorFromTypeInfo(typeInfo2));
        assertEquals(TypeInfoUtils.getStandardWritableObjectInspectorFromTypeInfo(typeInfo1), TypeInfoUtils.getStandardWritableObjectInspectorFromTypeInfo(typeInfo2));
        // null
        assertNull(uoi1.getField(null));
        assertEquals(-1, uoi1.getTag(null));
        // Union
        UnionObject union = new StandardUnion((byte) 0, 1);
        assertEquals(0, uoi1.getTag(union));
        assertEquals(1, uoi1.getField(union));
        assertEquals("{0:1}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 0, 1), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(1));
        union = new StandardUnion((byte) 1, "two");
        assertEquals(1, uoi1.getTag(union));
        assertEquals("two", uoi1.getField(union));
        assertEquals("{1:\"two\"}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 1, "two"), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals("two"));
        union = new StandardUnion((byte) 2, true);
        assertEquals(2, uoi1.getTag(union));
        assertEquals(true, uoi1.getField(union));
        assertEquals("{2:true}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 2, true), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(true));
        ArrayList<Integer> iList = new ArrayList<Integer>();
        iList.add(4);
        iList.add(5);
        union = new StandardUnion((byte) 3, iList);
        assertEquals(3, uoi1.getTag(union));
        assertEquals(iList, uoi1.getField(union));
        assertEquals("{3:[4,5]}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 3, iList.clone()), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(iList));
        HashMap<Integer, String> map = new HashMap<Integer, String>();
        map.put(6, "six");
        map.put(7, "seven");
        map.put(8, "eight");
        union = new StandardUnion((byte) 4, map);
        assertEquals(4, uoi1.getTag(union));
        assertEquals(map, uoi1.getField(union));
        assertEquals("{4:{6:\"six\",7:\"seven\",8:\"eight\"}}", SerDeUtils.getJSONString(union, uoi1));
        Throwable th = null;
        try {
            ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 4, map.clone()), uoi2, null);
        } catch (Throwable t) {
            th = t;
        }
        assertNotNull(th);
        assertEquals("Compare on map type not supported!", th.getMessage());
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(map));
        ArrayList<Object> struct = new ArrayList<Object>(2);
        struct.add(9.0);
        struct.add(10L);
        union = new StandardUnion((byte) 5, struct);
        assertEquals(5, uoi1.getTag(union));
        assertEquals(struct, uoi1.getField(union));
        assertEquals("{5:{\"mydouble\":9.0,\"mylong\":10}}", SerDeUtils.getJSONString(union, uoi1));
        assertEquals(0, ObjectInspectorUtils.compare(union, uoi1, new StandardUnion((byte) 5, struct.clone()), uoi2));
        assertTrue(ObjectInspectorUtils.copyToStandardObject(union, uoi1).equals(struct));
    } catch (Throwable e) {
        e.printStackTrace();
        throw e;
    }
}
Also used : HashMap(java.util.HashMap) ArrayList(java.util.ArrayList) TypeInfo(org.apache.hadoop.hive.serde2.typeinfo.TypeInfo) UnionTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.UnionTypeInfo) StandardUnion(org.apache.hadoop.hive.serde2.objectinspector.StandardUnionObjectInspector.StandardUnion) UnionTypeInfo(org.apache.hadoop.hive.serde2.typeinfo.UnionTypeInfo)

Aggregations

ObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector)21 UnionObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.UnionObjectInspector)19 ArrayList (java.util.ArrayList)13 StructObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector)13 ListObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.ListObjectInspector)12 MapObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.MapObjectInspector)12 HiveCharObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveCharObjectInspector)12 HiveVarcharObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveVarcharObjectInspector)12 List (java.util.List)11 BinaryObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.BinaryObjectInspector)11 ByteObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.ByteObjectInspector)11 DateObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.DateObjectInspector)11 DoubleObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.DoubleObjectInspector)11 FloatObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.FloatObjectInspector)11 HiveDecimalObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.HiveDecimalObjectInspector)11 IntObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.IntObjectInspector)11 LongObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.LongObjectInspector)11 ShortObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.ShortObjectInspector)11 StringObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.StringObjectInspector)11 TimestampObjectInspector (org.apache.hadoop.hive.serde2.objectinspector.primitive.TimestampObjectInspector)11