Examples with DateMilliExtractor com.amazonaws.athena.connector.lambda.data.writers.extractors.DateMilliExtractor used on opensource projects

Example 1 with DateMilliExtractor

use of com.amazonaws.athena.connector.lambda.data.writers.extractors.DateMilliExtractor in project aws-athena-query-federation by awslabs.

the class ElasticsearchTypeUtilsTest method testField.

/**
 * Uses the correct field extractor to extract values from a document.
 * @param mapping is the metadata definitions of the document being processed.
 * @param document contains the values to be extracted.
 * @return a map of the field names and their associated values extracted from the document.
 * @throws Exception
 */
private Map<String, Object> testField(Schema mapping, Map<String, Object> document) throws Exception {
    Map<String, Object> results = new HashMap<>();
    for (Field field : mapping.getFields()) {
        Extractor extractor = typeUtils.makeExtractor(field);
        if (extractor instanceof VarCharExtractor) {
            NullableVarCharHolder holder = new NullableVarCharHolder();
            ((VarCharExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof BigIntExtractor) {
            NullableBigIntHolder holder = new NullableBigIntHolder();
            ((BigIntExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof IntExtractor) {
            NullableIntHolder holder = new NullableIntHolder();
            ((IntExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof SmallIntExtractor) {
            NullableSmallIntHolder holder = new NullableSmallIntHolder();
            ((SmallIntExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof TinyIntExtractor) {
            NullableTinyIntHolder holder = new NullableTinyIntHolder();
            ((TinyIntExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof Float8Extractor) {
            NullableFloat8Holder holder = new NullableFloat8Holder();
            ((Float8Extractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof Float4Extractor) {
            NullableFloat4Holder holder = new NullableFloat4Holder();
            ((Float4Extractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof DateMilliExtractor) {
            NullableDateMilliHolder holder = new NullableDateMilliHolder();
            ((DateMilliExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        } else if (extractor instanceof BitExtractor) {
            NullableBitHolder holder = new NullableBitHolder();
            ((BitExtractor) extractor).extract(document, holder);
            assertEquals("Could not extract value for: " + field.getName(), 1, holder.isSet);
            results.put(field.getName(), holder.value);
        }
    }
    return results;
}

Example 2 with DateMilliExtractor

use of com.amazonaws.athena.connector.lambda.data.writers.extractors.DateMilliExtractor in project aws-athena-query-federation by awslabs.

the class GeneratedRowWriter method makeFieldWriter.

private FieldWriter makeFieldWriter(FieldVector vector) {
    Field field = vector.getField();
    String fieldName = field.getName();
    Types.MinorType fieldType = Types.getMinorTypeForArrowType(field.getType());
    Extractor extractor = extractors.get(fieldName);
    ConstraintProjector constraint = constraints.get(fieldName);
    FieldWriterFactory factory = fieldWriterFactories.get(fieldName);
    if (factory != null) {
        return factory.create(vector, extractor, constraint);
    }
    if (extractor == null) {
        throw new IllegalStateException("Missing extractor for field[" + fieldName + "]");
    }
    switch(fieldType) {
        case INT:
            return new IntFieldWriter((IntExtractor) extractor, (IntVector) vector, constraint);
        case BIGINT:
            return new BigIntFieldWriter((BigIntExtractor) extractor, (BigIntVector) vector, constraint);
        case DATEMILLI:
            return new DateMilliFieldWriter((DateMilliExtractor) extractor, (DateMilliVector) vector, constraint);
        case DATEDAY:
            return new DateDayFieldWriter((DateDayExtractor) extractor, (DateDayVector) vector, constraint);
        case TINYINT:
            return new TinyIntFieldWriter((TinyIntExtractor) extractor, (TinyIntVector) vector, constraint);
        case SMALLINT:
            return new SmallIntFieldWriter((SmallIntExtractor) extractor, (SmallIntVector) vector, constraint);
        case FLOAT4:
            return new Float4FieldWriter((Float4Extractor) extractor, (Float4Vector) vector, constraint);
        case FLOAT8:
            return new Float8FieldWriter((Float8Extractor) extractor, (Float8Vector) vector, constraint);
        case DECIMAL:
            return new DecimalFieldWriter((DecimalExtractor) extractor, (DecimalVector) vector, constraint);
        case BIT:
            return new BitFieldWriter((BitExtractor) extractor, (BitVector) vector, constraint);
        case VARCHAR:
            return new VarCharFieldWriter((VarCharExtractor) extractor, (VarCharVector) vector, constraint);
        case VARBINARY:
            return new VarBinaryFieldWriter((VarBinaryExtractor) extractor, (VarBinaryVector) vector, constraint);
        default:
            throw new RuntimeException(fieldType + " is not supported");
    }
}

Example 3 with DateMilliExtractor

use of com.amazonaws.athena.connector.lambda.data.writers.extractors.DateMilliExtractor in project aws-athena-query-federation by awslabs.

the class ExampleRecordHandler method makeExtractor.

/**
 * Creates an Extractor for the given field. In this example the extractor just creates some random data.
 */
private Extractor makeExtractor(Field field, RowContext rowContext) {
    Types.MinorType fieldType = Types.getMinorTypeForArrowType(field.getType());
    // they need to match the split otherwise filtering will brake in unexpected ways.
    if (field.getName().equals("year")) {
        return (IntExtractor) (Object context, NullableIntHolder dst) -> {
            dst.isSet = 1;
            dst.value = rowContext.getYear();
        };
    } else if (field.getName().equals("month")) {
        return (IntExtractor) (Object context, NullableIntHolder dst) -> {
            dst.isSet = 1;
            dst.value = rowContext.getMonth();
        };
    } else if (field.getName().equals("day")) {
        return (IntExtractor) (Object context, NullableIntHolder dst) -> {
            dst.isSet = 1;
            dst.value = rowContext.getDay();
        };
    }
    switch(fieldType) {
        case INT:
            return (IntExtractor) (Object context, NullableIntHolder dst) -> {
                dst.isSet = 1;
                dst.value = ((RowContext) context).seed * (((RowContext) context).negative ? -1 : 1);
            };
        case DATEMILLI:
            return (DateMilliExtractor) (Object context, NullableDateMilliHolder dst) -> {
                dst.isSet = 1;
                dst.value = ((RowContext) context).seed * (((RowContext) context).negative ? -1 : 1);
            };
        case DATEDAY:
            return (DateDayExtractor) (Object context, NullableDateDayHolder dst) -> {
                dst.isSet = 1;
                dst.value = ((RowContext) context).seed * (((RowContext) context).negative ? -1 : 1);
            };
        case TINYINT:
            return (TinyIntExtractor) (Object context, NullableTinyIntHolder dst) -> {
                dst.isSet = 1;
                dst.value = (byte) ((((RowContext) context).seed % 4) * (((RowContext) context).negative ? -1 : 1));
            };
        case SMALLINT:
            return (SmallIntExtractor) (Object context, NullableSmallIntHolder dst) -> {
                dst.isSet = 1;
                dst.value = (short) ((((RowContext) context).seed % 4) * (((RowContext) context).negative ? -1 : 1));
            };
        case FLOAT4:
            return (Float4Extractor) (Object context, NullableFloat4Holder dst) -> {
                dst.isSet = 1;
                dst.value = ((float) ((RowContext) context).seed) * 1.1f * (((RowContext) context).negative ? -1f : 1f);
            };
        case FLOAT8:
            return (Float8Extractor) (Object context, NullableFloat8Holder dst) -> {
                dst.isSet = 1;
                dst.value = ((double) ((RowContext) context).seed) * 1.1D;
            };
        case DECIMAL:
            return (DecimalExtractor) (Object context, NullableDecimalHolder dst) -> {
                dst.isSet = 1;
                double d8Val = ((RowContext) context).seed * 1.1D * (((RowContext) context).negative ? -1d : 1d);
                BigDecimal bdVal = new BigDecimal(d8Val);
                dst.value = bdVal.setScale(((ArrowType.Decimal) field.getType()).getScale(), RoundingMode.HALF_UP);
            };
        case BIT:
            return (BitExtractor) (Object context, NullableBitHolder dst) -> {
                dst.isSet = 1;
                dst.value = ((RowContext) context).seed % 2;
            };
        case BIGINT:
            return (BigIntExtractor) (Object context, NullableBigIntHolder dst) -> {
                dst.isSet = 1;
                dst.value = ((RowContext) context).seed * 1L * (((RowContext) context).negative ? -1 : 1);
            };
        case VARCHAR:
            return (VarCharExtractor) (Object context, NullableVarCharHolder dst) -> {
                dst.isSet = 1;
                dst.value = "VarChar" + ((RowContext) context).seed;
            };
        case VARBINARY:
            return (VarBinaryExtractor) (Object context, NullableVarBinaryHolder dst) -> {
                dst.isSet = 1;
                dst.value = ("VarChar" + ((RowContext) context).seed).getBytes(Charsets.UTF_8);
            };
        default:
            return null;
    }
}

Example 4 with DateMilliExtractor

use of com.amazonaws.athena.connector.lambda.data.writers.extractors.DateMilliExtractor in project aws-athena-query-federation by awslabs.

the class ElasticsearchTypeUtils method makeDateMilliExtractor.

/**
 * Create an DATEMILLI field extractor to extract a date value from a Document. The Document value can be returned
 * as a numeric value, a String, or a List. For the latter, extract the first element only.
 * For dates extracted as a string, the ISO_ZONED_DATE_TIME format will be attempted first, followed by the
 * ISO_LOCAL_DATE_TIME format if the previous one fails. Examples of formats that will work:
 * 1) "2020-05-18T10:15:30.123456789"
 * 2) "2020-05-15T06:50:01.123Z"
 * 3) "2020-05-15T06:49:30.123-05:00".
 * Numeric dates values should be a long numeric value representing epoch milliseconds (e.g. 1589525370001)
 * Nanoseconds will be rounded to the nearest millisecond.
 * @param field is used to determine which extractor to generate based on the field type.
 * @return a field extractor.
 */
private Extractor makeDateMilliExtractor(Field field) {
    return (DateMilliExtractor) (Object context, NullableDateMilliHolder dst) -> {
        Object fieldValue = ((Map) context).get(field.getName());
        dst.isSet = 1;
        if (fieldValue instanceof String) {
            try {
                long epochSeconds;
                double nanoSeconds;
                try {
                    ZonedDateTime zonedDateTime = ZonedDateTime.parse((String) fieldValue, DateTimeFormatter.ISO_ZONED_DATE_TIME.withResolverStyle(ResolverStyle.SMART));
                    epochSeconds = zonedDateTime.toEpochSecond();
                    nanoSeconds = zonedDateTime.getNano();
                } catch (DateTimeParseException error) {
                    LocalDateTime localDateTime = LocalDateTime.parse((String) fieldValue, DateTimeFormatter.ISO_LOCAL_DATE_TIME.withResolverStyle(ResolverStyle.SMART));
                    epochSeconds = localDateTime.toEpochSecond(ZoneOffset.UTC);
                    nanoSeconds = localDateTime.getNano();
                }
                dst.value = epochSeconds * 1000 + Math.round(nanoSeconds / 1000000);
            } catch (DateTimeParseException error) {
                logger.warn("Error parsing localDateTime: {}.", error.getMessage());
                dst.isSet = 0;
            }
        } else if (fieldValue instanceof Number) {
            dst.value = ((Number) fieldValue).longValue();
        } else if (fieldValue instanceof List) {
            Object value = ((List) fieldValue).get(0);
            if (value instanceof String) {
                try {
                    long epochSeconds;
                    double nanoSeconds;
                    try {
                        ZonedDateTime zonedDateTime = ZonedDateTime.parse((String) value, DateTimeFormatter.ISO_ZONED_DATE_TIME.withZone(ZoneId.of("UTC")).withResolverStyle(ResolverStyle.SMART));
                        epochSeconds = zonedDateTime.toEpochSecond();
                        nanoSeconds = zonedDateTime.getNano();
                    } catch (DateTimeParseException error) {
                        LocalDateTime localDateTime = LocalDateTime.parse((String) value, DateTimeFormatter.ISO_LOCAL_DATE_TIME.withResolverStyle(ResolverStyle.SMART));
                        epochSeconds = localDateTime.toEpochSecond(ZoneOffset.UTC);
                        nanoSeconds = localDateTime.getNano();
                    }
                    dst.value = epochSeconds * 1000 + Math.round(nanoSeconds / 1000000);
                } catch (DateTimeParseException error) {
                    logger.warn("Error parsing localDateTime: {}.", error.getMessage());
                    dst.isSet = 0;
                }
            } else if (value instanceof Number) {
                dst.value = ((Number) value).longValue();
            } else {
                dst.isSet = 0;
            }
        } else {
            dst.isSet = 0;
        }
    };
}