Class DefaultConversionProcessor

java.lang.Object
com.univocity.parsers.common.DefaultConversionProcessor
All Implemented Interfaces:
ConversionProcessor
Direct Known Subclasses:
AbstractObjectProcessor, BeanConversionProcessor, ObjectRowWriterProcessor

public abstract class DefaultConversionProcessor extends Object implements ConversionProcessor
The base class for RowProcessor and RowWriterProcessor implementations that support value conversions provided by Conversion instances.
  • Field Details

    • conversionsByType

      private Map<Class<?>,Conversion[]> conversionsByType
    • conversions

      protected FieldConversionMapping conversions
    • conversionsInitialized

      private boolean conversionsInitialized
    • fieldIndexes

      private int[] fieldIndexes
    • reverseFieldIndexes

      private int[] reverseFieldIndexes
    • fieldsReordered

      private boolean fieldsReordered
    • errorHandler

      ProcessorErrorHandler errorHandler
    • context

      Context context
  • Constructor Details

    • DefaultConversionProcessor

      public DefaultConversionProcessor()
  • Method Details

    • convertIndexes

      public final FieldSet<Integer> convertIndexes(Conversion... conversions)
      Description copied from interface: ConversionProcessor
      Applies a set of Conversion objects over indexes of a record.

      The idiom to define which indexes should have these conversions applies is as follows:


      
       processor.convertIndexes(Conversions.trim(), Conversions.toUpperCase()).add(2, 5); // applies trim and uppercase conversions to fields in indexes 2 and 5
       

      Specified by:
      convertIndexes in interface ConversionProcessor
      Parameters:
      conversions - The sequence of conversions to be executed in a set of field indexes.
      Returns:
      A FieldSet for indexes.
    • convertAll

      public final void convertAll(Conversion... conversions)
      Description copied from interface: ConversionProcessor
      Applies a set of Conversion objects over all elements of a record
      Specified by:
      convertAll in interface ConversionProcessor
      Parameters:
      conversions - The sequence of conversions to be executed in all elements of a record
    • convertFields

      public final FieldSet<String> convertFields(Conversion... conversions)
      Description copied from interface: ConversionProcessor
      Applies a set of Conversion objects over fields of a record by name.

      The idiom to define which fields should have these conversions applied is as follows:


      
       processor.convertFields(Conversions.trim(), Conversions.toUpperCase()).add("name", "position"); // applies trim and uppercase conversions to fields with headers "name" and "position"
       

      Specified by:
      convertFields in interface ConversionProcessor
      Parameters:
      conversions - The sequence of conversions to be executed in a set of field indexes.
      Returns:
      A FieldSet for field names.
    • getConversions

      private FieldConversionMapping getConversions()
    • initializeConversions

      protected void initializeConversions(String[] row, Context context)
    • applyConversions

      public final Object[] applyConversions(String[] row, Context context)
      Executes the sequences of conversions defined using convertFields(Conversion...), convertIndexes(Conversion...) and convertAll(Conversion...), for every field in the given row.

      Each field will be transformed using the Conversion.execute(Object) method.

      In general the conversions will process a String and convert it to some object value (such as booleans, dates, etc).

      Parameters:
      row - the parsed record with its individual records as extracted from the original input.
      context - the current state of the parsing process.
      Returns:
      an row of Object instances containing the values obtained after the execution of all conversions.

      Fields that do not have any conversion defined will just be copied to the object array into their original positions.

    • populateReverseFieldIndexes

      private void populateReverseFieldIndexes()
    • validateAllValues

      private boolean validateAllValues(Object[] objectRow)
    • reverseConversions

      public final boolean reverseConversions(boolean executeInReverseOrder, Object[] row, NormalizedString[] headers, int[] indexesToWrite)
      Executes the sequences of reverse conversions defined using convertFields(Conversion...), convertIndexes(Conversion...) and convertAll(Conversion...), for every field in the given row.

      Each field will be transformed using the Conversion.revert(Object) method.

      In general the conversions will process an Object (such as a Boolean, Date, etc), and convert it to a String representation.

      Parameters:
      executeInReverseOrder - flag to indicate whether the conversion sequence should be executed in the reverse order of its declaration.
      row - the row of objects that will be converted
      headers - All field names used to produce records in a given destination. May be null if no headers have been defined in CommonSettings.getHeaders()
      indexesToWrite - The indexes of the headers that are actually being written. May be null if no fields have been selected using CommonSettings.selectFields(String...) or CommonSettings.selectIndexes(Integer...)
      Returns:
      true if the the row should be discarded
    • applyConversionsByType

      private boolean applyConversionsByType(boolean reverse, Object[] row, boolean[] convertedFlags)
    • handleConversionError

      protected final boolean handleConversionError(Throwable ex, Object[] row, int column)
      Handles an error that occurred when applying conversions over a value. If the user defined a ProcessorErrorHandler the user will receive the exception and is able to determine whether or not processing should continue, discarding the record. If the error handler is an instance of RetryableErrorHandler, the user can provide a default value to use in place of the one that could not be converted, and decide whether or not the record should be kept with the use of the RetryableErrorHandler.keepRecord() method.
      Parameters:
      ex - the exception that occurred when applying a conversion
      row - the record being processed at the time the exception happened.
      column - the column if the given row, whose value could not be converted. If negative, it's not possible to keep the record.
      Returns:
      true if the error has been handled by the user and the record can still be processed, otherwise false if the record should be discarded.
    • toDataProcessingException

      protected DataProcessingException toDataProcessingException(Throwable ex, Object[] row, int column)
    • convertType

      public final void convertType(Class<?> type, Conversion... conversions)
      Description copied from interface: ConversionProcessor
      Applies a sequence of conversions over values of a given type. Used for writing.
      Specified by:
      convertType in interface ConversionProcessor
      Parameters:
      type - the type over which a sequence of conversions should be applied
      conversions - the sequence of conversions to apply over values of the given type.
    • applyTypeConversion

      private Object applyTypeConversion(boolean revert, Object input)