Package app.babylon.table.plans

Class TablePlanRead

java.lang.Object
app.babylon.table.plans.TablePlanCommon<TablePlanRead>
app.babylon.table.plans.TablePlanRead
All Implemented Interfaces:
TablePlan
public class TablePlanRead extends TablePlanCommon<TablePlanRead>
Reads rows into a TableColumnar.

Example: 


 StreamSource streamSource = StreamSources.fromString("Code,Amount\nabc,10.5\n", "values.csv");
 RowSource rowSource = RowSourceCsv.builder().withStreamSource(streamSource).withSeparator(',').build();

 TableColumnar table = new TablePlanRead().withTableName(TableName.of("Trades")).execute(rowSource);

There are two useful layers for column typing when reading:

- source-side types from RowCursor.columns() choose the low-level builder before rows are consumed and are therefore the best place to avoid intermediate string materialization when that direct parser is genuinely worthwhile

- plan-side types configured on this class describe the desired output type and can override the source-side type when appropriate

  • Constructor Details

    • TablePlanRead

      public TablePlanRead()

  • Method Details

    • withIncludeResourceName

      public TablePlanRead withIncludeResourceName(ColumnName resourceName)

    • getResourceName

      public ColumnName getResourceName()

    • withLocalDateFormat

      public TablePlanRead withLocalDateFormat(DateFormat localDateFormat)

    • getLocalDateFormat

      public DateFormat getLocalDateFormat()

    • withTransform

      public TablePlanRead withTransform(Transform transform)

    • withTransforms

      public TablePlanRead withTransforms(Transform... transforms)

    • getTransforms

      public List<Transform> getTransforms()

    • withColumnType

      public TablePlanRead withColumnType(ColumnName columnName, Column.Type columnType)
      Specifies the desired output type for a column in the resulting table.

      This is best thought of as plan-side type intent. If the source cursor also exposes a type for the same column, that source-side type can still matter earlier because it may choose the low-level builder before rows are read.

      In other words:

      - source-side types are the right place to request direct builder parsing and bypass intermediate strings when the parser has a real advantage

      - plan-side types are the right place to describe the desired result schema or override the source metadata

      Parameters:
      columnName - the output column name
      columnType - the desired output type
      Returns:
      this plan

    • withColumnTypes

      public TablePlanRead withColumnTypes(Column.Type columnType, ColumnName... columnNames)

    • getColumnType

      public Column.Type getColumnType(ColumnName columnName)

    • getColumnTypes

      public Map<ColumnName,Column.Type> getColumnTypes()

    • execute

      public TableColumnar execute(TableName tableName, Column... columns)

    • execute

      public TableColumnar execute(TableName tableName, TableDescription tableDescription, Column... columns)

    • execute

      public TableColumnar execute(TableColumnar table)

    • execute

      public TableColumnar execute(RowCursor rowCursor)
      Description copied from interface: TablePlan
      Executes this plan against an already-open row supplier.

      Use this lower-level entry point when the caller already owns the live source resource and wants to control its lifetime explicitly.

      CSV example: 

      
 StreamSource streamSource = ...;
 TablePlanRead plan = new TablePlanRead();

 try (InputStream inputStream = streamSource.openStream();
         RowCursor rowCursor = RowCursorCsv.builder()
                 .withSeparator(';')
                 .build(inputStream))
 {
     TableColumnar table = plan.execute(rowCursor);
 }
      JDBC example: 
      
 PreparedStatement preparedStatement = ...;
 TablePlanRead plan = new TablePlanRead();

 try (ResultSet resultSet = preparedStatement.executeQuery();
         RowCursor rowCursor = new RowCursorResultSet(resultSet))
 {
     TableColumnar table = plan.execute(rowCursor);
 }
      Parameters:
      rowCursor - open row supplier to consume
      Returns:
      the resulting table

    • execute

      public TableColumnar execute(RowSource rowSource)
      Description copied from interface: TablePlan
      Executes this plan against a configured row source.

      This is the simplest high-level entry point when the source should handle opening and closing its own row supplier internally.

      CSV example: 

      
 StreamSource streamSource = ...;
 TablePlanRead plan = new TablePlanRead();

 TableColumnar table = plan.execute(RowSourceCsv.builder()
         .withStreamSource(streamSource)
         .withSeparator(';')
         .build());
      JDBC example: 
      
 PreparedStatement preparedStatement = ...;
 TablePlanRead plan = new TablePlanRead();

 TableColumnar table = plan.execute(RowSourceResultSet.builder()
         .withPreparedStatement(preparedStatement)
         .build());
      Parameters:
      rowSource - configured source that opens rows for this execution
      Returns:
      the resulting table

    • self

      protected TablePlanRead self()
      Specified by:
      self in class TablePlanCommon<TablePlanRead>