Functions that implement IO for Array objects

Collaboration diagram for Functions that implement IO for Array objects:

Topics
	Implementation details for functions that implement IO for Array objects

Files
file	read_data.h
	declarations of stir::read_data() functions for reading Arrays from file
file	read_data.inl
	Implementation of stir::read_data() functions for reading stir::Array's from file.
file	write_data.h
	declarations of stir::write_data() functions for writing Arrays to file
file	write_data.inl
	Implementation of stir::write_data() functions for writing stir::Array's to file.

Functions
template<int num_dimensions, class IStreamT, class elemT>
Succeeded	stir::read_data (IStreamT &s, ArrayType< num_dimensions, elemT > &data, const ByteOrder byte_order=ByteOrder::native)
	Read the data of an Array from file.
template<int num_dimensions, class IStreamT, class elemT, class InputType, class ScaleT>
Succeeded	stir::read_data (IStreamT &s, ArrayType< num_dimensions, elemT > &data, NumericInfo< InputType > input_type, ScaleT &scale_factor, const ByteOrder byte_order=ByteOrder::native)
	Read the data of an Array from file as a different type.
template<int num_dimensions, class IStreamT, class elemT, class ScaleT>
Succeeded	stir::read_data (IStreamT &s, ArrayType< num_dimensions, elemT > &data, NumericType type, ScaleT &scale, const ByteOrder byte_order=ByteOrder::native)
	Read the data of an Array from file as a different type.
template<int num_dimensions, class OStreamT, class elemT>
Succeeded	stir::write_data (OStreamT &s, const ArrayType< num_dimensions, elemT > &data, const ByteOrder byte_order=ByteOrder::native, const bool can_corrupt_data=false)
	Write the data of an Array to file.
template<int num_dimensions, class OStreamT, class elemT, class OutputType, class ScaleT>
Succeeded	stir::write_data (OStreamT &s, const ArrayType< num_dimensions, elemT > &data, NumericInfo< OutputType > output_type, ScaleT &scale_factor, const ByteOrder byte_order=ByteOrder::native, const bool can_corrupt_data=false)
	Write the data of an Array to file as a different type.
template<int num_dimensions, class OStreamT, class elemT, class OutputType, class ScaleT>
Succeeded	stir::write_data_with_fixed_scale_factor (OStreamT &s, const ArrayType< num_dimensions, elemT > &data, NumericInfo< OutputType > output_type, const ScaleT scale_factor, const ByteOrder byte_order=ByteOrder::native, const bool can_corrupt_data=false)
	Write the data of an Array to file as a different type but using a given scale factor.
template<int num_dimensions, class OStreamT, class elemT, class ScaleT>
Succeeded	stir::write_data (OStreamT &s, const ArrayType< num_dimensions, elemT > &data, NumericType type, ScaleT &scale, const ByteOrder byte_order=ByteOrder::native, const bool can_corrupt_data=false)
	Write the data of an Array to file as a different type.

Detailed Description

Function Documentation

read_data() [1/3]

template<int num_dimensions, class IStreamT, class elemT>

Succeeded stir::read_data ( IStreamT & s, ArrayType< num_dimensions, elemT > & data, const ByteOrder byte_order = ByteOrder::native )

inline

Read the data of an Array from file.

Only the data will be written, not the dimensions, start indices, nor byte-order. Hence, this should only used for low-level IO.

IStreamT is supposed to be stream or file type (see implementations for read_data_1d()).

Warning

When an error occurs, the function immediately returns. However, the data might have been partially read from s.

References read_data().

Referenced by stir::ProjDataFromStream::get_bin_value(), stir::ProjDataFromStream::get_segment_by_sinogram(), stir::ProjDataFromStream::get_segment_by_view(), stir::ProjDataFromStream::get_sinogram(), stir::ProjDataFromStream::get_viewgram(), read_data(), read_data(), read_data(), read_interfile_dynamic_image(), read_interfile_image(), read_interfile_parametric_image(), and stir::GE::RDF_HDF5::InputStreamWithRecordsFromHDF5< stir::GE::RDF_HDF5::CListRecordGEHDF5 >::set_saved_get_positions().

read_data() [2/3]

template<int num_dimensions, class IStreamT, class elemT, class InputType, class ScaleT>

Succeeded stir::read_data ( IStreamT & s, ArrayType< num_dimensions, elemT > & data, NumericInfo< InputType > input_type, ScaleT & scale_factor, const ByteOrder byte_order = ByteOrder::native )

inline

Read the data of an Array from file as a different type.

This function essentially first calls convert_data() to construct an array with elements of type InputType, and then calls read_data(IStreamT&, const ArrayType<num_dimensions,elemT>&, const ByteOrder, const bool).

See also

read_data(IStreamT&, const ArrayType<num_dimensions,elemT>&, const ByteOrder, const bool)

find_scale_factor() for the meaning of scale_factor.

References convert_array(), and read_data().

read_data() [3/3]

template<int num_dimensions, class IStreamT, class elemT, class ScaleT>

Succeeded stir::read_data ( IStreamT & s, ArrayType< num_dimensions, elemT > & data, NumericType type, ScaleT & scale, const ByteOrder byte_order = ByteOrder::native )

inline

Read the data of an Array from file as a different type.

See also

read_data(IStreamT&, const ArrayType<num_dimensions,elemT>&, NumericInfo<InputType>, ScaleT&, const ByteOrder, const bool) The only difference is that the input type is now specified using NumericType.

References read_data(), and warning().

write_data() [1/3]

template<int num_dimensions, class OStreamT, class elemT>

Succeeded stir::write_data ( OStreamT & s, const ArrayType< num_dimensions, elemT > & data, const ByteOrder byte_order = ByteOrder::native, const bool can_corrupt_data = false )

inline

Write the data of an Array to file.

Only the data will be written, not the dimensions, start indices, nor byte-order. Hence, this should only used for low-level IO.

OStreamT is supposed to be stream or file type (see implementations for detail::write_data_1d()).

If can_corrupt_data = true, the data in the array can be byte-swapped after writing, and hence should no longer be used. This is useful for saving (a little bit of) time.

Warning

When an error occurs, the function immediately returns. However, the data might have been partially written to s.

For deficient compilers (e.g. VC 6.0), work-arounds are necessary to get this to compile. In particular, this function is then not templated in num_dimensions, but explicitly defined for a few dimensions (see the source).

References write_data(), and write_data_with_fixed_scale_factor().

Referenced by stir::ProjDataFromStream::set_bin_value(), stir::ProjDataFromStream::set_segment(), stir::ProjDataFromStream::set_segment(), stir::ProjDataFromStream::set_sinogram(), stir::ColsherFilter::set_up(), stir::ProjDataFromStream::set_viewgram(), write_data(), write_data(), and write_data().

write_data() [2/3]

template<int num_dimensions, class OStreamT, class elemT, class OutputType, class ScaleT>

Succeeded stir::write_data ( OStreamT & s, const ArrayType< num_dimensions, elemT > & data, NumericInfo< OutputType > output_type, ScaleT & scale_factor, const ByteOrder byte_order = ByteOrder::native, const bool can_corrupt_data = false )

inline

Write the data of an Array to file as a different type.

This function essentially first calls convert_data() to construct an array with elements of type OutputType, and then calls write_data(OstreamT&, const ArrayType<num_dimensions,elemT>&, const ByteOrder, const bool).

See also

write_data(OstreamT&, const ArrayType<num_dimensions,elemT>&, const ByteOrder, const bool)

find_scale_factor() for the meaning of scale_factor.

Warning

For deficient compilers (e.g. VC 6.0), work-arounds are necessary to get this to compile. In particular, this function is then not templated in num_dimensions, but explicitly defined for a few dimensions (see the source).

References find_scale_factor(), write_data(), and write_data_with_fixed_scale_factor().

write_data_with_fixed_scale_factor()

template<int num_dimensions, class OStreamT, class elemT, class OutputType, class ScaleT>

Succeeded stir::write_data_with_fixed_scale_factor ( OStreamT & s, const ArrayType< num_dimensions, elemT > & data, NumericInfo< OutputType > output_type, const ScaleT scale_factor, const ByteOrder byte_order = ByteOrder::native, const bool can_corrupt_data = false )

inline

Write the data of an Array to file as a different type but using a given scale factor.

If scale_factor is such that the data/scale_factor does not fit in the range for OutputType, the writing will fail. However, data might have been partially written to file anyway.

See also

write_data(OStreamT&, const ArrayType<num_dimensions,elemT>&, NumericInfo<OutputType>, ScaleT&, const ByteOrder, const bool)

Warning

For deficient compilers (e.g. VC 6.0), work-arounds are necessary to get this to compile. In particular, this function is then not templated in num_dimensions, but explicitly defined for a few dimensions (see the source).

References write_data_with_fixed_scale_factor().

Referenced by write_data(), write_data(), and write_data_with_fixed_scale_factor().

write_data() [3/3]

template<int num_dimensions, class OStreamT, class elemT, class ScaleT>

Succeeded stir::write_data ( OStreamT & s, const ArrayType< num_dimensions, elemT > & data, NumericType type, ScaleT & scale, const ByteOrder byte_order = ByteOrder::native, const bool can_corrupt_data = false )

inline

Write the data of an Array to file as a different type.

See also

write_data(OStreamT&, const ArrayType<num_dimensions,elemT>&, NumericInfo<OutputType>, ScaleT&, const ByteOrder, const bool) The only difference is that the output type is now specified using NumericType.

Warning

For deficient compilers (e.g. VC 6.0), work-arounds are necessary to get this to compile. In particular, this function is then not templated in num_dimensions, but explicitly defined for a few dimensions (see the source).

References warning(), and write_data().