nabu.io.writer

source module nabu.io.writer

Classes

Writer — Base class for all writers.
NXProcessWriter — A class to write Nexus file with a processing result.
NXVolVolume — An interface to NXProcessWriter with the same API than tomoscan.esrf.volume.
HSTVolWriter — A writer to mimic PyHST2 ".vol" files
HSTVolVolume — An interface to HSTVolWriter with the same API than tomoscan.esrf.volume. This is really not ideal, see nabu:#381

Functions

get_datetime — Function used by some writers to indicate the current date.
h5_write_object
export_dict_to_h5 — Wrapper on top of silx.io.dictdump.dicttoh5 replacing None with "None"
create_virtual_layout — Create a HDF5 virtual layout.
merge_hdf5_files — Parameters

files_or_pattern: str or list A list of file names, or a wildcard pattern. If a list is provided, it will not be sorted! This will have to be done before calling this function. h5_path: str Path inside the HDF5 input file(s) output_file: str Path of the output file process_name: str Name of the process output_entry: str, optional Output HDF5 root entry (default is "/entry") output_filemode: str, optional File mode for output file. Default is "a" (append) processing_index: int, optional Processing index for the output file. Default is 0. config: dict, optional Dictionary describing the configuration needed to get the results. base_dir: str, optional Base directory when using relative file names. axis: int, optional Data axis to merge. Default is 0. overwrite: bool, optional Whether to overwrite already existing data in the final file. Default is False.

source get_datetime()

Function used by some writers to indicate the current date.

source class Writer(fname)

Base class for all writers.

Methods

get_filename

source method Writer.get_filename()

source h5_write_object(h5group, key, value, overwrite=False, default_val=None)

Raises

OSError

source class NXProcessWriter(fname, entry=None, filemode='a', overwrite=False)

Bases : Writer

A class to write Nexus file with a processing result.

Initialize a NXProcessWriter.

Parameters

fname : str — Path to the HDF5 file.
entry : str, optional — Entry in the HDF5 file. Default is "entry"

Methods

write — Write the result in the current NXProcess group.

source method NXProcessWriter.write(result, process_name, processing_index=0, config=None, data_name='data', is_frames_stack=True, direct_access=True)

Write the result in the current NXProcess group.

Parameters

result : numpy.ndarray — Array containing the processing result
process_name : str — Name of the processing
processing_index : int — Index of the processing (in a pipeline)
config : dict, optional — Dictionary containing the configuration.

source class NXVolVolume(**kwargs)

Bases : NXProcessWriter

An interface to NXProcessWriter with the same API than tomoscan.esrf.volume.

NX files are written in two ways

Partial files containing sub-volumes
Final volume: master file with virtual dataset pointing to partial files

Methods

source method NXVolVolume.save()

Raises

ValueError

source method NXVolVolume.save_metadata()

source method NXVolVolume.browse_data_files()

source export_dict_to_h5(dic, h5file, h5path, overwrite_data=True, mode='a')

Wrapper on top of silx.io.dictdump.dicttoh5 replacing None with "None"

Parameters

dic : dict — Dictionary containing the options
h5file : str — File name
h5path : str — Path in the HDF5 file
overwrite_data : bool, optional — Whether to overwrite data when writing HDF5. Default is True
mode : str, optional — File mode. Default is "a" (append).

source create_virtual_layout(files_or_pattern, h5_path, base_dir=None, axis=0, dtype='f')

Create a HDF5 virtual layout.

Parameters

files_or_pattern : str or list — A list of file names, or a wildcard pattern. If a list is provided, it will not be sorted! This will have to be done before calling this function.
h5_path : str — Path inside the HDF5 input file(s)
base_dir : str, optional — Base directory when using relative file names.
axis : int, optional — Data axis to merge. Default is 0.

Raises

ValueError

source merge_hdf5_files(files_or_pattern, h5_path, output_file, process_name, output_entry=None, output_filemode='a', data_name='data', processing_index=0, config=None, base_dir=None, axis=0, overwrite=False, dtype='f')

Parameters

files_or_pattern: str or list A list of file names, or a wildcard pattern. If a list is provided, it will not be sorted! This will have to be done before calling this function. h5_path: str Path inside the HDF5 input file(s) output_file: str Path of the output file process_name: str Name of the process output_entry: str, optional Output HDF5 root entry (default is "/entry") output_filemode: str, optional File mode for output file. Default is "a" (append) processing_index: int, optional Processing index for the output file. Default is 0. config: dict, optional Dictionary describing the configuration needed to get the results. base_dir: str, optional Base directory when using relative file names. axis: int, optional Data axis to merge. Default is 0. overwrite: bool, optional Whether to overwrite already existing data in the final file. Default is False.

source class HSTVolWriter(fname, append=False, **kwargs)