convert: HDF5 conversion

This module provides classes and function to convert file formats supported by silx into HDF5 file. Currently, SPEC file and fabio images are the supported formats.

Read the documentation of silx.io.spech5 and silx.io.fabioh5 for information on the structure of the output HDF5 files.

Strings are written to the HDF5 datasets as fixed-length ASCII (NumPy S type). This is done in order to produce files that have maximum compatibility with other HDF5 libraries, as recommended in the h5py documentation.

If you read the files back with h5py in Python 3, you will recover strings as bytes, which you should decode to transform them into python strings:

>>> import h5py
>>> f = h5py.File("myfile.h5")
>>> f["/1.1/instrument/specfile/scan_header"][0]
b'#S 94  ascan  del -0.5 0.5  20 1'
>>> f["/1.1/instrument/specfile/scan_header"][0].decode()
'#S 94  ascan  del -0.5 0.5  20 1'

Arrays of strings, such as file and scan headers, are stored as fixed-length strings. The length of all strings in an array is equal to the length of the longest string. Shorter strings are right-padded with blank spaces.

Note

This module has a dependency on the h5py library, which is not a mandatory dependency for silx. You might need to install it if you don’t already have it.

silx.io.convert.write_to_h5(infile, h5file, h5path='/', mode='a', overwrite_data=False, link_type='soft', create_dataset_args=None, min_size=500)[source]

Write content of a h5py-like object into a HDF5 file.

Parameters:
  • infile – Path of input file, or commonh5.File object or commonh5.Group object
  • h5file – Path of output HDF5 file or HDF5 file handle (h5py.File object)
  • h5path (str) – Target path in HDF5 file in which scan groups are created. Default is root ("/")
  • mode (str) – Can be "r+" (read/write, file must exist), "w" (write, existing file is lost), "w-" (write, fail if exists) or "a" (read/write if exists, create otherwise). This parameter is ignored if h5file is a file handle.
  • overwrite_data (bool) – If True, existing groups and datasets can be overwritten, if False they are skipped. This parameter is only relevant if file_mode is "r+" or "a".
  • link_type (str) – “soft” (default) or “hard”
  • create_dataset_args (dict) – Dictionary of args you want to pass to h5py.File.create_dataset. This allows you to specify filters and compression parameters. Don’t specify name and data. These arguments are only applied to datasets larger than 1MB.
  • min_size (int) – Minimum number of elements in a dataset to apply chunking and compression. Default is 500.

The structure of the spec data in an HDF5 file is described in the documentation of silx.io.spech5.

silx.io.convert.convert(infile, h5file, mode='w-', create_dataset_args=None)[source]

Convert a supported file into an HDF5 file, write scans into the root group (/).

This is a convenience shortcut to call:

write_to_h5(h5like, h5file, h5path='/',
            mode="w-", link_type="soft")
Parameters:
  • infile – Path of input file or commonh5.File object or commonh5.Group object
  • h5file – Path of output HDF5 file, or h5py.File object
  • mode – Can be "w" (write, existing file is lost), "w-" (write, fail if exists). This is ignored if h5file is a file handle.
  • create_dataset_args – Dictionary of args you want to pass to h5py.File.create_dataset. This allows you to specify filters and compression parameters. Don’t specify name and data.