Examples

Example:

Creation of a new file named 'test_write.h5' on the folder 'C:\temp'.
Add a new dataset sames 'some_name' in the root group /.
This datased is an unidimensional vector of a compound type, with this structure:

and filled with an array with two values:
15 3.0 2.0 1.0
2 1.1 2.2 3.3
Add another dataset named 'other_name', with a bidimensional array of floats, in 2 columns, and filled with:
15.1 3.0
22.0 1.4
6.5 2.8
(then the number of rows is 3)

Datasets are compressed with maximum compression level=9
Finally the file is closed

package require hdf5
set h hdf2
hdf5 $h {C:\temp\test_write.h5}
set compress_level 9
$h set -vtype compound -fields {index uint v1 float v2 float v3 float} -compress $compress_level -ncolumns 1 /some_name {15 3.0 2.0 1.0 2 1.1 2.2 3.3}
$h set -vtype float -compress $compress_level -ncolumns 2 /other_name {15.1 3.0 22.0 1.4 6.5 2.8}
$h close

Example:

Open an existent file named'test_read.h5' for reading.
get the attribute named /mesh/ntc1_2/all and print all its names and values
get the attribute named /mesh/ntc1_2/all and print the value of the attribute named 'type'
get the dimensions of the dataset named /mesh/ntc1_2/all/selectorOnMesh/GC011 and print them
print the data stored in this datased
get all groups child of the root group (named always \), and for each group get its datasets and print its data.

package require hdf5
set h hdf1
hdf5 $h {C:\temo\test_read.h5}
puts [$h get_attribute /mesh/ntc1_2/all]
puts [$h get_attribute /mesh/ntc1_2/all type]
set dataset /mesh/ntc1_2/all/selectorOnMesh/GC011
puts [$h get -dimensions $dataset] ;#get the dimensions only
puts [$h get $dataset] ;#get the data
set groups [$h glob -directory \ -types group *]
foreach group $groups {
    set datasets [$h glob -directory \$group -types dataset *]
    foreach dataset $datasets {
       set data [$h get $dataset]
       puts $data
    }
}
$h close