h5tojson.py to report correct string padding

[ghost]

Currently, it reports H5T_STR_NULLPAD when h5dump says it's a H5T_STR_NULLTERM.

[jreadey]

Related to this proposed h5py feature: https://github.com/h5py/h5py/issues/429.

[ajelenak]

This is no more the case with the current h5json master branch. Example:

• h5dump output

  HDF5 "data/hdf5/vlen_string_nullterm_dset.h5" {
  GROUP "/" {
  DATASET "DS1" {
    DATATYPE  H5T_STRING {
       STRSIZE H5T_VARIABLE;
       STRPAD H5T_STR_NULLTERM;
       CSET H5T_CSET_ASCII;
       CTYPE H5T_C_S1;
    }
    DATASPACE  SIMPLE { ( 4 ) / ( 4 ) }
    DATA {
    (0): "This", "is a", "nullterm", "string."
    }
  }
  }
  }

• h5tojsonoutput

  {
  "apiVersion": "1.1.1",
  "datasets": {
      "ab8aa27a-996b-11ec-89a6-8c8590747994": {
          "alias": [
              "/DS1"
          ],
          "creationProperties": {
              "allocTime": "H5D_ALLOC_TIME_LATE",
              "fillTime": "H5D_FILL_TIME_ALLOC",
              "layout": {
                  "class": "H5D_CONTIGUOUS"
              }
          },
          "shape": {
              "class": "H5S_SIMPLE",
              "dims": [
                  4
              ]
          },
          "type": {
              "charSet": "H5T_CSET_ASCII",
              "class": "H5T_STRING",
              "length": "H5T_VARIABLE",
              "strPad": "H5T_STR_NULLTERM"
          },
          "value": [
              "This",
              "is a",
              "nullterm",
              "string."
          ]
      }
  },
  "groups": {
      "ab895e06-996b-11ec-89a6-8c8590747994": {
          "alias": [
              "/"
          ],
          "links": [
              {
                  "class": "H5L_TYPE_HARD",
                  "collection": "datasets",
                  "id": "ab8aa27a-996b-11ec-89a6-8c8590747994",
                  "title": "DS1"
              }
          ]
      }
  },
  "root": "ab895e06-996b-11ec-89a6-8c8590747994"
  }