API: Update for dataset_description

[chinchien-lin]

• Refactoring. e.g. rename/update set_field to add_value/add_values
• Removing the need to specify row for dataset_descriptions metadata

Describing the dataset

dataset.list_elements()
dataset_description = dataset.get_category()
values = [“bob, builder”]
dataset_description.add_values(“element_name_string”, values)
dataset_description.remove_values(“element_name_string”, value)
dataset_description.clear_values(“field_string”)
values = dataset_description.get_values(“field_string”)

[LinkunGao]

For Tu2 steps:

1. create dataset from template
2. edit dataset metadata
   • description_metadata values: CRUD
3. save the dataset template
4. add data to dataset template

[LinkunGao]

Developed a MetadataEditor class for editing all metadata files(code_description, code_parameters, dataset_description, manifest, performances, resources, samples, subjects, submission). we can use MetadataEditor to edit all of them at the same time with these functions:

• metadataEditor.add_values(*values, field_name='', header='', append=False)

  • this function can accept multiple values for each rows or cols.
  • for the code_description and dataset_description we probably want set values by each rows, and start a Value column, so we can set the row name for field_name, ana the default value for header is Value column (you also can set others if you like) . metadataEditor.add_values(*values, field_name='contributor role')
  • in terms of others metadata files, we can set the values for them by rows or by each columns:
    • by rows: we can create a value list for that metadata file by metadataEditor.add_values(*values), in this way we don't need to consider the field name and header name. if we want add more rows values, we need set append = True. Then you can recall this function again. the second value list will be added into the next row.
    • by each columns: we also can create a value list for metadata files by their column header name. In this way, we can add values for a specific header, we can also choose to if append values or not. metadataEditor.add_values(*values, header='name', append=True)
• metadataEditor.get_values(field_name)
  • as for the code_description and dataset_description files, we need to provide the row name to get the values of that row.
  • others, we need to specify the header name to get all values of that column
• metadataEditor.remove_values(*values, field_name)
  • prepare the remove_value list for the specific row or col
  • as for the code_description and dataset_description files, we need to set the row name for the field_name
  • in terms of others, we need to set header name for field_name
• metadataEditor.clear_values(field_name='')
  • if we don't provide the field_name, this function will reset the entire file to the default status.
  • same as before, for the code_description and dataset_description files, we can provide the row name for field name to reset the values in that row to default status.
  • others, we can provide the header/column name.
• validate function for entire MetadataEditor:
  • I also developed a function to validate user inputs for the header/row_name/elements names typing. it can ignore the Case and white space. E.g: if we want use Contributor name row_name, we can ignore user types like this Contributor name, Contributor Name ,CONtributor naMe , contributorname .etc, these inputs are all match the Contributor Name .