{
  "context": "This criteria evaluates how well the engineer uses h5netcdf's attributes API to manage metadata on netCDF4 files and variables. The focus is on proper usage of the .attrs dictionary interface for setting, reading, and deleting attributes at both file and variable levels.",
  "type": "weighted_checklist",
  "checklist": [
    {
      "name": "File object creation",
      "description": "Uses h5netcdf.File() to open netCDF4 files with appropriate mode ('r+', 'a', or 'w') for reading and writing attributes",
      "max_score": 15
    },
    {
      "name": "File-level attributes access",
      "description": "Accesses file-level (global) attributes using the .attrs property on the File object (e.g., f.attrs['title'] = value)",
      "max_score": 20
    },
    {
      "name": "Variable attributes access",
      "description": "Accesses variable attributes using the .attrs property on Variable objects (e.g., var.attrs['units'] = value)",
      "max_score": 20
    },
    {
      "name": "Dictionary-style attribute setting",
      "description": "Sets attributes using dictionary-style assignment with proper types (strings, numeric arrays) via the attrs interface",
      "max_score": 15
    },
    {
      "name": "Dictionary-style attribute reading",
      "description": "Reads attributes using dictionary-style access (attrs['key']) or iteration methods (attrs.keys(), attrs.items(), attrs.values())",
      "max_score": 15
    },
    {
      "name": "Attribute deletion",
      "description": "Deletes attributes using dictionary-style deletion (del attrs['key']) from the attrs interface",
      "max_score": 10
    },
    {
      "name": "Resource management",
      "description": "Properly manages file resources using context managers (with statement) or explicit close() calls to ensure attributes are persisted",
      "max_score": 5
    }
  ]
}