The format and location of SMART / health data varies across protocols.  To
simplify the output, this application uses a Dumb Unified Model of SMART
Buffers. In this model, SMART data is located in one or more log pages.  Each
page contains one or more values ("attributes") differentiated by an ID.  Note
that ID's are only unique within a log page. Thus, the application outputs:

<Log Page ID> <Attribute ID> <Attribute value>

for each selected attribute.

See the shell scripts atasmart, nvmesmart, and scsismart for examples of parsing
the output.

WWW: https://github.com/ctuffli/smart