The body component is a single table of data. The other components of a dataset are information that complements that table (metadata, structure information, column types and descriptions, etc.), but body is the table itself.
Body is closely linked to the structure component, which stores information about column types, validation rules, and more. Qri depends on structure to be able to interpret the body. If you create a dataset from a CSV, the structure is inferred on import.
Body is the component of a Qri Dataset that can get BIG. Qri is intended for use with datasets whose body component is less than 1GB in size. Qri uses pagination to load small sections of the body on demand.
The body may contain invalid data. For example, the structure component may define a column as numeric, but some of the column's values in the body contain text. Qri can be configured to enforce validity when committing changes using strict mode.
Each dataset component has its own tab in Qri Desktop's sidebar. You can click the body tab to view a table of the body.
You can use
qri get to inspect a dataset's body, which will write the meta to the terminal as yaml.
$ qri get body --all foo/simple_csvfoo,bar,baz1," one"," true"2," two"," false"
Likewise, you can commit changes to the body using
qri save with the
$ qri save --body simple.csvusing dataset [me/my_dataset]dataset saved: foo/[email protected]/ipfs/QmXciFZ7CZj3PfauaXSpKd6amyUpWh4qiPhPGywFbzjhWa
Qri Desktop is optimized for CSV bodies, but Qri's underlying technology can handle other formats, including json, xlsx, and cbor.