Different consumers want differently shaped JSON. Excel2JSON ships five JSON layouts plus CSV/TSV; this page shows the same two-row table in each.
Product Units Widget 42 Gadget 7
Objects (the default) #
One object per row, headers as keys. What an API endpoint or a config file usually wants. Turns on nested mode when your headers describe structure.
[
{ "Product": "Widget", "Units": 42 },
{ "Product": "Gadget", "Units": 7 }
]
NDJSON: one object per line #
The data-pipeline format: each line is a complete, compact JSON object. Stream it, grep it, feed it to log tooling or bulk-import endpoints.
{"Product":"Widget","Units":42}
{"Product":"Gadget","Units":7}
Rows as arrays #
Positional data without keys, headers included as the first row of your selection if you keep them selected.
[["Widget", 42], ["Gadget", 7]]
Headers + rows #
Keys once, data as arrays: compact for wide tables, and chart libraries like it.
{ "headers": ["Product", "Units"], "data": [["Widget", 42], ["Gadget", 7]] }
Cell map (A1 to value) #
Every cell keyed by its real sheet address; the grid is origin-anchored, so data starting at C5 emits true C5 addresses. Handy for spreadsheet tooling and audits.
{ "A1": "Product", "B1": "Units", "A2": "Widget", "B2": 42, "A3": "Gadget", "B3": 7 }
Shared options #
- Headers: first row, first column (the table is transposed), or none, where keys become
col1,col2, … - Skip blanks: omits empty cells instead of writing
null. - Pretty print: indented or compact (NDJSON is always compact by definition).
- More formats: CSV and TSV live at the bottom of the format list.
