Quick reference
| Setting | Value / behavior |
|---|---|
| Headers | union of object keys |
| Missing value | empty cell |
| Nested value | JSON text |
Input and conversion rules
Start with a non-empty array of objects, such as a list of records returned by an API. The converter collects the union of keys in encounter order, making later-only properties visible as columns. Missing and null values become empty cells. Nested objects and arrays become compact JSON text inside a cell rather than being flattened into dotted column names. This preserves a useful representation, but nested data is usually better modeled as separate tables for analysis.
Important edge cases
Fields containing commas, quotes or newlines are wrapped in double quotes, and embedded double quotes are doubled. Formula protection is enabled by default: cells starting with =, +, -, @, tab or carriage return receive an apostrophe. This can alter numeric-looking negative strings, but helps stop spreadsheet applications interpreting untrusted data as executable formulas. Turn it off only for data you trust and when the receiving workflow requires exact cell contents. JSON number precision limits still apply.
A reliable workflow
First run the built-in example to understand the output structure. Replace it with a small representative sample from your own workflow and inspect every field that has special meaning: identifiers, currency, dates, missing values and nested collections. Then process the complete input, up to two million characters. The result can be copied or downloaded. Keep a separate original because conversion is not always reversible, especially when the target format cannot represent all the distinctions in the source.
Privacy and verification
The transformation runs in your browser using local JavaScript. There is no data-upload request and no account requirement. The page does not save your input to browser storage or place it in a shareable URL. Clearing the form removes the displayed input and result; it does not delete the original file or erase your operating system’s clipboard. Check the receiving program with a small import before loading a complete dataset. If the input is untrusted, retain formula protection where available and avoid running exported commands without review.
Frequently asked questions
Are my values uploaded?
No. The conversion is performed in the page.
Is the conversion lossless?
Not always. Review the edge cases above, especially missing values, nested data and number precision.
Why did validation fail?
Check syntax, expected input shape and the specific row or parser error shown by the tool.
About this tool
Built and maintained by Mihai at Void Fabric. This page documents the behavior of the current implementation. Found an error or an unclear assumption? Send a reproducible example with sensitive values removed.
Reference: Void Fabric field notes · Testing and editorial policy