Keep column meaning while changing the file format
A spreadsheet export does not carry the same type information as a database. The value 007 might be an employee code rather than the number seven. A long numeric ID might exceed the exact integer range used by common browser code. CSV to JSON therefore keeps every cell as a string. Empty cells remain empty strings, and words such as true and null remain text until you deliberately change them elsewhere.
CSV quoting matters whenever a field contains a comma, quotation mark or line break. A quoted address can span two physical lines while still belonging to one record. This parser tracks that distinction and reports inconsistent column counts instead of shifting cells into the wrong headings. Headers must be unique and nonempty. Without a header row, the tool assigns column_1, column_2 and subsequent names.
In the reverse direction, JSON must contain a nonempty array of objects. The export uses the union of their keys in first-seen order. Missing fields become empty cells; an explicit JSON null becomes the text null. Nested arrays and objects become compact JSON inside one quoted cell. This is useful for inspection, but CSV cannot preserve every distinction needed for a perfect round trip back to the original object structure.
Spreadsheet applications sometimes interpret a cell beginning with an equals sign as a formula. The default protection prefixes formula-like cells with an apostrophe and reports how many cells changed. This also affects negative values and applies to column names. Turn it off only when exact raw text is what the receiving application needs, and review how that application imports the file. The preview shows source values before this export protection is applied.