YAML to CSV

How the conversion actually works

The engine parses your YAML into records first, because YAML files that tabulate well are sequences of mappings — a top-level list whose every item has keys. Each record's nested values are flattened to dot-separated column paths, the union of those paths across all records becomes the header row, and each record fills the columns it has, leaving blanks where it does not.

• Parse: the YAML text is read as a sequence of mappings. A file whose top level is a single mapping gets a clear error asking for a list.
• Flatten: nested mappings and arrays become column paths — address.city, tags.0, tags.1.
• Union: the header row is the ordered union of every record's paths, first-seen first, so the columns read in a stable, predictable order.
• Emit: each record is written as one CRLF-terminated row, fields quoted per RFC 4180 only when the content forces it.

What the output looks like

The worked example below is the exact pair the converter's Load sample button produces: two people records, one with a nested address and an array of tags. Watch what happens to each shape — the address object grows two columns, the tags array enumerates by index, and the comma inside a note field triggers quoting on exactly that one field.

Edge cases this converter refuses rather than mangles

An honest converter has edges, and the worst thing it can do is pretend it does not. Three inputs are rejected with a specific message instead of being silently reshaped, because each one means the output would lie to you:

• A top-level mapping instead of a sequence — a single config object is not a table, and inventing a one-row table would hide the mistake.
• Sequence items that are scalars or arrays — a list of strings cannot tabulate into named columns.
• Homogeneous keys that differ wildly between records — the union still converts, but the preview shows you the ragged result so you can decide before saving.