JSON to CSV

Example

Input:

[
  { "id": 1, "user": { "name": "Alice", "email": "a@a.com" }, "tags": ["admin"] },
  { "id": 2, "user": { "name": "Bob",   "email": "b@b.com" }, "tags": ["user", "beta"] }
]

Output (flatten nested, join arrays):

id,user.name,user.email,tags
1,Alice,a@a.com,admin
2,Bob,b@b.com,"user,beta"

Delimiter choice

Quoting rules

A field must be quoted when it contains:

• the delimiter
• a double quote (which is escaped by doubling: " -> "")
• a newline

“Always quote” is safer for cross-tool compatibility; “quote only when needed” makes the file easier to read manually.

Nested object handling

• Flatten: { user: { name: "A" } } becomes a user.name column. Deep paths become a.b.c.d.
• JSON string: the nested object stays as a single-column JSON value: "{\"name\":\"A\"}". Useful when downstream tools can parse it.
• Stringify: joins with a separator, losing the structure.

Array handling

• Join: ["a", "b"] becomes "a,b" in a single column.
• Indexed columns: produces tags.0, tags.1, tags.2, works only when array length is consistent.
• First element only: takes tags[0] and ignores the rest.

Common mistakes

• Uneven keys across objects. If one object has phone and another does not, CSV gets a phone column with empty cells for the missing rows. Always validate that the key set is consistent.
• Forgetting BOM for Excel. Excel on Windows reads UTF-8 CSVs without BOM as Latin-1 and garbles accented characters. The tool offers a “prefix UTF-8 BOM” toggle for that case.
• Embedding newlines in values. Legal in RFC 4180 CSV, but some importers choke. If your target is fragile, replace \n with a space or \\n token.
• Using Excel as a round-trip format. Excel silently converts 007 to 7, truncates long numbers, and corrupts leading zeros in IDs. For data integrity, use CSV with a proper parser.