# Read newline-delimited JSON

Acquire / file

A variant of [Read a JSON file](https://meridian.online/operations/read-a-json-file).

Read a file holding one JSON object per line, as JSON Lines or NDJSON, into rows.

Page: <https://meridian.online/operations/read-a-json-file/read-newline-delimited-json>

## Status

Runs as a SQL step.

arcform carries it as a `sql:` step today.

## The DuckDB step

read a file with one JSON object per line

It reads the `orders` table the proofs share. Create it first:

```sql
CREATE TABLE orders AS SELECT * FROM (VALUES
  (1,  'Alice Ng',      'north', DATE '2026-01-03', 120.00, 'widget', 'a|b',   '{"channel":"web","items":[1,2]}'),
  (2,  'alice ng',      'north', DATE '2026-01-05',  80.50, 'gadget', 'b',     '{"channel":"shop","items":[3]}'),
  (3,  'Bob  Okafor',   'south', DATE '2026-01-05', 300.00, 'widget', 'a|c',   '{"channel":"web","items":[4,5,6]}'),
  (4,  'Carla Díaz',    'east',  DATE '2026-01-09',  NULL,  'gizmo',  NULL,    '{"channel":"web","items":[]}'),
  (5,  'Bob Okafor',    'south', DATE '2026-01-12',  45.25, 'gadget', 'c',     '{"channel":"phone","items":[7]}'),
  (6,  'Dev Patel',     'west',  DATE '2026-01-12', 510.00, 'widget', 'a',     '{"channel":"web","items":[8,9]}'),
  (7,  'Carla Diaz',    'east',  DATE '2026-01-20',  99.99, 'gizmo',  'b|c',   '{"channel":"shop","items":[10]}'),
  (8,  'Dev Patel',     'west',  DATE '2026-02-02', 510.00, 'widget', 'a',     '{"channel":"web","items":[8,9]}'),
  (9,  'Eve Larsen',    NULL,    DATE '2026-02-03',  15.00, 'gadget', NULL,    '{"channel":"web","items":[11]}'),
  (10, 'Alice Ng',      'north', DATE '2026-02-10', 220.00, 'gizmo',  'a|b|c', '{"channel":"phone","items":[12,13]}')
) t(order_id, customer, region, order_date, amount, product, tags, meta);
```

Then run the statement:

```sql
COPY (SELECT order_id, customer, order_date, amount, string_split(tags, '|') AS tags FROM orders) TO '<your_scratch_dir>/lines.ndjson' (FORMAT json);
CREATE TEMP TABLE result AS
SELECT * FROM read_json('<your_scratch_dir>/lines.ndjson', format = 'newline_delimited');
SELECT * FROM result;
SELECT CASE WHEN ((SELECT count(*) FROM read_text('<your_scratch_dir>/lines.ndjson') t, unnest(string_split(trim(content, chr(10)), chr(10)))) = 10 AND (SELECT count(*) = 10 AND count(amount) = 9 AND round(sum(amount), 2) = 1900.74 AND count(tags) = 8 FROM result) AND (SELECT list(column_type) FROM (DESCRIBE result)) = ['BIGINT', 'VARCHAR', 'DATE', 'DOUBLE', 'VARCHAR[]'] AND (SELECT tags FROM result WHERE order_id = 10) = ['a', 'b', 'c']) THEN 'as expected' ELSE error('the file did not read as one row for each of its 10 lines, with the tags a list and the date a DATE') END AS verdict;
```

Replace `<your_scratch_dir>` with a directory of your own that you can write to before running the statement.

The statement writes `lines.ndjson` itself, in `<your_scratch_dir>`.

It needs nothing installed beyond DuckDB itself.

## What other tools call it

Meridian's classification; an independent blind rating agreed on 95.7% of 324 rows (variant, version 3).

- Apache DataFusion: [`JSON`](https://datafusion.apache.org/user-guide/features.html) (read 2026-09-27)
- DuckDB: `read_ndjson` (read 2026-09-27)
- DuckDB: `read_ndjson_auto` (read 2026-09-27)
- Duckle: [`JSONL / NDJSON`](https://github.com/slothflowlabs/duckle/blob/f85f8960edfb6efbf4cb89d58a7b32e16df0e867/website/docs/components.html) (read 2026-09-25)
- Polars: [`read_ndjson`](https://docs.pola.rs/api/python/stable/reference/io.html) (read 2026-09-27)
- Polars: [`scan_ndjson`](https://docs.pola.rs/api/python/stable/reference/io.html) (read 2026-09-27)
- VisiData: [`jsonl (loader)`](https://github.com/saulpw/visidata/blob/7609212ec2febafa9caba7183cbebfaf49627909/visidata/loaders/json.py#L33) (read 2026-09-23)
- VisiData: [`jsonla (loader)`](https://github.com/saulpw/visidata/blob/7609212ec2febafa9caba7183cbebfaf49627909/visidata/loaders/jsonla.py#L36) (read 2026-09-23)