JSON to NDJSON Converter

Turn an array into one JSON value per line for a log or a stream, and turn those lines back into an array.

Voltar para todas as ferramentas

Direcao Blank lines Keep valid lines

JSON

NDJSON

O resultado aparece aqui.

Batch validation

A report on the JSONL in the left pane, whichever direction is selected. Nothing here changes the document.

JSON Schema (optional)

Required fields

Duplicate ID field

Check that every record shares the common shape

Funciona inteiramente no seu navegador.

Sobre esta ferramenta

NDJSON is one JSON value per line, and that is the whole format. It exists because a reader can handle a line and forget it, which is what makes a log file or a large export streamable when a single enormous array is not. Any JSON value is allowed on a line, not only an object, and no line is indented, because a line break inside a value would end the record. This tool skips a completely blank line rather than failing on it, since almost every file ends with a newline, and it reports a line that does not parse by the number it has in your file rather than by a position in some rewritten document. That skip is the default; the Blank lines option can treat a blank line as a failure instead. Keep valid lines changes what a bad line does to the rest of the file: instead of stopping the conversion, every line is judged on its own, the lines that parse build the result, and the lines that do not are listed with their line number and reason. Both directions go through the token stream, so a twenty digit integer comes back with all twenty digits. Batch validation, below the converter, is a separate report rather than a conversion: it checks every record on its own against an optional JSON Schema, checks for required fields, finds duplicate values in one chosen ID field, and compares every record's keys and types against whichever shape most records actually have, without changing the document.

Como usar

Choose a direction, then paste your document into the left pane or load the sample.
Press Convert. Going to NDJSON needs an array, since that is the JSON shape that is a sequence of values.
Copy the result or download it. Any line that does not parse is reported with its line number.

Perguntas frequentes

What is NDJSON for?

Streaming and appending. A JSON array has to be complete before anything can read it, so the process writing one holds the whole thing in memory and the process reading one waits for the closing bracket. With one value per line, a writer appends a line and a reader handles it immediately, which is why logs, database exports and machine learning datasets are shipped this way. JSON Lines is the same format under another name, and the extensions .ndjson and .jsonl both mean this.

What happens to blank lines?

By default they are skipped, and the page tells you how many were skipped. Nearly every file ends with a newline, which strictly speaking means a blank last line, so failing on one would reject almost every real document. A line holding only spaces is treated the same way. The Blank lines option can switch this to Treat as error, for someone who wants every line in the file to hold a value. Nothing else is skipped: a line that has content on it and does not parse stops the conversion, or is added to the invalid line report when Keep valid lines is on.

Why does converting to NDJSON need an array?

Because NDJSON is a sequence of values, and an array is the only JSON shape that is one. An object is a set of named members, and there is no reading of it that becomes lines without the tool inventing a rule about whether each member becomes a line or the whole object does. Rather than pick one silently, the tool names the shape it was given and stops.

What does Keep valid lines do?

With Keep valid lines on, a single bad line no longer stops the conversion. Every line is judged on its own: the ones that parse go into the result array, in their original order, and the ones that do not are listed below the result with their line number, column and the reason they failed, so they can be pulled out and fixed separately. This is off by default, which stops at the first bad line the way the tool always has.

What does batch validation check, and what is the common shape it compares against?

Four things, each independent and each optional. A JSON Schema, when given, is applied to every record on its own, and every violation is reported with its line and the path inside the record it failed at. A required fields list checks that every record carries every name you gave, whatever its value. A duplicate ID field checks one field you name for values that repeat across records, and groups the line numbers that share a value. The consistency check has no schema to go by, so it builds its own baseline from the file: every record's own key list becomes a signature, the signature the most records share becomes the expected shape, and a record missing a key, carrying an extra one, or holding a different type for a shared key is reported against that baseline. A line that is not valid JSON at all is named and left out of every other check, since a value that never became a record cannot be validated as one.

JSON to NDJSON Converter

Atalhos de teclado

Dados locais

Arquivos abertos recentemente

Compartilhar via link

Abrir o link compartilhado em

JSON

Carregar de uma URL

NDJSON

Sobre esta ferramenta

Como usar