JSON Stringify

Why stringify JSON

Sometimes a whole JSON document has to travel as a single string value. Common cases:

• An environment variable or Kubernetes secret that holds a JSON config, such as GOOGLE_CREDENTIALS or a feature-flag payload.
• A field in another JSON document, such as an SQS or SNS message body, an AWS Step Functions input, a webhook that wraps its payload, or a JSON column exported to text.
• A test fixture or constant in source code, where the document must be a string literal in the language you write.

Stringifying turns {"a": 1} into "{\"a\":1}": quotes and backslashes inside the document are escaped and the whole thing is wrapped in quotes. With Minify first (on by default) the result equals what JSON.stringify(JSON.stringify(value)) produces, except that numbers are never re-parsed, so a 19-digit ID keeps every digit instead of being rounded. Turn it off to keep the document’s own line breaks, which then appear as \n escapes.

Language snippets

Output as picks the target. The escaping differs per language, and getting it wrong is a classic source of broken builds:

• JSON string literal: the plain "…" value, ready to use inside another JSON document or an .env file that accepts double quotes.
• JavaScript: a single-quoted string, which keeps the many double quotes of JSON readable, plus JSON.parse(json).
• Python: a single-quoted literal passed to json.loads. Control characters become \xHH escapes.
• Java and C#: double-quoted literals for String and string. Characters a compiler rejects raw (control characters, and the separators U+0085, U+2028 and U+2029, which C# treats as line breaks) are written as \uXXXX, and so are lone surrogates, which a UTF-8 source file cannot store.

Each snippet is tested by decoding it back and comparing with the original document, so it pastes without edits.

Parsing a stringified string

Parse string does the reverse. It is most useful for logs and message queues where a payload was encoded twice, so you see "{\"order\":{\"id\":7}}" or worse, "\"{\\\"order\\\"…\"". The tool decodes the first level, checks whether the result is again a quoted string, and keeps going until it reaches something that is not, up to 32 levels. The list under the panes shows every level found; select one to see that intermediate text. When the final text is a JSON document it is pretty-printed with two-space indentation, keeping exact numbers. Untick Pretty-print decoded JSON to see it as it was.

The first level may also be escaped text without its quotes, as copied from the middle of a log line. Plain JSON that is not stringified at all is recognised too and simply reported as having no levels.

Errors

In Stringify mode the input must be valid JSON, so mistakes are reported where they are, with the line and column, before anything is generated. To wrap arbitrary text such as an SQL query or a log message instead, use JSON Escape, which accepts any text. In Parse mode, a broken first level is located the same way; deeper levels simply stop at the last text that decodes cleanly. For a quick structure check of the decoded result, open it in the JSON validator.

All work happens in a background worker in your browser, so even multi-megabyte payloads do not freeze the page and nothing is uploaded.

Examples

{
  "featureFlags": { "newCheckout": true },
  "retries": 3,
  "region": "eu-west-1"
}

"{\"featureFlags\":{\"newCheckout\":true},\"retries\":3,\"region\":\"eu-west-1\"}"

{
  "name": "O'Brien",
  "tags": ["vip", "beta"]
}

Output as Python

import json

data = json.loads('{"name":"O\'Brien","tags":["vip","beta"]}')

"\"{\\\"id\\\":12345678901234567890,\\\"tags\\\":[\\\"a\\\",\\\"b\\\"]}\""

{
  "id": 12345678901234567890,
  "tags": [
    "a",
    "b"
  ]
}

Common errors and how to fix them

Related tools

• JSON Escape / Unescape
• JSON formatter
• json minifier
• json validator
• big integers in json
• json invalid escape