UI incorrect display of type: integer format: int64 as rounded int values in response body

[rgunn]

swagger.json correctly shows the data types and format but the values displayed in the response body are rounded int values.

[webron]

Rounded int values?

[rgunn]

swagger.json properties are correctly typed

"someKey":{"type":"integer","format":"int64"},
"anotherKey":{"type":"integer","format":"int64"},

The raw JSON returned from my REST service is correct but the values displayed in the swagger response body:

"someKey": 1253261109769874400,
"anotherKey": 212081872531357440,

Should be:

"someKey": 1253261109769874438,
"anotherKey": 212081872531357452,

Which is exactly the same result as running JavaScript's
parseInt("1253261109769874438")
/*
1253261109769874400
*/

I am new to swagger and have produced swagger 2 spec for approx 70 REST API endpoints and this is the only issue I have. Javascript only supports 53 bit integers and the UI does not process these values properly although I assumed it would according to the "format" type ?

[scott-orleman]

What news of this issue? Having the same problem.

[webron]

I believe this is a javascript limitation. Not sure there's much we can do about it, but ideas are welcome.

[code-ape]

I personally would like to see this resolved if possible. A project I've worked on has had to be rearchitectured because of this limitation of the values presented in the UI being rounded. We had been using 64 bit hashes.

[bobvanluijt]

Not sure if this is still an issue, but for other projects I've used this package to deal with large integers: https://www.npmjs.com/package/big-integer

[webron]

@bobvanluijt thanks for that, it's still an issue, yes. Would you consider submitting a PR that would add support for it?

[heldersepu]

This is caused by JSON.parse you can see it in response-body.jsx#L25

Try this on the browser console:

JSON.parse("{\"MyInt64\":1253261109769874438,\"MyData\":\"1253261109769874438\"}")

{MyInt64: 1253261109769874400, MyData: "1253261109769874438"}

[heldersepu]

Perhaps we should have a config option to bypass the prettification, and always render response as PlainText. Or a way to alternate between the RAW & Parsed views

Found this project: https://github.com/josdejong/lossless-json
looks like a nice candidate but the owner is very clear about the performance:
(Spoiler: lossless-json is much slower than native)

[shockey]

I think a Pretty/Raw switcher is the right answer here, as it offers a more generalized approach