Please add Jonathan

[rudolph-miller]

Jonathan is aJSON encoder and decoder.

https://github.com/Rudolph-Miller/jonathan

[quicklisp]

Is there any way the benefits the Jonathan can be incorporated into one of the many existing JSON libraries?

[rudolph-miller]

It seems to be the fastest. (I think the fastest library to parse JSON in Common Lisp was jsown.)
Benchmarks are here.
https://github.com/Rudolph-Miller/jonathan#to-json
https://github.com/Rudolph-Miller/jonathan#parse

[eudoxia0]

Suggestion: Yason has these macros like with-object, with-array, encode-array-element etc for writing stuff to JSON. It would be cool if Jonathan had these so we don't have to take the error-prone route of writing (write-char #\{), etc.

[rudolph-miller]

Yeah, I thought of that kind of macro to ease encoding original structure or class, and I will write some.

[quicklisp]

I am asking if Jonathan can be incorporated completely into something else. Is that possible?

[rudolph-miller]

No, it's totally different from other projects from the viewpoint of the codes and it makes faster than other projects.

[guicho271828]

it would catch more attentions if you add some keywords like "compatible with jsown" to the description.

[quicklisp]

If it's compatible with jsown, why not merge the projects so jsown is faster and users of jsown get the benefits without changing their code?

[rudolph-miller]

Ah, this is just optional.
I don't like the style of jsown object, and operations of jsown object.
I just wanted to handle plist mainly, and then I added compatible layer.
Moreover, it's not really compatible with jsown:parse.jsown:parse can raise the error with incomplete json string, and jonathan:parse behaves like junk-allowed of parse-integer.

[quicklisp]

To be clear: I will not reject Jonathan (or any other project) for covering similar ground as another project. However, I would like to see projects work together whenever possible. If it is not possible, that is unfortunate but not an insurmountable obstacle to inclusion in Quicklisp.

[rudolph-miller]

It's really reasonable, and I agree with you of course.
If I could make jsown faster and make jsown be able to handle plist and alist with some patches (in other words, with not replacement of whole codes), I would do so.