Now that would be very handy.

I could also see a large number of variables requiring consideration.

Suresh,

From my support days, I recall some customers who would use UDO to read in the JSON, convert it to XML, and then create XMAPS to map it to a dynamic array.  If one of them happens to be on the forum, perhaps they will share what they did.

I tend to look to Python as a better way of dealing with a JSON, please see the following:

UOPY Example - Working with JSON | Rocket U2 | UniVerse & UniData (rocketsoftware.com)

Hope that helps

Hi Suresh, UDO functions allows you to parse a JSON, and then extract values from properties or arrays. We use UDO functions thoroughly in our MV web framework to produce and parse JSON.

  Hope this helps,

  Enrique Murphy

Thanks. We'll give that a try in a few days, and keep you posted on the progress.

Hi Enrique,

I am not familiar with UDO functions, would you be able to assist me with this please, examples or documentation please?

Daniel,

This is an old Blog post but it should help.

https://www.rocketsoftware.com/blogs/parse-json-u2-dynamic-objects

Also, the documentation can be found at doc.rocketsoftware.com

UDO is in a few manuals, but I always go to the latest Basic Extensions Manual.

thanks very much Mike, will have a look through that all.

Hi ,

I share two subr to handle json clob from basic to extract values. The inside comments are in french but you can translate it easily

IN.JS.XPATH  

Allows you to extract a value in a json by expressing its path such as key.key.key….

key := a property or object name  

key := an array name followed by the item number: key[x]

x := a positive integer

root can be an array [] with values, keypair or object 

Syntax: CALL IN.JS.XPATH(RESULT[out], JSON[in], XPATH[in], ERR[out])

RESULT will be the value ;  if one key of the xpath refer a array without [indice] the complete array is returned with @am/@vm/@sm .. regarding the nest-level of the array
JSON is the json object to search for.
- it can be in string json format
- it can be in UDO format, usefull to request many times on the save udo
XPATH is the path to search:  key.key[x].key   or  [x].key.key  or key[x].key.key.key

ERR is error return
ERR<1> = 0 no error
ERR<2> = 1; the requested xpath does not exist; RESULT contains 'empty'
ERR<1> = 1 error
ERR<2> the error message

see IN.JS.XPATH.SAMPLE attached

IN.JS.XPATH.LIST 

Allows you to list all the xpaths of a json.

Syntax: CALL IN.JS.XPATH.LIST(RESULT[out], ROOT[in], JSON[in])

RESULT: am'ed with each AM in the format: xpath key[x].key.key = value (type)
type := string, number, true, false, null
ROOT is either 'empty' or the word to use to say the first key
ROOT = root → RESULT<x> ⇒ root.key.key…
ROOT = '' → RESULT<x> ⇒ key.key…
JSON: is the json object,
- it can be in string json format
- it can be in UDO format

see IN.JS.XPATH.LIST.SAMPLE attached

i hope this help

Here are some of examples of how the UDO functions work.  Might assist in the understanding. 

https://github.com/RocketSoftware/multivalue-lab/tree/master/U2/Demos/UDO/JSON

We do this all the time with our Red Leaf web framework.  Based on that project, we have a series of API routines to extract JSON data into name-value pairs.  The problem with JSON (and XML) is that the nesting can go deeper than three levels.  So the API routines allow an X-Path kind of specifier to identify what to extract so anything can be extracted, even if it's deeper than attributes/values/subvalues.

Hi Kevin,

that sounds good, could you send me samples or anything that i can look at to see if i can use it pls?

Daniel, send me your email address and I'll send you the code and docs.

daniel.conradienz@gmail.com

Thanks very much Kevin