Re: JSON un-named arrays

Hi Greg,

I don't know, if it helps, but you can walk the whole JSON document with SQL without problems.

select *
from json_table(
'{...insert.your.json.here...}',
'lax $'
columns (
records integer path '$.records',
fields varchar(1000) path '$.fields',
nested path 'lax $.data[*]'
columns (
trano integer path '$[0]',
ref varchar(50) path '$[1]',
duedate decimal(8, 0) path '$[2]',
name varchar(50) path '$[3]',
amount decimal(10, 2) path '$[4]'
)
)
) as json_doc;

As you can see, the named parts are referenced by ".name", whereas the "unnamed" JSON elements have no named reference, but only the "[index]". The nested path reference is the JSON equivalent of an table JOIN.

Of course - if the "data" part of the JSON is dynamic, and you have to detect the columns be the "fields" element, it can get a but complicated.

But even then, you can first split the "fields" list into its elements, an then construct a dynamic SQL to extract the "data" parts.

So from my POV - SQL is often the easier solution - especially because you can test it with Run SQL Scripts until everything is "perfect".

HTH and kind regards
Daniel

Am 17.07.2023 um 17:31 schrieb Greg Wilburn <gwilburn@xxxxxxxxxxxxxxxxxxxxxxx>:

Typically, I use DATA-INTO and the YAJLINTO parser... But I'm not sure if that's possible with the JSON below. The array "data" contains an un-named array of values.

Any suggestions on how handle reading this?
Do I need to use the YAJL "tree-based" method?

{
"records": 4,
"fields": "trano,ref,duedate,name,amount",
"data": [
[
1,
"40061433",
20230719,
"Mudpie",
187.2
],
[
2,
"40061349",
20230719,
"kiddywampus",
156.21
],
[
3,
"40062677",
20230721,
"Chase",
70.8
],
[
4,
"40061703",
20230723,
"Stamford Toys",
110.7
]
]
}
[Logo]<https://www.totalbizfulfillment.com/> Greg Wilburn
Director of IT
301.895.3792 ext. 1231
301.895.3895 direct
gwilburn@xxxxxxxxxxxxxxxxxxxxxxx<mailto:gwilburn@xxxxxxxxxxxxxxxxxxxxxxx>
1 Corporate Dr
Grantsville, MD 21536
www.totalbizfulfillment.com<http://www.totalbizfulfillment.com>