topic Re: Parse JSON with “empty key” in Talend Studio

Parse JSON with “empty key”

Anonymous — Sat, 16 Nov 2024 00:04:53 GMT

Hi Team,

We are having a scenario in which we have a JSON with “empty key”. We are unable to parse data for the corresponding key

Sample data as below.

{

"firstName": "John",

"lastName" : "doe",

"age" : 26,

"address" : {

"streetAddress": "naist street",

"city" : "Nara",

"postalCode" : "630-0192"

"phone.Numbers": [

{

"type" : "iPhone",

"number": "0123-4567-8888",

"" : "abcd",

{

"type" : "home",

"number": "0123-4567-8910",

"" : "pqr",

}

]

}

We are using JSON query for the same to parse in tFileInputJSON.

Please help if anyone came across with this scenario replace battery iPhone

Regards,

davillan

Re: Parse JSON with “empty key”

gjeremy1617088143 — Tue, 08 Jun 2021 06:47:52 GMT

Hi, your json is not valid :

after "abcd" and "pqr" you have comma , you have to delete them.

I can parse it :

I use a tfileinputraw to read the json file as a string (it create an object type) then i use a tconvertType to convert it as a string an then a textractjsonfield with read by JsonPath option:

Loop Jsonpathquery : "$"

Mapping column : json query :

firstname : "firstName"

type: "['phone.Numbers'][*].type"

empty : "['phone.Numbers'][*].['']"

Send me love and kudos

Re: Parse JSON with “empty key”

gjeremy1617088143 — Tue, 08 Jun 2021 07:44:12 GMT

to get all the values :

i read a first time the json with these query :

ex:

LoopJsonpath query : "$"

column : json query:

firstName : "firstName"

phone : "$.['phone.Numbers'][*]."

then i use a tflowtoiterate to stock the value in globalmap and iterate each phone infos of the contact.

and i read again with textractjsonfield the phone string :

ex:

LoopJsonpath query : "$.[*]"

column : json query:

type : "type"

number : "number"

empty : "['']"

and in a tjavarow i add the field i stock before in tflowtoiterate like firstName

ex : if i iterate row3 in my job :

output_row.firstName = ((String)globalMap.get("row3.firstName"));

output_row.type = input_row.type;

output_row.number = input_row.number;

output_row.empty = input_row.empty;