DATA-INTO - ERROR RNQ0356

[RINALDI_V]

I have a JSON of this type, and I am unable to structure the data structure (DS) to process it.

[
{
"NCampi": "1"
},
{
"CAMPO0": "CODICE_ARTICOLO"
},
{
"TIPOCAMPO0": "1"
},
{
"CODICE_ARTICOLO": "001012001"
}
]

Message: The document for the DATA-INTO operation does not match the RPG variable (C G D F)

Thank you in advance for the answer.

[jonboy49]

I’m viewing this on my phone so I may be misreading it but…

This is the type of json I hate because it is an array of elements but each element has a different name. Unless you know all the possible names it is tough to do with Date-Into - and even then it would be messy.

You might be better off using the raw yajl apis

[Scott Klement]

This problem is caused by the fact that many people use languages where a variable doesn't need to be known at compile-time. For example, PHP, JavaScript or Python are interpreted languages... they can read the JSON and build variables with the names in the JSON on-the-fly. RPG, however, is a compiled language so it needs to know the variable names at compile-time.

You may be thinking that this would somehow make working with a JavaScript/PHP/Python (etc) program to be easier than RPG. But it's not, really, when you get into the actual coding. For example, after the JSON is parsed, what are you going to do with the data? If it is JavaScript, you'd probably do something like this:

Code: Select all

if (typeof myvar.NCapi != "undefined") {
  //handle NCampi
else if (typeof myvar.CAMPO0 != "undefined") {
  // handle CAMPO0
}
... etc ...

So even though you don't need to hard-code the list of field names in a data structure definition, you'll still need to hard code them in the logic of the program. UNLESS they are using the key names as variable values rather than field names -- but this isn't the typical structure for that sort of thing, so I'm not thinking that's what's happening in this case.

Assuming you're willing to hard-code the list of field names, you can do so by creating an array of data strucrures where all of the fields are optional fields.

Code: Select all

**free

dcl-ds *N psds;          
  num_arr Int(20) Pos(372);
end-ds;                  

dcl-s json varchar(500);
dcl-s i    int(10);

dcl-ds el qualified;
  num_NCampi          int(10);
  NCampi              char(1);
  num_CAMPO0          int(10);
  CAMPO0              varchar(30);
  num_TIPOCAMPO0      int(10);
  TIPOCAMPO0          char(1);
  num_CODICE_ARTICOLO int(10);
  CODICE_ARTICOLO     varchar(20);
end-ds;

dcl-ds myArr likeds(el) dim(100);

json = 
 '[{+
   "NCampi": "1"+
  },+
  {+
    "CAMPO0": "CODICE_ARTICOLO"+
  },+
  {+
    "TIPOCAMPO0": "1"+
  },+
  {+
    "CODICE_ARTICOLO": "001012001"+
  }]';

DATA-INTO myarr %data(json: 'case=convert countprefix=num_')
                %parser('YAJLINTO');

for-each el in %subarr(myArr:1:num_arr);

  if el.num_NCampi = 1;
      // handle NCampi
  elseif el.num_CAMPO0 = 1;
    // handle NCampi
  endif;        

endfor;

*inlr = *on;  

So you see that's not very different from what you'd have to do in another language like JavaScript. On the other hand, if the keys are really variable values rather than field names (i.e. associative arrays), your best bet would be to use the YAJL subprocedures/routines...