ASN.1 Compiler Possible Improvement: Integer Constraint Check

[zhouxt1]

Hi,
I am testing the asn.1 part of the srsRAN. I am aware of that the compiler is an internal tool as metioned in #18. So I am only testing the unpack function of ul_dcch_msg_s. And during my testing I found the following issue:

First it seems that the parser does not check integer constraints. For example, in the ASN.1 definitions there is,

OffsetValue-r17::=         SEQUENCE {
    offsetValue-r17            INTEGER (-20000.. 20000),
    shift7dot5kHz-r17          BOOLEAN
}

But, given an input 20 88 13 f5 3d 8c db a2 80 66 80 20 01 e0 5a 01 a0 0a 86 00 66 87 29 7a 95 4c 3c e5 10 1b 00 01 00 00 1b 1b 1b 1b 1b 1b 40 09 20 2d f1 91 b6 55 e2 7e e4 to the ul-dcch-message unpack, we can obtain a result message that will have such an incorrect field (in JSON)

"UL-DCCH-Message": {
     "message": {
       "c1": {
         "rrcResumeComplete": {
...
           "offsetToDefault-r17": {
                       "offsetlist": [
                             {
                                "offsetValue-r17": -30240,
                                "shift7dot5kHz-r17": false
                             } ]  }
...
} } } }

We can see the offsetValue-r17 exceeds the correct value range. Is it a design choice? Or is there any problem with such output?