APER decoding failure for SEQUENCE

[oaitamrane]

@mouse07410 , @mhanna123 I am using the version https://github.com/mouse07410/asn1c.git suggested by @mouse07410 to encode/decode APER PDUs. I have performed a basic test with a SEQUENCE with Extension Marker contains 2 MANDATORY ENUMS with Extension Marker and 2 OPTIONAL Parameters that are not Present. All the SEQUENCE Parameters are in the Extension Root. Encoding is successful, but decoding fails. When debugging I have located a problem in SEQUENCE_decode_aper() function in constr_SEQUENCE.c dealing withe padding. After decoding the Extension Marker of the SEQUENCE (1bit) and the 2 Optional Parameters Indicators (2bits) decoding procedure assumes padding of 5 bits before the Extension Marker of the 1st ENUM while there should be no padding in this situation. I think that this problem has been addressed by @mhanna123 but he did not solve it completely. I attach a basic code to test this. Thank you in advance RequestType.ZIP

[mhanna123]

@oaitamrane this is the code I am using right now, constr_SEQUENCE.c at line 1320:

     if(edx != (td->elements_count-1))
      {
         if(!edx && specs->roms_count && (padding > 0))
            ASN_DEBUG(">>>>> not getting padding of %d bits for element:%d out of %d (specs->roms_count:%d)", padding, edx, td->elements_count, specs->roms_count);
         else
            per_get_few_bits(pd, padding);
      }
      else //
      {
         if(specs->roms_count && (padding > 0))
            ASN_DEBUG(">>>>> not getting padding of %d bits for element:%d out of %d (specs->roms_count:%d)", padding, edx, td->elements_count, specs->roms_count);
         else
            per_get_few_bits(pd, padding);
      }

I didn't report it in the previous issue cause I honestly don't remember the context of that fix. Anyway, you can try it and see if it solves your problem

[oaitamrane]

@mhanna123 Unfortuantely your proposed patch did not solve the problem. Can you execute the small program I sent you.

[mhanna123]

@oaitamrane in constr_SEQUENCE.c, at line 1324 or in the code in my previous comment, change if(!edx && specs->roms_count && (padding > 0))

with if( (edx < specs->roms_count) && (padding > 0))

[mouse07410]

@oaitamrane and @mhanna123 can you post here sample data for Request-Type.asn? I'd like to check whether the current vlm_master branch resolves this problem.

[mouse07410]

Actually, my tests show that the current vlm_master branch resolves this issue. Please post here or re-open if you have additional info or problems.

These are my test-cases:

Request-Type.asn

-- **************************************************************
--
-- Common definitions
--
-- **************************************************************

Messages

DEFINITIONS AUTOMATIC TAGS ::= 

BEGIN

-- **************************************************************
--
-- Request Type
--
-- **************************************************************
RequestType ::= SEQUENCE {
    event           RequestTypeEvent,
    reportArea      RequestTypeReportArea,
    horizontalaccuracyCode  RequestTypeAccuracyCode     OPTIONAL,
    iE-Extensions       NULL    OPTIONAL,
    ...
}

RequestTypeEvent ::= ENUMERATED {
    stop-change-of-service-area,
    direct,
    change-of-service-area,
    stop-direct,
    ...,
    periodic,
    stop-periodic
}

RequestTypeReportArea ::= ENUMERATED {
    service-area,
    geographical-area,
    ...
}

RequestTypeAccuracyCode ::= INTEGER (0..127)

END

sample.xer:

<RequestType>
  <event><direct/></event>
  <reportArea><geographical-area/></reportArea>
  <horizontalaccuracyCode>31</horizontalaccuracyCode>
</RequestType>

sample2.xer:

<RequestType>
  <event><direct/></event>
  <reportArea><geographical-area/></reportArea>
  <iE-Extensions></iE-Extensions>
</RequestType>