Ada code generated does not compile with SEQUENCE SIZE(XXX) OF CHOICE

[jmmery-dev]

Hi,

the XER code generated for Ada does not compile:

generated/bug_xer.adb:657:66: error: expected type "XString" defined at adaasn1rtl-encoding-xer.ads:10 ( TEST_CHOICE_XER_Encode_aux(val.Data(i1), null, bs, ret); )
/generated/bug_xer.adb:657:66: error: found type universal access

procedure LIST_OF_CHOICE_XER_Encode_aux(val : LIST_OF_CHOICE; xmlTag : adaasn1rtl.encoding.xer.XString; bs : in out adaasn1rtl.encoding.xer.CharStream; ret : OUT adaasn1rtl.ASN1_RESULT)
is
    i1:Integer;
begin
    ret := LIST_OF_CHOICE_IsConstraintValid(val);
    if ret.success then
        -- SEQUENCE OF Encode
        adaasn1rtl.encoding.xer.Xer_EncodeComplexElementStart(bs, xmlTag, ret, 0);
        if ret.Success then
        	i1 := 1;
        	while i1<= 3 and ret.Success loop
                        -- Here Is the Problem
        		TEST_CHOICE_XER_Encode_aux(val.Data(i1), null, bs, ret);
        		i1 := i1 + 1;
        	end loop;
            if ret.Success then
                adaasn1rtl.encoding.xer.Xer_EncodeComplexElementEnd(bs, xmlTag, ret, 0);
        		ret.ErrorCode := (if ret.Success then 0 else ERR_XER_ENCODE_LIST_OF_CHOICE);
            end if;
        end if;
    end if;
end LIST_OF_CHOICE_XER_Encode_aux;

when doing tests with sequence with named attributes no problem

my version of asn1scc is build from sources : tag 4.6.0.9-17
my gnat is GNAT 13.3.0

here is my test project

bugtrackasn1scc_4.tar.gz

thx for you works

Regards

Jean-Michel

[jmmery-dev]

Hi, are there some news on these two bugs ? #352 and #353 ?

Jean-Michel

[usr3-1415]

Hi Jean-Michel,

Thank you for your message. Both issues are noted and will be reviewed in the coming period. Since they concern XER encoding, which is not among the most commonly used features, their review might take a bit longer.

[jmmery-dev]

ok thanks, that was only to know if they were taken into account. Thanks for you great job ! Regards Le mar. 4 nov. 2025 à 10:21, usr3-1415 ***@***.***> a écrit :

…

*usr3-1415* left a comment (esa/asn1scc#353) <#353 (comment)> Hi Jean-Michel, Thank you for your message. Both issues are noted and will be reviewed in the coming period. Since they concern XER encoding, which is not among the most commonly used features, their review might take a bit longer. — Reply to this email directly, view it on GitHub <#353 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BP3JO55ZJO2ZLO6DS23LDKT33BV23AVCNFSM6AAAAACIWVPJEOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTIOBUHAZTKOBYHA> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[maxime-esa]

In the meantime, a suggestion, you may have a look at the template that generates this piece of code.
In your installation folder of asn1scc you have a file named xer_a.stg

This file contains the rules for generating the XER encoding and decoding functions. From the error you pointed out, I think these two lines are the one responsible for emitting the null that cause a compilation issue.
You may try to fix it directly here.

   458:call_base_type_func_encode(p, soXmlTag, sFuncName) ::= "<sFuncName>_aux(<p>, <if(soXmlTag)><soXmlTag><else>null<endif>, bs, ret);"
   459:call_base_type_func_decode(p, soXmlTag, sFuncName) ::= "<sFuncName>_aux(<p>, <if(soXmlTag)><soXmlTag><else>null<endif>, bs, ret);"

However, I am not sure in the specific case if this will be sufficient (I assume that the null works in some cases).

[maxime-esa]

Since the parameter is a string, replacing with empty quotes seems to work. Try this modification:

call_base_type_func_encode(p, soXmlTag, sFuncName) ::= <<<sFuncName>_aux(<p>, <if(soXmlTag)><soXmlTag><else>""<endif>, bs, ret);>>
call_base_type_func_decode(p, soXmlTag, sFuncName) ::= <<<sFuncName>_aux(<p>, <if(soXmlTag)><soXmlTag><else>""<endif>, bs, ret);>>

With this change your application works, and the output is:

ubuntu@taste:~/work/asn1/bug/bugtrackasn1scc_4$ ./obj/main
LIST_OF_CHOICE Size statique =  216 octets
Programme principal

Could you confirm that this is the expected output ?