# ADLParser ArchetypeInternalRef

**Category:** [Technical (archive)](https://discourse.openehr.org/c/technical-archive/156)
**Created:** 2014-03-21 10:43 UTC
**Views:** 1
**Replies:** 2
**URL:** https://discourse.openehr.org/t/adlparser-archetypeinternalref/15613

---

## Post #1 by @system

Hi,

I think I found something strange in the ADL-parser.

See following ADL:

`attribute3 matches {`
`use_node SECTION /items[at0001]`
`use_node SECTION /items[at0002]`
`}`
`attribute4 cardinality matches {0..*} matches {`
`use_node SECTION[at0006] /items[at0002]`
`use_node SECTION[at0005] /items[at0001]`
`}`

The nodes where pointed to, exist, they look like this

`items cardinality matches {0..*} matches {`
`SECTION[at0001] occurrences matches {0..1} matches {*}`
`SECTION[at0002] occurrences matches {1..2} matches {*}`
`}`

When I parse this, I come to following questions

At /attribute3, it only occurs once in the Archetype's->pathNodeMap.
I think that one entry is overwritten by the next, because the path's are the same
This is a faulty entry in the archetype.

But /attribute4 is parsed wrong I think.
It should occur twice in the pathNodeMap because the path's shoudl be different.
But they do not. And that is because the archetype_node_id is not taken into account.

There should, in my opinion, have been two paths
/attribute4[at0006]
/attribute4[at0007]

So if I am right, there is a bug in the code in ADL.jj which parses internal-refs, it should add the node_id's

Can someone please comment on this. Repairing may not be that hard

Thanks
Bert Verhees

---

## Post #2 by @yampeku

I think that could be related to the parser assuming that Internal References have no atxxxx code (which they can according to the specifications)

---

## Post #3 by @system

The NodeId is detected but it was forgotten to add it to the path, like is done in c_complex_object. It was taken into account that there can be a nodeId, and it is read also, but it is forgotten to add it to the path, for example, like is done in c_complex_object in adl.jj As I expected, easy to repair. The function archetype_internal_ref in adl.jj should be rewritten like this. I tested it. I see that I have the right to do so. So I think, I just do so, if someone complains, it is easy to restore the old version. ArchetypeInternalRef archetype_internal_ref(String path, CAttribute parent) : { String type; Interval<Integer> occurrences = new Interval<Integer>(1, 1); String target; String nodeID = null; } { <SYM_USE_NODE> type = type_identifier() [ nodeID = constraint_ref() ] [ occurrences = c_occurrences() ] { if(path == null) { path = "/"; } else { path += (nodeID == null ? "" : "[" + nodeID + "]"); } } target = absolute_path() { return new ArchetypeInternalRef(path, type, occurrences, nodeID, parent, target); } } Thanks Diego I see that

---


**Canonical:** https://discourse.openehr.org/t/adlparser-archetypeinternalref/15613
**Original content:** https://discourse.openehr.org/t/adlparser-archetypeinternalref/15613