Differences between the toCubeShape used in Cube Creator vs b59-cube

[tpluscode]

As discussed, in all cases the sh:in gets separated between the sh:or alternative shapes.

This would make it difficult for consumers to analyse the cube contents

In the cases when there are multiple sh:in lists, they should be merged and appear in the property shape itself.

Dimension with ""^^:cube:Undefined

--- before.ttl	2024-05-20 10:31:34
+++ after.ttl	2024-05-20 10:32:16
@@ -2,17 +2,19 @@
     sh:or (
               [
                   sh:datatype xsd:string ;
+                  sh:in (
+                            "incomplete series"
+                        ) ;
               ]
               [
                   sh:datatype <https://cube.link/Undefined> ;
+                  sh:in (
+                            ""^^<https://cube.link/Undefined>
+                        ) ;
               ]
           ) ;
-    sh:in (
-              "incomplete series"
-              ""^^<https://cube.link/Undefined>
-          ) ;
     sh:maxCount 1 ;
     sh:minCount 1 ;
     sh:nodeKind sh:Literal ;
     sh:path rdfs:comment ;
 ]

Numeric dimension with ""^^cube:Undefined literals

--- before.ttl	2024-05-20 10:40:23
+++ after.ttl	2024-05-20 10:39:30
@@ -2,15 +2,18 @@
     sh:or (
               [
                   sh:datatype <https://cube.link/Undefined> ;
+                  sh:in (
+                            ""^^<https://cube.link/Undefined>
+                        ) ;
               ]
               [
+                  sh:maxInclusive 500 ;
+                  sh:minInclusive "1"^^xsd:int ;
                   sh:datatype xsd:int ;
               ]
           ) ;
-    sh:maxInclusive "500"^^xsd:int ;
-    sh:minInclusive "1"^^xsd:int ;
     sh:maxCount 1 ;
     sh:minCount 1 ;
     sh:nodeKind sh:Literal ;
     sh:path <https://environment.ld.admin.ch/foen/ubd/28/dimension/limitvalue> ;
 ]

Dimension with mixed IRIs and Literals

--- before.ttl	2024-05-20 10:37:26
+++ after.ttl	2024-05-20 10:37:41
@@ -1,19 +1,28 @@
 [
-    sh:in (
-              <https://cube.link/Undefined>
-              unit:MicroGM-PER-M3
-              "count"
-              "ppb"
-              "mg/m3"
-              "1/cm3"
-              "ng/m3"
-              "mg/(m²·d)"
-              "µg/(m²·d)"
-              "µg CH4/m3"
-              "ppm·h"
+    sh:or (
+              [
+                  sh:datatype xsd:string ;
+                  sh:in (
+                            "count"
+                            "ppb"
+                            "mg/m3"
+                            "1/cm3"
+                            "ng/m3"
+                            "mg/(m²·d)"
+                            "µg/(m²·d)"
+                            "µg CH4/m3"
+                            "ppm·h"
+                        ) ;
+              ]
+              [
+                  sh:in (
+                            <https://cube.link/Undefined>
+                            unit:MicroGM-PER-M3
+                        ) ;
+              ]
           ) ;
+    sh:nodeKind sh:IRIOrLiteral ;
     sh:maxCount 1 ;
     sh:minCount 1 ;
-    sh:nodeKind sh:IRI ;
     sh:path <https://environment.ld.admin.ch/foen/ubd/28/unit> ;
 ]

[tpluscode]

Here's what we could expect respectively

1. Same as before
2. Alternative between numeric range and other values: https://s.zazuko.com/7yhdfz
3. Top level sh:in and alternative between datatype and node kind https://s.zazuko.com/3m3mUg9

What do you think?