some problems with embeded dADL within cADL

Technical (archive)
1

Hello all!

I try to send some messages to Problem Report List to crs_prs@openehr.org as described on http://www.openehr.org/developer/t_PR.htm and
got failure notice - so I send this here. Most of this problems is around embeded dADL section within cADL part of ADL.

1. -

    C_QUANTITY sample in "Predefined Type Libraries" section of ADL revision 1.2 book is obsolete dADL syntax. Why?

2. -

    dADL serialiser for embeded cADL section "C_QUANTITY" makes valid dADL2 syntax, that can't be rereaded by
    validator. For example see file:
    $OPENEHR/archetype-dev/adl/archetypes/test/clinical_types/adl-test-entry.c_quantity.draft.adl
    ...
    clinical_quantity_attr_1 matches {
                        C_QUANTITY <
                                property = <"temperature">
                                list(1) = <
                                        units = <"C">
                                        magnitude = <|>=4.0|>
                                >
                                list(2) = <
                                        units = <"F">
                                        magnitude = <|>=40.0|>
                                >
                        >
                }
        ...
        see result after validate/serialise with utility "adl_validator" in file
        $OPENEHR/archetype-dev/adl/archetypes/test/clinical_types/adl-test-entry.c_quantity.draft.adl2 looks like:
        ...
        clinical_quantity_attr_1 matches {
                        C_QUANTITY <
                                property = <"temperature">
                                list = <
                                        ["1"] = <
                                                units = <"C">
                                                magnitude = <|>= 4.0|>
                                        >
                                        ["2"] = <
                                                units = <"F">
                                                magnitude = <|>= 40.0|>
                                        >
                                >
                        >
                }
       ...
        and this raise error when revalidate with message:
            -------- reparsing --------
            Archetype $OPENEHR/archetype-dev/adl/archetypes/test/clinical_types/adl-test-entry.c_quantity.draft.adl2 is NOT VALIDATED; reasons:
            line 40: invalid dADL section; error: line 26: parse error [last token = V_TYPE_IDENTIFIER]
            [last token = ERR_C_DOMAIN_TYPE]
            (Parse failed)

3. -
    classes C_STRING, C_DATE, INTERVAL has intercrosses with same classes in new EiffelStudio 5.5 workbench - this rise compile errors when adl_validator is rebulded

thanks,
Edgar Roze

2

Edgars Roze wrote:

Hello all!

I try to send some messages to Problem Report List to crs_prs@openehr.org as described on http://www.openehr.org/developer/t_PR.htm and
got failure notice - so I send this here.

We are currently implementing a web-based problem reporting facility...

Most of this problems is around embeded dADL section within cADL part of ADL.

1. -

   C_QUANTITY sample in "Predefined Type Libraries" section of ADL revision 1.2 book is obsolete dADL syntax. Why?

er...no good reason - that's an error in the document...CR-000125 has been raised to fix this.

2. -

   dADL serialiser for embeded cADL section "C_QUANTITY" makes valid dADL2 syntax, that can't be rereaded by
   validator. For example see file:
   $OPENEHR/archetype-dev/adl/archetypes/test/clinical_types/adl-test-entry.c_quantity.draft.adl

   ...
   clinical_quantity_attr_1 matches {
                       C_QUANTITY <
                               property = <"temperature">
                               list(1) = <
                                       units = <"C">
                                       magnitude = <|>=4.0|>
                               >
                               list(2) = <
                                       units = <"F">
                                       magnitude = <|>=40.0|>
                               >
                       >
               }
       ...
       see result after validate/serialise with utility "adl_validator" in file
       $OPENEHR/archetype-dev/adl/archetypes/test/clinical_types/adl-test-entry.c_quantity.draft.adl2 looks like:
       ...
       clinical_quantity_attr_1 matches {
                       C_QUANTITY <
                               property = <"temperature">
                               list = <
                                       ["1"] = <
                                               units = <"C">
                                               magnitude = <|>= 4.0|>
                                       >
                                       ["2"] = <
                                               units = <"F">
                                               magnitude = <|>= 40.0|>
                                       >
                               >
                       >
               }
      ...
       and this raise error when revalidate with message:
           -------- reparsing --------
           Archetype $OPENEHR/archetype-dev/adl/archetypes/test/clinical_types/adl-test-entry.c_quantity.draft.adl2 is NOT VALIDATED; reasons:
           line 40: invalid dADL section; error: line 26: parse error [last token = V_TYPE_IDENTIFIER]
           [last token = ERR_C_DOMAIN_TYPE]
           (Parse failed)

I've just done this with version 1.31 (build timestamp Oct 21 2004) - and it works fine - can you indicate which version you have of the work bench (use the "About" menu option)?

3. -
   classes C_STRING, C_DATE, INTERVAL has intercrosses with same classes in new EiffelStudio 5.5 workbench - this rise compile errors when adl_validator is rebulded

this is a problem of a class name clash with the inbuilt Eiffel classes, and is a top-priority problem for us to fix. If you email me I will send a workaround (which I won't post publicly, as we hope that the proper fix will remove the need for any workaround).

hope this helps,

- thomas beale