| 1 | {{{ |
| 2 | > now I just got back to the CMDI profiles once again, trying to |
| 3 | > incorporate some of the components or DCs the BAS and/or Tübingen |
| 4 | > profiles use to assure we at least have something in common although |
| 5 | > nobody knows how that part works yet:) Anyway, again I stumbled upon |
| 6 | > something: The component cmdi-access used in cmdi-COLLECTION is - I |
| 7 | > guess - one of those old "hand-made" components (in fact there are a |
| 8 | > lot of those old cmdi-something components) and for the element |
| 9 | > Availability, the cardinality is 1-1 in the component registry as View |
| 10 | > or XML, but 1-unbounded if I download the XSD for e.g. the |
| 11 | > SpeechCorpusProfile or the media-corpus-profile using this component. |
| 12 | Hi Hana, |
| 13 | |
| 14 | Good observation. In this case it is because the multilinguality flag of |
| 15 | Availability is true. This always makes an element repeatable, even if |
| 16 | MaxOccurrences is set to a fixed value (like 1). I added a note about |
| 17 | this to the relevant FAQ: |
| 18 | |
| 19 | http://www.clarin.eu/faq/3490 |
| 20 | |
| 21 | > The same thing applies for the cmdi-totalsize component's Element |
| 22 | > SizeUnit. In this case however the cardinality 1-1 seems to be more |
| 23 | > adequate, i.e. the one displayed in the component registry, not the |
| 24 | > one in the schema. |
| 25 | This is the same case. |
| 26 | |
| 27 | > For all but the Validation element in |
| 28 | > cmdi-validation, the component registry says 0-1 and the schema says |
| 29 | > 0-unbounded. I suppose in the end the schema counts, but I was just |
| 30 | > wondering if someone has checked all these cmdi-something components |
| 31 | > for these old cardinality errors (if that's even the problem here, I |
| 32 | > can't really remember if this is the same behaviour). I haven't, and |
| 33 | > these might be extremely rare cases or not, I don't know. If these old |
| 34 | > components still need some correcting maybe someone should have a look |
| 35 | > at it before we start creating CMDI instances and/or CMDI generating |
| 36 | > stylesheets at the centres? |
| 37 | The most serious result of enabling the multilinguality option is indeed |
| 38 | that a field might be repeated, even if the profile creator fixed it to |
| 39 | a max nr. |
| 40 | |
| 41 | @Twan: maybe we can grey out the MaxCardinality in the component editor |
| 42 | when someone selects multilingual? That would alert the modeller |
| 43 | immediately about the effects. |
| 44 | |
| 45 | }}} |
| 46 | |
| 47 | related question, same source: |
| 48 | |
| 49 | {{{ |
| 50 | > another thing I was wondering about is the fact that some components |
| 51 | > (OriginLocation in the cmdi-COLLECTION component, the TimeCoverage |
| 52 | > component in the cmdi-generalinfo and the BirthCountry component of |
| 53 | > cmdi-actor) don't seem to have component ids, and can't be found when |
| 54 | > searching the registry. Are they somehow "private" components that |
| 55 | > shouldn't be reused on their own or has the component id simply been |
| 56 | > lost somewhere along the way? I wanted to reuse the BirthCountry |
| 57 | > component, but *not* the entire "parent" Actor component, but I don't |
| 58 | > know if this should be possible (I suppose I could save the cmdi-actor |
| 59 | > as our new component and the adapt this one, but this includes a bit |
| 60 | > more work than I wanted, since I already have a component...). And the |
| 61 | > lack of component ids might be problematic for future |
| 62 | > harvesting/searching? |
| 63 | Hi Hana, |
| 64 | |
| 65 | these are embedded components; it is indeed possible to include a |
| 66 | component that does not exists as a separate entity in another |
| 67 | component. If you want to use such a component nevertheless, you will |
| 68 | need to make a copy of the containing component, remove the useless |
| 69 | elements/components and store it as a new component. |
| 70 | |
| 71 | }}} |