Allow DE to work without the meta.contentLanguages property
As Daniele,
I want the DE to optionally ignore the non-required meta.contentLanguages
property in the SDMX-JSON messages,
So that it can correctly display localised names (labels of concepts and codes) retrieved from back-end services not relying on the ESTAT NSIWS (e.g. Fusion Registry from Metadata Technology).
Requirements
For organisations such as UNICEF using the Fusion Registry of Metadata Technology, the DE visualisation pages only display identifiers, and names (labels of concepts and codes) are not shown at all, e.g.:
The reason is that the Fusion Registry implementation doesn't present the “meta.contentLanguages” property in SDMX-JSON data messages for indicating the available languages. MT is not planning to change this in the near future.
Technical background
The localised SDMX-JSON response has 2 ways of obtaining the localised elements:
-
through the singular form of the elements: it contains the first best language match according to the user’s preferred language choices expressed through the HTTP content negotiation (Accept-Language header parameter).
Example:
"name": "Frequency",
-
through the plural form of the elements: it contains all available language matches according to the user’s preferred language choices expressed through the HTTP content negotiation (Accept-Language header parameter).
Example:
"names": { "en": "Frequency", "fr": "Fréquence" },
In the DE, data requests are localised according to the current locale. The DE could thus trust the server and use the first best language match contained in the localised elements of the first singular form, that doesn't require the knowledge of the languages contained in the data message through the “meta.contentLanguages”
property.