Localization Refactor (API & Blazor Client cross-concern)

[mluepkes]

Right now, all localization files on API and Blazor client will have to be maintained manually, which is tedious and error-prone. Following a discussion on Discord, starting here and here (sorry for the channel switching) and referencing issue #401 I have summarized the following points:

• The goal is to automatize the creation of localization resource files for API and Blazor client alike.
• Switch client and server over to Localization with .po files, so we can automatically extract localization strings with something like PoExtractor No similar tools could be found for .json or .resx. Furthermore, .po files are widely accepted across different software stacks for localization purposes. Numerous translation tools are available. Done
• In order to do that, all instances of IStringLocalizer localizer and the like have to be changed follow the conventions of OrchardCore meaning IStringLocalizer S Done
• If we want to display localized validation error messages on the client, we will have to rely on server side error messages for now, as Blazor does not support localized Data Annotations for now as per Microsoft Docs
• Basic translation of localization could be automated by using something like the Azure Cognitive Services Translator leveraging the corresponding Github Action

Topics like the feasibility of automatically running of PoExtractor via some kind of Github Action should be evaluated (see e.g. here)