Can Ecore::EString: be used in an Xtext grammar?

Question

I have generated an Xtext grammar from an ecore model and the grammar that I got is the following:

generate myDsl "http://www.xtext.org/example/mydsl16/MyDsl"
import "http://www.eclipse.org/emf/2002/Ecore" as ecore

WorldMap returns WorldMap:
    'WorldMap'
    '{'
        'WorldMap' WorldMap=EString
        'countrycorona'  countrycorona+=CountryCorona ( "," countrycorona+=CountryCorona)*  
    '}';


CountryCorona returns CountryCorona:
    'CountryCorona'
    '{'
        'CountryCorona' CountryCorona=EString
        'population' population=EInt
        'continent' continent=EString
        'infectedPopulation' infectedPopulation=EInt
        'zonelabel' zonelabel=zonelabelcolour
    
    '}';

EString returns ecore::EString:
    STRING | ID;

EInt returns ecore::EInt:
    '-'? INT;

enum zonelabelcolour returns zonelabelcolour:
                red = 'red' | orange = 'orange' | yellow = 'yellow' | green = 'green' | white = 'white'; 

My issue is that when I try write a model for this grammar, I usually use CTRL Space, to get the suggestion for the following token. However after I write WorldMap { WorldMap, I get

no viable alternative at input ''

Furthermore, I do not get a suggestion that the next token should be an EString. Refering to Xtext documentation I should be able to use this EString, but apparently I can't. Do you know what might be the issue?

Thank you!

Answer 1

Xtext content assist does not do a deep analysis of datatype rules. Thus you have to implement the content assist rule in the proposal provider yourself e.g.

public class MyDslProposalProvider extends AbstractMyDslProposalProvider {
    
    @Override
    public void complete_EString(EObject model, RuleCall ruleCall, ContentAssistContext context,
            ICompletionProposalAcceptor acceptor) {
        acceptor.accept(createCompletionProposal("NiceID", context));
        acceptor.accept(createCompletionProposal("\"NiceString\"", context));
    }
}