Incomplete unicode escape in template literal gets parsed without errors

Hello
Is it intentional that incomplete unicode escape sequences in template literals cause no errors? because things like \u get parsed without errors:

var l = require('esprima')
console.log(l. parse('`\\u`').body[0].expression)

Furthermore, I came across a test case, in which this very idiosyncrasy has been put to actual test (and looks to have passed it).

I'm not sure whether this kind of behavior has been mentioned in the spec; if it is, then I'd be glad should you help me find it; but in case it's not, it would be great to have it expressly mentioned (somewhere.)

Start by reproducing the behavior with the l.parse('\\u') entry point and read test/fixtures/ES6/template-literals/literal-escape-sequences.source.js. Check the applicable ECMAScript rules to determine whether the incomplete escape should be accepted. Done means the behavior and expected outcome are established, with the relevant test or documentation reflecting that decision.