Translation
Analyzer-internal diagnostics raised when code that parsed successfully cannot be lowered into the tool’s internal representation. Parsing recovers the syntax; translation is the next stage, which converts that syntax tree into the normalized form the analyses actually operate on — and when a construct cannot be expressed in that form, the unit is dropped before any check runs.
A failure here reflects a gap in the analyzer rather than a fault in the program: a language feature or construct the front end accepted but the lowering stage does not yet model. The practical consequence is the same as a parsing failure — that piece of code goes unanalysed.