You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now, the reference parser uses the repeat combinator for Attributes, which yields a Success even when it succeeds 0 times. This has the interesting consequence of Messages and Terms parsing just fine up until a broken Attribute is encountered.
The text was updated successfully, but these errors were encountered:
We talked about this in the today's meeting. We decided broken Attributes should not reject the whole Message (nor Term) they're defined in.
Attributes can be checked for existence by static analysis tools and/or by the bindings on the runtime. Producing a valid Message even if one of its attributes is broken is more in line with Fluent's comprehensive error recovery strategy. It will also make it easier to edit Fluent files; starting a new attribute will not break the entire message.
Right now, the reference parser uses the
repeat
combinator forAttributes
, which yields aSuccess
even when it succeeds 0 times. This has the interesting consequence ofMessages
andTerms
parsing just fine up until a brokenAttribute
is encountered.The text was updated successfully, but these errors were encountered: