It would be useful to have a partially written block of text parsed as much as possible and provide information that can be used for autocomplete purposes. For my specific purposes I'm writing a query language that would want to be able to see name= and provide suggestions for the values of names that it sees are possible in the set of values to be searched.
It seems like there might be a way to do it between the tracing api and error recovery api(since the string will probably be only partially complete so invalid in most cases). Any pointers on how to start putting together a proof of concept for this?
It would be useful to have a partially written block of text parsed as much as possible and provide information that can be used for autocomplete purposes. For my specific purposes I'm writing a query language that would want to be able to see
name=
and provide suggestions for the values of names that it sees are possible in the set of values to be searched.It seems like there might be a way to do it between the tracing api and error recovery api(since the string will probably be only partially complete so invalid in most cases). Any pointers on how to start putting together a proof of concept for this?