Fix regression with parsing of scalar value documents #90
Conversation
With the ondemand parser, the document to value conversion fails with the error SCALAR_DOCUMENT_AS_VALUE if the document is a single null, string, bool, or number value.
Since we are now using a template type, this is no longer inferable by an IDE. Adding explicit types also reduces the number of `.value()` conversions required.
|
This solution seems like a bit of a hack, but it seems to be the way to have the least code changes in |
|
@tobil4sk How could it be less of a hack? My C++ isn't the best. I'm willing to dig in (or at least create an issue for the future) if you had any suggestions. |
|
I'm not sure there is another way of doing this at the moment, unless we just copy the function implementation to have one for documents and one for values, however, that is worse than using the template. I've tried to look a bit into what other bindings are trying to do, though not many have switched to the ondemand parser yet. cysimdjson has a PR open which has duplicate implementations for document and value: https://github.com/TeskaLabs/cysimdjson/pull/13/files#diff-083a29ee978c1780e06e7653801bd953dfb9e92ce0ee95155534b51428f355a9. This might be something that could be improved on |
|
Ahh I see. Well, I'll get this merged for now. I'll probably also update to 3.10.1 before cutting a new release. |
FourierTransformer
merged commit 789d08a
into
FourierTransformer:master
|
Thanks! |
This fixes a regression from 0f609dd where lua-simdjson switched to the ondemand parser. The ondemand::document type refuses to be convert to ondemand::value if the document is a single scalar value, e.g.
null,10,helloortrue.It would give this error:
I've also added relevant test cases.