Just tried that. No effect.

if you are using testdata still, then maybe clearing your db might help.. Next time I run into it I'll try to make sure I know why and create an issue on the inspect github. I've seen more people run into it here.

because it is not literal data but the template form, which fulcro replaces with :root/people (comp/get-initial-state PersonList) - see http://book.fulcrologic.com/#_template_mode

makes sense, thanks!

Perhaps you need to provide your own row component if you need more info than the formatter gets

Well according to the docstrings, that info is supposed to be available. So I wonder if I did something wrong.

According to the code you only get the report instance and value https://github.com/fulcrologic/fulcro-rad/blob/develop/src/main/com/fulcrologic/rad/report.cljc#L620

could be a regression. It’s a lot of spinning wheels to keep track of

all things in demo should work, so if something is not working pls open issue on demo. thx

look through demo for a usage of the pattern, see if it works, and if so, let me know if the docstring is just wrong.

It isn't used in the demo, but it is in the docstring of ro/field-formatters

My usecase is that I'd like to have a dropdown on the report. With the functionality described in the docstring, that would be possible without defining a separate row component.

dropdown on the overall Report (shown in the demo), or on a Report Row? I would suggest you use the source. The rendering is not terribly complex, and it is intended that you extend it as needed. A dropdown on a row might be harder if you want the options to somehow be filled from a server query…for that you’ll probably want to use hooks/floating roots.

A dropdown in a row cell, with pretty much the technique described in the ro/field-formatters docstring. Making the dropdown and having the current value pre-selected is easy, but since, inside a function called by the field formatting functionality, I have no way to know which row I am on, it is impossible to use the dropdown to actually change the value

ah, I see…so we have a function signature problem

ugh, yeah, that function signature should be different…and that is a breaking change

I guess I could invent a new name and deprecate that one

cell-formatter?

or, I could add additional arguments to the current one, though that is kinda ugly….`(fn [report-instance value row-props attr] …)`

value in that case is superfluous, since it is in row-props, but since js won’t care about arity at runtime, it would probably just work.

I’ll do the latter for the moment…will push a snapshot in a moment @U2APCNHCN

Awesome! 😄 Yes, cell-formatter sounds like a reasonable name

0.0.14-alpha-SNAPSHOT on clojars

same old name, more args

👍

I also fixed the attribute level override…it wasn’t being honored