mapstructure VS TinyGo

Maybe https://github.com/mitchellh/mapstructure can do what you want? It has some options for Remainder Values and Omit Empty

I've tried using mapstructure to then marshal the map fields into a struct which mostly works (it struggles with times and custom time types which requires a workaround for each case), but this doesn't feel very idiomatic and requires two passes at marshaling.

Although some orders may have more records maybe a superset can be indentified that you can actually create a struct of it and after gathereing first all values into a map then convert it to a struct maybe using a library like https://github.com/mitchellh/mapstructure . this way you can at least isolate the non structured data only on the data extraction part and the rest of your application can work with well formed structs.

Alternatively, you could try https://github.com/mitchellh/mapstructure if you don't know what your incoming structure is

I'm doing it by validating a map[string]any first then putting it into a structure using mapstructure. It covers most use-cases and offers the most flexibility, at the expense of a bit of performance.

That said, it is possible to do this with JSON using something like https://github.com/tidwall/gjson or if you are fine with the switch statement but don't want to marshal and unmarshal again: https://github.com/mitchellh/mapstructure

https://github.com/mitchellh/mapstructure because we have JSON documents which contain rugged arrays ;-)

I've been using mapstructure.Decode to great effect, but currently can't figure out why a given mapping doesn't work. I'd appreciate it if someone could point out wtf I'm doing wrong or at least in the right direction:

What do you mean by "decode"? I've used https://github.com/mitchellh/mapstructure but that doesn't quite look like what you're doing.

Reading through the documentation, you will understand why this plugin is called "tiny," i.e., the SDK uses the TinyGo compiler instead of the official Go compiler. You can read more about why this is the case on the SDK\'s overview page, but the TLDR version is that the Go compiler can only produce Wasm binaries that run in the browser.

Not only you can fit Go into a kernel, there is at least two products that do so.

TamaGo, used to write the firmware used in USB armory.

https://www.withsecure.com/en/solutions/innovative-security-...

TinyGo, which even has official Arduino and ARM support, and is sponsored by Google

https://tinygo.org/

Ah but that isn't proper Go! Well neither is the C code that is allowed to be used in typical kernel code, almost nothing from ISO C standard library is available, and usually plenty of compiler specific language extensions are used instead.

Reminds me of https://tinygo.org/ - a project that brings Golang to embedded devices, browser (wasm) contexts. Do you converge or diverge from that project?

You should also consider TinyGo. It can compile Go for the Pico, and is starting to get good device support.

Thankfully some folks completly ignored whatever the rest of the world thinks system programming is all about and created:

- TinyGo (https://tinygo.org/), which is acknowledged by people in the industry[0][1]

- TamaGo unikernel on USB Armory secure key (https://www.withsecure.com/de/solutions/innovative-security-...)

And then there is the question if writing compilers, assemblers, linkers is systems programming or not.

[0]-https://www.cnx-software.com/2019/08/28/tinygo-go-compiler-f...

[1]-https://twitter.com/ArmSoftwareDev/status/131680481331796787...

Have you seen TinyGo? In the case of embedded system I would probably still chose C over Rust if the system didn't support dynamic memory allocation, and most embedded systems do not.

>I really hate how for microcontrollers the only two choices are either C++ or Micropython

There's TinyGo as well. https://tinygo.org/

Gist link fixed, thanks. Compared to TinyGo, Go with GOOS=wasip1 will probably generate larger artifacts (at least, for now). This is because it bundles the entire Go runtime. The benefit is that it fully supports goroutine scheduling and non-blocking I/O. TinyGo (I believe) still uses a custom asyncify pass and does not support non-blocking I/O nor basic WASI networking (e.g. https://github.com/tinygo-org/tinygo/pull/2748 never landed, but GOOS=wasip1 supports it).