I currently have a js library that will parse code and return either the html (render) or the AST (abstract syntax tree). The AST action per node is modifiable. I'd like use it in the following way:
For a single node type in the AST, I want to process it (rather than just rendering it to Html) For every other node type, I want it to render Html, which the library does by default. I'm curious about how to do the integration with Elm here. Should I use ports, managed effect module (like Time module), or a native module (like Http module)? The guide discourages managed effects. If that's the solution here, I'm open to doing it. Writing native modules seem like reaching into private interfaces and will cause me trouble down the line when new versions of elm come out. But in all the Html and Markdown modules I've seen, it just uses Native modules, presumably to avoid pitfall #2 below. Hence, I think I need to use Ports. The easy part is sending a String of code into the port for parsing. I see two pitfalls with this approach: 1. When I get back the single node type in the AST, I'd need to pass it back to Elm as a JSON, and write a decoder to parse it. That seems like a lot of work (Decoders take me a long time to write). 2. For AST nodes rendered as Html, if I don't pass them back into Elm, I assume that Elm wouldn't be able to do shadow dom diffing on it for rendering. If I do pass them back into Elm, I'd have to write an Elm Decoder from JSON into Html Msg, and that seems like a lot of work. So what's the best way to tackle this? Is parsing a side effect? Would it make sense if this was a managed effect? Wil -- You received this message because you are subscribed to the Google Groups "Elm Discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to elm-discuss+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
