@RikedyP woah thats cool, time to solve this codegolf.stackexchange.com/questions/248070/… in like 3 bytes lmao

ok wait... its post on april 1, am i getting fooled here?

i hope not, this is genuinely a cool feature...

seems kinda niche tho

You are 1 day late, otherwise i think would be a great answer

so the blog post is a prank??

@AidenChow Yeah sorry - April fools!

@finooiigee Also ⎕JSON can take a matrix input similar to ⎕XML, so you can do text → ⎕XML → ⎕JSON → namespace

See Import → Import as Matrix in help.dyalog.com/latest/#Language/System%20Functions/json.htm

Thanks!

I think ⎕HTML could be a nice addition. One issue is that most HTML parsers are much too forgiving :)

@finooiigee In principle, we already have that. It should be possible to use HTMLRenderer to convert dirty HTML into strict/XHTML, and get that back for processing with ⎕XML.

@finooiigee I have a utility that attempts to convert HTML to XHTML. It's in no way comprehensive but seems to handle reasonably structured HTML. The result can then be passed to ⎕XML. There are other online and downloadable HTML to XHTML conversion tools.

I keep extending the utility as I find new HTML anomalies that it doesn't currently handle.

Last time I find someone made an epub file (which is essentially a zipped folder of XHTMLs) I cannot open, and using validation tool I realized they didn't escape & to &, likely the result of poor conversion script.

@RikedyP I'm intrigued by the idea, which never occurred to me, of using ⎕JSON to replace recursion when creating a namespace tree DOM from the result of ⎕XML. I assume you envision manipulating the matrix and then using ⎕JSON twice, first to convert to JSON, and then a second time to convert the JSON string to a namespace, correct?

@Adám yeah I know I just can't think of when that would be useful