Michael Färber

> 🥳 nice, i gave it a try. with this jqjq changes i can get it to lex but parsing seems to bail out in the precedence climbing > code,...

@pkoppstein, thanks for offering your help. I will first await the merging of the current state of the manual, before augmenting it with more content. So let us discuss the...

Hi @wader, just as I was setting out to look into jqjq myself right now, I find that you have beaten me to it. :) Thanks for uncovering this bug....

Perhaps it is interesting to mention at this point that jaq's destructuring follows the same rules as its indexing, which diverge from jq by failing to index any null value....

Thanks for mentioning the jaq branch of jqjq. I found that quite a few things already work, using: jaq -n -L . 'include "jqjq"; eval($FILTER; {}; {}) This evaluates the...

Concerning the lexing speed: jaq -L . -n 'include "jqjq"; _builtins_src | lex' is indeed slower than jq right now (1.2 seconds in jaq vs. 0.4 seconds in jq). If...

Hi @wader, amazing work! Thanks so much for it. > Doh, yes you correct, i build without `--release`. Now i've updated to latest main and built with release target, much...

> Without any changes it seem expressions like `1+2` just output `1`, but if i add `... | debug` or `... | .` to both it seem to work. Some...

> Any plan to add `--args`, would make it possible to use the jqjq shell wrapper with jaq, `./jqjq --jq jaq ...`. That should actually not be that hard to...

Hi @gwpl, thanks a lot for the idea and your work on the sixel branch! Funnily, just today, before seeing your post, I thought about precisely the same functionality (showing...