Hi Nilesh, On Fri, 2023-11-17 at 16:45 +0530, Nilesh Patra wrote: > On Fri, Nov 17, 2023 at 06:18:04PM +0800, Maytham Alsudany wrote: > > golang-github-go-logfmt-logfmt doesn't have this Handler type that's being > > used, > > and humanlog already uses go-logfmt as much as it can > > > > I've made an issue upstream at [3] regarding this and I've patched out the > > Handler type completely from humanlog for now. > > > > Would you like me to close this RFS bug and the ITP? > > If this is needed for some actual functionality in humanlog, I will upload > kr-logfmt > which would otherwise block your work.
There's only one thing, a Go interface, in humanlog that uses kr-logfmt
functionality, but it is possible to patch out (as lazygit doesn't use it).
humanlog/handler.go[4]:
import (
"github.com/humanlogio/humanlog/internal/pkg/config"
"github.com/kr/logfmt"
)
// Handler can recognize it's log lines, parse them and prettify them.
type Handler interface {
CanHandle(line []byte) bool
Prettify(skipUnchanged bool) []byte
logfmt.Handler
}
This is the only place kr-logfmt is used, and humanlog's Handler interface is
not used anywhere else within humanlog.
> From your mail, it is not clear to me whether or not it is needed.
> Please let me know.
Yes, it is needed, but I've let humanlog upstream and the lazygit packaging
effort know about your concerns.
I see two options:
1. Patch out kr-logfmt out of humanlog, removing the single Handler interface
that uses it, and when a package needs the Handler interface from humanlog,
we revist the issue.
2. Upload kr-logfmt and package humanlog as usual.
> > [1]: https://bugs.debian.org/1055740
> > [2]:
> > https://salsa.debian.org/go-team/packages/golang-github-humanlogio-humanlog/
> > [3]: https://github.com/humanlogio/humanlog/issues/71
[4]:
https://github.com/humanlogio/humanlog/blob/eceedbf5df8383202e34d3aa4cb3d8dc182b3600/handler.go
Kind regards,
Maytham
signature.asc
Description: This is a digitally signed message part
