Sebastian Thiel

The "using a fuzzer to generate test-cases for invariants" definitely goes into the tool-belt :). Thanks for pointing it out.

Should we close this PR :)?

Thanks for sharing and for explaining the situation so well. There is already a way to pass option to the conversion engine and I'd wonder if a flag can be...

I think this is closely related to #16 - maybe you can join the people who start talking to `pulldown-cmark` to make not degenerating information a priority. Otherwise there is...

This means you can [try for yourself](https://github.com/Byron/pulldown-cmark-to-cmark/issues/16#issuecomment-668886290) how pulldown-cmark maintains information about backticks. If it does indeed not maintain any, then there is nothing we can do here.

> (BTW, why the result of cargo run --example stupicat -- sample.md | pulldown-cmark is broken? Where is the last ````? I don't know either, but I don't entirely know...

Fantastic. With this it is certainly possible to make the requested change. I think it should be quite straightforward.

Good point! I think it could be done by optionally passing the input buffer as part of the [Options](https://docs.rs/pulldown-cmark-to-cmark/6.0.0/pulldown_cmark_to_cmark/fn.cmark_with_options.html). There using the range information should be a breeze.

That sounds like a post-process entirely unrelated to this crate, or am I missing something? Maybe knowing more about this would help to understand why you would discard [a proposal](https://github.com/Byron/pulldown-cmark-to-cmark/issues/20#issuecomment-762226030)...

@yumetodo > I'm concern that the index information in the range is valid when the passed event stream is modified from the original. > pulldown-cmark maintainer's suggestion code requests source...