Allow way for hoooks like pre_format_html_hook to know the content-type or other headers before deciding how to rewrite HTML (or non-HTML)

[Moult]

Visit https://au.hardfought.org/userdata/M/Moult/nethack/dumplog/1734863710.nh.txt and you'll see the <-- You were here text appear like this: <- You were here.

[Moult]

Interesting, inversely if you visit https://github.com/rkd77/elinks/issues/356 (i.e. this bugreport) I see this:

[rkd77]

Cannot reproduce it. I can see Level 51: <- You were here. without entities.

[Moult]

I think this is because I have? HTML rewriting with Python. It assumes the content type is HTML and escapes it. How do I check for content type in hooks before rewriting? Dion Moult Sent from Proton Mail Android

…

-------- Original Message --------

On 9/1/25 3:22 am, rkd77 wrote: Cannot reproduce it. I can see Level 51: <- You were here. without entities. — Reply to this email directly, [view it on GitHub](#356 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/AAAVR3V4CC6ZNU5VAKJHB6T2JVGD3AVCNFSM6AAAAABUYXCNPWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKNZYGA4DONJSGM). You are receiving this because you authored the thread.Message ID: ***@***.***>

[rkd77]

AFAIK, there is no such feature. There was an idea to read headers as API call, and set arbitrary headers as another API call, but no code yet.

[Moult]

OK I added this as a temporary workaround:

if url.endswith("txt"):
    return html

However what about this inverse issue? #356 (comment)

[rkd77]

Yes, libdom converted & to & and old elinks renderer converted concatenated generated & and lt; as <.
Added workaround. Check now.