Re: Good html to text rendering

[Stefan Hagen <sh+gopher@ptrace.org>]

On 2018-08-13 14:33, Mike Gran wrote:
> On Sun, Aug 12, 2018 at 08:41:47PM +0100, Nuno Silva wrote:
> > On 2018-08-12 18:44 +0000, Mike Gran wrote:
> > > I was looking for a good html to text renderer to put in, but, the ones I've tried
> > > (like html2text in Python) are kind of boring and do a bad job with tables.
> > >
> > > Right now, the best one I've found is piping the text through lynx.
> > >
> > > What is the greatest html to text rendering chain?
> >
> > Do you have some example whose link you can share? Something with 
> > tables
> > and other items you want to come out right.
>
> I don't really have anything to show yet.  I had made a couple of
> simple test entries with the HTML I wanted to support, and then tried
> Python's text2html on them, which made valid Markdown but wasn't the
> best visually.
>
> I probably will just go with piping html through lynx for now. My only
> argument against that, which was pretty weak, wass that I didn't want
> the hassle of putting the lynx executable in my Docker container.

I was looking for a solution to bring content from random websites to 
gopher.
I ended up using readability. It extracts the content part of many 
websites.

https://github.com/cantino/ruby-readability

I brought hacker news to gopher here:
gopher://codevoid.de/1/hn

On some entries, you see a selector "text version". These are created 
via
readability.

I'm not sure if this is helping you, but I thought I throw it into the 
mix.
It's probably overkill. But it would allow you to create a gopher menu 
that
uses the same text as the HTML version.

Personally, I would suggest to do it the other way around. Create gopher 
and
proxy it to HTML. My gopher hole is also accessible via 
https://codevoid.de

Works great, looks great (I think) and it saves me from all the 
conversion hassle.

Best Regards,
Stefan