[Edbrowse-dev] Possible tidy5 alternative

[Edbrowse-dev] Possible tidy5 alternative

[Adam Thompson]

[-- Attachment #1: Type: text/plain, Size: 303 bytes --]

Hi all,

Following my previous email, I managed to locate another html5 parsing library
from google namely:
Gumbo <https://github.com/google/gumbo-parser>

If I can get some code compiled to use it I'll evaluate it against the buggy
pages to see if it's any better than tidy5.

Cheers,
Adam.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]

Re: [Edbrowse-dev] Possible tidy5 alternative

[Chris Brannon]

Adam Thompson <arthompson1990@gmail.com> writes:

> Hi all,
>
> Following my previous email, I managed to locate another html5 parsing library
> from google namely:
> Gumbo <https://github.com/google/gumbo-parser>

I've also had libhubbub suggested to me.  It's actively being used in
another independent browser project (netsurf), so it could be promising.

-- Chris

Re: [Edbrowse-dev] Possible tidy5 alternative

[Adam Thompson]

[-- Attachment #1: Type: text/plain, Size: 1176 bytes --]

On Fri, Sep 11, 2015 at 02:08:58PM -0700, Chris Brannon wrote:
> Adam Thompson <arthompson1990@gmail.com> writes:
> 
> > Hi all,
> >
> > Following my previous email, I managed to locate another html5 parsing library
> > from google namely:
> > Gumbo <https://github.com/google/gumbo-parser>
> 
> I've also had libhubbub suggested to me.  It's actively being used in
> another independent browser project (netsurf), so it could be promising.

My issue with libhubbub is exactly the fact that it's tied into netsurf's
development, so much so that the Debian package for it was removed because it's
now packaged as part of netsurf. I appreciate the fact that being part of
another browser means it's battle-tested,
but after spidermonkey I'd rather have a library which is explicitly supposed
to be a library rather than something which can be used as a library but is
developed as part of another browser.

Cheers,
Adam.
ps: If none of the html options out there turns out to be of any help,
at least html5 has standardised the parsing algorithm so we can make our parser
algorithmically correct, and we have a more maintainable design out of all this.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]