Hi,
I can't reproduce that this works with standalone pygmentize: Pygments also
generates an error token for the snippet you posted.
Insofar, if it is valid TCL, this is a bug in Pygments.
cheers,
Georg
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.18 (GNU/Linux)
iEYEARECAAYFAk6MXycACgkQN9GcIYhpnLAixACdF4O4Apk58s7NlJpsR+PT/rsK
GxEAn3Jr2TPYIjSI60T9VNqmmQu3CTGG
=17Ts
-----END PGP SIGNATURE-----
On 10/05/11 16:02, bulletmark wrote:
> I was running pygmentize to my terminal and after your comment I have
> realized that some of the output is underlined so I guess that means
> pygmentize believes it is invalid syntax?
Yep, this is how the default style shows Pygments "Error" tokens.
> Either way, why does sphinx just fail to render the output at all if the
> highlighter trips over some syntax? Is there any message where a user can
> see the reason why his code is not highlighting?
It's because then you can keep the default highlight language even if you have
a shell/text/whatever sample inbetween.
> I spent ages on this problem thinking I had an environment problem. The
> code section is not even being line numbered at all. More particularly, can
> a user tell sphinx to ignore syntax errors and render (as best it can)
> anyway? I imagine there will always be different language dialects which a
> general purpose highlighter will not completely accommodate.
Yep, we're planning an option to customize this in 1.1.
Georg
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.18 (GNU/Linux)
iEYEARECAAYFAk6MbXgACgkQN9GcIYhpnLCwswCeNecHcgiZiBPKj5lcp35AKhUJ
FcAAoJwLXB2+t7L6ghsVtB5Wx6+g93lK
=Z0cv
-----END PGP SIGNATURE-----
Yep, we're planning an option to customize this in 1.1.