need to validate site before makes it live

Gerard Beekmans gerard at
Fri Oct 24 09:13:30 PDT 2003

On Fri, 2003-10-24 at 09:55, Jeroen Coumans wrote:
> Or if that takes too long you could restrict it to the dynamically 
> generated files.

Maybe all of the site is better just to catch all errors in there are
any. I'm running into some problems though.

First of all, the online books can't be validated, things don't work
quite well (  not being defined is a common error) so that needs to
be ignored by find. Then there are issues with the copyright symbol you
put in the files. It generates errors like:

./news/hints/2003/08.html:95: error: Input is not proper UTF-8, indicate
encoding !
                <li>© 1998-2003  Gerard Beekmans.</li>

I'll see what I can do about encoding.

Then the 'old' directory needs to be ignored, no big deal there.

Then there are errors like:

./alfs/archive-top.html:40: error: Premature end of data in tag div line
./alfs/archive-top.html:40: error: Premature end of data in tag html
line 2
./alfs/archive-bottom.html:2: error: xmlParseStartTag: invalid element
</div> <!-- end content -->

also happens in news/archive*.html and others.

./blfs/faq.html:100: error: xmlParseEntityRef: expecting ';'
s a <a

./blfs/faq.html:101: error: xmlParseEntityRef: expecting ';'
ot <a

archive-top.html and -bottom.html don't seem to be used anywhere in the
site but it should be fixed else xmllint craps out, if it can be fixed.
If not, adding to find's prune statement is an option (pruning takes
place anyways so it's no biggie, just let me know if you can fix it).

Jeroen, maybe you want to run xmllint yourself on all the files to see
what you can do with it. I copied the dir into my home
directory and removed some things like lfs/download lfs/view
blfs/download blfs/view blfs/edguide (easier that way than writing out
the prune regex).

Gerard Beekmans

/* Linux Consultant --- OSDN / DevChannel *
 * Technical Writer --- CheapBytes        */

/* If Linux doesn't have the solution, you have the wrong problem */

More information about the website mailing list