Re: doctype declaration

On Monday 22 July 2002 19.04, PsychoTekk .de wrote:
> > Not at all. What we should do is provide the error details (maybe
> > by removing the email address from the %W. and using %w%W in the
> > mailto:) in a separate %tag, then the sysadmin making a form just
> > uses:
> >
> > <HIDDEN NAME='errordetails' value="%W" />

What is wrong with simply using the existing % formats for such forms?

> but what about the meta tags and the charset info?

Different question.

> i think it would be okay to define a default charset
> in squid.conf (because most err docs are okay in iso-8859-1
> and don't contain a meta tag), but this default should be
> overridden if a meta tag specifies a different charset.

Scanning for the META tag is not too difficult. Needs to be done in
errorBuildReply.

> the other solution would be something similar to tcn, a .var file
> that contains the different charsets for all the err doc languages
> so that squid knows which charset belongs to which language.
>
> thinking of this, what about utf-8?

not yet... lets try to address the charset issue first.

Once Squid knows the charset adding utf-8 rendration of the error
strings is not a big deal (all error strings can be assumed to be in
iso-8859-1, as there is no UTF-8 in the HTTP protocol yet..).

> about the %tag for <HIDDEN NAME='errordetails' value="%W" />:
> it's just parsed once so %tag cannot contain another tag
> that has to be parsed.
> thus i think it'd be better to have strInfo that is handled by
> errorpage.c
> 1) (if strInfp != '') %W = "&body=", strInfo;
> 2) %tag = "<HIDDEN NAME='errordetails' value=\"", strInfo, "\"
> />:";

Having % codes that may contain % codes is not a problem, but I do not
see any problem with HTML forms.. If you do HTML forms you most
likely want to have it split in different fields for different data
so the existing % codes is just fine and there is not really any need
for a new %W code..

Regards
Henrik
Received on Sat Aug 10 2002 - 17:36:47 MDT