«

In article <>, rvtol+
says...
> You are including a character >127 in a posting that has no Content-Type
>

The HTML page I generated contains "Content-type: text/html;
charset=iso-8859-1"

Yohan N. Leder schreef:

> The HTML page I generated contains "Content-type: text/html;
> charset=iso-8859-1"

As you have been told before, an as you can check on the
wikipedia.org-page about that charset, ISO-8859-1 does not contain a
Euro-sign.

--
Affijn, Ruud

"Gewoon is een tijger."

In article <>, rvtol+
says...
> Yohan N. Leder schreef:
>
> > The HTML page I generated contains "Content-type: text/html;
> > charset=iso-8859-1"
>
> As you have been told before, an as you can check on the
> wikipedia.org-page about that charset, ISO-8859-1 does not contain a
> Euro-sign.
>
>

Agreed, but the fact is that with some browser (a known IE bug as stated
by someone in alt.html), more than just send an unrepresentable
character it corrupt the full posted content.

Because of this and after some discuss here in two threads (including
this current one) and a one in alt.html, I've added some trivial
checkings on the multipart/form-data content found at the arrival in
STDIN. Thus, even if the bug stays (corrupted content due to the
presence of outside charset char + uncheckd checkbox), I'm not trying to
parse a malformed multipart content. Result of these basics checkings
are visible in the sentence starting with "Integrity :" after submission
of form here : <http://yohannl.tripod.com/cgi-bin/form2dump.pl>.

So, now, I have to choose to use the ISO-8859-15 charset or go to UTF-8
(sounds a nightmare with Perl 5.00503 I have to keep in mind ; eben if
I'm developing under 5.6 and 5.... to be able to accept this euro
sign.

What do you think about ISO-8859-15 ? What major drawbacks if any ?

Yohan N. Leder schreef:

> Result of these basics
> checkings are visible in the sentence starting with "Integrity :"
> after submission of form here :
> <http://yohannl.tripod.com/cgi-bin/form2dump.pl>.

That page isn't valid html.

<http://validator.w3.org/check?uri=ht...om/cgi-bin/for
m2dump.pl&verbose=1&doctype=Inline>

--
Affijn, Ruud

"Gewoon is een tijger."

In article <>, rvtol+
says...
> That page isn't valid html.
>
> <http://validator.w3.org/check?uri=ht...om/cgi-bin/for
> m2dump.pl&verbose=1&doctype=Inline>
>

OK, solved (added DOCTYPE and html/head/body tags) : however, it doesn't
change anything about form submission behavior.

Yohan N. Leder schreef:
> rvtol:

>> That page isn't valid html.
>>
>>
<http://validator.w3.org/check?uri=ht...om/cgi-bin/for
>> m2dump.pl&verbose=1&doctype=Inline>
>
> OK, solved (added DOCTYPE and html/head/body tags) : however, it
> doesn't change anything about form submission behavior.

The response page should use "&euro;" to disambiguate the character
0xA4, because in the active encoding of the page (ISO-8859-1), the
character 0xA4 is the currency-character "&curren;" and not specifically
the Euro-sign.

See http://en.wikipedia.org/wiki/ISO_8859-15 about the 8 differences
between it and Latin1.

--
Affijn, Ruud

"Gewoon is een tijger."

In article <>, rvtol+
says...
> The response page should use "&euro;" to disambiguate the character
> 0xA4, because in the active encoding of the page (ISO-8859-1), the
> character 0xA4 is the currency-character "&curren;" and not specifically
> the Euro-sign.
>
> See http://en.wikipedia.org/wiki/ISO_8859-15 about the 8 differences
> between it and Latin1.
>

OK, thanks I'll read-it asap. Nevertheless, what do you think ? Awaiting
to go to UTF-8 using a more recent Perl interpreter, do you think it's
better to use ISO-8859-1 and convert euro sign as html entity on receive
(ie. after reading of STDIN), or use ISO-8859-15 instead (without html
entity conversion) ?

Yohan N. Leder schreef:
> rvtol:

>> The response page should use "&euro;" to disambiguate the character
>> 0xA4, because in the active encoding of the page (ISO-8859-1), the
>> character 0xA4 is the currency-character "&curren;" and not
>> specifically the Euro-sign.
>>
>> See http://en.wikipedia.org/wiki/ISO_8859-15 about the 8 differences
>> between it and Latin1.
>
> OK, thanks I'll read-it asap. Nevertheless, what do you think ?
> Awaiting to go to UTF-8 using a more recent Perl interpreter, do you
> think it's better to use ISO-8859-1 and convert euro sign as html
> entity on receive (ie. after reading of STDIN), or use ISO-8859-15
> instead (without html entity conversion) ?

In this specific (border) case, I would certainly convert from 0xA4 to
"&euro;" or even to "EUR&nbsp;". I have never seen semantic usage of the
"&curren;"; it is mainly used for ASCII-art (and of course in texts and
tables about HTML).

--
Affijn, Ruud

"Gewoon is een tijger."