Re: lynx-dev another instance when automatic space insertion is not wanted

[Leonid Pauzner]

14-Apr-2003 12:16 Henry Nelson wrote:

> Anyway, I've just come across another case when it's a bad idea to blindly
> insert a space.  As you all know harvesting for e-mail addresses has become
> downright vicious these days to the point that you don't want to put your
> e-mail anywhere in a Web page.  Unfortunately, sometimes you HAVE to put
> it in a page, so to get a modicum of protection, you might encode it.
> For example, you might encode "mailto:address@hidden"; as:
> <a href="
> &#109;&#97;&#105;&#108;&#116;&#111;:&#115;&#111;&#109;&#101;&#111;
> &#110;&#101;@&#115;&#111;&#109;&#101;&#119;&#104;&#101;&#114;&#101;
> &#46;&#99;&#111;&#109;
">>Someone</a>
> (two line breaks).  Well, that's fine if you're using MSIE or some such
> beast, but Lynx will turn that into "mailto:someo address@hidden .com".  Note
> the spaces after the "o" and the "e".  Lynx will attempt to send a letter
> to three people, "someo" "address@hidden" and ".com", if you hit return on

Uhh! This is a bug.
Lynx treats newlines differently in #fragment and other parts of URL.
Apparently, LYLegitimizeHREF() search for '#' as a start of #fragment,
and this is before expanding entities with TRANSLATE_AND_UNESCAPE...


> that link.  The only way to not break it in Lynx is to concatenate the line.

> Ah ..., just FYI.  No time here to do anything about it.

> __Henry

> ; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden



; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to address@hidden