[emacs-w3m:11319] Re: Failure to render latin1-encoded page properly.

--=-=-=
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable

Cc: [email protected]
(Note: the w3m-dev list is not open to the public.)

=C5=A0t=C4=9Bp=C3=A1n N=C4=9Bmec wrote:
> I can confirm it works now, but there is still a small glitch: emacs-w3m
> seems to eat the space following the bullets, i.e. instead of "word1
> <bullet> word2" it displays "word1 <bullet>word2" (observed on the same
> URL as the original bug). Could that be fixed as well?

Unfortunately we seem to have nothing to do[1] for it in
emacs-w3m since it's w3m's deed:

$ echo 'word1 &#149; word2'| w3m -halfdump -T text/html
word1 &#149;word2
<internal>
</internal>

This one does the same:

$ echo 'word1        &#149;        word2'| w3m -halfdump -T text/html

Firefox, Google Chrome, Konqueror, and IE8 display those examples
as "word1 <bullet> word2".  But the w3m solo doesn't display the
<bullet> symbol for "&#149;"; the result is "word1  word2".

An html file containing those examples is attached.  To try
halfdump, do:

$ cat test.html| w3m -halfdump -T text/html

I would highly appreciate the w3m team coping with this minor
problem.

Regards,

[1] Though we can replace whitespace after "&#149;" with "&nbsp;"
 in an html source as a workaround before passing it to w3m, I'm
 not sure it is the right thing.

--=-=-=
Content-Type: application/octet-stream
Content-Disposition: attachment; filename=test.html
Content-Transfer-Encoding: base64

PGh0bWw+CiAgPGJvZHk+CiAgICB3b3JkMSAmIzE0OTsgd29yZDI8YnI+CiAgICB3b3JkMSAgICAg
ICAgJiMxNDk7ICAgICAgICB3b3JkMjxicj4KICAgIHdvcmQxICZidWxsOyB3b3JkMjxicj4KICAg
IHdvcmQxICAgICAgICAmYnVsbDsgICAgICAgIHdvcmQyPGJyPgogIDwvYm9keT4KPC9odG1sPgo=
--=-=-=--