Re: [Bug-wget] Unexpected character on a downloaded page

[Angel Tsankov]

On 06/15/14 23:28, Ángel González wrote:
> On 14/06/14 20:31, Angel Tsankov wrote:
> > Why does wget 1.15 (and 1.12) insert  in several places in the copy
> > it makes of the following page:
> >
> > http://www.helloquizzy.com/results/helen-fisher-personality-type-test/?var_Explorer=1&var_Negotiator=1&var_Director=1&var_Builder=1&fromCGI=1
> Short answer: because that's what is at that page.
>
> Long answer: That page contains several non-breaking spaces (ASCII 160,
> U+00A0) which when encoded as UTF-8 result in the bytes C2 A0. If you
> read the page as if it was iso-8859, you will view instead the byte C2
> as the glyph Â.
>
> The page correctly states it's in utf-8:
> Content-Type: text/html; charset=utf-8
> so it should be read in utf-8 mode.
>
> (wget is doing nothing here, it's just receiving bytes and storing in
> the file as-is)

Indeed, the browser (Firefox 27.0.1) displays the original page in UTF-8 
and the downloaded page in Windows-1252 (which turned out to be the 
fallback encoding for pages that do not declare their encoding).  But if 
"wget is doing nothing here" why does the browser think that only the 
original page declares its encoding?


Regards,

Angel Tsankov