Converting text and detecting encoding

[Flix]

Hello.

What I want to do is simple: correctly reading a text file whose encoding is
not known (it can be Ascii,UTF7,UTF8 or Unicode).

I'm thinking of something like that:

1) Read the text as Ascii:

string text="";
System.Text.Enc oding encoding=System .Text.Encoding. ASCII;
using (StreamReader sr=new StreamReader(fi lePath,encoding ))
{
text=sr.ReadToE nd();
}

2)Implement some kind of static methods like the following:

public static System.Text.Enc oding GetEncodingFrom Text(string text)
{
[...]
}

3)Convert the string "text" into the correct encoding.
I got no idea on how to implement points 2 and 3.
Any suggestion is welcome.

[Barry Kelly]

"Flix" <wr***@newsgrou p.comwrote:

What I want to do is simple: correctly reading a text file whose encoding is
not known (it can be Ascii,UTF7,UTF8 or Unicode).

It's not really that simple. Text can be UTF-16, in little-endian or
big-endian, without a BOM (byte order marker), for example. Check out
the IsTextUnicode() Win32 API - the functionality in Windows essentially
uses heuristics and guesses.

Check out Encoding.GetPre amble() in the docs for other possible clues.

-- Barry

--
http://barrkel.blogspot.com/

[Flix]

"Barry Kelly" <ba***********@ gmail.comha scritto nel messaggio
news:mg******** *************** *********@4ax.c om...

"Flix" <wr***@newsgrou p.comwrote:

>What I want to do is simple: correctly reading a text file whose encoding
is
not known (it can be Ascii,UTF7,UTF8 or Unicode).

It's not really that simple. Text can be UTF-16, in little-endian or
big-endian, without a BOM (byte order marker), for example. Check out
the IsTextUnicode() Win32 API - the functionality in Windows essentially
uses heuristics and guesses.

Check out Encoding.GetPre amble() in the docs for other possible clues.

-- Barry

Thank you for your reply.

[Jon Skeet [C# MVP]]

Flix <wr***@newsgrou p.comwrote:

What I want to do is simple: correctly reading a text file whose encoding is
not known (it can be Ascii,UTF7,UTF8 or Unicode).

That's not simple - it's impossible. Every UTF7 file is a valid ASCII
file, for instance.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too