From: "Wolfgang Nádasi-Donner" <ed.odanow@...>
Date: 2007-10-25T08:26:27+09:00
Subject: Re: Hopefully Complete List of Possible Encoding Specifications - Existing Ones

Yukihiro Matsumoto schrieb:
> Wolfgang N��dasi-Donner <ed.odanow@wonado.de> writes:
> |
> |Yukihiro Matsumoto schrieb:
> |> Wolfgang N��dasi-Donner <ed.odanow@wonado.de> writes:
> |> |There is only one file specific encoding identification. If a file
> |> |starts with the Byte sequence 0xEF 0xBB 0xBF it will be identified
> |> |as encoded in utf-8 and the the bytes will be ignored.
> |> Currently, I am not going to add magic BOM support.  Not for 1.9.1 at
> |> least.
> |
> |I don't understand. In the moment, at least on Windows (where I make my tests) 
> |the BOM will be recognized and used for identifying a source as utf-8 encoded. 
> |When I don't use any command line option or magic coding comment, but the file 
> |will start with 0xEF 0xBB 0xBF, these bytes are ignored and the String methods 
> |will work fine on utf-8 encoded strings. Do you want to remove this feature? - 
> |This may give problems for Windows envirinments, because the tools working there 
> |on basis of these bytes.
> 
> You meant programs.  Sorry, I meant data files.  open(path) does not
> read BOM at the head, parser does.

Oh - the whole list is produced for program files. I'm testing actually only 
programs by combining command line options, BOM/noBOM, and magic coding 
comments. When reading files I haven't any success with encoding directives now.

Wolfgang N��dasi-Donner