Chcp 65001 and a .bat file

batch-filecmdcommand linewindows-vista

I have a problem with chcp 65001 command in Windows shell.

I need to generate a list of files in a folder.
So I ran cmd.exe, typed

cd folder
dir /B /O:N > list_of_files.txt

It worked, but I had a problem with special, non-ASCII characters which are in some file names.
So I added
chcp 65001

Everything worked, but when I put these commands into a .bat file, the script doesn't work.

So

cd folder
chcp 65001
dir /B /O:N > list_of_files.txt

doesn't generate the list.

and

cd folder
chcp 65001 && dir /B /O:N > list_of_files.txt

as well as

cd folder
chcp 65001 > nul && dir /B /O:N > list_of_files.txt

generates the list, but with the default encoding :/.

Everything works in cmd.exe, but not in .bat files.

I've read the topic: stackoverflow.com/questions/2182568/batch-script-is-not-executed-if-chcp-was-called, but it didn't help.

EDIT:
I partially solved my problem, changing chcp 65001 to chcp 1250 because all characters were in this encoding. But actually this doesn't answer the question.

Best Answer

Use cmd /U. See http://ss64.com/nt/cmd.html:

Most common text files are ANSI, use these switches when you need to convert the character set. These options will affect piping or redirecting to a file:

  • /A Output ANSI characters
  • /U Output UNICODE characters (UCS-2 Little Endian)

Here's my attempt (launch it under cmd /A, of course):

@ECHO OFF >NUL
SETLOCAL EnableExtensions

:: create a UNICODE file with Byte Order Mark using `wmic` 
chcp 852 >NUL
>list_of_files.txt wmic os get localdatetime

:: store a line with BOM to a variable
:: although FINDSTR does not support UTF-16 files
:: it will read first three bytes at least
for /F "delims=" %%G in ('
    findstr "^" list_of_files.txt
  ') do set "UTF8BOM=%%G"

:: write BOM only* to a file (* echo writes hexadecimal value FFFE0D0A)
:: the `<NUL set /p =text` trick does not work: chokes down leading `FF`  
>list_of_files.txt echo(%UTF8BOM:~0,2%

chcp 65001 >NUL
:: add CRLF in  Unicode (hexadecimal 0D000A00)
>>list_of_files.txt cmd /U /C echo(

:: add result of `dir /B /O:N` in Unicode 
>>list_of_files.txt cmd /U /C dir /B /O:N

:: check the result: still invalid first line, see output
type list_of_files.txt
chcp 852 >NUL

Output. Still invalid first line (that hexadecimal 0D0A), sorry; use another method to get pure Utf-8 byte order mark:

==>cmd /A /C D:\bat\SO\UTF8BOM32182619.bat
਍
cpANSI_OoCcSsUu.txt
cpANSI_ÖöÇ窺Üü.txt
escrzyaie.txt
ěščřžýáíé.txt
list_of_files.txt

==>
Related Topic