In previous implementation, a UnicodeDecodeError exception will be
raised if multi-byte encoded characters are printed by the subprocess.
As an example, the following command will fail in an en_US.UTF-8
environment because wget quotes its saving destination with '‘'(0xE2
0x80 0x98), while just the first byte is provided for decoding:
devtool add recipe http://example.com/source.tar.xz
The patch fixes the issue by avoiding such kind of incomplete decoding.
Signed-off-by: Jiajie Hu <jiajie.hu@intel.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
import subprocess
import logging
import re
+import codecs
logger = logging.getLogger('devtool')
cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT, **options
)
+ reader = codecs.getreader('utf-8')(process.stdout)
buf = ''
while True:
- out = process.stdout.read(1)
- out = out.decode('utf-8')
+ out = reader.read(1, 1)
if out:
sys.stdout.write(out)
sys.stdout.flush()