Контекст менеджер по регистрации, эквивалентной contextmanager => выход на стандартный вывод

Question

библиотека Я использую принимает функцию так:

@contextlib.contextmanager 
def stdout_cm(): 
    yield sys.stdout 

Теперь вместо foo(stdout_cm), как я могу отправить через logging, например, что tee s в несколько обработчиков? (Стандартный вывод и StringIO)

Попытка:

from sys import stdout, stderr 
from cStringIO import StringIO 
from logging import getLogger, basicConfig, StreamHandler 
from subprocess import check_call 

basicConfig(format='%(asctime)s %(name)-12s %(levelname)-8s %(message)s', 
      datefmt='%m-%d %H:%M', level='INFO') 

log = getLogger(__name__) 
console = StreamHandler(StringIO()) 

log.addHandler(console) 
log.addHandler(StreamHandler(stdout)) 

class LoggerContext(object): 
    def __init__(self, logger, level=None, close=True): 
     self.logger = logger 
     self.level = level 
     self.close = close 

    def __call__(self): 
     return self 

    def __enter__(self): 
     if self.level is not None: 
      self.old_level = self.logger.level 
      self.logger.setLevel(self.level) 

    def __exit__(self, et, ev, tb): 
     if not self.close: 
      return 
     for handler in self.logger.handlers: 
      handler.close() 

Упрощенная функция:

def foo(f): 
    with f() as fh: 
     check_call(["printf '\n\n' | wc -l"], shell=True, stderr=fh, stdout=fh) 

foo(LoggerContext(log)) #, level='WARN') 


print console.stream.getvalue() 

Answer 1

Прелюдия:

import os 
import logging 
import subprocess 

from io import IOBase 
from sys import stdout 
from select import select 
from threading import Thread 
from time import sleep 

log = logging.getLogger(__name__) 

Класс (на основе https://stackoverflow.com/a/4838875):

class StreamLogger(IOBase): 
    _run = None 

    def __init__(self, logger_obj, level): 
     super(StreamLogger, self).__init__() 
     self.logger_obj = logger_obj 
     self.level = level 
     self.pipe = os.pipe() 
     self.thread = Thread(target=self._flusher) 
     self.thread.start() 

    def __call__(self): return self 

    def _flusher(self): 
     self._run = True 
     buf = b'' 
     while self._run: 
      for fh in select([self.pipe[0]], [], [], 1)[0]: 
       buf += os.read(fh, 1024) 
       while b'\n' in buf: 
        data, buf = buf.split(b'\n', 1) 
        self.write(data.decode()) 
      sleep(1) 
     self._run = None 

    def write(self, data): return self.logger_obj.log(self.level, data) 
    def fileno(self): return self.pipe[1] 

    def close(self): 
     if self._run: 
      self._run = False 
      while self._run is not None: 
       sleep(1) 
      os.close(self.pipe[0]) 
      os.close(self.pipe[1]) 

Использование:

with StreamLogger(log, logging.INFO) as out, StreamLogger(log, 
                  logging.ERROR) as err: 
    subprocess.Popen('ls 1>&2', stderr=err, shell=True)