better-exceptions 可以在python异常发生时显示更多信息,帮助我们更好的发现问题.

简介

介绍这个库前,我们先看一个简单的例子(example.py),代码如下:

# -*- encoding: utf-8 -*-

def get_student_infos(logs):
student_infos = []
for log in logs:
name, catgory, grade = log.split(‘ ‘)
student_infos.append({
‘name’: name,
‘catgory’: catgory,
‘grade’: grade,

})
return student_infos

if __name__ == ‘__main__’:
exam_logs = [
‘zhangsan math 60’,
‘lisi english 80’,
‘wangwu chinese 90’,
‘qianliu music’
]
get_student_infos(exam_logs)

运行输出:

Traceback (most recent call last):
File “example.py”, line 23, in <module>
get_student_infos(exam_logs)
File “example.py”, line 7, in get_student_infos
name, catgory, grade = log.split(‘ ‘)
ValueError: not enough values to unpack (expected 3, got 2)

如果我们给example.py增加一行代码import better_exceptions,再执行,输入如下:

Traceback (most recent call last):
File “example.py”, line 24, in <module>
get_student_infos(exam_logs)
│                 └ [‘zhangsan math 60’, ‘lisi english 80’, ‘wangwu chinese 90’, ‘qianliu music’]
└ <function get_student_infos at 0x7f147344ee18>
File “example.py”, line 8, in get_student_infos
name, catgory, grade = log.split(‘ ‘)
│     │        │       └ ‘qianliu music’
│     │        └ ’90’
│     └ ‘chinese’
└ ‘wangwu’
ValueError: not enough values to unpack (expected 3, got 2)

看到了吗,加上import better_exceptions后, 异常时会将调用栈每一层用的变量值打印出来, 和普通异常时输出有比有什么好处呢,然我们来回忆一下,

• 没有better_exceptions时是这样对待异常的:
  • 看到一个异常, 能看到异常的类型, 但是无法直接给出错误原因
  • 追踪异常时, 往往得修改代码,将关键变量打印出来. 如果是线上环境, 貌似没什么好办法,只能上线上改一改.
• 有better_exceptions时:
  • 从异常时打印出来的信息可以清晰的看到异常发生时变量的值,可很容易定位问题.

是不是有小伙伴会想, 要是某些变量特别大(比如有几万个元素的列表),这样会造成日志很大的.确实存在这样的问题,不过这个库的作者已经够给解决方案了: 加上这句better_exceptions.MAX_LENGTH = 字符数控制字符的个数, 对于上面的例子加上better_exceptions.MAX_LENGTH = 20这句输出如下:

Traceback (most recent call last):
File “example.py”, line 27, in <module>
get_student_infos(exam_logs)
│                 └ [‘zhangsan math 60’,…
└ <function get_studen…
File “example.py”, line 10, in get_student_infos
name, catgory, grade = log.split(‘ ‘)
│     │        │       └ ‘qianliu music’
│     │        └ ’90’
│     └ ‘chinese’
└ ‘wangwu’
ValueError: not enough values to unpack (expected 3, got 2)

看到了吗,只是简单的再代码上加上一句import better_exceptions就有如此神奇的效果. 但是, 这个库目前只会在控制台打印出这样的错误信息, 下面说一下怎么与logging、django集成.

logging接入

先简单说一下,:

• logging模块在输出一个异常的时候,会调用handler的formater的formatException函数来格式化异常.
• better_exceptions有个format_exception方法会将异常调用栈变量值输出,就是 上面例子的那种输出.

看代码:

# -*- encoding: utf-8 -*-
import logging
from better_exceptions import format_exception

logger = logging.getLogger()
formatter = logging.Formatter(‘%(asctime)s – %(name)s – %(levelname)s – %(message)s’)
formatter.formatException = lambda exc_info: format_exception(*exc_info)

file_handler = logging.FileHandler(“example.log”)
file_handler.setLevel(logging.DEBUG)
file_handler.setFormatter(formatter)
logger.addHandler(file_handler)
def get_student_infos(logs):
student_infos = []
for log in logs:
name, catgory, grade = log.split(‘ ‘)
student_infos.append({
‘name’: name,
‘catgory’: catgory,
‘grade’: grade,

})
return student_infos

if __name__ == ‘__main__’:
exam_logs = [
‘zhangsan math 60’,
‘lisi english 80’,
‘wangwu chinese 90’,
‘qianliu music’
]
try:
get_student_infos(exam_logs)
except Exception as e:
logger.exception(e)

查看example.log文件输出:

2017-09-03 22:07:47,121 – root – ERROR – not enough values to unpack (expected 3, got 2)
Traceback (most recent call last):
File “log_example.py”, line 39, in <module>
get_student_infos(exam_logs)
│                 └ [‘zhangsan math 60’, ‘lisi english 80’, ‘wangwu chinese 90’, ‘qianliu music’]
└ <function get_student_infos at 0x7fd2e7fe46a8>
File “log_example.py”, line 22, in get_student_infos
name, catgory, grade = log.split(‘ ‘)
│     │        │       └ ‘qianliu music’
│     │        └ ’90’
│     └ ‘chinese’
└ ‘wangwu’
ValueError: not enough values to unpack (expected 3, got 2)

django接入

思路和logging接入一样的. 例如有如下django项目:

☁  test_better_exceptions_django  tree
.
├── db.sqlite3
├── manage.py
└── test_better_exceptions_django
├── fm.py
├── __init__.py
├── settings.py
├── urls.py
├── views.py
└── wsgi.py

两处修改:

• 增加一个py文件:
• import logging
  from better_exceptions import format_exception
  class ExceptionFormatter(logging.Formatter):
  def formatException(self, ei):
  return format_exception(*ei)
• 修改py的LOGGING配置:
• LOGGING = {
  ‘version’: 1,
  ‘disable_existing_loggers’: True,
  ‘formatters’: {
  ‘simple’: {
  ‘()’: ‘test_better_exceptions_django.fm.ExceptionFormatter’,
  ‘format’: ‘%(levelname)s %(message)s’
  },
  },
  ‘handlers’: {
  ‘console’: {
  ‘level’: ‘INFO’,
  ‘class’: ‘logging.StreamHandler’,
  ‘formatter’: ‘simple’
  },
  ‘file’: {
  ‘level’: ‘ERROR’,
  ‘class’: ‘logging.FileHandler’,
  ‘filename’: os.path.join(BASE_DIR, “error.log”),
  ‘formatter’: ‘simple’}
  },
  ‘loggers’: {
  ‘django’: {
  ‘handlers’: [‘console’, ‘file’],

  }
  }
  }

• 这里主要是自定义了一个formatters.

我是在请求里故意出发了一个异常,代码如下:

• py:
• fromconf.urls import url
  from django.contrib import admin
  from . import viewsurlpatterns = [
  url(r’^$’, views.index, name=’index’),
  ]
• py:
• fromhttp import HttpResponse
  import logging
  logger = logging.getLogger(‘django’)
  def index(request):
  a, b = 2, 0
  try:
  a / b
  except Exception as e:
  logger.exception(e)
  return HttpResponse(“You have an excepion, see console and log”)

接着打开浏览器,可以看到根目录下error.log文件内容:

☁  test_better_exceptions_django  cat error.log
ERROR division by zero
Traceback (most recent call last):
File “/vagrant_data/better_exceptions/test_better_exceptions_django/test_better_exceptions_django/views.py”, line 12, in index
a / b
│   └ 0
└ 2
ZeroDivisionError: division by zero