Catching a specific IO error

[John Machin]

On 25/04/2007 4:06 AM, Steven Howe wrote:
> Steve Holden wrote:
>> Thomas Krüger wrote:
>>  
>>> Tina I schrieb:
>>>    
>>>> Now, this works but of course it catches every IOError, and I can not
>>>> figure out how to restrict it to only catch the "[Errno 2]"?
>>>>       
>>> There's an example that uses the error number:
>>> http://docs.python.org/tut/node10.html#SECTION0010300000000000000000
>>>
>>>     
>> So what you'll need to do is catch all IOError exceptions, then test 
>> to see if you've got (one of) the particular one(s) you are interested 
>> in. If not then you can re-raise the same error with a bare "raise" 
>> statement, and any containing exception handlers will be triggered. If 
>> there are none then you will see the familiar traceback termination 
>> message.
>>
>> regards
>>   Steve
>>   
> you could also use some pre-testing of the filename  os.path.isfile, 
> os.path.isdir, os.path.split are good
> functions to test file/directory existence. 

In general, this is laborious, tedious, and possibly even platform 
dependent. Then you still need to wrap the open call in try/accept. Why 
bother?

In particular, (1) please explain how os.path.split helps with existence 
testing:

Python 2.5.1 (r251:54863, Apr 18 2007, 08:51:08) [MSC v.1310 32 bit 
(Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
| >>> import os.path
| >>> os.path.split(r'no\such\path\nothing.nix')
('no\\such\\path', 'nothing.nix')

(2) please explain why you avoided mentioning os.path.exists.