Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.python > #94177

Re: Is this a good way to work with init and exception

From Cecil Westerhof <Cecil@decebal.nl>
Newsgroups comp.lang.python
Subject Re: Is this a good way to work with init and exception
Date 2015-07-20 01:27 +0200
Organization Decebal Computing
Message-ID <87vbdfwwwd.fsf@Equus.decebal.nl> (permalink)
References (3 earlier) <mailman.733.1437329515.3674.python-list@python.org> <87io9gx8tb.fsf@Equus.decebal.nl> <mailman.743.1437340103.3674.python-list@python.org> <87615fyem2.fsf@Equus.decebal.nl> <mailman.748.1437345644.3674.python-list@python.org>

Show all headers | View raw


On Monday 20 Jul 2015 00:40 CEST, Chris Angelico wrote:

> On Mon, Jul 20, 2015 at 8:19 AM, Cecil Westerhof <Cecil@decebal.nl> wrote:
>>> If two modules import the same module, they get two references to
>>> that same module, not two separate module instances. Since your
>>> parameters appear only to affect the initialization itself, this
>>> is not likely to be a problem (it's not like you'll need to
>>> authenticate with two different sets of credentials, for
>>> instance), but it will mean that the second one will import an
>>> already-initialized module. That's why I suggested the try_init
>>> function which would quietly return an immediate success if the
>>> module had already been initialized. But if this isn't going to be
>>> an issue, then your code's fine.
>>
>> Good to know. I would expect two different instances.
>>
>> I agree that in my case it would not be a problem, but I put the
>> code on GitHub:
>> https://github.com/CecilWesterhof/PythonLibrary/blob/master/twitterDecebal.py
>> I should do my best to circumvent nasty surprises for users of the
>> code. Someone else could use several Twitter accounts at the same
>> time. Is there a way to do this?
>
> Does the instantiation of Core() involve authentication? Is it
> possible to call Core() more than once and use different accounts?
> Your send_message() takes an account identifier, so it might be you
> don't need separate accounts. But if, just very occasionally, you do
> need multiple, here's a possible design style: Have init() return
> the Core as well as stashing it in _core, and then have
> send_message() take an optional keyword argument (in 3.x,
> keyword-only) to choose a different core. That way, it'll by default
> use the most recently initialized core, but you can create multiple
> and manage them yourself if you so choose. (Obviously you'd use
> reinit_allowed=True for all the initializations.)

You are right: core is a general initialisation, so in this case
nothing to worry about. :-D

When I write something where it could make a difference, I should use
your tip.

-- 
Cecil Westerhof
Senior Software Engineer
LinkedIn: http://www.linkedin.com/in/cecilwesterhof

Back to comp.lang.python | Previous | NextPrevious in thread | Find similar | Unroll thread


Thread

Is this a good way to work with init and exception Cecil Westerhof <Cecil@decebal.nl> - 2015-07-19 12:35 +0200
  Re: Is this a good way to work with init and exception Chris Angelico <rosuav@gmail.com> - 2015-07-19 22:59 +1000
    Re: Is this a good way to work with init and exception Cecil Westerhof <Cecil@decebal.nl> - 2015-07-19 18:46 +0200
      Re: Is this a good way to work with init and exception Chris Angelico <rosuav@gmail.com> - 2015-07-20 04:11 +1000
        Re: Is this a good way to work with init and exception Cecil Westerhof <Cecil@decebal.nl> - 2015-07-19 21:10 +0200
          Re: Is this a good way to work with init and exception Chris Angelico <rosuav@gmail.com> - 2015-07-20 07:08 +1000
            Re: Is this a good way to work with init and exception Cecil Westerhof <Cecil@decebal.nl> - 2015-07-20 00:19 +0200
              Re: Is this a good way to work with init and exception Chris Angelico <rosuav@gmail.com> - 2015-07-20 08:40 +1000
                Re: Is this a good way to work with init and exception Cecil Westerhof <Cecil@decebal.nl> - 2015-07-20 01:27 +0200

csiph-web