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

Groups > comp.lang.python > #74823 > unrolled thread

What's the proper style for a library string function?

Back to article view | Back to comp.lang.python

Contents

  What's the proper style for a library string function? "C.D. Reimer" <chris@cdreimer.com> - 2014-07-19 10:38 -0700
    Re: What's the proper style for a library string function? Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-07-20 01:41 +0000

#74823 — What's the proper style for a library string function?

Greetings,

I typically write a Python 2.7 string function in my library like this:

     def getCompletedTime(start, end): return "Time completed:", str(end 
- start)

And called it like this:

     print getCompletedTime(start, end)

Since every Python script I write is executed from the command line, I 
rewrote the string function like this:

     def getCompletedTime(start, end): print "Time completed:", str(end 
- start)

And call it like this:

     getCompletedTime(start, end)

The first version is what I'm familiar with having reluctantly learned 
Java at community college, which couldn't afford a Microsoft site 
license for Visual C++ and taught every class in Java. (The Linux 
instructor rebelled against this policy by teaching basic C/C++ and 
shell scripting in his classes.) I recently read an article that Python 
is replacing Java as a teaching language.

The second version is more straight forward but seems less readable 
(i.e., "print getCompletedTime(start, end)" vs. "getCompletedTime(start, 
end)") from the calling script.

Alternatively, I thought about rewriting the string function to accept 
an extra parameter to do either and default to the print statement.

     def getCompletedTime(start, end, type = 'p'):
         string = "Time completed: " + str(end - start)
         if type == 'p':
             print string
         else:
             return string

I'm curious as to what the proper Python style would be for this.

Thank you,

Chris Reimer

[toc] | [next] | [standalone]

#74855

On Sat, 19 Jul 2014 10:38:47 -0700, C.D. Reimer wrote:

> Greetings,
> 
> I typically write a Python 2.7 string function in my library like this:
> 
>      def getCompletedTime(start, end): 
>          return "Time completed:", str(end - start)
> 
> And called it like this:
> 
>      print getCompletedTime(start, end)
> 
> Since every Python script I write is executed from the command line, I
> rewrote the string function like this:
> 
>      def getCompletedTime(start, end): 
>           print "Time completed:", str(end - start)
> 
> And call it like this:
> 
>      getCompletedTime(start, end)

In general, I much prefer the first version. The general principle I 
follow is, as much as possible, to separate calculation from display. 
Advantages include:

- You can easily re-use the getCompletedTime function in situations where 
you don't want to display the time, or pass the result on to another 
function for further processing before display.

- It makes automated testing of the functions much easier.

- It makes it easier to vary the type of display. Today you want to print 
to the standard output. Tomorrow you might want to output to a file, or 
display in a GUI window.

- It makes it easier to understand the large-scale structure of your 
code. Seeing "print getCompletedTime(a, b)" makes it pretty clear that 
you are printing the result of the calculation, whereas 
"getCompletedTime(a, b)" does not.



-- 
Steven

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web