[Python-3000] Support for PEP 3131

[Baptiste Carvello]

James Y Knight a écrit :
> If another developer is planning to write code in English, this whole  
> debate is moot. So, let's take as a given that he is going to write a  
> program in his own non-English language. Now, will he write in a  
> asciified form of his language, or using the proper character set?  
> Right now, the only option is the first. The PEP proposes to also  
> allow the second.
> 
that's a very nice summary of the situation.

> So, your question should be: is it easier to understand an ASCIIified  
> form of another language, or the actual language itself? For me (who  
> doesn't speak said langauge, nor perhaps even know its character  
> set), I'm pretty sure the answer is still going to be the second: I'd  
> rather a program written in Chinese use Chinese characters, rather  
> than a transliteration of Chinese into ASCII. 
>
This is where we strongly disagree. If an identifier is written in
transliterated chinese, I cannot understand what it means, but I can recognise
it when it is used in the code. I will then find out the meaning from the
context. By contrast, with chineses identifiers, I will not recognise them from
one another. So I won't be able to make any sense from the code without going
through the complex task of translating everything.

> because it is actually  
> feasible for me to do automatic translation of Chinese into something  
> resembling English. And of course, that's even more true when talking  
> about a language like French, which uses an alphabet quite familiar  
> to me, but yet online translators still fail to function if it's been  
> transliterated into ASCII.
> 
dream on! Automatic translation won't work. For example, if you actually try
feeding python code to a french-to-english translator, you might be surprised by
what happens to the keyword "if" (just try it:-). You would have to translate
the identifiers one by one, which is not practical.

Cheers,
BC