Help me cythonize a python routine!

[BartC]

On 09/11/2016 21:25, breamoreboy at gmail.com wrote:
> On Wednesday, November 9, 2016 at 7:34:41 PM UTC, BartC wrote:

> However according to your mindset nothing matters provided it's fast,
 > accuracy does not matter to users.

> Hence your recent comment on another thread about converting invalid integer entries into zero.

That's by design. And it's not unprecedented; using C's 'atoi()' 
function (convert a string to an integer), then inputs of "" and "ABC" 
both return 0. Input of "123ABC" returns 123.

And in the context that was being discussed, it was desired that hitting 
end-of-line while attempting to read more items should return null 
values such as 0, 0.0 or "".

For any more refined behaviour, values can be read differently (read as 
strings then converted with routines that do more validation).

Effectively, the conversion looks at the entire line buffer contents 
remaining, and converts the first integer encountered, if any, then 
removes that and any terminator from the buffer.

If I try something like that in Python, trying to convert an integer at 
the beginning of a string, it fails: int("123,456"), int("123 456"),
int("63924 the"). It will need processing first so that the parameter 
comprises only the thing we're trying to read.

Anyway the comparison between Python and that particular /lower-level/ 
language is moot as Python apparently doesn't have a way of directly 
reading items from an input line, either from a console or from a file.

You have to read a line as one string then apply DIY conversions. Which 
the other language could equally do.

So it's a detail. The other language could also have chosen to use 
exceptions to signal such events (no integer following, end of input, 
invalid terminator etc). I chose to keep it simple and to make it work 
the same way it's always done.

-- 
Bartc