Module talk:Formatnum
Appearance
![]() | This module was considered for deletion on 2018 September 21. The result of the discussion was "keep". |
language:formatNum(tonumber(number),…)
If I read the code correctly, then tonumber(number)
is already computed and stored in the variable named “value”. Incnis Mrsi (talk) 09:55, 22 October 2019 (UTC)
- @Incnis Mrsi: It's true that the variable value stores the numeric value of the parameter passed as number. However, if you read line 44, the
mw.language:formatNum()
call doesn't handle exponents properly. The lines 41 to 52 convert number to a string, then if there is an exponent, it stores the significand as a string back into number and the exponent in exponent. So by the time the code reaches lines 84 or 86 the string value in number may or may not be the same as that passed into the function. So it's necessary to again convert that string back to a number in order to usemw.language:formatNum()
. Writinglanguage:formatNum(value, ...)
wouldn't work properly if the number passed contained an exponent. Hope that makes sense. --RexxS (talk) 16:18, 22 October 2019 (UTC)- Ah, got it – the variable may be reassigned with
number = string.sub(number, 1, pos - 1)
. Incnis Mrsi (talk) 17:11, 22 October 2019 (UTC)
- Ah, got it – the variable may be reassigned with
Zero precision not working
If I use fn.formatNum(5435.12345,"en",2)
I get the result 5,435.12
as expected. Similarly with other positive values for precision. But with fn.formatNum(5435.12345,"en",0)
the number is returned with all the post decimal point digits, i.e. 5,435.12345
). Looking at the code there are a couple of if prec > 0 then ... elseif prec < 0 then .. end
blocks which suggests zero is ignored. Is this intentional or an oversight? Jts1882 | talk 15:35, 9 December 2019 (UTC)