July 09, 2017 Lazy range, extract only Nth element, set range size constraint? | ||||
|---|---|---|---|---|
| ||||
Following is the code for a more generalized Fibonacci range.
Questions:
1. How do I get only the value of the Nth (i.e. N = 25) element in an idiomatic way?
2. Can I set constraints in the range so that user gets warning if he asks for Nth element greater than a limit, say N> 30; or if the actually range value at N is greater than datatype limit (e.g. max long)? Maybe this should be done outside of the range, i.e. do check before accessing the range?
#!/usr/bin/env rdmd
import std.stdio : writeln;
long multifactor = 4;
int elemth = 25;
struct FibonacciRange
{
long first = 1;
long second = 1;
bool empty() const @property
{
// how to stop at n = 30?
return false;
}
void popFront()
{
long tmp = 0;
tmp = first*multifactor + second;
first = second;
second = tmp;
}
long front() const @property
{
return first;
}
}
void main()
{
import std.range : take;
import std.array : array;
FibonacciRange fib;
auto fib10 = take(fib, elemth);
long[] the10Fibs = array(fib10);
}
| ||||
July 10, 2017 Re: Lazy range, extract only Nth element, set range size constraint? | ||||
|---|---|---|---|---|
| ||||
Posted in reply to biocyberman | On 07/09/2017 11:51 PM, biocyberman wrote: > Following is the code for a more generalized Fibonacci range. > > Questions: > > 1. How do I get only the value of the Nth (i.e. N = 25) element in an idiomatic way? As you've only got an input range, you have to popFront the 24 values that come before. You can use std.range.drop: ---- import std.range : drop; auto twentyfifth = fib.drop(24).front; ---- But if you're not sure that the range actually has 25 elements, you should check `empty`, of course: ---- import std.range : popFrontN; fib.popFrontN(24); /* or fib = fib.drop(24); */ if (!fib.empty) { auto twentyfifth = fib.front; } ---- > 2. Can I set constraints in the range so that user gets warning if he asks for Nth element greater than a limit, say N> 30; You can keep track of N in FibonacciRange and when it hits 30 you throw an exception or print a message or just set `empty` to true. You can't make it a compilation warning/error as far as I can tell. > or if the actually range value at N is greater than datatype limit (e.g. max long)? You can use std.experimental.checkedint to detect it at run time: ---- private bool _empty = false; bool empty() const @property { return _empty; } void popFront() { import std.experimental.checkedint : checked, Throw; long tmp = 0; try tmp = (checked!Throw(first)*multifactor + second).get; catch (Throw.CheckFailure e) _empty = true; first = second; second = tmp; } ---- (There may be a smarter way than making the operation throw an exception and catching that.) | |||
Copyright © 1999-2021 by the D Language Foundation
Permalink
Reply