Detect duplicated elements of a string slice happening in order - rust

Question

I need to detect and list string characters of slice that repeat themselves in order more or equal than N times. I managed to write non-higher-order-function solution in Rust already (below), but I wonder if this can be simplified to chaining iter methods.
The idea:
let v = "1122253225";
let n = 2;
Output:
There are 2 repetition of '1'
There are 3 repetition of '2'
There are 2 repetition of '2'
Indexes where repetition happens are not important. Repetition must happen in order (ie. 3 repetition of '2' separated by other values from the other 2 repetition of '2' counts as separate output lines).
My non-iterator solution:
let mut cur_ch = '\0';
let mut repeat = 0;
for ch in v.chars() {
if ch == cur_ch {
repeat = repeat + 1;
}
else {
if repeat >= n {
printf!("There are {} repetition of '{}'", repeat, cur_ch);
}
cur_ch = ch;
repeat = 1;
}
}
if repeat >= n {
printf!("There are {} repetition of '{}'", repeat, cur_ch);
}
It works, but is there a better way to do so with chaining iter methods?

Answer 1

Here is a solution that uses scan and filter_map:
fn main() {
let s = "112225322555";
let n = 2;
let i = s
.chars()
.map(|v| Some(v))
.chain(std::iter::once(None))
.scan((0, None), |(count, ch), v| match ch {
Some(c) if *c == v => {
*count += 1;
Some((None, *count))
}
_ => Some((ch.replace(v), std::mem::replace(count, 1))),
})
.filter_map(|(ch, count)| match ch {
Some(Some(ch)) if count >= n => Some((ch, count)),
_ => None,
});
for (ch, num) in i {
println!("There are {} repititions of {}", num, ch);
}
}
Playground Link
The first step is to use scan to count the number of adjacent characters. The first argument to scan is a state variable, which gets passed to each call of the closure that you pass as the second argument. In this case the state variable is a tuple containing the current character and the number of times it has been seen.
Note:
We need to extend the iteration one beyond the end of the string we are analyzing (otherwise we would miss the case where the end of the string contained a run of characters meeting the criteria). We do this by mapping the iteration into Option<char> and then chaining on a single None. This is better than special-casing a character such as \0, so that we could even count \0 characters.
For the same reason, we use Option<char> as the current character within the state tuple.
The return value of scan is an iterator over (Option<Option<char>>, i32). The first value in the tuple will be None for each repeated character in the iterator, whereas at each boundary where the character changes it will be Some(Some(char))
We use replace to both return the current character and count, at the same time as setting the state tuple to new values
The last step is to use filter_map to:
remove the (None, i32) variants, which indicate no change in the incoming character
filter out the cases where the count does not reach the limit n.

Answer 2

Here's one attempt at using filter_map():
fn foo(v: &str, n: usize) -> impl Iterator<Item = (usize, char)> + '_ {
let mut cur_ch = '\0';
let mut repeat = 0;
v.chars().chain(std::iter::once('\0')).filter_map(move |ch| {
if ch == cur_ch {
repeat += 1;
return None;
}
let val = if repeat >= n {
Some((repeat, cur_ch))
} else {
None
};
cur_ch = ch;
repeat = 1;
val
})
}
fn main() {
for (repeat, ch) in foo("1122253225", 2) {
println!("There are {} repetition of '{}'", repeat, ch);
}
}
And then you can generalize this to something like this:
fn foo<'i, I, T>(v: I, n: usize) -> impl Iterator<Item = (usize, T)> + 'i
where
I: Iterator<Item = T> + 'i,
T: Clone + Default + PartialEq + 'i,
{
let mut cur = T::default();
let mut repeat = 0;
v.chain(std::iter::once(T::default()))
.filter_map(move |i| {
if i == cur {
repeat += 1;
return None;
}
let val = if repeat >= n {
Some((repeat, cur.clone()))
} else {
None
};
cur = i;
repeat = 1;
val
})
}
This would be higher-order, but not sure if it's actually much simpler than just using a for loop!