idiomatic way to do stuff like `val words = input.split(" ").map { it.trim() }`?

# announcements

[Zimri Leisher]

idiomatic way to do stuff like val words = input.split(" ").map { it.trim() }?

[ephemient]

.split() already excludes the splitter from the results

[Zimri Leisher]

oh, really? great. for some reason i was under the impression that it didn't

[Animesh Sahu]

You can input.split(" ") { it.trim() } Oh wait, its for joinToString actually...

[nanodeath]

not that you needed to in this case, per @ephemient, but I'd use splitToSequence here, avoid that extra list allocation. input.splitToSequence(" ").map { it.trim() }.toList. also...trim removes all whitespace, not just the space character, so strictly speaking it's not redundant. depends on whether you're expecting \t and friends to show up.

[Ahmed Mourad]

split also accepts a regex so you could do

input.split(Regex("\\s+"))

which makes the trimming step unnecessary, and also prevents having empty string entries when you have three or more consecutive whitespaces.

[Zimri Leisher]

@nanodeath super helpful thanks, and @Ahmed Mourad you too 🙂

[nanodeath]

@Ahmed Mourad but now you're compiling regexes 🙃

[Vampire]

Well, it heavily depends on what you want to achieve

"oh  \tmy   god".split(" ")
res2: kotlin.collections.List<kotlin.String> = [oh, , <tab>my, , , god]

"oh  \tmy   god".split(" ").map { it.trim() }
res3: kotlin.collections.List<kotlin.String> = [oh, , my, , , god]

"oh  \tmy   god".split("""\s+""".toRegex())
res4: kotlin.collections.List<kotlin.String> = [oh, my, god]

[Animesh Sahu]

Hm,

"oh  \tmy   god".splitToSequence(" ").map { it.trim() }.filter { it.isNotEmpty() }.toList()
res0: kotlin.collections.List<kotlin.String> = [oh, my, god]

blob thinking upside down
Edit: I checked regex version is faster (too much) -> 1ms vs 0.06ms 😮

[Animesh Sahu]

Go with regex!

[Vampire]

Of course it is faster, Regex is meant for text processing, having a 4 step sequence processing needs time. 😄
You can probably go even faster if you pre-compile the regex, at least if the code is called multiple times.

[Zimri Leisher]

good to know 🙂