Search for uppercased substring

Question

I need to take out the longest substring of uppercased characters. So out of the string:

"aaBBBBcBBdDDD"

I need to get "BBBB".

Is there a convenient Ruby method for that or a regexp of some kind? I tried:

string.scan(/[[:upper:]]/)

and that's almost it, only it gives ALL capital characters, not the longest sequence.

That's not something regex can do. (At the very least, it's not something regex is designed to do.) — Aran-Fey, Dec 29 '14 at 13:31
As @Rawing said, that's not what Regexp are for. Regular expressions are great at finding things that *look* like something, but they're terrible at finding something that is exactly something when presented with multiple choices, especially if there is any wiggle room. — the Tin Man, Dec 29 '14 at 21:00

Yu Hao · Answer 1 · 2014-12-30T01:09:18.610

8

Use regex to get an array of uppercase words, then use Enumerable#max_by to find the longest:

"aaBBBBcBBdDDD".scan(/[[:upper:]]+/).max_by {|x| x.length}
# => "BBBB"

or simpler:

"aaBBBBcBBdDDD".scan(/[[:upper:]]+/).max_by(&:length)
# => "BBBB"

edited Dec 30 '14 at 01:09

answered Dec 29 '14 at 13:37

Yu Hao

119,891
44
235
294

score 1 · Accepted Answer · edited Dec 29 '14 at 19:21

1

You can't find the string which has the maximum length only through regex. You need to use some built-in Ruby functions.

> m = "aaBBBBcBBdDDD".scan(/[[:upper:]]+/)
=> ["BBBB", "BB", "DDD"]
> vc = m.sort{|a,b| b.size <=> a.size}
=> ["BBBB", "DDD", "BB"]
> vc.delete_if{|a| a.size < vc.first.size}
=> ["BBBB"]

edited Dec 29 '14 at 19:21

the Tin Man

158,662
42
215
303

answered Dec 29 '14 at 13:34

Avinash Raj

172,303
28
230
274

yep, `+` matches the previous token one or more times. – Avinash Raj Dec 29 '14 at 13:38
What if the string looks like this: `aaBBBBAABcBBdDDD` ? It would return `BBBBAAB` – hek2mgl Dec 29 '14 at 13:38
yep , it returns `BBBBAAB` – Avinash Raj Dec 29 '14 at 13:40
I though sequences of *identical* uppercased characters should be returned. – hek2mgl Dec 29 '14 at 13:45
Don't use `sort{|a,b| b.size <=> a.size}`. `sort_by` would be faster because it remembers the intermediate value, instead of `sort` having to compute the `size` of `a` and `b` each pass through the loop. – the Tin Man Dec 29 '14 at 19:22
@theTinMan so you are suggesting `.sort_by{|a| -a.length}` Or `.sort_by(&:length).reverse)`? Or are you suggesting changing the last line to be `.delete_if{|a| a.length < vc.last.length}`. When benchmarking `sort_by` seems to be slower than `sort` in this case. Also in regards to this method right now it is determining the max on each loop through `delete_if` by referencing itself. I don't think this seems like the best approach. – engineersmnky Dec 29 '14 at 20:37
I'd suggest reading "[Sorting an array in descending order in Ruby](http://stackoverflow.com/questions/2642182/sorting-an-array-in-descending-order-in-ruby/2651028#2651028)". – the Tin Man Dec 29 '14 at 20:55
Using `delete_if` isn't the best choice. I'd use `max_by` personally, but others MMV. – the Tin Man Dec 29 '14 at 20:57

score 0 · Answer 3 · answered Dec 29 '14 at 13:32

0

([A-Z]+)

Try this.Capture all the groups and the one with maximum length is your answer.See demo.

https://regex101.com/r/gX5qF3/11

answered Dec 29 '14 at 13:32

vks

67,027
10
91
124

score 0 · Answer 4 · answered Dec 29 '14 at 17:51

0

You did not specify the expected result for more than 1 string of the same max length.

@AvinashRaj's answer will handle this while @YuHao's will not. If you want only 1 result I would suggest @YuHao's answer if you want all the results I would change @AvinashRaj's answer to something like this.

"aaBBBBcBBdDDDD".scan(/[[:upper:]]+/).tap do |a| 
   max_length = a.map(&:length).max
   a.delete_if{|x| x.length < max_length } 
end
#=> ["BBBB","DDDD"]

answered Dec 29 '14 at 17:51

engineersmnky

25,495
2
36
52

I don't think `tap` buys you anything here. It would be clearer, imo, to write: `a = "aaBBBBcBBdDDDD".scan(/[[:upper:]]+/); max_length = a.map(&:length).max; a.select { |s| s.size == max_length }`. – Cary Swoveland Dec 29 '14 at 19:44
@CarySwoveland thank you for the response. There are probably a dozen more simply conceived options for this. I just don't see a need for multiple local variables that will not be utilized anywhere else. Benchmarking them shows not significant difference in performance either they are all within a 10% +/- – engineersmnky Dec 29 '14 at 20:31

Search for uppercased substring

4 Answers4