Getting elements from a list by an index list - list

Question

I have two lists in Haskell.
Original list containing string values:
["Hello", "HELLO", "", "WORLD", "xxx", "world"]
Index list containing integer values where the strings are all upper case in the original list:
[1,3]
I incremented all the values in the index list with a function I created and make index2 list, overall it looks like this:
My code:
import Data.List
import Data.Char
import Data.Maybe
main = do
contents <- readFile "h.txt"
let original = lines (contents)
let allUpper = lines (map toUpper contents)
let onlyUpper = filter(/="") (intersect original allUpper)
let upperIndex = findIndices ('elem' onlyUpper) original
let index2 = (map increment upperIndex)
print index2
increment :: Int -> Int
increment x = x+1
I have managed to came this far with the help of yours. However, since I am a beginner I do not seem to understand how iteration over lists works.
The thing I want to accomplish is to check whether the corresponding index values (in index2) are empty or not in the original list, if they are empty, I want to remove them in index2.

Answer 1

Filtering empty elements
The thing I want to accomplish is to check whether the corresponding
index values (in index2) are empty or not in the original list, if
they are empty, I want to remove them in index2.
The code already filters out empty elements! Look at the following line:
let onlyUpper = filter(/="") (intersect original allUpper)
This line does two things:
it keeps only elements which are constituted only with uppercase letters(intersect original allUpper),
it filters out empty elements (filter(/="")).
If by empty elements you mean strings which contains only space characters or nothing, you can use instead:
filter (all isSpace)
Iterating over lists
I do not seem to understand how iteration over lists works.
In Haskell, lists are single chained lists: each element contains a value and a reference to the next value.
Therefore lists are not indexed: the !! operator have to go through each element to access a specific element making lists completely inefficient when dealing with direct access.
When you’re submitting a list to a function, you simply give it the first element.
With these considerations, when you work on lists, you have to avoid accessing elements via their index.
The idea is to create functions which do their job on simple values and mapping them to list of elements. Take a look at the toUpper function:
toUpper :: Char -> Char
It takes a Char and returns its uppercased version (also a Char).
Haskell does not have a toUpper function which works on String, you have to use something like map or <$> to apply toUpper to a list of char (a String):
map toUpper "ab" -- "AB"
toUpper <$> "ab" -- "AB"
The idea is to have functions which does only one specific thing. Upercasing and iterating over a list are two different things. Does the toUpper function need to know the index of the element it will uppercase? No!
Iterating over a list with index
You may ask: but what if my function REALLY need to consider the index of the elements? (ie: for filtering out even or odd elements).
You have two way of considering it:
a List is not the type you need to work with. Maybe Data.Map, Data.IntMap or Data.Vector are better suited for the task (see these modules for more information),
you need to use an intermediate type which will hold the index.
For example:
let string = "abcde"
let indexedString = zip [1..] string
print indexedString -- [(1, 'a'), (2, 'b), (3, 'c), (4, 'd), (5, 'e)]
Note that this also solves your need of an increment function since the index is started at whatever value you want.
To go back to the original string, you write:
map snd indexedString -- "abcde"
You need to use the fst and snd functions to work with the intermediate type, or to use pattern matching:
filter (\x -> snd x == 'b') indexedString -- [(2, 'b')]
map (\(i,s) -> (i, toUpper s)) indexedString -- [(1,'A'),(2,'B'),(3,'C'),(4,'D'),(5,'E')]
Taking the index into account:
let string = "abcde"
indexedString = zip [1..] string
upperEven (i, c) | even i = (i, toUpper c)
| otherwise = (i, c)
print $ map upperEven indexedString -- [(1,'a'),(2,'B'),(3,'c'),(4,'D'),(5,'e')]
print $ map snd $ map upperEven indexedString -- "aBcDe"
Notes
The increment function already exists in Haskell, it’s called succ (it is also a more generic function which works on every types supporting the Enum class like Int, Char…)

Answer 2

Why not use words :: String -> [String] on the contents you get from a file? Using lines :: String -> [String] would be an alternative if you had one word per line.
Then if i get your problem right, you could write the following to solve your problem:
import Data.List (findIndices)
import Data.Char (isUpper)
allUpperOneBasedIndices :: String -> [Int]
allUpperOneBasedIndices = map succ . findIndices (all isUpper) . words