I want to write a programm which would print out only one integer after applying some functions to the lines in text file, so far I have :
main = do
c <- getLine
let plot = plots (split ',' (change c))
print plot
main
where plots, split and change are the functions that convert input from string to int removing non-integer chars and then applying some calculations, the problem is that my input file has a lot of lines and I only managed to write a program which applies those functions to every line separately and prints out the result of every line,(I get as much output lines as there are input lines), but I want that this programm would sum up the results of every line and would print out only that number, where should I start or maybe somebody knows the solution? I am new in Haskell so please don't judge :/
Assuming that your conversion functions work for multi-line inputs, you can probably get away with simply replacing getLine with getContents and removing the recursive call to main. Although if you actually want to read a file, using readFile is probably cleaner than using getContents, since the latter is typically used for reading from the command line.
To expand upon what Paul wrote, getContents is like getLine, but it gives you the entire file in a single string. You can then get a list of lines using lines. For example:
main :: IO ()
main = do
contents <- getContents
print (lines contents)
When given the file
Haskell
is
neat!
will print ["Haskell","is","neat!"]. From here, you can map your line-handling function over that list to get a list of integers, and finally print (sum resultList).
Related
This question already has answers here:
How can I print a newline properly in Haskell?
(2 answers)
Closed 6 years ago.
I am a complete novice in Haskell and have the following question:
I intend to create the function, which puts three Strings on different lines. Here is the code:
onThreeLines :: String -> String -> String -> String
onThreeLines a b c = a++"\n"++b++"\n"++c
Here is what I run:
onThreeLines "Life" "is" "wonderful"
And what I get:
"Life\nis\nwonderful"
I have also tried the following character, but it doesn't work as well.
"'\n'"
Your function works. If you’re running this in GHCi or using print, you might be confused by the fact that it calls show on the result of your computation, which formats a value as a Haskell term for debugging. For strings, that means including quotes and escapes.
putStrLn (onThreeLines "Life" "is" "wonderful") should do exactly what you expect.
Executing it like this should make it work:
main :: IO ()
main = putStrLn $ onThreeLines "hello" "world" "test"
Executing the program, I get:
$ ./test.hs
hello
world
test
The reason you are getting "Life\nis\nwonderful" is because the Show instance is being used for displaying which will escape the newline.
λ> putStrLn "hello\nworld"
hello
world
λ> print "hello\nworld"
"hello\nworld"
Note that print uses the Show instance for displaying.
There's nothing wrong with your function. "Life\nis\nwonderful" is the resulting String you want. Just remember that if you want the newlines rendered correctly, pass it to a function like putStrLn
putStrLn (onThreeLines "Life" "is" "wonderful")
Also, be sure to check out the unlines function which concatenates a list of strings, separating each element with a newline character.
How can I write a string that contains newlines ("\n") to a file so that each string is on a new line in the file?
I have an accumulator function that iterates over some data and incrementally constructs a string (that contains information) for each element of the data. I don't want to write to the file every step so I'm appending the strings in each step. I do this so I can write the string in one time and limit the amount of IO.
Adding a newline to the string via str ++ "\n" doesn't work, hPrint h str will just print "\n" instead of starting on a new line.
I've tried accumulating a list of strings, instead of one big string, and iterating over the list and printing each string via hPrint. This works for the newlines but it also prints the quotation marks around each string on every line.
Don't use hPrint to write the strings to the file. Just like regular print it outputs the result of show, which produces a debugging-friendly version of the string with control characters and line endings escaped (and the surrounding quotes).
Use hPutStr or hPutStrLn instead. They will write the string to the file as-is (well, the latter adds a newline at the end).
The probably idiomatic solution to what you try to do is to simply aggregate the resulting strings in a list. Then, use the unlines prelude function which has the signature unlines :: [String] -> String and does your \n business for you.
Then, writing the string to disk can be done with help of writeFile which has the signature: writeFile :: FilePath -> String -> IO ().
Haskell is lazy. As such, it sometimes helps to think of Haskell lists as enumerators (C# like IEnumerable). This means here, that trying to compute line wise, then build the string manually and write it line by line is not really necessary. Just as readFile works lazily, so then does e.g. lines. In other words, you gain nothing if you try to "optimize" code which looks in its genuine form similar to this:
main = do
input <- readFile "infile"
writeFile "outfile" ((unlines . process) (lines input))
where
process inputLines = -- whatever you do
How do you print Strings in Haskell with newline characters?
printString :: String -> String -> String -> String
printString s1 s2 s3 = (s1++"\n"++s2++"\n"++s3)
When using the function it prints the entire line including the newline characters as well
As already commented, your code has nothing to do with printing. (It shouldn't have print in it's name either!) Printing is a side-effectful action (it makes something appear observably on your screen!) and hence can't be done without binding in the IO type.
As for the actual task of generating a single three-line string, that is perfectly fulfilled by your suggested solution. "This\nis\ntest" is a three-line string, as witnessed by
Prelude> lines $ printString "This" "is" "test"
["This","is","test"]
The reason why GHCi doesn't actually output three separate lines when you just write
Prelude> "This\nis\ntest"
"This\nis\ntest"
is that the print function that's used for this purpose always guarantees a format that's safe to use again in Haskell code, hence it puts strings in quotes and escapes all tricky characters including newlines.
If you simply want to dump a string to the terminal as-is, use putStrLn instead of print.
Prelude> putStrLn $ printString "This" "is" "test"
This
is
test
In ghci, you can just type
putStrLn "line1\nline2"
If you want to write a program to do this, you need to make sure that putStrLn runs in the IO monad, for instance, by putting it in main
main = do
<do stuff>
putStrLn "line1\nline2"
<do other stuff>
Your printString function does not print a string, it simply returns a string. Because you're running this in GHCi, it gets printed out. But it's GHCi that's printing it; your function itself prints nothing.
If you were to compile some code that calls printString, you would discover that nothing gets printed.
By default, GHCi prints stuff as expressions. If you want to write the string to the console unaltered, you need to use putStrLn, as the other answers suggest. Compare:
Prelude> print "1\n2"
"1\n2"
Prelude> putStrLn "1\n2"
1
2
Basically when you write an expression in GHCi, it automatically calls print on the result if possible. Because if you're executing something, you probably want to see what the result was. Compiled code doesn't do this; it only prints what you explicitly tell it to.
Why doesn't the following program print my input? It seems that putStr is not taking the input. How does getContents work?
main = do
contents <- getContents
when (length contents < 10) $ putStr contents
However, this program prints the input line by line:
main = do
contents <- getContents
putStr contents
getContents gets everything. Linebuffering makes it line-by-line
getContents gets the entire input from a handle (eg file or user input), so your program
main = do
contents <- getContents
putStr contents
reads the entire contents of the standard input and prints it. The only reason you see this line by line is that it's using linebuffering on the terminal, so getContents gets its incoming String a line at the time.
Lazy evaluation
Haskell uses lazy evaluation, which means it only calculates something when it has to - it doesn't need calculate the end of the string you're printing to print the start of it, so it doesn't bother, and just prints what it has now. This lazy functions are able to return partial results when they can, rather than having to calculate everything first.
A line at a time
You appear in comments to want to only print lines when they're short like this session:
don't print this it's long
print this
print this
this is also too long
boo!
boo!
but since getContents is all the input, it won't print anything unless the total length is less than 10. What you were after was something where it treats each line separately, more like
main = do
contents <- getContents
putStrLn . unlines . filter short . lines $ contents
short xs = length xs <= 10
getContents reads all the stdin, but it works lazily. It returns a thunk, it is a promise to return some value when you ask it (force the thunk).
putStr asks for one char at a time, forcing the thunk to return value. For lists (note, String is a list of Chars), thunk when forced returns either "end of of list" or pair of ("next char", "thunk for the rest of the list"). So the second example works because putStr outputs chars when they become available. You enter the next line -- putStr outputs it char-by-char, tries to force the next thunk, but blocks because the next char is not available yet.
The first example forces the thunk until it returns "end of the list", because it is not possible to know string length until it is available.
If you're doing an interactive application, getting the length of getContents is a bad idea. That's because the length the standard input can only be calculated when the stream is closed. Which means that you'd have to either use Ctrl+D on Linux, Ctrl+Z on Windows or close the application, before you got to see the results.
I was wondering what the preferred method is to convert lazy text into a list of strict texts split up by line endings. I came up with something like the following, but am not sure that I'm guaranteed that each strict text element will be a complete line (or if there are other problems with it):
import qualified Data.Text.Lazy as LT
readLines filePath = do
contents <- Data.Text.Lazy.IO.readFile filePath
let lines = concat (map LT.toChunks (LT.lines contents))
return lines
readLines filePath = do
contents <- Data.Text.Lazy.IO.readFile filePath
let lines = concat (map LT.toChunks (LT.lines contents))
return lines
doesn't guarantee that each strict chunk is a complete line. In fact, every time a chunk boundary of the lazy Text does not coincide with a line boundary, you get a line having part in at least two strict chunks.
readLines filePath = do
contents <- Data.Text.Lazy.IO.readFile filePath
let lines = map (T.concat . LT.toChunks) (LT.lines contents)
return lines
however, concatenates each line into one strict chunk. Doing the concatenation may however be slower than working with each line as a list of strict chunks resp. a lazy Text.