Data.List.Utils

This module provides various helpful utilities for dealing with lists.

Written by John Goerzen, [e-mail]

Merging

merge :: Ord a => [a] -> [a] -> [a]

Merge two sorted lists into a single, sorted whole.

Example:

merge [1,3,5] [1,2,4,6] -> [1,1,2,3,4,5,6]

QuickCheck test property:

prop_merge xs ys = merge (sort xs) (sort ys) == sort (xs ++ ys) where types = xs :: [Int]

mergeBy :: (a -> a -> Ordering) -> [a] -> [a] -> [a]

Merge two sorted lists using into a single, sorted whole, allowing the programmer to specify the comparison function.

QuickCheck test property:

prop_mergeBy xs ys = mergeBy cmp (sortBy cmp xs) (sortBy cmp ys) == sortBy cmp (xs ++ ys) where types = xs :: [ (Int, Int) ] cmp (x1,_) (x2,_) = compare x1 x2

Tests

startswith :: Eq a => [a] -> [a] -> Bool

Returns true if the given list starts with the specified elements; false otherwise. (This is an alias for isPrefixOf.)

Example:

startswith "He" "Hello" -> True

endswith :: Eq a => [a] -> [a] -> Bool

Returns true if the given list ends with the specified elements; false otherwise. (This is an alias for isSuffixOf.)

Example:

endswith "lo" "Hello" -> True

contains :: Eq a => [a] -> [a] -> Bool

Returns true if the given parameter is a sublist of the given list; false otherwise. Alias for isInfixOf.

Example:

contains "Haskell" "I really like Haskell." -> True
contains "Haskell" "OCaml is great." -> False

hasAny

Arguments

:: Eq a

=> [a]

List of elements to look for

-> [a]

List to search

-> Bool

Result

Returns true if the given list contains any of the elements in the search list.

Association List Utilities

These functions are designed to augment the association list functions in Data.List and provide an interface similar to Data.FiniteMap or Data.Map for association lists.

addToAL :: Eq key => [(key, elt)] -> key -> elt -> [(key, elt)]

Adds the specified (key, value) pair to the given list, removing any existing pair with the same key already present.

delFromAL :: Eq key => [(key, a)] -> key -> [(key, a)]

Removes all (key, value) pairs from the given list where the key matches the given one.

flipAL :: (Eq key, Eq val) => [(key, val)] -> [(val, [key])]

Flips an association list. Converts (key1, val), (key2, val) pairs to (val, [key1, key2]).

keysAL :: [(key, a)] -> [key]

Returns the keys that comprise the (key, value) pairs of the given AL.

Same as:

map fst

valuesAL :: [(a, value)] -> [value]

Returns the values the comprise the (key, value) pairs of the given AL.

Same as:

map snd

hasKeyAL :: Eq a => a -> [(a, b)] -> Bool

Indicates whether or not the given key is in the AL.

Association List Conversions

strFromAL :: (Show a, Show b) => [(a, b)] -> String

Converts an association list to a string. The string will have one pair per line, with the key and value both represented as a Haskell string.

This function is designed to work with [(String, String)] association lists, but may work with other types as well.

strToAL :: (Read a, Read b) => String -> [(a, b)]

The inverse of strFromAL, this function reads a string and outputs the appropriate association list.

Like strFromAL, this is designed to work with [(String, String)] association lists but may also work with other objects with simple representations.

Conversions

split :: Eq a => [a] -> [a] -> [[a]]

Given a delimiter and a list (or string), split into components.

Example:

split "," "foo,bar,,baz," -> ["foo", "bar", "", "baz", ""]

split "ba" ",foo,bar,,baz," -> [",foo,","r,,","z,"]

join :: [a] -> [[a]] -> [a]

Given a delimiter and a list of items (or strings), join the items by using the delimiter. Alias for intercalate.

Example:

join "|" ["foo", "bar", "baz"] -> "foo|bar|baz"

replace :: Eq a => [a] -> [a] -> [a] -> [a]

Given a list and a replacement list, replaces each occurance of the search list with the replacement list in the operation list.

Example:

replace "," "." "127,0,0,1" -> "127.0.0.1"

This could logically be thought of as:

replace old new l = join new . split old $ l

genericJoin :: Show a => String -> [a] -> String

Like join, but works with a list of anything showable, converting it to a String.

Examples:

genericJoin ", " [1, 2, 3, 4] -> "1, 2, 3, 4"
genericJoin "|" ["foo", "bar", "baz"] -> "\"foo\"|\"bar\"|\"baz\""

takeWhileList :: ([a] -> Bool) -> [a] -> [a]

Similar to takeWhile, takes elements while the func is true. The function is given the remainder of the list to examine.

dropWhileList :: ([a] -> Bool) -> [a] -> [a]

Similar to dropWhile, drops elements while the func is true. The function is given the remainder of the list to examine.

spanList :: ([a] -> Bool) -> [a] -> ([a], [a])

Similar to span, but performs the test on the entire remaining list instead of just one element.

spanList p xs is the same as (takeWhileList p xs, dropWhileList p xs)

breakList :: ([a] -> Bool) -> [a] -> ([a], [a])

Similar to break, but performs the test on the entire remaining list instead of just one element.

Advanced Conversions

newtype WholeFunc a b

The type used for functions for wholeMap. See wholeMap for details.

Constructors

WholeFunc ([a] -> (WholeFunc a b, [a], [b]))

wholeMap :: WholeFunc a b -> [a] -> [b]

This is an enhanced version of the concatMap or map functions in Data.List.

Unlike those functions, this one:

• Can consume a varying number of elements from the input list during each iteration
• Can arbitrarily decide when to stop processing data
• Can return a varying number of elements to insert into the output list
• Can actually switch processing functions mid-stream
• Is not even restricted to processing the input list intact

The function used by wholeMap, of type WholeFunc, is repeatedly called with the input list. The function returns three things: the function to call for the next iteration (if any), what remains of the input list, and the list of output elements generated during this iteration. The return value of wholeMap is the concatenation of the output element lists from all iterations.

Processing stops when the remaining input list is empty. An example of a WholeFunc is fixedWidth.

fixedWidth :: [Int] -> WholeFunc a [a]

A parser designed to process fixed-width input fields. Use it with wholeMap.

The Int list passed to this function is the list of the field widths desired from the input. The result is a list of those widths, if possible. If any of the input remains after processing this list, it is added on as the final element in the result list. If the input is less than the sum of the requested widths, then the result list will be short the appropriate number of elements, and its final element may be shorter than requested.

Examples:

wholeMap (fixedWidth [1, 2, 3]) "1234567890"
 --> ["1","23","456","7890"]
wholeMap (fixedWidth (repeat 2)) "123456789"
 --> ["12","34","56","78","9"]
wholeMap (fixedWidth []) "123456789"
 --> ["123456789"]
wholeMap (fixedWidth [5, 3, 6, 1]) "Hello, This is a test."
 --> ["Hello",", T","his is"," ","a test."]

Fixed-Width and State Monad Utilities

grab :: Int -> State [a] [a]

Helps you pick out fixed-width components from a list.

Example:

conv :: String -> (String,String)
conv = runState $
        do f3 <- grab 3
           n2 <- grab 2
           return $ f3 ++ "," ++ n2

main = print $ conv "TestIng"

Prints:

("Tes,tI","ng")

Miscellaneous

countElem :: Eq a => a -> [a] -> Int

Returns a count of the number of times the given element occured in the given list.

elemRIndex :: Eq a => a -> [a] -> Maybe Int

Returns the rightmost index of the given element in the given list.

alwaysElemRIndex :: Eq a => a -> [a] -> Int

Like elemRIndex, but returns -1 if there is nothing found.

seqList :: [a] -> [a]

Forces the evaluation of the entire list.

subIndex :: Eq a => [a] -> [a] -> Maybe Int

Similar to Data.List.elemIndex. Instead of looking for one element in a list, this function looks for the first occurance of a sublist in the list, and returns the index of the first element of that occurance. If there is no such list, returns Nothing.

If the list to look for is the empty list, will return Just 0 regardless of the content of the list to search.

Examples:

subIndex "foo" "asdfoobar" -> Just 3
subIndex "foo" [] -> Nothing
subIndex "" [] -> Just 0
subIndex "" "asdf" -> Just 0
subIndex "test" "asdftestbartest" -> Just 4
subIndex [(1::Int), 2] [0, 5, 3, 2, 1, 2, 4] -> Just 4

uniq :: Eq a => [a] -> [a]

Given a list, returns a new list with all duplicate elements removed. For example:

uniq "Mississippi" -> "Misp"

You should not rely on this function necessarily preserving order, though the current implementation happens to.

This function is not compatible with infinite lists.

This is presently an alias for nub.