Data.List.Split

Split on the given sublist. Equivalent to split . dropDelims . onSublist.

>>> splitOn ":" "12:35:07"["12","35","07"]

>>> splitOn "x" "axbxc"["a","b","c"]

>>> splitOn "x" "axbxcx"["a","b","c",""]

>>> splitOn ".." "a..b...c....d.."["a","b",".c","","d",""]

In some parsing combinator frameworks this is also known as sepBy.

Note that this is the right inverse of the intercalate function from Data.List, that is,

intercalate x . splitOn x === id

splitOn x . intercalate x is the identity on certain lists, but it is tricky to state the precise conditions under which this holds. (For example, it is not enough to say that x does not occur in any elements of the input list. Working out why is left as an exercise for the reader.)

Split on any of the given elements. Equivalent to split . dropDelims . oneOf.

>>> splitOneOf ";.," "foo,bar;baz.glurk"["foo","bar","baz","glurk"]

Split on elements satisfying the given predicate. Equivalent to split . dropDelims . whenElt.

>>> splitWhen (<0) [1,3,-4,5,7,-9,0,2][[1,3],[5,7],[0,2]]

>>> splitWhen (<0) [1,-2,3,4,-5,-6,7,8,-9][[1],[3,4],[],[7,8],[]]

Split into chunks terminated by the given subsequence. Equivalent to split . dropFinalBlank . dropDelims . onSublist.

>>> endBy ".;" "foo.;bar.;baz.;"["foo","bar","baz"]

Note also that the lines function from Data.List is equivalent to endBy "\n".

Split into chunks terminated by one of the given elements. Equivalent to split . dropFinalBlank . dropDelims . oneOf.

>>> endByOneOf ";," "foo;bar,baz;"["foo","bar","baz"]

Split into "words", with word boundaries indicated by the given predicate. Satisfies words === wordsBy isSpace; equivalent to split . dropBlanks . dropDelims . whenElt.

>>> wordsBy (`elem` ",;.?! ") "Hello there, world! How?"["Hello","there","world","How"]

>>> wordsBy (=='x') "dogxxxcatxbirdxx"["dog","cat","bird"]

Split into "lines", with line boundaries indicated by the given predicate. Satisfies lines === linesBy (=='n'); equivalent to split . dropFinalBlank . dropDelims . whenElt.

>>> linesBy (==';') "foo;bar;;baz;"["foo","bar","","baz"]

>>> linesBy (=='x') "dogxxxcatxbirdxx"["dog","","","cat","bird",""]

chunksOf n splits a list into length-n pieces. The last piece will be shorter if n does not evenly divide the length of the list. If n <= 0, chunksOf n l returns an infinite list of empty lists.

>>> chunksOf 3 [1..12][[1,2,3],[4,5,6],[7,8,9],[10,11,12]]

>>> chunksOf 3 "Hello there"["Hel","lo ","the","re"]

>>> chunksOf 3 ([] :: [Int])[]

Note that chunksOf n [] is [], not [[]]. This is intentional, and satisfies the property that

chunksOf n xs ++ chunksOf n ys == chunksOf n (xs ++ ys)

whenever n evenly divides the length of xs.

Split a list into chunks of the given lengths.

>>> splitPlaces [2,3,4] [1..20][[1,2],[3,4,5],[6,7,8,9]]

>>> splitPlaces [4,9] [1..10][[1,2,3,4],[5,6,7,8,9,10]]

>>> splitPlaces [4,9,3] [1..10][[1,2,3,4],[5,6,7,8,9,10]]

If the input list is longer than the total of the given lengths, then the remaining elements are dropped. If the list is shorter than the total of the given lengths, then the result may contain fewer chunks than requested, and the last chunk may be shorter than requested.

Split a list into chunks of the given lengths. Unlike splitPlaces, the output list will always be the same length as the first input argument. If the input list is longer than the total of the given lengths, then the remaining elements are dropped. If the list is shorter than the total of the given lengths, then the last several chunks will be shorter than requested or empty.

>>> splitPlacesBlanks [2,3,4] [1..20][[1,2],[3,4,5],[6,7,8,9]]

>>> splitPlacesBlanks [4,9] [1..10][[1,2,3,4],[5,6,7,8,9,10]]

>>> splitPlacesBlanks [4,9,3] [1..10][[1,2,3,4],[5,6,7,8,9,10],[]]

Notice the empty list in the output of the third example, which differs from the behavior of splitPlaces.

A useful recursion pattern for processing a list to produce a new list, often used for "chopping" up the input list. Typically chop is called with some function that will consume an initial prefix of the list and produce a value and the rest of the list.

For example, many common Prelude functions can be implemented in terms of chop:

group :: (Eq a) => [a] -> [[a]]
group = chop (\ xs@(x:_) -> span (==x) xs)

words :: String -> [String]
words = filter (not . null) . chop (break isSpace . dropWhile isSpace)

Divides up an input list into a set of sublists, according to n and m input specifications you provide. Each sublist will have n items, and the start of each sublist will be offset by m items from the previous one.

>>> divvy 5 5 [1..15][[1,2,3,4,5],[6,7,8,9,10],[11,12,13,14,15]]

>>> divvy 5 2 [1..15][[1,2,3,4,5],[3,4,5,6,7],[5,6,7,8,9],[7,8,9,10,11],[9,10,11,12,13],[11,12,13,14,15]]

In the case where a source list's trailing elements do no fill an entire sublist, those trailing elements will be dropped.

>>> divvy 5 2 [1..10][[1,2,3,4,5],[3,4,5,6,7],[5,6,7,8,9]]

As an example, you can generate a moving average over a list of prices:

type Prices = [Float]
type AveragePrices = [Float]

average :: [Float] -> Float
average xs = sum xs / (fromIntegral $ length xs)

simpleMovingAverage :: Prices -> AveragePrices
simpleMovingAverage = map average . divvy 20 1

A splitting strategy.

The default splitting strategy: keep delimiters in the output as separate chunks, don't condense multiple consecutive delimiters into one, keep initial and final blank chunks. Default delimiter is the constantly false predicate.

Note that defaultSplitter should normally not be used; use oneOf, onSublist, or whenElt instead, which are the same as the defaultSplitter with just the delimiter overridden.

The defaultSplitter strategy with any delimiter gives a maximally information-preserving splitting strategy, in the sense that (a) taking the concat of the output yields the original list, and (b) given only the output list, we can reconstruct a Splitter which would produce the same output list again given the original input list. This default strategy can be overridden to allow discarding various sorts of information.

Split a list according to the given splitting strategy. This is how to "run" a Splitter that has been built using the other combinators.