Haskell中用递归与模式匹配实现连续重复列表元素去重

Let's start by breaking down what's not working with your current code. Right now, your implementation uses a list comprehension that filters out all occurrences of x from the recursive call result—not just the consecutive duplicates. That means if you ran unrepeat [1,1,2,1], you'd get [1,2] instead of the expected [1,2,1]—the non-consecutive 1 gets incorrectly removed too.

The Correct Recursive Approach with Pattern Matching #

We need to focus only on adjacent, consecutive duplicates. Here's how to structure the recursion properly by checking adjacent elements directly in the pattern matches:

unrepeat :: Eq a => [a] -> [a]
-- Base case: empty list stays empty
unrepeat [] = []
-- Base case: single element list stays unchanged
unrepeat [x] = [x]
-- Handle cases with at least two elements
unrepeat (x:y:xs)
  | x == y    = unrepeat (y:xs)  -- Skip x (it's a consecutive duplicate), keep processing the rest
  | otherwise = x : unrepeat (y:xs)  -- Keep x, then process the rest starting with y

How This Works (Walkthrough of Your Examples) #

Let’s verify with your sample inputs to see the logic in action:

• For unrepeat [True,True,True,True]:

  1. Start with True:True:[True,True] → x and y are equal, so call unrepeat (True:[True,True])
  2. Next, True:True:[True] → again equal, call unrepeat (True:[True])
  3. Then True:True:[] → equal, call unrepeat [True]
  4. Single element returns [True] → final result is [True]

• For unrepeat [1,1,2,1,3,3,3]:

  1. 1:1:[2,1,3,3,3] → equal, call unrepeat (1:[2,1,3,3,3])
  2. 1:2:[1,3,3,3] → not equal, so 1 : unrepeat (2:[1,3,3,3])
  3. 2:1:[3,3,3] → not equal, so 2 : unrepeat (1:[3,3,3])
  4. 1:3:[3,3] → not equal, so 1 : unrepeat (3:[3,3])
  5. 3:3:[3] → equal, call unrepeat (3:[3])
  6. 3:3:[] → equal, call unrepeat [3]
  7. Single element returns [3] → building back up gives [1,2,1,3]

Why Your Original Code Failed #

Your line x : [ k | k <- unrepeat(xs), k /=x] was removing every instance of x from the rest of the list, not just consecutive ones. For example, in [1,1,2,1], after processing the first 1, it would filter out the final 1 from the recursive result of unrepeat [1,2,1], leading to an incorrect output.

内容的提问来源于stack exchange，提问作者Coding Elf