2 {-# OPTIONS_GHC -funbox-strict-fields #-}
3 {-# LANGUAGE NoImplicitPrelude, NoBangPatterns #-}
4 {-# OPTIONS_HADDOCK hide #-}
5 -----------------------------------------------------------------------------
8 -- Copyright : (c) The University of Glasgow, 1994-2000
9 -- License : see libraries/base/LICENSE
11 -- Maintainer : cvs-ghc@haskell.org
12 -- Stability : internal
13 -- Portability : non-portable (GHC extensions)
15 -- GHC\'s array implementation.
17 -----------------------------------------------------------------------------
35 %*********************************************************
37 \subsection{The @Ix@ class}
39 %*********************************************************
42 -- | The 'Ix' class is used to map a contiguous subrange of values in
43 -- a type onto integers. It is used primarily for array indexing
44 -- (see the array package).
46 -- The first argument @(l,u)@ of each of these operations is a pair
47 -- specifying the lower and upper bounds of a contiguous subrange of values.
49 -- An implementation is entitled to assume the following laws about these
52 -- * @'inRange' (l,u) i == 'elem' i ('range' (l,u))@
54 -- * @'range' (l,u) '!!' 'index' (l,u) i == i@, when @'inRange' (l,u) i@
56 -- * @'map' ('index' (l,u)) ('range' (l,u))) == [0..'rangeSize' (l,u)-1]@
58 -- * @'rangeSize' (l,u) == 'length' ('range' (l,u))@
60 -- Minimal complete instance: 'range', 'index' and 'inRange'.
62 class (Ord a) => Ix a where
63 -- | The list of values in the subrange defined by a bounding pair.
65 -- | The position of a subscript in the subrange.
66 index :: (a,a) -> a -> Int
67 -- | Like 'index', but without checking that the value is in range.
68 unsafeIndex :: (a,a) -> a -> Int
69 -- | Returns 'True' the given subscript lies in the range defined
71 inRange :: (a,a) -> a -> Bool
72 -- | The size of the subrange defined by a bounding pair.
73 rangeSize :: (a,a) -> Int
74 -- | like 'rangeSize', but without checking that the upper bound is
76 unsafeRangeSize :: (a,a) -> Int
78 -- Must specify one of index, unsafeIndex
80 -- 'index' is typically over-ridden in instances, with essentially
81 -- the same code, but using indexError instead of hopelessIndexError
82 -- Reason: we have 'Show' at the instances
83 {-# INLINE index #-} -- See Note [Inlining index]
84 index b i | inRange b i = unsafeIndex b i
85 | otherwise = hopelessIndexError
87 unsafeIndex b i = index b i
89 rangeSize b@(_l,h) | inRange b h = unsafeIndex b h + 1
90 | otherwise = 0 -- This case is only here to
91 -- check for an empty range
92 -- NB: replacing (inRange b h) by (l <= h) fails for
93 -- tuples. E.g. (1,2) <= (2,1) but the range is empty
95 unsafeRangeSize b@(_l,h) = unsafeIndex b h + 1
98 Note that the following is NOT right
99 rangeSize (l,h) | l <= h = index b h + 1
102 Because it might be the case that l<h, but the range
103 is nevertheless empty. Consider
105 Here l<h, but the second index ranges from 2..1 and
108 %*********************************************************
110 \subsection{Instances of @Ix@}
112 %*********************************************************
114 Note [Inlining index]
115 ~~~~~~~~~~~~~~~~~~~~~
116 We inline the 'index' operation,
118 * Partly because it generates much faster code
119 (although bigger); see Trac #1216
121 * Partly because it exposes the bounds checks to the simplifier which
124 If you make a per-instance index method, you may consider inlining it.
126 Note [Double bounds-checking of index values]
127 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
128 When you index an array, a!x, there are two possible bounds checks we might make:
130 (A) Check that (inRange (bounds a) x) holds.
132 (A) is checked in the method for 'index'
134 (B) Check that (index (bounds a) x) lies in the range 0..n,
135 where n is the size of the underlying array
137 (B) is checked in the top-level function (!), in safeIndex.
139 Of course it *should* be the case that (A) holds iff (B) holds, but that
140 is a property of the particular instances of index, bounds, and inRange,
141 so GHC cannot guarantee it.
143 * If you do (A) and not (B), then you might get a seg-fault,
144 by indexing at some bizarre location. Trac #1610
146 * If you do (B) but not (A), you may get no complaint when you index
147 an array out of its semantic bounds. Trac #2120
149 At various times we have had (A) and not (B), or (B) and not (A); both
150 led to complaints. So now we implement *both* checks (Trac #2669).
152 For 1-d, 2-d, and 3-d arrays of Int we have specialised instances to avoid this.
154 Note [Out-of-bounds error messages]
155 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
156 The default method for 'index' generates hoplelessIndexError, because
157 Ix doesn't have Show as a superclass. For particular base types we
158 can do better, so we override the default method for index.
161 -- Abstract these errors from the relevant index functions so that
162 -- the guts of the function will be small enough to inline.
164 {-# NOINLINE indexError #-}
165 indexError :: Show a => (a,a) -> a -> String -> b
167 = error (showString "Ix{" . showString tp . showString "}.index: Index " .
168 showParen True (showsPrec 0 i) .
169 showString " out of range " $
170 showParen True (showsPrec 0 rng) "")
172 hopelessIndexError :: Int -- Try to use 'indexError' instead!
173 hopelessIndexError = error "Error in array index"
175 ----------------------------------------------------------------------
176 instance Ix Char where
180 {-# INLINE unsafeIndex #-}
181 unsafeIndex (m,_n) i = fromEnum i - fromEnum m
183 {-# INLINE index #-} -- See Note [Out-of-bounds error messages]
184 -- and Note [Inlining index]
185 index b i | inRange b i = unsafeIndex b i
186 | otherwise = indexError b i "Char"
188 inRange (m,n) i = m <= i && i <= n
190 ----------------------------------------------------------------------
191 instance Ix Int where
193 -- The INLINE stops the build in the RHS from getting inlined,
194 -- so that callers can fuse with the result of range
197 {-# INLINE unsafeIndex #-}
198 unsafeIndex (m,_n) i = i - m
200 {-# INLINE index #-} -- See Note [Out-of-bounds error messages]
201 -- and Note [Inlining index]
202 index b i | inRange b i = unsafeIndex b i
203 | otherwise = indexError b i "Int"
205 {-# INLINE inRange #-}
206 inRange (I# m,I# n) (I# i) = m <=# i && i <=# n
208 ----------------------------------------------------------------------
209 instance Ix Integer where
213 {-# INLINE unsafeIndex #-}
214 unsafeIndex (m,_n) i = fromInteger (i - m)
216 {-# INLINE index #-} -- See Note [Out-of-bounds error messages]
217 -- and Note [Inlining index]
218 index b i | inRange b i = unsafeIndex b i
219 | otherwise = indexError b i "Integer"
221 inRange (m,n) i = m <= i && i <= n
223 ----------------------------------------------------------------------
224 instance Ix Bool where -- as derived
228 {-# INLINE unsafeIndex #-}
229 unsafeIndex (l,_) i = fromEnum i - fromEnum l
231 {-# INLINE index #-} -- See Note [Out-of-bounds error messages]
232 -- and Note [Inlining index]
233 index b i | inRange b i = unsafeIndex b i
234 | otherwise = indexError b i "Bool"
236 inRange (l,u) i = fromEnum i >= fromEnum l && fromEnum i <= fromEnum u
238 ----------------------------------------------------------------------
239 instance Ix Ordering where -- as derived
243 {-# INLINE unsafeIndex #-}
244 unsafeIndex (l,_) i = fromEnum i - fromEnum l
246 {-# INLINE index #-} -- See Note [Out-of-bounds error messages]
247 -- and Note [Inlining index]
248 index b i | inRange b i = unsafeIndex b i
249 | otherwise = indexError b i "Ordering"
251 inRange (l,u) i = fromEnum i >= fromEnum l && fromEnum i <= fromEnum u
253 ----------------------------------------------------------------------
256 range ((), ()) = [()]
257 {-# INLINE unsafeIndex #-}
258 unsafeIndex ((), ()) () = 0
259 {-# INLINE inRange #-}
260 inRange ((), ()) () = True
262 {-# INLINE index #-} -- See Note [Inlining index]
263 index b i = unsafeIndex b i
265 ----------------------------------------------------------------------
266 instance (Ix a, Ix b) => Ix (a, b) where -- as derived
267 {-# SPECIALISE instance Ix (Int,Int) #-}
270 range ((l1,l2),(u1,u2)) =
271 [ (i1,i2) | i1 <- range (l1,u1), i2 <- range (l2,u2) ]
273 {-# INLINE unsafeIndex #-}
274 unsafeIndex ((l1,l2),(u1,u2)) (i1,i2) =
275 unsafeIndex (l1,u1) i1 * unsafeRangeSize (l2,u2) + unsafeIndex (l2,u2) i2
277 {-# INLINE inRange #-}
278 inRange ((l1,l2),(u1,u2)) (i1,i2) =
279 inRange (l1,u1) i1 && inRange (l2,u2) i2
281 -- Default method for index
283 ----------------------------------------------------------------------
284 instance (Ix a1, Ix a2, Ix a3) => Ix (a1,a2,a3) where
285 {-# SPECIALISE instance Ix (Int,Int,Int) #-}
287 range ((l1,l2,l3),(u1,u2,u3)) =
288 [(i1,i2,i3) | i1 <- range (l1,u1),
292 unsafeIndex ((l1,l2,l3),(u1,u2,u3)) (i1,i2,i3) =
293 unsafeIndex (l3,u3) i3 + unsafeRangeSize (l3,u3) * (
294 unsafeIndex (l2,u2) i2 + unsafeRangeSize (l2,u2) * (
295 unsafeIndex (l1,u1) i1))
297 inRange ((l1,l2,l3),(u1,u2,u3)) (i1,i2,i3) =
298 inRange (l1,u1) i1 && inRange (l2,u2) i2 &&
301 -- Default method for index
303 ----------------------------------------------------------------------
304 instance (Ix a1, Ix a2, Ix a3, Ix a4) => Ix (a1,a2,a3,a4) where
305 range ((l1,l2,l3,l4),(u1,u2,u3,u4)) =
306 [(i1,i2,i3,i4) | i1 <- range (l1,u1),
311 unsafeIndex ((l1,l2,l3,l4),(u1,u2,u3,u4)) (i1,i2,i3,i4) =
312 unsafeIndex (l4,u4) i4 + unsafeRangeSize (l4,u4) * (
313 unsafeIndex (l3,u3) i3 + unsafeRangeSize (l3,u3) * (
314 unsafeIndex (l2,u2) i2 + unsafeRangeSize (l2,u2) * (
315 unsafeIndex (l1,u1) i1)))
317 inRange ((l1,l2,l3,l4),(u1,u2,u3,u4)) (i1,i2,i3,i4) =
318 inRange (l1,u1) i1 && inRange (l2,u2) i2 &&
319 inRange (l3,u3) i3 && inRange (l4,u4) i4
321 -- Default method for index
323 instance (Ix a1, Ix a2, Ix a3, Ix a4, Ix a5) => Ix (a1,a2,a3,a4,a5) where
324 range ((l1,l2,l3,l4,l5),(u1,u2,u3,u4,u5)) =
325 [(i1,i2,i3,i4,i5) | i1 <- range (l1,u1),
331 unsafeIndex ((l1,l2,l3,l4,l5),(u1,u2,u3,u4,u5)) (i1,i2,i3,i4,i5) =
332 unsafeIndex (l5,u5) i5 + unsafeRangeSize (l5,u5) * (
333 unsafeIndex (l4,u4) i4 + unsafeRangeSize (l4,u4) * (
334 unsafeIndex (l3,u3) i3 + unsafeRangeSize (l3,u3) * (
335 unsafeIndex (l2,u2) i2 + unsafeRangeSize (l2,u2) * (
336 unsafeIndex (l1,u1) i1))))
338 inRange ((l1,l2,l3,l4,l5),(u1,u2,u3,u4,u5)) (i1,i2,i3,i4,i5) =
339 inRange (l1,u1) i1 && inRange (l2,u2) i2 &&
340 inRange (l3,u3) i3 && inRange (l4,u4) i4 &&
343 -- Default method for index
346 %*********************************************************
348 \subsection{The @Array@ types}
350 %*********************************************************
353 type IPr = (Int, Int)
355 -- | The type of immutable non-strict (boxed) arrays
356 -- with indices in @i@ and elements in @e@.
357 -- The Int is the number of elements in the Array.
358 data Ix i => Array i e
359 = Array !i -- the lower bound, l
360 !i -- the upper bound, u
361 !Int -- a cache of (rangeSize (l,u))
362 -- used to make sure an index is
364 (Array# e) -- The actual elements
366 -- | Mutable, boxed, non-strict arrays in the 'ST' monad. The type
367 -- arguments are as follows:
369 -- * @s@: the state variable argument for the 'ST' type
371 -- * @i@: the index type of the array (should be an instance of 'Ix')
373 -- * @e@: the element type of the array.
376 = STArray !i -- the lower bound, l
377 !i -- the upper bound, u
378 !Int -- a cache of (rangeSize (l,u))
379 -- used to make sure an index is
381 (MutableArray# s e) -- The actual elements
382 -- No Ix context for STArray. They are stupid,
383 -- and force an Ix context on the equality instance.
385 -- Just pointer equality on mutable arrays:
386 instance Eq (STArray s i e) where
387 STArray _ _ _ arr1# == STArray _ _ _ arr2# =
388 sameMutableArray# arr1# arr2#
392 %*********************************************************
394 \subsection{Operations on immutable arrays}
396 %*********************************************************
399 {-# NOINLINE arrEleBottom #-}
401 arrEleBottom = error "(Array.!): undefined array element"
403 -- | Construct an array with the specified bounds and containing values
404 -- for given indices within these bounds.
406 -- The array is undefined (i.e. bottom) if any index in the list is
407 -- out of bounds. The Haskell 98 Report further specifies that if any
408 -- two associations in the list have the same index, the value at that
409 -- index is undefined (i.e. bottom). However in GHC's implementation,
410 -- the value at such an index is the value part of the last association
411 -- with that index in the list.
413 -- Because the indices must be checked for these errors, 'array' is
414 -- strict in the bounds argument and in the indices of the association
415 -- list, but nonstrict in the values. Thus, recurrences such as the
416 -- following are possible:
418 -- > a = array (1,100) ((1,1) : [(i, i * a!(i-1)) | i <- [2..100]])
420 -- Not every index within the bounds of the array need appear in the
421 -- association list, but the values associated with indices that do not
422 -- appear will be undefined (i.e. bottom).
424 -- If, in any dimension, the lower bound is greater than the upper bound,
425 -- then the array is legal, but empty. Indexing an empty array always
426 -- gives an array-bounds error, but 'bounds' still yields the bounds
427 -- with which the array was constructed.
430 => (i,i) -- ^ a pair of /bounds/, each of the index type
431 -- of the array. These bounds are the lowest and
432 -- highest indices in the array, in that order.
433 -- For example, a one-origin vector of length
434 -- '10' has bounds '(1,10)', and a one-origin '10'
435 -- by '10' matrix has bounds '((1,1),(10,10))'.
436 -> [(i, e)] -- ^ a list of /associations/ of the form
437 -- (/index/, /value/). Typically, this list will
438 -- be expressed as a comprehension. An
439 -- association '(i, x)' defines the value of
440 -- the array at index 'i' to be 'x'.
443 = let n = safeRangeSize (l,u)
444 in unsafeArray' (l,u) n
445 [(safeIndex (l,u) n i, e) | (i, e) <- ies]
447 {-# INLINE unsafeArray #-}
448 unsafeArray :: Ix i => (i,i) -> [(Int, e)] -> Array i e
449 unsafeArray b ies = unsafeArray' b (rangeSize b) ies
451 {-# INLINE unsafeArray' #-}
452 unsafeArray' :: Ix i => (i,i) -> Int -> [(Int, e)] -> Array i e
453 unsafeArray' (l,u) n@(I# n#) ies = runST (ST $ \s1# ->
454 case newArray# n# arrEleBottom s1# of
456 foldr (fill marr#) (done l u n marr#) ies s2#)
459 fill :: MutableArray# s e -> (Int, e) -> STRep s a -> STRep s a
460 fill marr# (I# i#, e) next s1# =
461 case writeArray# marr# i# e s1# of { s2# ->
465 done :: Ix i => i -> i -> Int -> MutableArray# s e -> STRep s (Array i e)
466 done l u n marr# s1# =
467 case unsafeFreezeArray# marr# s1# of
468 (# s2#, arr# #) -> (# s2#, Array l u n arr# #)
470 -- This is inefficient and I'm not sure why:
471 -- listArray (l,u) es = unsafeArray (l,u) (zip [0 .. rangeSize (l,u) - 1] es)
472 -- The code below is better. It still doesn't enable foldr/build
473 -- transformation on the list of elements; I guess it's impossible
474 -- using mechanisms currently available.
476 -- | Construct an array from a pair of bounds and a list of values in
478 {-# INLINE listArray #-}
479 listArray :: Ix i => (i,i) -> [e] -> Array i e
480 listArray (l,u) es = runST (ST $ \s1# ->
481 case safeRangeSize (l,u) of { n@(I# n#) ->
482 case newArray# n# arrEleBottom s1# of { (# s2#, marr# #) ->
483 let fillFromList i# xs s3# | i# ==# n# = s3#
484 | otherwise = case xs of
486 y:ys -> case writeArray# marr# i# y s3# of { s4# ->
487 fillFromList (i# +# 1#) ys s4# } in
488 case fillFromList 0# es s2# of { s3# ->
489 done l u n marr# s3# }}})
491 -- | The value at the given index in an array.
493 (!) :: Ix i => Array i e -> i -> e
494 arr@(Array l u n _) ! i = unsafeAt arr $ safeIndex (l,u) n i
496 {-# INLINE safeRangeSize #-}
497 safeRangeSize :: Ix i => (i, i) -> Int
498 safeRangeSize (l,u) = let r = rangeSize (l, u)
499 in if r < 0 then negRange
502 -- Don't inline this error message everywhere!!
503 negRange :: Int -- Uninformative, but Ix does not provide Show
504 negRange = error "Negative range size"
506 {-# INLINE[1] safeIndex #-}
507 -- See Note [Double bounds-checking of index values]
508 -- Inline *after* (!) so the rules can fire
509 safeIndex :: Ix i => (i, i) -> Int -> i -> Int
510 safeIndex (l,u) n i = let i' = index (l,u) i
511 in if (0 <= i') && (i' < n)
513 else badSafeIndex i' n
515 -- See Note [Double bounds-checking of index values]
517 "safeIndex/I" safeIndex = lessSafeIndex :: (Int,Int) -> Int -> Int -> Int
518 "safeIndex/(I,I)" safeIndex = lessSafeIndex :: ((Int,Int),(Int,Int)) -> Int -> (Int,Int) -> Int
519 "safeIndex/(I,I,I)" safeIndex = lessSafeIndex :: ((Int,Int,Int),(Int,Int,Int)) -> Int -> (Int,Int,Int) -> Int
522 lessSafeIndex :: Ix i => (i, i) -> Int -> i -> Int
523 -- See Note [Double bounds-checking of index values]
524 -- Do only (A), the semantic check
525 lessSafeIndex (l,u) _ i = index (l,u) i
527 -- Don't inline this long error message everywhere!!
528 badSafeIndex :: Int -> Int -> Int
529 badSafeIndex i' n = error ("Error in array index; " ++ show i' ++
530 " not in range [0.." ++ show n ++ ")")
532 {-# INLINE unsafeAt #-}
533 unsafeAt :: Ix i => Array i e -> Int -> e
534 unsafeAt (Array _ _ _ arr#) (I# i#) =
535 case indexArray# arr# i# of (# e #) -> e
537 -- | The bounds with which an array was constructed.
538 {-# INLINE bounds #-}
539 bounds :: Ix i => Array i e -> (i,i)
540 bounds (Array l u _ _) = (l,u)
542 -- | The number of elements in the array.
543 {-# INLINE numElements #-}
544 numElements :: Ix i => Array i e -> Int
545 numElements (Array _ _ n _) = n
547 -- | The list of indices of an array in ascending order.
548 {-# INLINE indices #-}
549 indices :: Ix i => Array i e -> [i]
550 indices (Array l u _ _) = range (l,u)
552 -- | The list of elements of an array in index order.
554 elems :: Ix i => Array i e -> [e]
555 elems arr@(Array _ _ n _) =
556 [unsafeAt arr i | i <- [0 .. n - 1]]
558 -- | The list of associations of an array in index order.
559 {-# INLINE assocs #-}
560 assocs :: Ix i => Array i e -> [(i, e)]
561 assocs arr@(Array l u _ _) =
562 [(i, arr ! i) | i <- range (l,u)]
564 -- | The 'accumArray' deals with repeated indices in the association
565 -- list using an /accumulating function/ which combines the values of
566 -- associations with the same index.
567 -- For example, given a list of values of some index type, @hist@
568 -- produces a histogram of the number of occurrences of each index within
569 -- a specified range:
571 -- > hist :: (Ix a, Num b) => (a,a) -> [a] -> Array a b
572 -- > hist bnds is = accumArray (+) 0 bnds [(i, 1) | i<-is, inRange bnds i]
574 -- If the accumulating function is strict, then 'accumArray' is strict in
575 -- the values, as well as the indices, in the association list. Thus,
576 -- unlike ordinary arrays built with 'array', accumulated arrays should
577 -- not in general be recursive.
578 {-# INLINE accumArray #-}
580 => (e -> a -> e) -- ^ accumulating function
581 -> e -- ^ initial value
582 -> (i,i) -- ^ bounds of the array
583 -> [(i, a)] -- ^ association list
585 accumArray f initial (l,u) ies =
586 let n = safeRangeSize (l,u)
587 in unsafeAccumArray' f initial (l,u) n
588 [(safeIndex (l,u) n i, e) | (i, e) <- ies]
590 {-# INLINE unsafeAccumArray #-}
591 unsafeAccumArray :: Ix i => (e -> a -> e) -> e -> (i,i) -> [(Int, a)] -> Array i e
592 unsafeAccumArray f initial b ies = unsafeAccumArray' f initial b (rangeSize b) ies
594 {-# INLINE unsafeAccumArray' #-}
595 unsafeAccumArray' :: Ix i => (e -> a -> e) -> e -> (i,i) -> Int -> [(Int, a)] -> Array i e
596 unsafeAccumArray' f initial (l,u) n@(I# n#) ies = runST (ST $ \s1# ->
597 case newArray# n# initial s1# of { (# s2#, marr# #) ->
598 foldr (adjust f marr#) (done l u n marr#) ies s2# })
600 {-# INLINE adjust #-}
601 adjust :: (e -> a -> e) -> MutableArray# s e -> (Int, a) -> STRep s b -> STRep s b
602 adjust f marr# (I# i#, new) next s1# =
603 case readArray# marr# i# s1# of
605 case writeArray# marr# i# (f old new) s2# of
608 -- | Constructs an array identical to the first argument except that it has
609 -- been updated by the associations in the right argument.
610 -- For example, if @m@ is a 1-origin, @n@ by @n@ matrix, then
612 -- > m//[((i,i), 0) | i <- [1..n]]
614 -- is the same matrix, except with the diagonal zeroed.
616 -- Repeated indices in the association list are handled as for 'array':
617 -- Haskell 98 specifies that the resulting array is undefined (i.e. bottom),
618 -- but GHC's implementation uses the last association for each index.
620 (//) :: Ix i => Array i e -> [(i, e)] -> Array i e
621 arr@(Array l u n _) // ies =
622 unsafeReplace arr [(safeIndex (l,u) n i, e) | (i, e) <- ies]
624 {-# INLINE unsafeReplace #-}
625 unsafeReplace :: Ix i => Array i e -> [(Int, e)] -> Array i e
626 unsafeReplace arr ies = runST (do
627 STArray l u n marr# <- thawSTArray arr
628 ST (foldr (fill marr#) (done l u n marr#) ies))
630 -- | @'accum' f@ takes an array and an association list and accumulates
631 -- pairs from the list into the array with the accumulating function @f@.
632 -- Thus 'accumArray' can be defined using 'accum':
634 -- > accumArray f z b = accum f (array b [(i, z) | i <- range b])
637 accum :: Ix i => (e -> a -> e) -> Array i e -> [(i, a)] -> Array i e
638 accum f arr@(Array l u n _) ies =
639 unsafeAccum f arr [(safeIndex (l,u) n i, e) | (i, e) <- ies]
641 {-# INLINE unsafeAccum #-}
642 unsafeAccum :: Ix i => (e -> a -> e) -> Array i e -> [(Int, a)] -> Array i e
643 unsafeAccum f arr ies = runST (do
644 STArray l u n marr# <- thawSTArray arr
645 ST (foldr (adjust f marr#) (done l u n marr#) ies))
648 amap :: Ix i => (a -> b) -> Array i a -> Array i b
649 amap f arr@(Array l u n _) =
650 unsafeArray' (l,u) n [(i, f (unsafeAt arr i)) | i <- [0 .. n - 1]]
652 -- | 'ixmap' allows for transformations on array indices.
653 -- It may be thought of as providing function composition on the right
654 -- with the mapping that the original array embodies.
656 -- A similar transformation of array values may be achieved using 'fmap'
657 -- from the 'Array' instance of the 'Functor' class.
659 ixmap :: (Ix i, Ix j) => (i,i) -> (i -> j) -> Array j e -> Array i e
661 array (l,u) [(i, arr ! f i) | i <- range (l,u)]
663 {-# INLINE eqArray #-}
664 eqArray :: (Ix i, Eq e) => Array i e -> Array i e -> Bool
665 eqArray arr1@(Array l1 u1 n1 _) arr2@(Array l2 u2 n2 _) =
666 if n1 == 0 then n2 == 0 else
667 l1 == l2 && u1 == u2 &&
668 and [unsafeAt arr1 i == unsafeAt arr2 i | i <- [0 .. n1 - 1]]
670 {-# INLINE cmpArray #-}
671 cmpArray :: (Ix i, Ord e) => Array i e -> Array i e -> Ordering
672 cmpArray arr1 arr2 = compare (assocs arr1) (assocs arr2)
674 {-# INLINE cmpIntArray #-}
675 cmpIntArray :: Ord e => Array Int e -> Array Int e -> Ordering
676 cmpIntArray arr1@(Array l1 u1 n1 _) arr2@(Array l2 u2 n2 _) =
678 if n2 == 0 then EQ else LT
679 else if n2 == 0 then GT
680 else case compare l1 l2 of
681 EQ -> foldr cmp (compare u1 u2) [0 .. (n1 `min` n2) - 1]
684 cmp i rest = case compare (unsafeAt arr1 i) (unsafeAt arr2 i) of
688 {-# RULES "cmpArray/Int" cmpArray = cmpIntArray #-}
692 %*********************************************************
694 \subsection{Array instances}
696 %*********************************************************
699 instance Ix i => Functor (Array i) where
702 instance (Ix i, Eq e) => Eq (Array i e) where
705 instance (Ix i, Ord e) => Ord (Array i e) where
708 instance (Ix a, Show a, Show b) => Show (Array a b) where
710 showParen (p > appPrec) $
711 showString "array " .
712 showsPrec appPrec1 (bounds a) .
714 showsPrec appPrec1 (assocs a)
715 -- Precedence of 'array' is the precedence of application
717 -- The Read instance is in GHC.Read
721 %*********************************************************
723 \subsection{Operations on mutable arrays}
725 %*********************************************************
727 Idle ADR question: What's the tradeoff here between flattening these
728 datatypes into @STArray ix ix (MutableArray# s elt)@ and using
729 it as is? As I see it, the former uses slightly less heap and
730 provides faster access to the individual parts of the bounds while the
731 code used has the benefit of providing a ready-made @(lo, hi)@ pair as
732 required by many array-related functions. Which wins? Is the
733 difference significant (probably not).
735 Idle AJG answer: When I looked at the outputted code (though it was 2
736 years ago) it seems like you often needed the tuple, and we build
737 it frequently. Now we've got the overloading specialiser things
738 might be different, though.
741 {-# INLINE newSTArray #-}
742 newSTArray :: Ix i => (i,i) -> e -> ST s (STArray s i e)
743 newSTArray (l,u) initial = ST $ \s1# ->
744 case safeRangeSize (l,u) of { n@(I# n#) ->
745 case newArray# n# initial s1# of { (# s2#, marr# #) ->
746 (# s2#, STArray l u n marr# #) }}
748 {-# INLINE boundsSTArray #-}
749 boundsSTArray :: STArray s i e -> (i,i)
750 boundsSTArray (STArray l u _ _) = (l,u)
752 {-# INLINE numElementsSTArray #-}
753 numElementsSTArray :: STArray s i e -> Int
754 numElementsSTArray (STArray _ _ n _) = n
756 {-# INLINE readSTArray #-}
757 readSTArray :: Ix i => STArray s i e -> i -> ST s e
758 readSTArray marr@(STArray l u n _) i =
759 unsafeReadSTArray marr (safeIndex (l,u) n i)
761 {-# INLINE unsafeReadSTArray #-}
762 unsafeReadSTArray :: Ix i => STArray s i e -> Int -> ST s e
763 unsafeReadSTArray (STArray _ _ _ marr#) (I# i#)
764 = ST $ \s1# -> readArray# marr# i# s1#
766 {-# INLINE writeSTArray #-}
767 writeSTArray :: Ix i => STArray s i e -> i -> e -> ST s ()
768 writeSTArray marr@(STArray l u n _) i e =
769 unsafeWriteSTArray marr (safeIndex (l,u) n i) e
771 {-# INLINE unsafeWriteSTArray #-}
772 unsafeWriteSTArray :: Ix i => STArray s i e -> Int -> e -> ST s ()
773 unsafeWriteSTArray (STArray _ _ _ marr#) (I# i#) e = ST $ \s1# ->
774 case writeArray# marr# i# e s1# of
779 %*********************************************************
781 \subsection{Moving between mutable and immutable}
783 %*********************************************************
786 freezeSTArray :: Ix i => STArray s i e -> ST s (Array i e)
787 freezeSTArray (STArray l u n@(I# n#) marr#) = ST $ \s1# ->
788 case newArray# n# arrEleBottom s1# of { (# s2#, marr'# #) ->
789 let copy i# s3# | i# ==# n# = s3#
791 case readArray# marr# i# s3# of { (# s4#, e #) ->
792 case writeArray# marr'# i# e s4# of { s5# ->
793 copy (i# +# 1#) s5# }} in
794 case copy 0# s2# of { s3# ->
795 case unsafeFreezeArray# marr'# s3# of { (# s4#, arr# #) ->
796 (# s4#, Array l u n arr# #) }}}
798 {-# INLINE unsafeFreezeSTArray #-}
799 unsafeFreezeSTArray :: Ix i => STArray s i e -> ST s (Array i e)
800 unsafeFreezeSTArray (STArray l u n marr#) = ST $ \s1# ->
801 case unsafeFreezeArray# marr# s1# of { (# s2#, arr# #) ->
802 (# s2#, Array l u n arr# #) }
804 thawSTArray :: Ix i => Array i e -> ST s (STArray s i e)
805 thawSTArray (Array l u n@(I# n#) arr#) = ST $ \s1# ->
806 case newArray# n# arrEleBottom s1# of { (# s2#, marr# #) ->
807 let copy i# s3# | i# ==# n# = s3#
809 case indexArray# arr# i# of { (# e #) ->
810 case writeArray# marr# i# e s3# of { s4# ->
811 copy (i# +# 1#) s4# }} in
812 case copy 0# s2# of { s3# ->
813 (# s3#, STArray l u n marr# #) }}
815 {-# INLINE unsafeThawSTArray #-}
816 unsafeThawSTArray :: Ix i => Array i e -> ST s (STArray s i e)
817 unsafeThawSTArray (Array l u n arr#) = ST $ \s1# ->
818 case unsafeThawArray# arr# s1# of { (# s2#, marr# #) ->
819 (# s2#, STArray l u n marr# #) }