[project @ 2005-09-22 09:43:01 by ross]
[packages/random.git] / Text / PrettyPrint / HughesPJ.hs
1 -----------------------------------------------------------------------------
2 -- |
3 -- Module : Text.PrettyPrint.HughesPJ
4 -- Copyright : (c) The University of Glasgow 2001
5 -- License : BSD-style (see the file libraries/base/LICENSE)
6 --
7 -- Maintainer : libraries@haskell.org
8 -- Stability : provisional
9 -- Portability : portable
10 --
11 -- John Hughes's and Simon Peyton Jones's Pretty Printer Combinators
12 --
13 -- Based on /The Design of a Pretty-printing Library/
14 -- in Advanced Functional Programming,
15 -- Johan Jeuring and Erik Meijer (eds), LNCS 925
16 -- <http://www.cs.chalmers.se/~rjmh/Papers/pretty.ps>
17 --
18 -- Heavily modified by Simon Peyton Jones, Dec 96
19 --
20 -----------------------------------------------------------------------------
21
22 {-
23 Version 3.0 28 May 1997
24 * Cured massive performance bug. If you write
25
26 foldl <> empty (map (text.show) [1..10000])
27
28 you get quadratic behaviour with V2.0. Why? For just the same
29 reason as you get quadratic behaviour with left-associated (++)
30 chains.
31
32 This is really bad news. One thing a pretty-printer abstraction
33 should certainly guarantee is insensivity to associativity. It
34 matters: suddenly GHC's compilation times went up by a factor of
35 100 when I switched to the new pretty printer.
36
37 I fixed it with a bit of a hack (because I wanted to get GHC back
38 on the road). I added two new constructors to the Doc type, Above
39 and Beside:
40
41 <> = Beside
42 $$ = Above
43
44 Then, where I need to get to a "TextBeside" or "NilAbove" form I
45 "force" the Doc to squeeze out these suspended calls to Beside and
46 Above; but in so doing I re-associate. It's quite simple, but I'm
47 not satisfied that I've done the best possible job. I'll send you
48 the code if you are interested.
49
50 * Added new exports:
51 punctuate, hang
52 int, integer, float, double, rational,
53 lparen, rparen, lbrack, rbrack, lbrace, rbrace,
54
55 * fullRender's type signature has changed. Rather than producing a
56 string it now takes an extra couple of arguments that tells it how
57 to glue fragments of output together:
58
59 fullRender :: Mode
60 -> Int -- Line length
61 -> Float -- Ribbons per line
62 -> (TextDetails -> a -> a) -- What to do with text
63 -> a -- What to do at the end
64 -> Doc
65 -> a -- Result
66
67 The "fragments" are encapsulated in the TextDetails data type:
68
69 data TextDetails = Chr Char
70 | Str String
71 | PStr FAST_STRING
72
73 The Chr and Str constructors are obvious enough. The PStr
74 constructor has a packed string (FAST_STRING) inside it. It's
75 generated by using the new "ptext" export.
76
77 An advantage of this new setup is that you can get the renderer to
78 do output directly (by passing in a function of type (TextDetails
79 -> IO () -> IO ()), rather than producing a string that you then
80 print.
81
82
83 Version 2.0 24 April 1997
84 * Made empty into a left unit for <> as well as a right unit;
85 it is also now true that
86 nest k empty = empty
87 which wasn't true before.
88
89 * Fixed an obscure bug in sep that occassionally gave very weird behaviour
90
91 * Added $+$
92
93 * Corrected and tidied up the laws and invariants
94
95 ======================================================================
96 Relative to John's original paper, there are the following new features:
97
98 1. There's an empty document, "empty". It's a left and right unit for
99 both <> and $$, and anywhere in the argument list for
100 sep, hcat, hsep, vcat, fcat etc.
101
102 It is Really Useful in practice.
103
104 2. There is a paragraph-fill combinator, fsep, that's much like sep,
105 only it keeps fitting things on one line until it can't fit any more.
106
107 3. Some random useful extra combinators are provided.
108 <+> puts its arguments beside each other with a space between them,
109 unless either argument is empty in which case it returns the other
110
111
112 hcat is a list version of <>
113 hsep is a list version of <+>
114 vcat is a list version of $$
115
116 sep (separate) is either like hsep or like vcat, depending on what fits
117
118 cat behaves like sep, but it uses <> for horizontal conposition
119 fcat behaves like fsep, but it uses <> for horizontal conposition
120
121 These new ones do the obvious things:
122 char, semi, comma, colon, space,
123 parens, brackets, braces,
124 quotes, doubleQuotes
125
126 4. The "above" combinator, $$, now overlaps its two arguments if the
127 last line of the top argument stops before the first line of the
128 second begins.
129
130 For example: text "hi" $$ nest 5 (text "there")
131 lays out as
132 hi there
133 rather than
134 hi
135 there
136
137 There are two places this is really useful
138
139 a) When making labelled blocks, like this:
140 Left -> code for left
141 Right -> code for right
142 LongLongLongLabel ->
143 code for longlonglonglabel
144 The block is on the same line as the label if the label is
145 short, but on the next line otherwise.
146
147 b) When laying out lists like this:
148 [ first
149 , second
150 , third
151 ]
152 which some people like. But if the list fits on one line
153 you want [first, second, third]. You can't do this with
154 John's original combinators, but it's quite easy with the
155 new $$.
156
157 The combinator $+$ gives the original "never-overlap" behaviour.
158
159 5. Several different renderers are provided:
160 * a standard one
161 * one that uses cut-marks to avoid deeply-nested documents
162 simply piling up in the right-hand margin
163 * one that ignores indentation (fewer chars output; good for machines)
164 * one that ignores indentation and newlines (ditto, only more so)
165
166 6. Numerous implementation tidy-ups
167 Use of unboxed data types to speed up the implementation
168 -}
169
170 module Text.PrettyPrint.HughesPJ (
171
172 -- * The document type
173 Doc, -- Abstract
174
175 -- * Constructing documents
176 -- ** Converting values into documents
177 char, text, ptext,
178 int, integer, float, double, rational,
179
180 -- ** Simple derived documents
181 semi, comma, colon, space, equals,
182 lparen, rparen, lbrack, rbrack, lbrace, rbrace,
183
184 -- ** Wrapping documents in delimiters
185 parens, brackets, braces, quotes, doubleQuotes,
186
187 -- ** Combining documents
188 empty,
189 (<>), (<+>), hcat, hsep,
190 ($$), ($+$), vcat,
191 sep, cat,
192 fsep, fcat,
193 nest,
194 hang, punctuate,
195
196 -- * Predicates on documents
197 isEmpty,
198
199 -- * Rendering documents
200
201 -- ** Default rendering
202 render,
203
204 -- ** Rendering with a particular style
205 Style(..),
206 style,
207 renderStyle,
208
209 -- ** General rendering
210 fullRender,
211 Mode(..), TextDetails(..),
212
213 ) where
214
215
216 import Prelude
217
218 infixl 6 <>
219 infixl 6 <+>
220 infixl 5 $$, $+$
221
222 -- ---------------------------------------------------------------------------
223 -- The interface
224
225 -- The primitive Doc values
226
227 isEmpty :: Doc -> Bool; -- ^ Returns 'True' if the document is empty
228
229 -- | The empty document, with no height and no width.
230 -- 'empty' is the identity for '<>', '<+>', '$$' and '$+$', and anywhere
231 -- in the argument list for 'sep', 'hcat', 'hsep', 'vcat', 'fcat' etc.
232 empty :: Doc
233
234 semi :: Doc; -- ^ A ';' character
235 comma :: Doc; -- ^ A ',' character
236 colon :: Doc; -- ^ A ':' character
237 space :: Doc; -- ^ A space character
238 equals :: Doc; -- ^ A '=' character
239 lparen :: Doc; -- ^ A '(' character
240 rparen :: Doc; -- ^ A ')' character
241 lbrack :: Doc; -- ^ A '[' character
242 rbrack :: Doc; -- ^ A ']' character
243 lbrace :: Doc; -- ^ A '{' character
244 rbrace :: Doc; -- ^ A '}' character
245
246 -- | A document of height and width 1, containing a literal character.
247 char :: Char -> Doc
248
249 -- | A document of height 1 containing a literal string.
250 -- 'text' satisfies the following laws:
251 --
252 -- * @'text' s '<>' 'text' t = 'text' (s'++'t)@
253 --
254 -- * @'text' \"\" '<>' x = x@, if @x@ non-empty
255 --
256 -- The side condition on the last law is necessary because @'text' \"\"@
257 -- has height 1, while 'empty' has no height.
258 text :: String -> Doc
259
260 -- | An obsolete function, now identical to 'text'.
261 ptext :: String -> Doc
262
263 int :: Int -> Doc; -- ^ @int n = text (show n)@
264 integer :: Integer -> Doc; -- ^ @integer n = text (show n)@
265 float :: Float -> Doc; -- ^ @float n = text (show n)@
266 double :: Double -> Doc; -- ^ @double n = text (show n)@
267 rational :: Rational -> Doc; -- ^ @rational n = text (show n)@
268
269 parens :: Doc -> Doc; -- ^ Wrap document in @(...)@
270 brackets :: Doc -> Doc; -- ^ Wrap document in @[...]@
271 braces :: Doc -> Doc; -- ^ Wrap document in @{...}@
272 quotes :: Doc -> Doc; -- ^ Wrap document in @\'...\'@
273 doubleQuotes :: Doc -> Doc; -- ^ Wrap document in @\"...\"@
274
275 -- Combining @Doc@ values
276
277 -- | Beside.
278 -- '<>' is associative, with identity 'empty'.
279 (<>) :: Doc -> Doc -> Doc
280
281 -- | Beside, separated by space, unless one of the arguments is 'empty'.
282 -- '<+>' is associative, with identity 'empty'.
283 (<+>) :: Doc -> Doc -> Doc
284
285 -- | Above, except that if the last line of the first argument stops
286 -- at least one position before the first line of the second begins,
287 -- these two lines are overlapped. For example:
288 --
289 -- > text "hi" $$ nest 5 (text "there")
290 --
291 -- lays out as
292 --
293 -- > hi there
294 --
295 -- rather than
296 --
297 -- > hi
298 -- > there
299 --
300 -- '$$' is associative, with identity 'empty', and also satisfies
301 --
302 -- * @(x '$$' y) '<>' z = x '$$' (y '<>' z)@, if @y@ non-empty.
303 --
304 ($$) :: Doc -> Doc -> Doc
305
306 -- | Above, with no overlapping.
307 -- '$+$' is associative, with identity 'empty'.
308 ($+$) :: Doc -> Doc -> Doc
309
310 hcat :: [Doc] -> Doc; -- ^List version of '<>'.
311 hsep :: [Doc] -> Doc; -- ^List version of '<+>'.
312 vcat :: [Doc] -> Doc; -- ^List version of '$$'.
313
314 cat :: [Doc] -> Doc; -- ^ Either 'hcat' or 'vcat'.
315 sep :: [Doc] -> Doc; -- ^ Either 'hsep' or 'vcat'.
316 fcat :: [Doc] -> Doc; -- ^ \"Paragraph fill\" version of 'cat'.
317 fsep :: [Doc] -> Doc; -- ^ \"Paragraph fill\" version of 'sep'.
318
319 -- | Nest (or indent) a document by a given number of positions
320 -- (which may also be negative). 'nest' satisfies the laws:
321 --
322 -- * @'nest' 0 x = x@
323 --
324 -- * @'nest' k ('nest' k' x) = 'nest' (k+k') x@
325 --
326 -- * @'nest' k (x '<>' y) = 'nest' k z '<>' 'nest' k y@
327 --
328 -- * @'nest' k (x '$$' y) = 'nest' k x '$$' 'nest' k y@
329 --
330 -- * @'nest' k 'empty' = 'empty'@
331 --
332 -- * @x '<>' 'nest' k y = x '<>' y@, if @x@ non-empty
333 --
334 -- The side condition on the last law is needed because
335 -- 'empty' is a left identity for '<>'.
336 nest :: Int -> Doc -> Doc
337
338 -- GHC-specific ones.
339
340 -- | @hang d1 n d2 = sep [d1, nest n d2]@
341 hang :: Doc -> Int -> Doc -> Doc
342
343 -- | @punctuate p [d1, ... dn] = [d1 \<> p, d2 \<> p, ... dn-1 \<> p, dn]@
344 punctuate :: Doc -> [Doc] -> [Doc]
345
346
347 -- Displaying @Doc@ values.
348
349 instance Show Doc where
350 showsPrec prec doc cont = showDoc doc cont
351
352 -- | Renders the document as a string using the default 'style'.
353 render :: Doc -> String
354
355 -- | The general rendering interface.
356 fullRender :: Mode -- ^Rendering mode
357 -> Int -- ^Line length
358 -> Float -- ^Ribbons per line
359 -> (TextDetails -> a -> a) -- ^What to do with text
360 -> a -- ^What to do at the end
361 -> Doc -- ^The document
362 -> a -- ^Result
363
364 -- | Render the document as a string using a specified style.
365 renderStyle :: Style -> Doc -> String
366
367 -- | A rendering style.
368 data Style
369 = Style { mode :: Mode -- ^ The rendering mode
370 , lineLength :: Int -- ^ Length of line, in chars
371 , ribbonsPerLine :: Float -- ^ Ratio of ribbon length to line length
372 }
373
374 -- | The default style (@mode=PageMode, lineLength=100, ribbonsPerLine=1.5@).
375 style :: Style
376 style = Style { lineLength = 100, ribbonsPerLine = 1.5, mode = PageMode }
377
378 -- | Rendering mode.
379 data Mode = PageMode -- ^Normal
380 | ZigZagMode -- ^With zig-zag cuts
381 | LeftMode -- ^No indentation, infinitely long lines
382 | OneLineMode -- ^All on one line
383
384 -- ---------------------------------------------------------------------------
385 -- The Doc calculus
386
387 -- The Doc combinators satisfy the following laws:
388
389 {-
390 Laws for $$
391 ~~~~~~~~~~~
392 <a1> (x $$ y) $$ z = x $$ (y $$ z)
393 <a2> empty $$ x = x
394 <a3> x $$ empty = x
395
396 ...ditto $+$...
397
398 Laws for <>
399 ~~~~~~~~~~~
400 <b1> (x <> y) <> z = x <> (y <> z)
401 <b2> empty <> x = empty
402 <b3> x <> empty = x
403
404 ...ditto <+>...
405
406 Laws for text
407 ~~~~~~~~~~~~~
408 <t1> text s <> text t = text (s++t)
409 <t2> text "" <> x = x, if x non-empty
410
411 Laws for nest
412 ~~~~~~~~~~~~~
413 <n1> nest 0 x = x
414 <n2> nest k (nest k' x) = nest (k+k') x
415 <n3> nest k (x <> y) = nest k z <> nest k y
416 <n4> nest k (x $$ y) = nest k x $$ nest k y
417 <n5> nest k empty = empty
418 <n6> x <> nest k y = x <> y, if x non-empty
419
420 ** Note the side condition on <n6>! It is this that
421 ** makes it OK for empty to be a left unit for <>.
422
423 Miscellaneous
424 ~~~~~~~~~~~~~
425 <m1> (text s <> x) $$ y = text s <> ((text "" <> x)) $$
426 nest (-length s) y)
427
428 <m2> (x $$ y) <> z = x $$ (y <> z)
429 if y non-empty
430
431
432 Laws for list versions
433 ~~~~~~~~~~~~~~~~~~~~~~
434 <l1> sep (ps++[empty]++qs) = sep (ps ++ qs)
435 ...ditto hsep, hcat, vcat, fill...
436
437 <l2> nest k (sep ps) = sep (map (nest k) ps)
438 ...ditto hsep, hcat, vcat, fill...
439
440 Laws for oneLiner
441 ~~~~~~~~~~~~~~~~~
442 <o1> oneLiner (nest k p) = nest k (oneLiner p)
443 <o2> oneLiner (x <> y) = oneLiner x <> oneLiner y
444
445 You might think that the following verion of <m1> would
446 be neater:
447
448 <3 NO> (text s <> x) $$ y = text s <> ((empty <> x)) $$
449 nest (-length s) y)
450
451 But it doesn't work, for if x=empty, we would have
452
453 text s $$ y = text s <> (empty $$ nest (-length s) y)
454 = text s <> nest (-length s) y
455 -}
456
457 -- ---------------------------------------------------------------------------
458 -- Simple derived definitions
459
460 semi = char ';'
461 colon = char ':'
462 comma = char ','
463 space = char ' '
464 equals = char '='
465 lparen = char '('
466 rparen = char ')'
467 lbrack = char '['
468 rbrack = char ']'
469 lbrace = char '{'
470 rbrace = char '}'
471
472 int n = text (show n)
473 integer n = text (show n)
474 float n = text (show n)
475 double n = text (show n)
476 rational n = text (show n)
477 -- SIGBJORN wrote instead:
478 -- rational n = text (show (fromRationalX n))
479
480 quotes p = char '\'' <> p <> char '\''
481 doubleQuotes p = char '"' <> p <> char '"'
482 parens p = char '(' <> p <> char ')'
483 brackets p = char '[' <> p <> char ']'
484 braces p = char '{' <> p <> char '}'
485
486
487 hcat = foldr (<>) empty
488 hsep = foldr (<+>) empty
489 vcat = foldr ($$) empty
490
491 hang d1 n d2 = sep [d1, nest n d2]
492
493 punctuate p [] = []
494 punctuate p (d:ds) = go d ds
495 where
496 go d [] = [d]
497 go d (e:es) = (d <> p) : go e es
498
499 -- ---------------------------------------------------------------------------
500 -- The Doc data type
501
502 -- A Doc represents a *set* of layouts. A Doc with
503 -- no occurrences of Union or NoDoc represents just one layout.
504
505 -- | The abstract type of documents.
506 -- The 'Show' instance is equivalent to using 'render'.
507 data Doc
508 = Empty -- empty
509 | NilAbove Doc -- text "" $$ x
510 | TextBeside TextDetails !Int Doc -- text s <> x
511 | Nest !Int Doc -- nest k x
512 | Union Doc Doc -- ul `union` ur
513 | NoDoc -- The empty set of documents
514 | Beside Doc Bool Doc -- True <=> space between
515 | Above Doc Bool Doc -- True <=> never overlap
516
517 type RDoc = Doc -- RDoc is a "reduced Doc", guaranteed not to have a top-level Above or Beside
518
519
520 reduceDoc :: Doc -> RDoc
521 reduceDoc (Beside p g q) = beside p g (reduceDoc q)
522 reduceDoc (Above p g q) = above p g (reduceDoc q)
523 reduceDoc p = p
524
525
526 data TextDetails = Chr Char
527 | Str String
528 | PStr String
529 space_text = Chr ' '
530 nl_text = Chr '\n'
531
532 {-
533 Here are the invariants:
534
535 * The argument of NilAbove is never Empty. Therefore
536 a NilAbove occupies at least two lines.
537
538 * The arugment of @TextBeside@ is never @Nest@.
539
540
541 * The layouts of the two arguments of @Union@ both flatten to the same
542 string.
543
544 * The arguments of @Union@ are either @TextBeside@, or @NilAbove@.
545
546 * The right argument of a union cannot be equivalent to the empty set
547 (@NoDoc@). If the left argument of a union is equivalent to the
548 empty set (@NoDoc@), then the @NoDoc@ appears in the first line.
549
550 * An empty document is always represented by @Empty@. It can't be
551 hidden inside a @Nest@, or a @Union@ of two @Empty@s.
552
553 * The first line of every layout in the left argument of @Union@ is
554 longer than the first line of any layout in the right argument.
555 (1) ensures that the left argument has a first line. In view of
556 (3), this invariant means that the right argument must have at
557 least two lines.
558 -}
559
560 -- Arg of a NilAbove is always an RDoc
561 nilAbove_ p = NilAbove p
562
563 -- Arg of a TextBeside is always an RDoc
564 textBeside_ s sl p = TextBeside s sl p
565
566 -- Arg of Nest is always an RDoc
567 nest_ k p = Nest k p
568
569 -- Args of union are always RDocs
570 union_ p q = Union p q
571
572
573 -- Notice the difference between
574 -- * NoDoc (no documents)
575 -- * Empty (one empty document; no height and no width)
576 -- * text "" (a document containing the empty string;
577 -- one line high, but has no width)
578
579
580 -- ---------------------------------------------------------------------------
581 -- @empty@, @text@, @nest@, @union@
582
583 empty = Empty
584
585 isEmpty Empty = True
586 isEmpty _ = False
587
588 char c = textBeside_ (Chr c) 1 Empty
589 text s = case length s of {sl -> textBeside_ (Str s) sl Empty}
590 ptext s = case length s of {sl -> textBeside_ (PStr s) sl Empty}
591
592 nest k p = mkNest k (reduceDoc p) -- Externally callable version
593
594 -- mkNest checks for Nest's invariant that it doesn't have an Empty inside it
595 mkNest k _ | k `seq` False = undefined
596 mkNest k (Nest k1 p) = mkNest (k + k1) p
597 mkNest k NoDoc = NoDoc
598 mkNest k Empty = Empty
599 mkNest 0 p = p -- Worth a try!
600 mkNest k p = nest_ k p
601
602 -- mkUnion checks for an empty document
603 mkUnion Empty q = Empty
604 mkUnion p q = p `union_` q
605
606 -- ---------------------------------------------------------------------------
607 -- Vertical composition @$$@
608
609 p $$ q = Above p False q
610 p $+$ q = Above p True q
611
612 above :: Doc -> Bool -> RDoc -> RDoc
613 above (Above p g1 q1) g2 q2 = above p g1 (above q1 g2 q2)
614 above p@(Beside _ _ _) g q = aboveNest (reduceDoc p) g 0 (reduceDoc q)
615 above p g q = aboveNest p g 0 (reduceDoc q)
616
617 aboveNest :: RDoc -> Bool -> Int -> RDoc -> RDoc
618 -- Specfication: aboveNest p g k q = p $g$ (nest k q)
619
620 aboveNest _ _ k _ | k `seq` False = undefined
621 aboveNest NoDoc g k q = NoDoc
622 aboveNest (p1 `Union` p2) g k q = aboveNest p1 g k q `union_`
623 aboveNest p2 g k q
624
625 aboveNest Empty g k q = mkNest k q
626 aboveNest (Nest k1 p) g k q = nest_ k1 (aboveNest p g (k - k1) q)
627 -- p can't be Empty, so no need for mkNest
628
629 aboveNest (NilAbove p) g k q = nilAbove_ (aboveNest p g k q)
630 aboveNest (TextBeside s sl p) g k q = k1 `seq` textBeside_ s sl rest
631 where
632 k1 = k - sl
633 rest = case p of
634 Empty -> nilAboveNest g k1 q
635 other -> aboveNest p g k1 q
636
637
638 nilAboveNest :: Bool -> Int -> RDoc -> RDoc
639 -- Specification: text s <> nilaboveNest g k q
640 -- = text s <> (text "" $g$ nest k q)
641
642 nilAboveNest _ k _ | k `seq` False = undefined
643 nilAboveNest g k Empty = Empty -- Here's why the "text s <>" is in the spec!
644 nilAboveNest g k (Nest k1 q) = nilAboveNest g (k + k1) q
645
646 nilAboveNest g k q | (not g) && (k > 0) -- No newline if no overlap
647 = textBeside_ (Str (spaces k)) k q
648 | otherwise -- Put them really above
649 = nilAbove_ (mkNest k q)
650
651 -- ---------------------------------------------------------------------------
652 -- Horizontal composition @<>@
653
654 p <> q = Beside p False q
655 p <+> q = Beside p True q
656
657 beside :: Doc -> Bool -> RDoc -> RDoc
658 -- Specification: beside g p q = p <g> q
659
660 beside NoDoc g q = NoDoc
661 beside (p1 `Union` p2) g q = (beside p1 g q) `union_` (beside p2 g q)
662 beside Empty g q = q
663 beside (Nest k p) g q = nest_ k (beside p g q) -- p non-empty
664 beside p@(Beside p1 g1 q1) g2 q2
665 {- (A `op1` B) `op2` C == A `op1` (B `op2` C) iff op1 == op2
666 [ && (op1 == <> || op1 == <+>) ] -}
667 | g1 == g2 = beside p1 g1 (beside q1 g2 q2)
668 | otherwise = beside (reduceDoc p) g2 q2
669 beside p@(Above _ _ _) g q = beside (reduceDoc p) g q
670 beside (NilAbove p) g q = nilAbove_ (beside p g q)
671 beside (TextBeside s sl p) g q = textBeside_ s sl rest
672 where
673 rest = case p of
674 Empty -> nilBeside g q
675 other -> beside p g q
676
677
678 nilBeside :: Bool -> RDoc -> RDoc
679 -- Specification: text "" <> nilBeside g p
680 -- = text "" <g> p
681
682 nilBeside g Empty = Empty -- Hence the text "" in the spec
683 nilBeside g (Nest _ p) = nilBeside g p
684 nilBeside g p | g = textBeside_ space_text 1 p
685 | otherwise = p
686
687 -- ---------------------------------------------------------------------------
688 -- Separate, @sep@, Hughes version
689
690 -- Specification: sep ps = oneLiner (hsep ps)
691 -- `union`
692 -- vcat ps
693
694 sep = sepX True -- Separate with spaces
695 cat = sepX False -- Don't
696
697 sepX x [] = empty
698 sepX x (p:ps) = sep1 x (reduceDoc p) 0 ps
699
700
701 -- Specification: sep1 g k ys = sep (x : map (nest k) ys)
702 -- = oneLiner (x <g> nest k (hsep ys))
703 -- `union` x $$ nest k (vcat ys)
704
705 sep1 :: Bool -> RDoc -> Int -> [Doc] -> RDoc
706 sep1 g _ k ys | k `seq` False = undefined
707 sep1 g NoDoc k ys = NoDoc
708 sep1 g (p `Union` q) k ys = sep1 g p k ys
709 `union_`
710 (aboveNest q False k (reduceDoc (vcat ys)))
711
712 sep1 g Empty k ys = mkNest k (sepX g ys)
713 sep1 g (Nest n p) k ys = nest_ n (sep1 g p (k - n) ys)
714
715 sep1 g (NilAbove p) k ys = nilAbove_ (aboveNest p False k (reduceDoc (vcat ys)))
716 sep1 g (TextBeside s sl p) k ys = textBeside_ s sl (sepNB g p (k - sl) ys)
717
718 -- Specification: sepNB p k ys = sep1 (text "" <> p) k ys
719 -- Called when we have already found some text in the first item
720 -- We have to eat up nests
721
722 sepNB g (Nest _ p) k ys = sepNB g p k ys
723
724 sepNB g Empty k ys = oneLiner (nilBeside g (reduceDoc rest))
725 `mkUnion`
726 nilAboveNest False k (reduceDoc (vcat ys))
727 where
728 rest | g = hsep ys
729 | otherwise = hcat ys
730
731 sepNB g p k ys = sep1 g p k ys
732
733 -- ---------------------------------------------------------------------------
734 -- @fill@
735
736 fsep = fill True
737 fcat = fill False
738
739 -- Specification:
740 -- fill [] = empty
741 -- fill [p] = p
742 -- fill (p1:p2:ps) = oneLiner p1 <#> nest (length p1)
743 -- (fill (oneLiner p2 : ps))
744 -- `union`
745 -- p1 $$ fill ps
746
747 fill g [] = empty
748 fill g (p:ps) = fill1 g (reduceDoc p) 0 ps
749
750
751 fill1 :: Bool -> RDoc -> Int -> [Doc] -> Doc
752 fill1 g _ k ys | k `seq` False = undefined
753 fill1 g NoDoc k ys = NoDoc
754 fill1 g (p `Union` q) k ys = fill1 g p k ys
755 `union_`
756 (aboveNest q False k (fill g ys))
757
758 fill1 g Empty k ys = mkNest k (fill g ys)
759 fill1 g (Nest n p) k ys = nest_ n (fill1 g p (k - n) ys)
760
761 fill1 g (NilAbove p) k ys = nilAbove_ (aboveNest p False k (fill g ys))
762 fill1 g (TextBeside s sl p) k ys = textBeside_ s sl (fillNB g p (k - sl) ys)
763
764 fillNB g _ k ys | k `seq` False = undefined
765 fillNB g (Nest _ p) k ys = fillNB g p k ys
766 fillNB g Empty k [] = Empty
767 fillNB g Empty k (y:ys) = nilBeside g (fill1 g (oneLiner (reduceDoc y)) k1 ys)
768 `mkUnion`
769 nilAboveNest False k (fill g (y:ys))
770 where
771 k1 | g = k - 1
772 | otherwise = k
773
774 fillNB g p k ys = fill1 g p k ys
775
776
777 -- ---------------------------------------------------------------------------
778 -- Selecting the best layout
779
780 best :: Mode
781 -> Int -- Line length
782 -> Int -- Ribbon length
783 -> RDoc
784 -> RDoc -- No unions in here!
785
786 best OneLineMode w r p
787 = get p
788 where
789 get Empty = Empty
790 get NoDoc = NoDoc
791 get (NilAbove p) = nilAbove_ (get p)
792 get (TextBeside s sl p) = textBeside_ s sl (get p)
793 get (Nest k p) = get p -- Elide nest
794 get (p `Union` q) = first (get p) (get q)
795
796 best mode w r p
797 = get w p
798 where
799 get :: Int -- (Remaining) width of line
800 -> Doc -> Doc
801 get w _ | w==0 && False = undefined
802 get w Empty = Empty
803 get w NoDoc = NoDoc
804 get w (NilAbove p) = nilAbove_ (get w p)
805 get w (TextBeside s sl p) = textBeside_ s sl (get1 w sl p)
806 get w (Nest k p) = nest_ k (get (w - k) p)
807 get w (p `Union` q) = nicest w r (get w p) (get w q)
808
809 get1 :: Int -- (Remaining) width of line
810 -> Int -- Amount of first line already eaten up
811 -> Doc -- This is an argument to TextBeside => eat Nests
812 -> Doc -- No unions in here!
813
814 get1 w _ _ | w==0 && False = undefined
815 get1 w sl Empty = Empty
816 get1 w sl NoDoc = NoDoc
817 get1 w sl (NilAbove p) = nilAbove_ (get (w - sl) p)
818 get1 w sl (TextBeside t tl p) = textBeside_ t tl (get1 w (sl + tl) p)
819 get1 w sl (Nest k p) = get1 w sl p
820 get1 w sl (p `Union` q) = nicest1 w r sl (get1 w sl p)
821 (get1 w sl q)
822
823 nicest w r p q = nicest1 w r 0 p q
824 nicest1 w r sl p q | fits ((w `minn` r) - sl) p = p
825 | otherwise = q
826
827 fits :: Int -- Space available
828 -> Doc
829 -> Bool -- True if *first line* of Doc fits in space available
830
831 fits n p | n < 0 = False
832 fits n NoDoc = False
833 fits n Empty = True
834 fits n (NilAbove _) = True
835 fits n (TextBeside _ sl p) = fits (n - sl) p
836
837 minn x y | x < y = x
838 | otherwise = y
839
840 -- @first@ and @nonEmptySet@ are similar to @nicest@ and @fits@, only simpler.
841 -- @first@ returns its first argument if it is non-empty, otherwise its second.
842
843 first p q | nonEmptySet p = p
844 | otherwise = q
845
846 nonEmptySet NoDoc = False
847 nonEmptySet (p `Union` q) = True
848 nonEmptySet Empty = True
849 nonEmptySet (NilAbove p) = True -- NoDoc always in first line
850 nonEmptySet (TextBeside _ _ p) = nonEmptySet p
851 nonEmptySet (Nest _ p) = nonEmptySet p
852
853 -- @oneLiner@ returns the one-line members of the given set of @Doc@s.
854
855 oneLiner :: Doc -> Doc
856 oneLiner NoDoc = NoDoc
857 oneLiner Empty = Empty
858 oneLiner (NilAbove p) = NoDoc
859 oneLiner (TextBeside s sl p) = textBeside_ s sl (oneLiner p)
860 oneLiner (Nest k p) = nest_ k (oneLiner p)
861 oneLiner (p `Union` q) = oneLiner p
862
863
864 -- ---------------------------------------------------------------------------
865 -- Displaying the best layout
866
867 renderStyle style doc
868 = fullRender (mode style)
869 (lineLength style)
870 (ribbonsPerLine style)
871 string_txt
872 ""
873 doc
874
875 render doc = showDoc doc ""
876 showDoc doc rest = fullRender PageMode 100 1.5 string_txt rest doc
877
878 string_txt (Chr c) s = c:s
879 string_txt (Str s1) s2 = s1 ++ s2
880 string_txt (PStr s1) s2 = s1 ++ s2
881
882
883 fullRender OneLineMode _ _ txt end doc = easy_display space_text txt end (reduceDoc doc)
884 fullRender LeftMode _ _ txt end doc = easy_display nl_text txt end (reduceDoc doc)
885
886 fullRender mode line_length ribbons_per_line txt end doc
887 = display mode line_length ribbon_length txt end best_doc
888 where
889 best_doc = best mode hacked_line_length ribbon_length (reduceDoc doc)
890
891 hacked_line_length, ribbon_length :: Int
892 ribbon_length = round (fromIntegral line_length / ribbons_per_line)
893 hacked_line_length = case mode of { ZigZagMode -> maxBound; other -> line_length }
894
895 display mode page_width ribbon_width txt end doc
896 = case page_width - ribbon_width of { gap_width ->
897 case gap_width `quot` 2 of { shift ->
898 let
899 lay k _ | k `seq` False = undefined
900 lay k (Nest k1 p) = lay (k + k1) p
901 lay k Empty = end
902
903 lay k (NilAbove p) = nl_text `txt` lay k p
904
905 lay k (TextBeside s sl p)
906 = case mode of
907 ZigZagMode | k >= gap_width
908 -> nl_text `txt` (
909 Str (multi_ch shift '/') `txt` (
910 nl_text `txt` (
911 lay1 (k - shift) s sl p)))
912
913 | k < 0
914 -> nl_text `txt` (
915 Str (multi_ch shift '\\') `txt` (
916 nl_text `txt` (
917 lay1 (k + shift) s sl p )))
918
919 other -> lay1 k s sl p
920
921 lay1 k _ sl _ | k+sl `seq` False = undefined
922 lay1 k s sl p = Str (indent k) `txt` (s `txt` lay2 (k + sl) p)
923
924 lay2 k _ | k `seq` False = undefined
925 lay2 k (NilAbove p) = nl_text `txt` lay k p
926 lay2 k (TextBeside s sl p) = s `txt` (lay2 (k + sl) p)
927 lay2 k (Nest _ p) = lay2 k p
928 lay2 k Empty = end
929 in
930 lay 0 doc
931 }}
932
933 cant_fail = error "easy_display: NoDoc"
934 easy_display nl_text txt end doc
935 = lay doc cant_fail
936 where
937 lay NoDoc no_doc = no_doc
938 lay (Union p q) no_doc = {- lay p -} (lay q cant_fail) -- Second arg can't be NoDoc
939 lay (Nest k p) no_doc = lay p no_doc
940 lay Empty no_doc = end
941 lay (NilAbove p) no_doc = nl_text `txt` lay p cant_fail -- NoDoc always on first line
942 lay (TextBeside s sl p) no_doc = s `txt` lay p no_doc
943
944 -- OLD version: we shouldn't rely on tabs being 8 columns apart in the output.
945 -- indent n | n >= 8 = '\t' : indent (n - 8)
946 -- | otherwise = spaces n
947 indent n = spaces n
948
949 multi_ch 0 ch = ""
950 multi_ch n ch = ch : multi_ch (n - 1) ch
951
952 -- (spaces n) generates a list of n spaces
953 --
954 -- It should never be called with 'n' < 0, but that can happen for reasons I don't understand
955 -- Here's a test case:
956 -- ncat x y = nest 4 $ cat [ x, y ]
957 -- d1 = foldl1 ncat $ take 50 $ repeat $ char 'a'
958 -- d2 = parens $ sep [ d1, text "+" , d1 ]
959 -- main = print d2
960 -- I don't feel motivated enough to find the Real Bug, so meanwhile we just test for n<=0
961 spaces n | n <= 0 = ""
962 | otherwise = ' ' : spaces (n - 1)
963
964 {- Comments from Johannes Waldmann about what the problem might be:
965
966 In the example above, d2 and d1 are deeply nested, but `text "+"' is not,
967 so the layout function tries to "out-dent" it.
968
969 when I look at the Doc values that are generated, there are lots of
970 Nest constructors with negative arguments. see this sample output of
971 d1 (obtained with hugs, :s -u)
972
973 tBeside (TextDetails_Chr 'a') 1 Doc_Empty) (Doc_NilAbove (Doc_Nest
974 (-241) (Doc_TextBeside (TextDetails_Chr 'a') 1 Doc_Empty)))))
975 (Doc_NilAbove (Doc_Nest (-236) (Doc_TextBeside (TextDetails_Chr 'a') 1
976 (Doc_NilAbove (Doc_Nest (-5) (Doc_TextBeside (TextDetails_Chr 'a') 1
977 Doc_Empty)))))))) (Doc_NilAbove (Doc_Nest (-231) (Doc_TextBeside
978 (TextDetails_Chr 'a') 1 (Doc_NilAbove (Doc_Nest (-5) (Doc_TextBeside
979 (TextDetails_Chr 'a') 1 (Doc_NilAbove (Doc_Nest (-5) (Doc_TextBeside
980 (TextDetails_Chr 'a') 1 Doc_Empty))))))))))) (Doc_NilAbove (Doc_Nest
981 -}