Match.lhs 34.8 KB
Newer Older
1
%
Simon Marlow's avatar
Simon Marlow committed
2
% (c) The University of Glasgow 2006
3
% (c) The GRASP/AQUA Project, Glasgow University, 1992-1998
4
%
Simon Marlow's avatar
Simon Marlow committed
5 6

The @match@ function
7 8

\begin{code}
9
module Match ( match, matchEquations, matchWrapper, matchSimply, matchSinglePat ) where
10

11
#include "HsVersions.h"
12

13 14
import {-#SOURCE#-} DsExpr (dsLExpr)

Simon Marlow's avatar
Simon Marlow committed
15
import DynFlags
sof's avatar
sof committed
16
import HsSyn		
Simon Marlow's avatar
Simon Marlow committed
17 18
import TcHsSyn
import Check
19
import CoreSyn
Simon Marlow's avatar
Simon Marlow committed
20 21
import Literal
import CoreUtils
22
import MkCore
23
import DsMonad
Simon Marlow's avatar
Simon Marlow committed
24 25
import DsBinds
import DsGRHSs
26
import DsUtils
Simon Marlow's avatar
Simon Marlow committed
27 28 29 30 31
import Id
import DataCon
import MatchCon
import MatchLit
import Type
32
import Coercion
Simon Marlow's avatar
Simon Marlow committed
33 34 35 36 37 38
import TysWiredIn
import ListSetOps
import SrcLoc
import Maybes
import Util
import Name
39
import Outputable
40
import FastString
41

42
import Control.Monad( when )
43
import qualified Data.Map as Map
44 45
\end{code}

46 47 48 49 50 51 52 53
This function is a wrapper of @match@, it must be called from all the parts where 
it was called match, but only substitutes the firs call, ....
if the associated flags are declared, warnings will be issued.
It can not be called matchWrapper because this name already exists :-(

JJCQ 30-Nov-1997

\begin{code}
54 55 56
matchCheck ::  DsMatchContext
	    -> [Id]	        -- Vars rep'ing the exprs we're matching with
            -> Type             -- Type of the case expression
57 58 59
            -> [EquationInfo]   -- Info about patterns, etc. (type synonym below)
            -> DsM MatchResult  -- Desugared result!

60 61 62
matchCheck ctx vars ty qs
  = do { dflags <- getDOptsDs
       ; matchCheck_really dflags ctx vars ty qs }
63

64 65 66 67 68 69
matchCheck_really :: DynFlags
                  -> DsMatchContext
                  -> [Id]
                  -> Type
                  -> [EquationInfo]
                  -> DsM MatchResult
70 71 72 73 74 75 76
matchCheck_really dflags ctx@(DsMatchContext hs_ctx _) vars ty qs
  = do { when shadow (dsShadowWarn ctx eqns_shadow)
       ; when incomplete (dsIncompleteWarn ctx pats)
       ; match vars ty qs }
  where 
    (pats, eqns_shadow) = check qs
    incomplete = incomplete_flag hs_ctx && (notNull pats)
77
    shadow     = wopt Opt_WarnOverlappingPatterns dflags
78 79 80
              	 && notNull eqns_shadow

    incomplete_flag :: HsMatchContext id -> Bool
81 82
    incomplete_flag (FunRhs {})   = wopt Opt_WarnIncompletePatterns dflags
    incomplete_flag CaseAlt       = wopt Opt_WarnIncompletePatterns dflags
83

84 85 86
    incomplete_flag LambdaExpr    = wopt Opt_WarnIncompleteUniPatterns dflags
    incomplete_flag PatBindRhs    = wopt Opt_WarnIncompleteUniPatterns dflags
    incomplete_flag ProcExpr      = wopt Opt_WarnIncompleteUniPatterns dflags
87

88
    incomplete_flag RecUpd        = wopt Opt_WarnIncompletePatternsRecUpd dflags
89 90 91 92 93 94

    incomplete_flag ThPatQuote    = False
    incomplete_flag (StmtCtxt {}) = False  -- Don't warn about incomplete patterns
    		    	      	    	   -- in list comprehensions, pattern guards
					   -- etc.  They are often *supposed* to be
					   -- incomplete 
95 96
\end{code}

97 98
This variable shows the maximum number of lines of output generated for warnings.
It will limit the number of patterns/equations displayed to@ maximum_output@.
99

sof's avatar
sof committed
100 101
(ToDo: add command-line option?)

102
\begin{code}
103
maximum_output :: Int
104 105 106
maximum_output = 4
\end{code}

107
The next two functions create the warning message.
108 109 110

\begin{code}
dsShadowWarn :: DsMatchContext -> [EquationInfo] -> DsM ()
111
dsShadowWarn ctx@(DsMatchContext kind loc) qs
112
  = putSrcSpanDs loc (warnDs warn)
113 114
  where
    warn | qs `lengthExceeds` maximum_output
Ian Lynagh's avatar
Ian Lynagh committed
115
         = pp_context ctx (ptext (sLit "are overlapped"))
116
		      (\ f -> vcat (map (ppr_eqn f kind) (take maximum_output qs)) $$
Ian Lynagh's avatar
Ian Lynagh committed
117
		      ptext (sLit "..."))
118
	 | otherwise
Ian Lynagh's avatar
Ian Lynagh committed
119
         = pp_context ctx (ptext (sLit "are overlapped"))
120
	              (\ f -> vcat $ map (ppr_eqn f kind) qs)
sof's avatar
sof committed
121

122 123

dsIncompleteWarn :: DsMatchContext -> [ExhaustivePat] -> DsM ()
124
dsIncompleteWarn ctx@(DsMatchContext kind loc) pats 
125
  = putSrcSpanDs loc (warnDs warn)
126
	where
Ian Lynagh's avatar
Ian Lynagh committed
127 128
	  warn = pp_context ctx (ptext (sLit "are non-exhaustive"))
                            (\_ -> hang (ptext (sLit "Patterns not matched:"))
129 130 131 132
		                   4 ((vcat $ map (ppr_incomplete_pats kind)
						  (take maximum_output pats))
		                      $$ dots))

Ian Lynagh's avatar
Ian Lynagh committed
133
	  dots | pats `lengthExceeds` maximum_output = ptext (sLit "...")
sof's avatar
sof committed
134
	       | otherwise                           = empty
135

136
pp_context :: DsMatchContext -> SDoc -> ((SDoc -> SDoc) -> SDoc) -> SDoc
137
pp_context (DsMatchContext kind _loc) msg rest_of_msg_fun
Ian Lynagh's avatar
Ian Lynagh committed
138 139
  = vcat [ptext (sLit "Pattern match(es)") <+> msg,
	  sep [ptext (sLit "In") <+> ppr_match <> char ':', nest 4 (rest_of_msg_fun pref)]]
140 141 142
  where
    (ppr_match, pref)
	= case kind of
143
	     FunRhs fun _ -> (pprMatchContext kind, \ pp -> ppr fun <+> pp)
144
             _            -> (pprMatchContext kind, \ pp -> pp)
145

146
ppr_pats :: Outputable a => [a] -> SDoc
147
ppr_pats pats = sep (map ppr pats)
148

149
ppr_shadow_pats :: HsMatchContext Name -> [Pat Id] -> SDoc
sof's avatar
sof committed
150
ppr_shadow_pats kind pats
Ian Lynagh's avatar
Ian Lynagh committed
151
  = sep [ppr_pats pats, matchSeparator kind, ptext (sLit "...")]
152 153 154 155

ppr_incomplete_pats :: HsMatchContext Name -> ExhaustivePat -> SDoc
ppr_incomplete_pats _ (pats,[]) = ppr_pats pats
ppr_incomplete_pats _ (pats,constraints) =
Ian Lynagh's avatar
Ian Lynagh committed
156
	                 sep [ppr_pats pats, ptext (sLit "with"), 
157
	                      sep (map ppr_constraint constraints)]
158

159
ppr_constraint :: (Name,[HsLit]) -> SDoc
Ian Lynagh's avatar
Ian Lynagh committed
160
ppr_constraint (var,pats) = sep [ppr var, ptext (sLit "`notElem`"), ppr pats]
161

162
ppr_eqn :: (SDoc -> SDoc) -> HsMatchContext Name -> EquationInfo -> SDoc
163
ppr_eqn prefixF kind eqn = prefixF (ppr_shadow_pats kind (eqn_pats eqn))
164 165 166
\end{code}


167 168 169 170 171 172
%************************************************************************
%*									*
		The main matching function
%*									*
%************************************************************************

173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191
The function @match@ is basically the same as in the Wadler chapter,
except it is monadised, to carry around the name supply, info about
annotations, etc.

Notes on @match@'s arguments, assuming $m$ equations and $n$ patterns:
\begin{enumerate}
\item
A list of $n$ variable names, those variables presumably bound to the
$n$ expressions being matched against the $n$ patterns.  Using the
list of $n$ expressions as the first argument showed no benefit and
some inelegance.

\item
The second argument, a list giving the ``equation info'' for each of
the $m$ equations:
\begin{itemize}
\item
the $n$ patterns for that equation, and
\item
192
a list of Core bindings [@(Id, CoreExpr)@ pairs] to be ``stuck on
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214
the front'' of the matching code, as in:
\begin{verbatim}
let <binds>
in  <matching-code>
\end{verbatim}
\item
and finally: (ToDo: fill in)

The right way to think about the ``after-match function'' is that it
is an embryonic @CoreExpr@ with a ``hole'' at the end for the
final ``else expression''.
\end{itemize}

There is a type synonym, @EquationInfo@, defined in module @DsUtils@.

An experiment with re-ordering this information about equations (in
particular, having the patterns available in column-major order)
showed no benefit.

\item
A default expression---what to evaluate if the overall pattern-match
fails.  This expression will (almost?) always be
215
a measly expression @Var@, unless we know it will only be used once
216 217 218
(as we do in @glue_success_exprs@).

Leaving out this third argument to @match@ (and slamming in lots of
219
@Var "fail"@s) is a positively {\em bad} idea, because it makes it
220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244
impossible to share the default expressions.  (Also, it stands no
chance of working in our post-upheaval world of @Locals@.)
\end{enumerate}

Note: @match@ is often called via @matchWrapper@ (end of this module),
a function that does much of the house-keeping that goes with a call
to @match@.

It is also worth mentioning the {\em typical} way a block of equations
is desugared with @match@.  At each stage, it is the first column of
patterns that is examined.  The steps carried out are roughly:
\begin{enumerate}
\item
Tidy the patterns in column~1 with @tidyEqnInfo@ (this may add
bindings to the second component of the equation-info):
\begin{itemize}
\item
Remove the `as' patterns from column~1.
\item
Make all constructor patterns in column~1 into @ConPats@, notably
@ListPats@ and @TuplePats@.
\item
Handle any irrefutable (or ``twiddle'') @LazyPats@.
\end{itemize}
\item
Ian Lynagh's avatar
Ian Lynagh committed
245
Now {\em unmix} the equations into {\em blocks} [w\/ local function
246 247 248 249
@unmix_eqns@], in which the equations in a block all have variable
patterns in column~1, or they all have constructor patterns in ...
(see ``the mixture rule'' in SLPJ).
\item
250
Call @matchEqnBlock@ on each block of equations; it will do the
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
appropriate thing for each kind of column-1 pattern, usually ending up
in a recursive call to @match@.
\end{enumerate}

We are a little more paranoid about the ``empty rule'' (SLPJ, p.~87)
than the Wadler-chapter code for @match@ (p.~93, first @match@ clause).
And gluing the ``success expressions'' together isn't quite so pretty.

This (more interesting) clause of @match@ uses @tidy_and_unmix_eqns@
(a)~to get `as'- and `twiddle'-patterns out of the way (tidying), and
(b)~to do ``the mixture rule'' (SLPJ, p.~88) [which really {\em
un}mixes the equations], producing a list of equation-info
blocks, each block having as its first column of patterns either all
constructors, or all variables (or similar beasts), etc.

@match_unmixed_eqn_blks@ simply takes the place of the @foldr@ in the
Wadler-chapter @match@ (p.~93, last clause), and @match_unmixed_blk@
corresponds roughly to @matchVarCon@.

\begin{code}
Ian Lynagh's avatar
Ian Lynagh committed
271
match :: [Id]		  -- Variables rep\'ing the exprs we\'re matching with
272 273 274 275 276
      -> Type             -- Type of the case expression
      -> [EquationInfo]	  -- Info about patterns, etc. (type synonym below)
      -> DsM MatchResult  -- Desugared result!

match [] ty eqns
277
  = ASSERT2( not (null eqns), ppr ty )
278
    return (foldr1 combineMatchResults match_results)
279
  where
280 281 282 283 284 285 286 287
    match_results = [ ASSERT( null (eqn_pats eqn) ) 
		      eqn_rhs eqn
		    | eqn <- eqns ]

match vars@(v:_) ty eqns
  = ASSERT( not (null eqns ) )
    do	{ 	-- Tidy the first pattern, generating
		-- auxiliary bindings if necessary
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
288
          (aux_binds, tidy_eqns) <- mapAndUnzipM (tidyEqnInfo v) eqns
289 290

		-- Group the equations and match each group in turn
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
291
        ; let grouped = groupEquations tidy_eqns
292 293

         -- print the view patterns that are commoned up to help debug
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
294
        ; ifDOptM Opt_D_dump_view_pattern_commoning (debug grouped)
295 296

	; match_results <- mapM match_group grouped
297 298 299 300 301 302 303
	; return (adjustMatchResult (foldr1 (.) aux_binds) $
		  foldr1 combineMatchResults match_results) }
  where
    dropGroup :: [(PatGroup,EquationInfo)] -> [EquationInfo]
    dropGroup = map snd

    match_group :: [(PatGroup,EquationInfo)] -> DsM MatchResult
304
    match_group [] = panic "match_group"
305
    match_group eqns@((group,_) : _)
306
        = case group of
307 308
            PgCon _    -> matchConFamily  vars ty (subGroup [(c,e) | (PgCon c, e) <- eqns])
            PgLit _    -> matchLiterals   vars ty (subGroup [(l,e) | (PgLit l, e) <- eqns])
309
            PgAny      -> matchVariables  vars ty (dropGroup eqns)
310
            PgN _      -> matchNPats      vars ty (dropGroup eqns)
311
            PgNpK _    -> matchNPlusKPats vars ty (dropGroup eqns)
312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329
            PgBang     -> matchBangs      vars ty (dropGroup eqns)
            PgCo _     -> matchCoercion   vars ty (dropGroup eqns)
            PgView _ _ -> matchView       vars ty (dropGroup eqns)

    -- FIXME: we should also warn about view patterns that should be
    -- commoned up but are not

    -- print some stuff to see what's getting grouped
    -- use -dppr-debug to see the resolution of overloaded lits
    debug eqns = 
        let gs = map (\group -> foldr (\ (p,_) -> \acc -> 
                                           case p of PgView e _ -> e:acc 
                                                     _ -> acc) [] group) eqns
            maybeWarn [] = return ()
            maybeWarn l = warnDs (vcat l)
        in 
          maybeWarn $ (map (\g -> text "Putting these view expressions into the same case:" <+> (ppr g))
                       (filter (not . null) gs))
330 331 332 333

matchVariables :: [Id] -> Type -> [EquationInfo] -> DsM MatchResult
-- Real true variables, just like in matchVar, SLPJ p 94
-- No binding to do: they'll all be wildcards by now (done in tidy)
334
matchVariables (_:vars) ty eqns = match vars ty (shiftEqns eqns)
335
matchVariables [] _ _ = panic "matchVariables"
336

337 338
matchBangs :: [Id] -> Type -> [EquationInfo] -> DsM MatchResult
matchBangs (var:vars) ty eqns
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
339 340
  = do	{ match_result <- match (var:vars) ty $
                          map (decomposeFirstPat getBangPat) eqns
341
	; return (mkEvalMatchResult var ty match_result) }
342
matchBangs [] _ _ = panic "matchBangs"
343 344 345

matchCoercion :: [Id] -> Type -> [EquationInfo] -> DsM MatchResult
-- Apply the coercion to the match variable and then match that
346
matchCoercion (var:vars) ty (eqns@(eqn1:_))
347
  = do	{ let CoPat co pat _ = firstPat eqn1
348
	; var' <- newUniqueId var (hsPatType pat)
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
349 350
	; match_result <- match (var':vars) ty $
                          map (decomposeFirstPat getCoPat) eqns
351
	; co' <- dsHsWrapper co
352 353
        ; let rhs' = co' (Var var)
	; return (mkCoLetMatchResult (NonRec var' rhs') match_result) }
354
matchCoercion _ _ _ = panic "matchCoercion"
355 356 357 358 359 360 361 362 363

matchView :: [Id] -> Type -> [EquationInfo] -> DsM MatchResult
-- Apply the view function to the match variable and then match that
matchView (var:vars) ty (eqns@(eqn1:_))
  = do	{ -- we could pass in the expr from the PgView,
         -- but this needs to extract the pat anyway 
         -- to figure out the type of the fresh variable
         let ViewPat viewExpr (L _ pat) _ = firstPat eqn1
         -- do the rest of the compilation 
364
	; var' <- newUniqueId var (hsPatType pat)
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
365 366
	; match_result <- match (var':vars) ty $
                          map (decomposeFirstPat getViewPat) eqns
367
         -- compile the view expressions
368
        ; viewExpr' <- dsLExpr viewExpr
369
	; return (mkViewMatchResult var' viewExpr' var match_result) }
370
matchView _ _ _ = panic "matchView"
371 372

-- decompose the first pattern and leave the rest alone
373
decomposeFirstPat :: (Pat Id -> Pat Id) -> EquationInfo -> EquationInfo
374 375
decomposeFirstPat extractpat (eqn@(EqnInfo { eqn_pats = pat : pats }))
	= eqn { eqn_pats = extractpat pat : pats}
376
decomposeFirstPat _ _ = panic "decomposeFirstPat"
377

simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
378 379 380 381 382 383 384
getCoPat, getBangPat, getViewPat :: Pat Id -> Pat Id
getCoPat (CoPat _ pat _)     = pat
getCoPat _                   = panic "getCoPat"
getBangPat (BangPat pat  )   = unLoc pat
getBangPat _                 = panic "getBangPat"
getViewPat (ViewPat _ pat _) = unLoc pat
getViewPat _                 = panic "getBangPat"
385 386
\end{code}

387 388 389 390 391 392
%************************************************************************
%*									*
		Tidying patterns
%*									*
%************************************************************************

393 394 395 396 397 398 399
Tidy up the leftmost pattern in an @EquationInfo@, given the variable @v@
which will be scrutinised.  This means:
\begin{itemize}
\item
Replace variable patterns @x@ (@x /= v@) with the pattern @_@,
together with the binding @x = v@.
\item
400
Replace the `as' pattern @x@@p@ with the pattern p and a binding @x = v@.
401 402 403
\item
Removing lazy (irrefutable) patterns (you don't want to know...).
\item
404 405
Converting explicit tuple-, list-, and parallel-array-pats into ordinary
@ConPats@. 
406 407
\item
Convert the literal pat "" to [].
408 409 410 411 412 413 414 415 416 417 418
\end{itemize}

The result of this tidying is that the column of patterns will include
{\em only}:
\begin{description}
\item[@WildPats@:]
The @VarPat@ information isn't needed any more after this.

\item[@ConPats@:]
@ListPats@, @TuplePats@, etc., are all converted into @ConPats@.

419 420
\item[@LitPats@ and @NPats@:]
@LitPats@/@NPats@ of ``known friendly types'' (Int, Char,
421
Float, 	Double, at least) are converted to unboxed form; e.g.,
422
\tr{(NPat (HsInt i) _ _)} is converted to:
423
\begin{verbatim}
424
(ConPat I# _ _ [LitPat (HsIntPrim i)])
425 426 427 428
\end{verbatim}
\end{description}

\begin{code}
429 430
tidyEqnInfo :: Id -> EquationInfo
	    -> DsM (DsWrapper, EquationInfo)
431
	-- DsM'd because of internal call to dsLHsBinds
432
	-- 	and mkSelectorBinds.
433 434
	-- "tidy1" does the interesting stuff, looking at
	-- one pattern and fiddling the list of bindings.
435 436 437 438 439 440 441 442 443
	--
	-- POST CONDITION: head pattern in the EqnInfo is
	--	WildPat
	--	ConPat
	--	NPat
	--	LitPat
	--	NPlusKPat
	-- but no other

444 445 446 447 448 449
tidyEqnInfo _ (EqnInfo { eqn_pats = [] }) 
  = panic "tidyEqnInfo"

tidyEqnInfo v eqn@(EqnInfo { eqn_pats = pat : pats })
  = do { (wrap, pat') <- tidy1 v pat
       ; return (wrap, eqn { eqn_pats = do pat' : pats }) }
450 451

tidy1 :: Id 			-- The Id being scrutinised
452
      -> Pat Id 		-- The pattern against which it is to be matched
453
      -> DsM (DsWrapper,	-- Extra bindings to do before the match
454
	      Pat Id) 		-- Equivalent pattern
455

456 457 458 459 460 461
-------------------------------------------------------
--	(pat', mr') = tidy1 v pat mr
-- tidies the *outer level only* of pat, giving pat'
-- It eliminates many pattern forms (as-patterns, variable patterns,
-- list patterns, etc) yielding one of:
--	WildPat
462
--	ConPatOut
463 464 465 466
--	LitPat
--	NPat
--	NPlusKPat

467 468
tidy1 v (ParPat pat)      = tidy1 v (unLoc pat) 
tidy1 v (SigPatOut pat _) = tidy1 v (unLoc pat) 
469
tidy1 _ (WildPat ty)      = return (idDsWrapper, WildPat ty)
470

471 472
	-- case v of { x -> mr[] }
	-- = case v of { _ -> let x=v in mr[] }
473
tidy1 v (VarPat var)
474
  = return (wrapBind var v, WildPat (idType var)) 
475

476 477
	-- case v of { x@p -> mr[] }
	-- = case v of { p -> let x=v in mr[] }
478 479 480
tidy1 v (AsPat (L _ var) pat)
  = do	{ (wrap, pat') <- tidy1 v (unLoc pat)
	; return (wrapBind var v . wrap, pat') }
481 482 483 484 485 486 487 488 489

{- now, here we handle lazy patterns:
    tidy1 v ~p bs = (v, v1 = case v of p -> v1 :
			v2 = case v of p -> v2 : ... : bs )

    where the v_i's are the binders in the pattern.

    ToDo: in "v_i = ... -> v_i", are the v_i's really the same thing?

490
    The case expr for v_i is just: match [v] [(p, [], \ x -> Var v_i)] any_expr
491 492
-}

493 494
tidy1 v (LazyPat pat)
  = do	{ sel_prs <- mkSelectorBinds pat (Var v)
495
	; let sel_binds =  [NonRec b rhs | (b,rhs) <- sel_prs]
496
	; return (mkCoreLets sel_binds, WildPat (idType v)) }
497

498
tidy1 _ (ListPat pats ty)
499
  = return (idDsWrapper, unLoc list_ConPat)
500
  where
501 502 503 504
    list_ty     = mkListTy ty
    list_ConPat = foldr (\ x y -> mkPrefixConPat consDataCon [x, y] list_ty)
	      	  	(mkNilPat list_ty)
	      	  	pats
505

506
-- Introduce fake parallel array constructors to be able to handle parallel
507
-- arrays with the existing machinery for constructor pattern
508
tidy1 _ (PArrPat pats ty)
509
  = return (idDsWrapper, unLoc parrConPat)
510 511
  where
    arity      = length pats
512
    parrConPat = mkPrefixConPat (parrFakeCon arity) pats (mkPArrTy ty)
513

514
tidy1 _ (TuplePat pats boxity ty)
515
  = return (idDsWrapper, unLoc tuple_ConPat)
516 517
  where
    arity = length pats
518
    tuple_ConPat = mkPrefixConPat (tupleCon boxity arity) pats ty
519

520
-- LitPats: we *might* be able to replace these w/ a simpler form
521
tidy1 _ (LitPat lit)
522
  = return (idDsWrapper, tidyLitPat lit)
523 524

-- NPats: we *might* be able to replace these w/ a simpler form
525
tidy1 _ (NPat lit mb_neg eq)
526
  = return (idDsWrapper, tidyNPat tidyLitPat lit mb_neg eq)
527

528 529 530 531 532 533 534 535 536 537 538 539 540 541
-- BangPatterns: Pattern matching is already strict in constructors,
-- tuples etc, so the last case strips off the bang for thoses patterns.
tidy1 v (BangPat (L _ (LazyPat p)))       = tidy1 v (BangPat p)
tidy1 v (BangPat (L _ (ParPat p)))        = tidy1 v (BangPat p)
tidy1 _ p@(BangPat (L _(VarPat _)))       = return (idDsWrapper, p)
tidy1 _ p@(BangPat (L _ (WildPat _)))     = return (idDsWrapper, p)
tidy1 _ p@(BangPat (L _ (CoPat _ _ _)))   = return (idDsWrapper, p)
tidy1 _ p@(BangPat (L _ (SigPatIn _ _)))  = return (idDsWrapper, p)
tidy1 _ p@(BangPat (L _ (SigPatOut _ _))) = return (idDsWrapper, p)
tidy1 v (BangPat (L _ (AsPat (L _ var) pat)))
  = do	{ (wrap, pat') <- tidy1 v (BangPat pat)
        ; return (wrapBind var v . wrap, pat') }
tidy1 v (BangPat (L _ p))                   = tidy1 v p

542
-- Everything else goes through unchanged...
543

544
tidy1 _ non_interesting_pat
545
  = return (idDsWrapper, non_interesting_pat)
546 547
\end{code}

548 549
\noindent
{\bf Previous @matchTwiddled@ stuff:}
550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637

Now we get to the only interesting part; note: there are choices for
translation [from Simon's notes]; translation~1:
\begin{verbatim}
deTwiddle [s,t] e
\end{verbatim}
returns
\begin{verbatim}
[ w = e,
  s = case w of [s,t] -> s
  t = case w of [s,t] -> t
]
\end{verbatim}

Here \tr{w} is a fresh variable, and the \tr{w}-binding prevents multiple
evaluation of \tr{e}.  An alternative translation (No.~2):
\begin{verbatim}
[ w = case e of [s,t] -> (s,t)
  s = case w of (s,t) -> s
  t = case w of (s,t) -> t
]
\end{verbatim}

%************************************************************************
%*									*
\subsubsection[improved-unmixing]{UNIMPLEMENTED idea for improved unmixing}
%*									*
%************************************************************************

We might be able to optimise unmixing when confronted by
only-one-constructor-possible, of which tuples are the most notable
examples.  Consider:
\begin{verbatim}
f (a,b,c) ... = ...
f d ... (e:f) = ...
f (g,h,i) ... = ...
f j ...       = ...
\end{verbatim}
This definition would normally be unmixed into four equation blocks,
one per equation.  But it could be unmixed into just one equation
block, because if the one equation matches (on the first column),
the others certainly will.

You have to be careful, though; the example
\begin{verbatim}
f j ...       = ...
-------------------
f (a,b,c) ... = ...
f d ... (e:f) = ...
f (g,h,i) ... = ...
\end{verbatim}
{\em must} be broken into two blocks at the line shown; otherwise, you
are forcing unnecessary evaluation.  In any case, the top-left pattern
always gives the cue.  You could then unmix blocks into groups of...
\begin{description}
\item[all variables:]
As it is now.
\item[constructors or variables (mixed):]
Need to make sure the right names get bound for the variable patterns.
\item[literals or variables (mixed):]
Presumably just a variant on the constructor case (as it is now).
\end{description}

%************************************************************************
%*									*
%*  matchWrapper: a convenient way to call @match@			*
%*									*
%************************************************************************
\subsection[matchWrapper]{@matchWrapper@: a convenient interface to @match@}

Calls to @match@ often involve similar (non-trivial) work; that work
is collected here, in @matchWrapper@.  This function takes as
arguments:
\begin{itemize}
\item
Typchecked @Matches@ (of a function definition, or a case or lambda
expression)---the main input;
\item
An error message to be inserted into any (runtime) pattern-matching
failure messages.
\end{itemize}

As results, @matchWrapper@ produces:
\begin{itemize}
\item
A list of variables (@Locals@) that the caller must ``promise'' to
bind to appropriate values; and
\item
638
a @CoreExpr@, the desugared output (main result).
639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657
\end{itemize}

The main actions of @matchWrapper@ include:
\begin{enumerate}
\item
Flatten the @[TypecheckedMatch]@ into a suitable list of
@EquationInfo@s.
\item
Create as many new variables as there are patterns in a pattern-list
(in any one of the @EquationInfo@s).
\item
Create a suitable ``if it fails'' expression---a call to @error@ using
the error-string input; the {\em type} of this fail value can be found
by examining one of the RHS expressions in one of the @EquationInfo@s.
\item
Call @match@ with all of this information!
\end{enumerate}

\begin{code}
658
matchWrapper :: HsMatchContext Name	-- For shadowing warning messages
659
	     -> MatchGroup Id		-- Matches being desugared
660
	     -> DsM ([Id], CoreExpr) 	-- Results
661
\end{code}
662

663 664
 There is one small problem with the Lambda Patterns, when somebody
 writes something similar to:
665
\begin{verbatim}
666
    (\ (x:xs) -> ...)
667
\end{verbatim}
668
 he/she don't want a warning about incomplete patterns, that is done with 
669 670 671 672 673 674 675 676 677 678 679 680
 the flag @opt_WarnSimplePatterns@.
 This problem also appears in the:
\begin{itemize}
\item @do@ patterns, but if the @do@ can fail
      it creates another equation if the match can fail
      (see @DsExpr.doDo@ function)
\item @let@ patterns, are treated by @matchSimply@
   List Comprension Patterns, are treated by @matchSimply@ also
\end{itemize}

We can't call @matchSimply@ with Lambda patterns,
due to the fact that lambda patterns can have more than
681 682 683
one pattern, and match simply only accepts one pattern.

JJQC 30-Nov-1997
684

685
\begin{code}
686
matchWrapper ctxt (MatchGroup matches match_ty)
simonpj@microsoft.com's avatar
simonpj@microsoft.com committed
687 688
  = ASSERT( notNull matches )
    do	{ eqns_info   <- mapM mk_eqn_info matches
689
	; new_vars    <- selectMatchVars arg_pats
690
	; result_expr <- matchEquations ctxt new_vars eqns_info rhs_ty
691
	; return (new_vars, result_expr) }
692
  where
693 694 695
    arg_pats    = map unLoc (hsLMatchPats (head matches))
    n_pats	= length arg_pats
    (_, rhs_ty) = splitFunTysN n_pats match_ty
696 697 698 699

    mk_eqn_info (L _ (Match pats _ grhss))
      = do { let upats = map unLoc pats
	   ; match_result <- dsGRHSs ctxt upats grhss rhs_ty
700
	   ; return (EqnInfo { eqn_pats = upats, eqn_rhs  = match_result}) }
701

702 703 704 705 706

matchEquations  :: HsMatchContext Name
		-> [Id]	-> [EquationInfo] -> Type
		-> DsM CoreExpr
matchEquations ctxt vars eqns_info rhs_ty
707 708
  = do	{ locn <- getSrcSpanDs
	; let   ds_ctxt   = DsMatchContext ctxt locn
709
		error_doc = matchContextErrString ctxt
710

711
	; match_result <- matchCheck ds_ctxt vars rhs_ty eqns_info
712

713
	; fail_expr <- mkErrorAppDs pAT_ERROR_ID rhs_ty error_doc
714
	; extractMatchResult match_result fail_expr }
715 716 717 718 719 720 721 722 723 724 725 726 727
\end{code}

%************************************************************************
%*									*
\subsection[matchSimply]{@matchSimply@: match a single expression against a single pattern}
%*									*
%************************************************************************

@mkSimpleMatch@ is a wrapper for @match@ which deals with the
situation where we want to match a single expression against a single
pattern. It returns an expression.

\begin{code}
728
matchSimply :: CoreExpr			-- Scrutinee
729 730
	    -> HsMatchContext Name	-- Match kind
	    -> LPat Id			-- Pattern it should match
731 732
	    -> CoreExpr			-- Return this if it matches
	    -> CoreExpr			-- Return this if it doesn't
733
	    -> DsM CoreExpr
734
-- Do not warn about incomplete patterns; see matchSinglePat comments
735 736
matchSimply scrut hs_ctx pat result_expr fail_expr = do
    let
737
      match_result = cantFailMatchResult result_expr
738 739 740
      rhs_ty       = exprType fail_expr
        -- Use exprType of fail_expr, because won't refine in the case of failure!
    match_result' <- matchSinglePat scrut hs_ctx pat rhs_ty match_result
741
    extractMatchResult match_result' fail_expr
742

743
matchSinglePat :: CoreExpr -> HsMatchContext Name -> LPat Id
744
	       -> Type -> MatchResult -> DsM MatchResult
745 746 747
-- Do not warn about incomplete patterns
-- Used for things like [ e | pat <- stuff ], where 
-- incomplete patterns are just fine
748 749 750 751 752 753 754 755 756 757
matchSinglePat (Var var) ctx (L _ pat) ty match_result 
  = do { locn <- getSrcSpanDs
       ; matchCheck (DsMatchContext ctx locn)
                    [var] ty  
                    [EqnInfo { eqn_pats = [pat], eqn_rhs  = match_result }] }

matchSinglePat scrut hs_ctx pat ty match_result
  = do { var <- selectSimpleMatchVarL pat
       ; match_result' <- matchSinglePat (Var var) hs_ctx pat ty match_result
       ; return (adjustMatchResult (bindNonRec var scrut) match_result') }
758 759
\end{code}

760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777

%************************************************************************
%*									*
		Pattern classification
%*									*
%************************************************************************

\begin{code}
data PatGroup
  = PgAny		-- Immediate match: variables, wildcards, 
			--		    lazy patterns
  | PgCon DataCon	-- Constructor patterns (incl list, tuple)
  | PgLit Literal	-- Literal patterns
  | PgN   Literal	-- Overloaded literals
  | PgNpK Literal	-- n+k patterns
  | PgBang		-- Bang patterns
  | PgCo Type		-- Coercion patterns; the type is the type
			--	of the pattern *inside*
778 779 780
  | PgView (LHsExpr Id) -- view pattern (e -> p):
                        -- the LHsExpr is the expression e
           Type         -- the Type is the type of p (equivalently, the result type of e)
781 782

groupEquations :: [EquationInfo] -> [[(PatGroup, EquationInfo)]]
783 784 785
-- If the result is of form [g1, g2, g3], 
-- (a) all the (pg,eq) pairs in g1 have the same pg
-- (b) none of the gi are empty
786
-- The ordering of equations is unchanged
787 788 789 790 791 792
groupEquations eqns
  = runs same_gp [(patGroup (firstPat eqn), eqn) | eqn <- eqns]
  where
    same_gp :: (PatGroup,EquationInfo) -> (PatGroup,EquationInfo) -> Bool
    (pg1,_) `same_gp` (pg2,_) = pg1 `sameGroup` pg2

793
subGroup :: Ord a => [(a, EquationInfo)] -> [[EquationInfo]]
794 795
-- Input is a particular group.  The result sub-groups the 
-- equations by with particular constructor, literal etc they match.
796 797 798
-- Each sub-list in the result has the same PatGroup
-- See Note [Take care with pattern order]
subGroup group 
799
    = map reverse $ Map.elems $ foldl accumulate Map.empty group
800
  where
801
    accumulate pg_map (pg, eqn)
802 803 804
      = case Map.lookup pg pg_map of
          Just eqns -> Map.insert pg (eqn:eqns) pg_map
          Nothing   -> Map.insert pg [eqn]      pg_map
805

806
    -- pg_map :: Map a [EquationInfo]
807 808
    -- Equations seen so far in reverse order of appearance
\end{code}
809

810 811 812 813 814 815 816 817 818
Note [Take care with pattern order]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
In the subGroup function we must be very careful about pattern re-ordering,
Consider the patterns [ (True, Nothing), (False, x), (True, y) ]
Then in bringing together the patterns for True, we must not 
swap the Nothing and y!


\begin{code}
819 820 821 822 823 824 825 826
sameGroup :: PatGroup -> PatGroup -> Bool
-- Same group means that a single case expression 
-- or test will suffice to match both, *and* the order
-- of testing within the group is insignificant.
sameGroup PgAny      PgAny      = True
sameGroup PgBang     PgBang     = True
sameGroup (PgCon _)  (PgCon _)  = True		-- One case expression
sameGroup (PgLit _)  (PgLit _)  = True		-- One case expression
827 828
sameGroup (PgN l1)   (PgN l2)   = l1==l2	-- Order is significant
sameGroup (PgNpK l1) (PgNpK l2) = l1==l2	-- See Note [Grouping overloaded literal patterns]
829
sameGroup (PgCo	t1)  (PgCo t2)  = t1 `eqType` t2
830 831 832 833
	-- CoPats are in the same goup only if the type of the
	-- enclosed pattern is the same. The patterns outside the CoPat
	-- always have the same type, so this boils down to saying that
	-- the two coercions are identical.
834 835 836
sameGroup (PgView e1 t1) (PgView e2 t2) = viewLExprEq (e1,t1) (e2,t2) 
       -- ViewPats are in the same gorup iff the expressions
       -- are "equal"---conservatively, we use syntactic equality
837
sameGroup _          _          = False
838

839
-- An approximation of syntactic equality used for determining when view
840
-- exprs are in the same group.
841
-- This function can always safely return false;
842 843
-- but doing so will result in the application of the view function being repeated.
--
844
-- Currently: compare applications of literals and variables
845 846 847 848 849 850 851
--            and anything else that we can do without involving other
--            HsSyn types in the recursion
--
-- NB we can't assume that the two view expressions have the same type.  Consider
--   f (e1 -> True) = ...
--   f (e2 -> "hi") = ...
viewLExprEq :: (LHsExpr Id,Type) -> (LHsExpr Id,Type) -> Bool
852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876
viewLExprEq (e1,_) (e2,_) = lexp e1 e2
  where
    lexp :: LHsExpr Id -> LHsExpr Id -> Bool
    lexp e e' = exp (unLoc e) (unLoc e')

    ---------
    exp :: HsExpr Id -> HsExpr Id -> Bool
    -- real comparison is on HsExpr's
    -- strip parens 
    exp (HsPar (L _ e)) e'   = exp e e'
    exp e (HsPar (L _ e'))   = exp e e'
    -- because the expressions do not necessarily have the same type,
    -- we have to compare the wrappers
    exp (HsWrap h e) (HsWrap h' e') = wrap h h' && exp e e'
    exp (HsVar i) (HsVar i') =  i == i' 
    -- the instance for IPName derives using the id, so this works if the
    -- above does
    exp (HsIPVar i) (HsIPVar i') = i == i' 
    exp (HsOverLit l) (HsOverLit l') = 
        -- Overloaded lits are equal if they have the same type
        -- and the data is the same.
        -- this is coarser than comparing the SyntaxExpr's in l and l',
        -- which resolve the overloading (e.g., fromInteger 1),
        -- because these expressions get written as a bunch of different variables
        -- (presumably to improve sharing)
877
        eqType (overLitType l) (overLitType l') && l == l'
878 879 880 881 882 883 884 885 886 887 888 889
    exp (HsApp e1 e2) (HsApp e1' e2') = lexp e1 e1' && lexp e2 e2'
    -- the fixities have been straightened out by now, so it's safe
    -- to ignore them?
    exp (OpApp l o _ ri) (OpApp l' o' _ ri') = 
        lexp l l' && lexp o o' && lexp ri ri'
    exp (NegApp e n) (NegApp e' n') = lexp e e' && exp n n'
    exp (SectionL e1 e2) (SectionL e1' e2') = 
        lexp e1 e1' && lexp e2 e2'
    exp (SectionR e1 e2) (SectionR e1' e2') = 
        lexp e1 e1' && lexp e2 e2'
    exp (ExplicitTuple es1 _) (ExplicitTuple es2 _) =
        eq_list tup_arg es1 es2
890
    exp (HsIf _ e e1 e2) (HsIf _ e' e1' e2') =
891 892 893 894 895 896 897 898 899 900
        lexp e e' && lexp e1 e1' && lexp e2 e2'

    -- Enhancement: could implement equality for more expressions
    --   if it seems useful
    -- But no need for HsLit, ExplicitList, ExplicitTuple, 
    -- because they cannot be functions
    exp _ _  = False

    ---------
    tup_arg (Present e1) (Present e2) = lexp e1 e2
901
    tup_arg (Missing t1) (Missing t2) = eqType t1 t2
902 903 904 905 906 907 908 909 910 911 912 913
    tup_arg _ _ = False

    ---------
    wrap :: HsWrapper -> HsWrapper -> Bool
    -- Conservative, in that it demands that wrappers be
    -- syntactically identical and doesn't look under binders
    --
    -- Coarser notions of equality are possible
    -- (e.g., reassociating compositions,
    --        equating different ways of writing a coercion)
    wrap WpHole WpHole = True
    wrap (WpCompose w1 w2) (WpCompose w1' w2') = wrap w1 w1' && wrap w2 w2'
914
    wrap (WpCast c)  (WpCast c')     = coreEqCoercion c c'
915
    wrap (WpEvApp et1) (WpEvApp et2) = ev_term et1 et2
916
    wrap (WpTyApp t) (WpTyApp t')    = eqType t t'
917 918 919 920 921 922 923
    -- Enhancement: could implement equality for more wrappers
    --   if it seems useful (lams and lets)
    wrap _ _ = False

    ---------
    ev_term :: EvTerm -> EvTerm -> Bool
    ev_term (EvId a)       (EvId b)       = a==b
924
    ev_term (EvCoercion a) (EvCoercion b) = coreEqCoercion a b
925 926 927 928 929 930 931 932
    ev_term _ _ = False	

    ---------
    eq_list :: (a->a->Bool) -> [a] -> [a] -> Bool
    eq_list _  []     []     = True
    eq_list _  []     (_:_)  = False
    eq_list _  (_:_)  []     = False
    eq_list eq (x:xs) (y:ys) = eq x y && eq_list eq xs ys
933

934 935 936 937 938
patGroup :: Pat Id -> PatGroup
patGroup (WildPat {})       	      = PgAny
patGroup (BangPat {})       	      = PgBang  
patGroup (ConPatOut { pat_con = dc }) = PgCon (unLoc dc)
patGroup (LitPat lit)		      = PgLit (hsLitKey lit)
939
patGroup (NPat olit mb_neg _)	      = PgN   (hsOverLitKey olit (isJust mb_neg))
940
patGroup (NPlusKPat _ olit _ _)	      = PgNpK (hsOverLitKey olit False)
941 942
patGroup (CoPat _ p _)		      = PgCo  (hsPatType p)	-- Type of innelexp pattern
patGroup (ViewPat expr p _)               = PgView expr (hsPatType (unLoc p))
943 944 945
patGroup pat = pprPanic "patGroup" (ppr pat)
\end{code}

946 947
Note [Grouping overloaded literal patterns]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
948 949 950 951 952 953 954
WATCH OUT!  Consider

	f (n+1) = ...
	f (n+2) = ...
	f (n+1) = ...

We can't group the first and third together, because the second may match 
955 956 957 958 959 960 961 962
the same thing as the first.  Same goes for *overloaded* literal patterns
	f 1 True = ...
	f 2 False = ...
	f 1 False = ...
If the first arg matches '1' but the second does not match 'True', we
cannot jump to the third equation!  Because the same argument might
match '2'!
Hence we don't regard 1 and 2, or (n+1) and (n+2), as part of the same group.
963