[GHC] #8710: Overlapping patterns warning misplaced

[GHC]

#8710: Overlapping patterns warning misplaced
-------------------------------------+-------------------------------------
        Reporter:  goldfire          |                Owner:
            Type:  bug               |               Status:  new
        Priority:  low               |            Milestone:
       Component:  Compiler          |              Version:  7.7
      Resolution:                    |             Keywords:
                                     |  PatternMatchWarnings
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Incorrect         |  Unknown/Multiple
  warning at compile-time            |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):  Phab:D1910
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by gkaracha):

 Replying to [comment:8 simonpj]:
 > > If we print a separate warning for every redundant clause it may
 become overwhelming
 >
 > Really? There are a few artificial examples in the testsuite, but I'd
 have thought that one error per redundant equation would be ok.

 Hmmm, true. The number of redundant clauses (and of those with
 inaccessible RHS) is bound by the
 number of clauses a match has (it's not the case with the uncovered
 though). In practice I have
 not seen matches with many redundant cases. Hence, I created Phab:D1920
 too, which implements this
 approach (one warning per redundant clause, each with its own location).

 >
 > But still, I'd really like the common case of one redundant equation to
 look like
 > {{{
 > Bug.hs:24:3: warning:                     -- Note correct line number
 >     Pattern match(es) are redundant
 >     In an equation for ‘show’:
 >         len _     = ... (at Bug.hs:24:3)
 > }}}
 > Or maybe even
 > {{{
 > Bug.hs:24:3: warning:
 >     Pattern match(es) are redundant
 >     In an equation for ‘show’:
 >         len _     = ...            -- No need to duplicate here
 > }}}
 > If there are more than one, then adding the individual locations is good
 > {{{
 > Bug.hs:24:3: warning:
 >     Pattern match(es) are redundant
 >     In an equation for ‘show’:
 >         len _     = ... (at Bug.hs:24:3)
 >         len _     = ... (at Bug.hs:25:3)
 > }}}
 > but the herald should mention the first.
 >
 > Can't be hard, I guess.

 Agreed on this one too, I have updated Phab:D1910 to behave like this.
 Merge at will the revision
 you prefer. :-)

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8710#comment:9>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler