[GHC] #15646: ghci takes super long time to find the type of large fractional number

[GHC]

#15646: ghci takes super long time to find the type of large fractional number
-------------------------------------+-------------------------------------
        Reporter:  Johannkokos       |                Owner:
                                     |  JulianLeviston
            Type:  bug               |               Status:  new
        Priority:  normal            |            Milestone:  8.6.1
       Component:  GHCi              |              Version:  8.4.3
      Resolution:                    |             Keywords:  newcomer
Operating System:  Unknown/Multiple  |         Architecture:
 Type of failure:  Compile-time      |  Unknown/Multiple
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:                    |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by JulianLeviston):

 I got it compiling and typechecking, then set about writing a test.
 Discussed with you (osa1) and a couple others in IRC about writing tests.
 Looked more into how the testing framework works, read up on writing
 tests. Pretty sure I can figure that out now (it's really not going to be
 hard, once I read all the test guides).

 So that's fine, but I had an issue compiling after doing a `git pull` (for
 some reason I had to recompile stage 1 of the compiler again), so I
 recompiled and thought I'd try out my changes. So I typechecked a large
 scientific notational literal, and it got `*** Exception: No match in
 record selector thfl_value` I realised immediately that was because I'd
 just made the types match in the desugarer and renamer by using the new TH
 part of the FractionalLit type because the desugarer talks bout pattern
 group matching on HsFractional and seemed to expect a Rational.

 So, the `PatGroup` data type, has a constructor `PgN :: Rational ->
 PatGroup` which I'm now going to refactor to be `PgN :: Integer -> Integer
 -> PatGroup`.

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15646#comment:39>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler