[GHC] #15115: `Numeric.showEFloat` does not honor precision with 0 digits

[GHC]

#15115: `Numeric.showEFloat` does not honor precision with 0 digits
-------------------------------------+-------------------------------------
           Reporter:  guibou         |             Owner:  (none)
               Type:  bug            |            Status:  new
           Priority:  normal         |         Milestone:  8.6.1
          Component:                 |           Version:  8.4.2
  libraries/base                     |
           Keywords:                 |  Operating System:  Unknown/Multiple
       Architecture:                 |   Type of failure:  None/Unknown
  Unknown/Multiple                   |
          Test Case:                 |        Blocked By:
           Blocking:                 |   Related Tickets:
Differential Rev(s):                 |         Wiki Page:
-------------------------------------+-------------------------------------
 Hello,

 `Numeric.showEFloat` should display a number of digit after the dot
 depending on the `Maybe Int` argument. It works for values > 0:

 {{{#!haskell
 Prelude Numeric> Numeric.showEFloat (Just 5) pi ""
 "3.14159e0"
 Prelude Numeric> Numeric.showEFloat (Just 1) pi ""
 "3.1e0"
 }}}

 But for `0` it shows the same result as with `1`:

 {{{#!haskell
 Prelude Numeric> Numeric.showEFloat (Just 0) pi ""
 "3.1e0"
 }}}

 I was expecting `3e0` or `3e0`.

 As a matter of comparison, here is the behavior with `Numeric.showFFloat`:

 {{{#!haskell
 Prelude Numeric> Numeric.showFFloat (Just 5) pi ""
 "3.14159"
 Prelude Numeric> Numeric.showFFloat (Just 1) pi ""
 "3.1"
 Prelude Numeric> Numeric.showFFloat (Just 0) pi ""
 "3"
 }}}

-- 
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/15115>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler