fixed 0.2.1.1 → 0.3

raw patch · 5 files changed

+11/−9 lines, 5 filesPVP ok

version bump matches the API change (PVP)

API changes (from Hackage documentation)

Files

• .travis.yml +0/−2
• CHANGELOG.markdown +4/−0
• README.markdown +1/−1
• fixed.cabal +2/−2
• src/Numeric/Fixed.hs +4/−4

@@ -4,8 +4,6 @@ env:  # we have to use CABALVER=1.16 for GHC<7.6 as well, as there's  # no package for earlier cabal versions in the PPA- - GHCVER=7.4.2 CABALVER=1.16- - GHCVER=7.6.3 CABALVER=1.16  - GHCVER=7.8.4 CABALVER=1.18  - GHCVER=7.10.1 CABALVER=1.22  - GHCVER=head CABALVER=1.22

@@ -1,3 +1,7 @@+0.3+---+* Fixed an outstanding bug in round.+ 0.2.1.1 ------- * Minor documentation improvements.

@@ -1,7 +1,7 @@ fixed ==== -[![Build Status](https://secure.travis-ci.org/ekmett/fixed.png?branch=master)](http://travis-ci.org/ekmett/fixed)+[![Hackage](https://img.shields.io/hackage/v/fixed.svg)](https://hackage.haskell.org/package/fixed) [![Build Status](https://secure.travis-ci.org/ekmett/fixed.png?branch=master)](http://travis-ci.org/ekmett/fixed)  This package supplies signed fixed-precision values 15 bits above the decimal, 16 bits below. 

@@ -1,6 +1,6 @@ name:          fixed category:      Numeric-version:       0.2.1.1+version:       0.3 license:       BSD3 cabal-version: >= 1.8 license-file:  LICENSE@@ -9,7 +9,7 @@ stability:     provisional homepage:      http://github.com/ekmett/fixed bug-reports:   http://github.com/ekmett/fixed/issues-copyright:     Copyright (C) 2014 Edward A. Kmett+copyright:     Copyright (C) 2014-2015 Edward A. Kmett build-type:    Simple tested-with:   GHC == 7.8.3 synopsis:      Signed 15.16 precision fixed point arithmetic

@@ -13,7 +13,7 @@ -- One sign bit, 15 bits to the left of the decimal place and 16 bits -- to the right packed into a 32-bit integer. ------------------------------------------------------------------------------module Numeric.Fixed +module Numeric.Fixed   ( Fixed(..)   , fromFixed   , toFixed@@ -72,13 +72,13 @@   toRational (Fixed i) = toInteger i % 65536  instance RealFrac Fixed where-  properFraction (Fixed a) +  properFraction (Fixed a)     | a >= 0 = (fromIntegral (unsafeShiftR a 16), Fixed (a .&. 0xffff))     | otherwise = (negate $ fromIntegral $ unsafeShiftR (negate a) 16, Fixed $ (a .&. 0xffff) - 0x10000)-  truncate (Fixed a) +  truncate (Fixed a)     | a >= 0 = fromIntegral (unsafeShiftR a 16)     | otherwise = negate $ fromIntegral $ unsafeShiftR (negate a) 16-  round (Fixed f)   = fromIntegral $ unsafeShiftR (f + 0x800) 16 +  round (Fixed f)   = fromIntegral $ unsafeShiftR (f + 0x8000) 16   ceiling (Fixed f) = fromIntegral $ unsafeShiftR (f + 0xffff) 16   floor (Fixed f)   = fromIntegral $ unsafeShiftR f 16