diff --git a/.travis.yml b/.travis.yml --- a/.travis.yml +++ b/.travis.yml @@ -1,19 +1,59 @@ -language: haskell -before_install: - # Uncomment whenever hackage is down. - # - mkdir -p ~/.cabal && cp travis/config ~/.cabal/config && cabal update - - cabal update +# NB: don't set `language: haskell` here - # Try installing some of the build-deps with apt-get for speed. - - travis/cabal-apt-install $mode +# See also https://github.com/hvr/multi-ghc-travis for more information +env: + # we have to use CABALVER=1.16 for GHC<7.6 as well, as there's + # no package for earlier cabal versions in the PPA + - GHCVER=7.4.2 CABALVER=1.16 + - GHCVER=7.6.3 CABALVER=1.16 + - GHCVER=7.8.4 CABALVER=1.18 + - GHCVER=7.10.1 CABALVER=1.22 + - GHCVER=head CABALVER=1.22 +matrix: + allow_failures: + - env: GHCVER=head CABALVER=1.22 + +# Note: the distinction between `before_install` and `install` is not +# important. +before_install: + - travis_retry sudo add-apt-repository -y ppa:hvr/ghc + - travis_retry sudo apt-get update + - travis_retry sudo apt-get install cabal-install-$CABALVER ghc-$GHCVER + - export PATH=/opt/ghc/$GHCVER/bin:/opt/cabal/$CABALVER/bin:$PATH + - cabal --version + install: - - cabal configure -flib-Werror $mode - - cabal build + - travis_retry cabal update + - cabal install --only-dependencies + - travis_retry sudo apt-get -q -y install hlint || cabal install hlint +# Here starts the actual work to be performed for the package under +# test; any command which exits with a non-zero exit code causes the +# build to fail. script: - - $script && hlint src --cpp-define HLINT + # -v2 provides useful information for debugging + - cabal configure -v2 --enable-tests + # this builds all libraries and executables + # (including tests/benchmarks) + - cabal build + - cabal test + + # tests that a source-distribution can be generated + - cabal sdist + - hlint src --cpp-define HLINT + + # check that the generated source-distribution can be built & installed + - export SRC_TGZ=$(cabal info . | awk '{print $2 ".tar.gz";exit}') ; + cd dist/; + if [ -f "$SRC_TGZ" ]; then + cabal install --force-reinstalls "$SRC_TGZ"; + else + echo "expected '$SRC_TGZ' not found"; + exit 1; + fi + notifications: irc: channels: @@ -21,6 +61,3 @@ skip_join: true template: - "\x0313hyperloglog\x03/\x0306%{branch}\x03 \x0314%{commit}\x03 %{build_url} %{message}" - -env: - - mode="--enable-tests" script="cabal test --show-details=always" diff --git a/CHANGELOG.markdown b/CHANGELOG.markdown --- a/CHANGELOG.markdown +++ b/CHANGELOG.markdown @@ -1,3 +1,14 @@ +0.3.1 +----- +* Added `insertHash`. This enables users of the 0.2 + era version of `hyperloglog` to manually pick the + hash used and load their old data. +* Builds warning-free on GHC 7.10 + +0.3.0.1 +--- +* Constraint bumps for `lens` and `generic-deriving` + 0.3 --- * Switched to `SipHash`, so the package actually works. diff --git a/LICENSE b/LICENSE --- a/LICENSE +++ b/LICENSE @@ -1,4 +1,4 @@ -Copyright 2013 Edward Kmett +Copyright 2013-2015 Edward Kmett All rights reserved. diff --git a/hyperloglog.cabal b/hyperloglog.cabal --- a/hyperloglog.cabal +++ b/hyperloglog.cabal @@ -1,6 +1,6 @@ name: hyperloglog category: Numeric -version: 0.3 +version: 0.3.1 license: BSD3 cabal-version: >= 1.8 license-file: LICENSE @@ -9,9 +9,9 @@ stability: provisional homepage: http://github.com/analytics/hyperloglog bug-reports: http://github.com/analytics/hyperloglog/issues -copyright: Copyright (C) 2013 Edward A. Kmett +copyright: Copyright (C) 2013-2015 Edward A. Kmett build-type: Custom -tested-with: GHC == 7.4.1, GHC == 7.6.1 +tested-with: GHC == 7.4.2, GHC == 7.6.3, GHC == 7.8.4, GHC == 7.10.0.20150307 synopsis: An approximate streaming (constant space) unique object counter description: This package provides an approximate streaming (constant space) unique object counter. @@ -58,10 +58,10 @@ comonad >= 4 && < 5, deepseq >= 1.3 && < 1.5, distributive >= 0.3 && < 1, - generic-deriving >= 1.4 && < 1.7, + generic-deriving >= 1.4 && < 1.8, hashable >= 1.1.2.3 && < 1.3, hashable-extras >= 0.1 && < 1, - lens >= 4 && < 4.5, + lens >= 4 && < 5, reflection >= 1.3 && < 2, semigroupoids >= 4 && < 5, semigroups >= 0.8.4 && < 1, diff --git a/src/Data/HyperLogLog.hs b/src/Data/HyperLogLog.hs --- a/src/Data/HyperLogLog.hs +++ b/src/Data/HyperLogLog.hs @@ -1,6 +1,6 @@ -------------------------------------------------------------------- -- | --- Copyright : (c) Edward Kmett 2013 +-- Copyright : (c) Edward Kmett 2013-2015 -- License : BSD3 -- Maintainer: Edward Kmett -- Stability : experimental @@ -17,6 +17,7 @@ , size , intersectionSize , insert + , insertHash , cast -- * Config , Config diff --git a/src/Data/HyperLogLog/Config.hs b/src/Data/HyperLogLog/Config.hs --- a/src/Data/HyperLogLog/Config.hs +++ b/src/Data/HyperLogLog/Config.hs @@ -1,15 +1,12 @@ {-# LANGUAGE CPP #-} -{-# LANGUAGE BangPatterns #-} {-# LANGUAGE DeriveGeneric #-} {-# LANGUAGE FlexibleContexts #-} {-# LANGUAGE NoMonomorphismRestriction #-} {-# LANGUAGE EmptyDataDecls #-} {-# LANGUAGE RankNTypes #-} -{-# LANGUAGE RecordWildCards #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE UndecidableInstances #-} {-# LANGUAGE MultiParamTypeClasses #-} -{-# LANGUAGE FunctionalDependencies #-} {-# LANGUAGE FlexibleInstances #-} {-# LANGUAGE TypeFamilies #-} {-# OPTIONS_GHC -fno-cse #-} @@ -34,7 +31,7 @@ {-# OPTIONS_GHC -fno-warn-missing-signatures #-} -------------------------------------------------------------------- -- | --- Copyright : (c) Edward Kmett 2013 +-- Copyright : (c) Edward Kmett 2013-2015 -- License : BSD3 -- Maintainer: Edward Kmett -- Stability : experimental diff --git a/src/Data/HyperLogLog/Type.hs b/src/Data/HyperLogLog/Type.hs --- a/src/Data/HyperLogLog/Type.hs +++ b/src/Data/HyperLogLog/Type.hs @@ -7,7 +7,6 @@ {-# LANGUAGE MultiParamTypeClasses #-} {-# LANGUAGE NoMonomorphismRestriction #-} {-# LANGUAGE RankNTypes #-} -{-# LANGUAGE RecordWildCards #-} {-# LANGUAGE ScopedTypeVariables #-} {-# LANGUAGE TypeFamilies #-} {-# LANGUAGE UndecidableInstances #-} @@ -26,7 +25,7 @@ -------------------------------------------------------------------- -- | --- Copyright : (c) Edward Kmett 2013 +-- Copyright : (c) Edward Kmett 2013-2015 -- License : BSD3 -- Maintainer: Edward Kmett -- Stability : experimental @@ -45,11 +44,15 @@ , HasHyperLogLog(..) , size , insert + , insertHash , intersectionSize , cast ) where +#if __GLASGOW_HASKELL__ < 710 import Control.Applicative +#endif + import Control.Lens import Control.Monad import Crypto.MAC.SipHash @@ -135,21 +138,25 @@ sipKey :: SipKey sipKey = SipKey 4 7 -siphash :: (Serial a) => a -> Word64 +siphash :: Serial a => a -> Word64 siphash a = h where !bs = runPutS (serialize a) (SipHash !h) = hash sipKey bs {-# INLINE siphash #-} insert :: (ReifiesConfig s, Serial a) => a -> HyperLogLog s -> HyperLogLog s -insert a m@(HyperLogLog v) = HyperLogLog $ V.modify (\x -> do +insert = insertHash . w32 . siphash +{-# INLINE insert #-} + +-- | Insert a value that has already been hashed by whatever user defined hash function you want. +insertHash :: ReifiesConfig s => Word32 -> HyperLogLog s -> HyperLogLog s +insertHash h m@(HyperLogLog v) = HyperLogLog $ V.modify (\x -> do old <- MV.read x bk when (rnk > old) $ MV.write x bk rnk ) v where - !h = w32 (siphash a) !bk = calcBucket m h !rnk = calcRank m h -{-# INLINE insert #-} +{-# INLINE insertHash #-} -- | Approximate size of our set size :: ReifiesConfig p => HyperLogLog p -> Approximate Int64