#10067: The Read Integer instance is too slow -------------------------------------+------------------------------------- Reporter: redneb | Owner: Type: feature request | Status: closed Priority: normal | Milestone: 7.12.1 Component: Core Libraries | Version: 7.11 Resolution: fixed | Keywords: Operating System: Unknown/Multiple | Architecture: Type of failure: Runtime | Unknown/Multiple performance bug | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Revisions: Phab:D645 -------------------------------------+------------------------------------- Changes (by bgamari): * priority: high => normal * status: new => closed * resolution: => fixed Comment: As far as I can tell the critical issue of quadratic parsing of integers is by Phab:D645, which has been merged. The only work that remains here is to specialize for `Natural` but as Edward points out this is now a reasonable cheap operation as we can show, {{{#!hs import GHC.Natural import Criterion.Main main = defaultMain [ bench "integer" $ nf (read :: String -> Integer) (take 1000000 $ cycle "1234567890") , bench "natural" $ nf (read :: String -> Natural) (take 1000000 $ cycle "1234567890") ] }}} {{{ $ ghc Test.hs -O -fforce-recomp [1 of 1] Compiling Main ( Test.hs, Test.o ) Linking Test ... $ ./Test benchmarking integer time 435.5 ms (410.6 ms .. 484.4 ms) 0.998 R² (0.997 R² .. 1.000 R²) mean 441.3 ms (434.3 ms .. 445.5 ms) std dev 6.383 ms (0.0 s .. 7.230 ms) variance introduced by outliers: 19% (moderately inflated) benchmarking natural time 428.2 ms (415.5 ms .. 446.6 ms) 1.000 R² (1.000 R² .. 1.000 R²) mean 434.6 ms (432.0 ms .. 436.0 ms) std dev 2.261 ms (0.0 s .. 2.387 ms) variance introduced by outliers: 19% (moderately inflated) }}} For this reason I"m going to close this issue.. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/10067#comment:15 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler