slaesvuo/hledger
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix a slowdown with report rendering in 1.19.1 (#1350)

Browse Source 
stripAnsi is called many times during rendering (by strWidth), so
should be fast. It was originally a regex replacement, and more
recently a custom parser. The parser was slower, particularly the one
in 1.19.1. See #1350, and this rough test:

time118ish = timeIt $ print $ length $ concat $ map (fromRight undefined . regexReplace (toRegex' "\ESC\\[([0-9]+;)*([0-9]+)?[ABCDHJKfmsu]") "") testdata
time119    = timeparser (many (takeWhile1P Nothing (/='\ESC') <|> "" <$ ansi))
time1191   = timeparser (many ("" <$ try ansi <|> pure <$> anySingle))
timeparser p = timeIt $ print $ length $ concat $ map (concat . fromJust . parseMaybe p) testdata
testdata = concat $ replicate 10000
    [ "2008-01-01 income               assets🏦checking            $1            $1"
    , "2008-06-01 gift                 assets🏦checking            $1            $2"
    , "2008-06-02 save                 assets🏦saving              $1            $3"
    , "                                assets🏦checking  ..m$-1\ESC[m\ESC[m            $2"
    , "2008-06-03 eat & shop           assets:cash           ..m$-2\ESC[m\ESC[m             0"
    , "2008-12-31 pay off              assets🏦checking  ..m$-1\ESC[m\ESC[m  ..m$-1\ESC[m\ESC[m"
    ]

ghci> time118ish
4560000
CPU time:   0.17s
ghci> time119
4560000
CPU time:   0.91s
ghci> time1191
4560000
CPU time:   2.76s

Possibly a more careful parser could beat regexReplace. Note the
latter does memoisation, which could be faster and/or could also use
more resident memory in some situations.

Ideally we would calculate all widths before adding ANSI colour codes,
so we wouldn't have to wastefully strip them.
This commit is contained in:
Simon Michael 2020-09-10 17:46:16 -07:00
parent 2d068662c1
commit f78dc639a5
1 changed files with 8 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
hledger-lib/Hledger/Utils/String.hs
View File

@ -48,15 +48,14 @@ module Hledger.Utils.String (
) where
import Data.Char (isDigit, isSpace, toLower, toUpper)
import Data.Char (isSpace, toLower, toUpper)
import Data.List (intercalate, transpose)
import Text.Megaparsec (Parsec, (<|>), (<?>), anySingle, between, many, noneOf,
oneOf, parseMaybe, sepBy, takeWhileP, try)
import Text.Megaparsec.Char (char, string)
import Text.Megaparsec ((<|>), between, many, noneOf, sepBy)
import Text.Megaparsec.Char (char)
import Text.Printf (printf)
import Hledger.Utils.Parse
import Hledger.Utils.Regex (toRegex', regexReplace)
-- | Take elements from the end of a list.
takeEnd n l = go (drop n l) l
@ -342,14 +341,10 @@ strWidth = maximum . (0:) . map (foldr (\a b -> charWidth a + b) 0) . lines . st
-- >>> stripAnsi "\ESC[31m-1\ESC[m"
-- "-1"
stripAnsi :: String -> String
stripAnsi s = case parseMaybe (many $ "" <$ try ansi <|> pure <$> anySingle) s of
Nothing -> error "Bad ansi escape" -- PARTIAL: should not happen
Just xs -> concat xs
stripAnsi s = either err id $ regexReplace ansire "" s
where
-- This parses lots of invalid ANSI escape codes, but that should be fine
ansi = string "\ESC[" *> digitSemicolons *> suffix <?> "ansi" :: Parsec CustomErr String Char
digitSemicolons = takeWhileP Nothing (\c -> isDigit c || c == ';')
suffix = oneOf ['A', 'B', 'C', 'D', 'H', 'J', 'K', 'f', 'm', 's', 'u']
err = error "stripAnsi: invalid replacement pattern" -- PARTIAL, shouldn't happen
ansire = toRegex' "\ESC\\[([0-9]+;)*([0-9]+)?[ABCDHJKfmsu]" -- PARTIAL, should succeed
-- | Get the designated render width of a character: 0 for a combining
-- character, 1 for a regular character, 2 for a wide character.