subtitleParser 0.2 → 0.3

raw patch · 5 files changed

+81/−43 lines, 5 filesdep ~attoparsecPVP ok

version bump matches the API change (PVP)

Dependency ranges changed: attoparsec

API changes (from Hackage documentation)

Files

• Text/Subtitles/SRT.hs +34/−11
• Text/Subtitles/SRT/Datatypes.hs +3/−23
• Text/Subtitles/SUB.hs +29/−6
• Text/Subtitles/SUB/Datatypes.hs +12/−0
• subtitleParser.cabal +3/−3

@@ -10,6 +10,9 @@  module Text.Subtitles.SRT    (+  -- * Don't use parseOnly!+  -- $noParseOnly +     -- * Terminology of the module   -- $example   @@ -18,16 +21,29 @@    -- * Main parsers   parseSRT,-  parseSingleLine+  parseSingleLine,++  -- * Temporal solutions+  parseOnly'   ) where  import Prelude hiding (takeWhile) import Control.Applicative-import Data.Attoparsec.Text +import Data.Attoparsec.Text hiding (parseOnly) import qualified Data.Text as T -- in project modules import Text.Subtitles.SRT.Datatypes +-- $noParseOnly+--+-- This module uses now peekChar in parseDialog, which replaces the ad-hoc+-- method used before. As a consequence it doesn't play well with+-- Data.Attoparsec.Text.parseOnly on some conditions.+--+-- You should use just 'parse' or the parseOnly' function I provide to avoid+-- problems until further notice. /Hopefully/ in the next version of attoparsec+-- this will be solved , so keep an eye for removal!.+ -- $example -- -- All the sections of a Line have their corresponding ADT in@@ -53,11 +69,11 @@ parseSRT :: Parser Subtitles parseSRT = many1 parseSingleLine --- |The individual Line parser. given the upper example return the+-- |The individual Line parser. Given the upper example return the -- corresponding Line representation parseSingleLine :: Parser Line parseSingleLine = -  Line <$> parseIndex <*> parseRange <* eol <*> parseDialog T.empty+  Line <$> parseIndex <*> parseRange <*> parseDialog T.empty  parseIndex :: Parser Int parseIndex = decimal <* eol@@ -65,9 +81,14 @@ eol :: Parser () eol = endOfLine  +-- |This version avoid the problems associated with peekChar and thus is safe to+-- use in this module. Subject to removal once parseOnly is fixed.+parseOnly' :: Parser a -> Text -> Either String a+parseOnly' p t = eitherResult $ feed (parse p t) T.empty+ {- Is clear that this just aplies parseTime breaking down the "-->" string -} parseRange :: Parser Range-parseRange = Range <$> parseTime <* arrowString <*> parseTime +parseRange = Range <$> parseTime <* arrowString <*> parseTime <* skipSpace   where     arrowString :: Parser Text     arrowString = string (T.pack " --> ") @@ -78,15 +99,17 @@     numDot :: Parser Int     numDot = decimal <* char ':' -{- return the dialog checking for newlines that could be in there. that why is- - written in a monad instead of applicative. more efficient version welcome -}+{- return the dialog checking for newlines that could be in there. That why is+ - written in a monad instead of applicative. More efficient version welcome -} parseDialog :: Text -> Parser Text parseDialog t = do    line <- takeWhile1 (not . isEndOfLine)   endOfLine-  let lineState = T.append t (T.snoc line '\n')-  next <- anyChar+  let lineState = T.append t (T.snoc line '\n') --takeWhile1 didn't consume \n+  next <- peekChar   case next of-    '\n' -> return lineState -    _    -> parseDialog (T.snoc lineState next)+    Nothing     -> return lineState +    (Just '\n') -> eol >> return lineState +    (Just _)    -> parseDialog lineState + 

@@ -31,12 +31,12 @@   , minutes :: Int   , seconds :: Int   , frame   :: Int-  } deriving (Eq, Ord)+  } deriving (Eq, Ord, Show)  data Range = Range    { from :: Time   , to   :: Time-  } deriving (Eq, Ord)+  } deriving (Eq, Ord, Show)  -- | The core of the parser. each one of the constructor representing one part -- of the Line@@ -44,27 +44,7 @@   { index  :: Int   , range  :: Range   , dialog   :: Text-  } deriving (Eq, Ord)+  } deriving (Eq, Ord, Show)  type Subtitles = [Line] -instance Show Time where-  show (Time h m s f) = concat [showT h, ":", showT m, ":", showT s, ",", showF f]--instance Show Range where-  show (Range f t) = concat [show f, " --> ", show t]--instance Show Line where-  show (Line i t s) = intercalate "\n" (show i : show t : [unpack s]) ++ "\n"--{- showT stands for showTime. In the parsing process "00" is read as "0", and- - that tailing 0 is lost unless we add it manually when showing -}-showT :: Int -> String-showT a | a < 10    = '0' : show a-        | otherwise = show a--{- showF stands for showFrame -}-showF :: Int -> String-showF a | a < 10    = '0':'0': show a-        | a < 100   = '0': show a-        | otherwise = show a

@@ -6,7 +6,7 @@ -- Maintainer  : ruben.astud@gmail.com -- Portability : unknown ----- A basic parser for .sub files (subtitles) based on 'Attoparsec' and 'Text'+-- A basic parser for .sub files (microDVD) based on 'Attoparsec' and 'Text'  module Text.Subtitles.SUB    (@@ -23,27 +23,50 @@  import Control.Applicative import Data.Attoparsec.Text+import Data.Text (pack, Text)+import Data.Maybe (Maybe)  import Text.Subtitles.SUB.Datatypes  -- $example --+-- I strongly recommend to understand the 'Line' Datatype , which is the+-- foundation of this module.+-- -- Refering to the parts of a single line of a subtitle file. ----- >{50}{100}Drama here!+-- >{50}{100}{y:i}Drama here! ----- the first to numbers correspond to the frame in which the text Drama here!--- is displayed. +-- The first to numbers correspond to the frame in which the text Drama here!+-- is displayed. Note that this implementation support the optional flags about+-- text format and color  frame :: Parser Frame-frame = char '{' *> decimal <* char '}'+frame = enclosed decimal  -- | Given the example return the corresponding Line representation. At the -- moment this not handles modifiers as underlines or bold text parseSingleLine :: Parser Line -parseSingleLine = Line <$> frame <*> frame <*> takeTill isEndOfLine+parseSingleLine = Line <$> frame <*> frame <*> parseProperty <*> takeTill isEndOfLine  -- | Main parser of .sub files, given a .sub file it return a list of the dialog -- lines parseSUB :: Parser Subtitles parseSUB = sepBy1 parseSingleLine endOfLine++parseProperty :: Parser (Maybe TextProperty) +parseProperty = option Nothing (Just <$> enclosed (choice allProp))+  where+    bold      = matchText "y:b" *> pure Bold+    italic    = matchText "y:i" *> pure Italic+    underline = matchText "y:u" *> pure UnderLine+    stroked   = matchText "y:s" *> pure Stroked+    color     = matchText "C:"  *> (C <$> takeWhile1 ('}' /=))+    allProp   = [bold, italic, underline, stroked, color]++-- All .sub command are enclosed in bracket. Worth defining.+enclosed :: Parser a -> Parser a+enclosed p = char '{' *> p <* char '}'++matchText :: String -> Parser Text+matchText s = string (pack s)

@@ -18,19 +18,31 @@   -- * Datatypes   Frame,   Subtitles,+  Color,+  TextProperty(..),   Line(..)   ) where  import Data.Text (Text) import Data.Attoparsec.Text (Parser)+import Data.Maybe (Maybe)  type Subtitles = [Line]  {- Should this be a newtype? or just a type? -} type Frame = Int +-- .sub utilize a non-standard RGB format, is better to keep them as Text +type Color = Text++-- |Optional property of text, the constructor should be clear.+data TextProperty = Italic | Bold | UnderLine | Stroked +                  | C Color -- ^ This is just text because is non-standard RGB+  deriving (Show)+ data Line = Line { startFrame :: Frame,                    finalFrame :: Frame,+                   property   :: Maybe TextProperty,                     dialog     :: Text }   deriving (Show) 

@@ -1,5 +1,5 @@ name:            subtitleParser-version:         0.2+version:         0.3 license:         BSD3 license-file:    LICENSE category:        Text, Parsing@@ -26,14 +26,14 @@   build-depends: base < 5,                  containers,                  text >= 0.11.1.5,-                 attoparsec+                 attoparsec >= 0.10.2.0    exposed-modules: Text.Subtitles.SRT                    Text.Subtitles.SRT.Datatypes                    Text.Subtitles.SUB                    Text.Subtitles.SUB.Datatypes -  ghc-options: -Wall+  ghc-options:     if flag(developer)     ghc-prof-options: -auto-all