使用Attoparsec读取Haskell中的POT文件

Question

我有一组POT文件，我想读取这些文件，以便在Haskell中对翻译进行后处理。由于POT文件非常大，我想使用attoparsec来获得良好的性能。

我尝试使用hgettext，但我的任务不是阅读翻译程序的翻译，而是解析POT文件。基本上，我想得到一个(comment, msgid, msgstr)元组的列表

Answer 1

假设要读取的POT文件存储在mypot.pot中，则可以使用可以直接在runghc中运行的代码。我已经使用一些Khan Academy PO文件成功地测试了这一点，但它目前只适用于简单的文件（没有复数等），但很容易扩展：

{-# LANGUAGE OverloadedStrings #-}
import Prelude hiding (takeWhile)
import Data.Attoparsec.ByteString
import Data.Attoparsec.ByteString.Char8 hiding (takeWhile, skipWhile)
import qualified Data.Text.Encoding as TE
import Data.Text(Text)
import Data.Word
import Data.ByteString (unsnoc, ByteString)
import qualified Data.ByteString.Char8 as B
import Control.Applicative
import Control.Monad
import Data.Maybe
import Data.Either
import qualified Data.Text as T
data PORecord = PORecord {
        poComment :: Text,
        poMsgid :: Text,
        poMsgstr :: Text
    } deriving (Show, Eq)
takeTillEOL :: Parser ByteString
takeTillEOL = takeWhile (not . isEndOfLine)
parseMsgidMsgstrLine :: ByteString -> Parser Text
parseMsgidMsgstrLine key = do
  void (string key) <?> "Line key"
  skipSpace
  char '"' <?> "Opening Qutotation mark"
  val <- takeTillEOL
  endOfLine <?> "EOL"
  return $ TE.decodeUtf8 $ fromMaybe "" $ (fst <$> unsnoc val)
msgidLine = parseMsgidMsgstrLine "msgid"
msgstrLine = parseMsgidMsgstrLine "msgstr"
escapedTextLine :: Parser Text
escapedTextLine = char '"' *> (TE.decodeUtf8 <$> takeTillEOL) <* endOfLine
--escapedTextLine = do
--    char '"'
--    val <- takeTillEOL
--    return $ (traceShow val ())
nameP :: String -> Parser a -> Parser a
nameP str p = p <?> str
commentLine :: Parser Text
commentLine = nameP "comment line" $ do
    char '#' <?> "Line start hash"
    -- Skip space but not newline
    void $ many (char ' ')
    txt <- TE.decodeUtf8 <$> takeTillEOL
    endOfLine <?> "EOF"
    return txt
emptyLine :: Parser ()
emptyLine = skipSpace <* endOfLine
poRecord :: Parser PORecord
poRecord = do
    comments <- many1 commentLine <?> "Comments"
    msgidPrimary <- msgidLine <?> "msgid"
    extraMsgid <- many escapedTextLine <?> "Extra msgid"
    msgstrPrimary <- msgstrLine <?> "msgstr"
    extraMsgstr <- many escapedTextLine <?> "Extra msgstr"
    endOfLine
    let comment = T.intercalate "n" comments
    let msgid = T.intercalate "n" $ msgidPrimary : extraMsgid
    let msgstr = T.intercalate "n" $ msgstrPrimary : extraMsgstr
    return $ PORecord comment msgid msgstr
poFile :: Parser [PORecord]
poFile =
    --let options = choice [emptyLine *> pure Nothing, ]
    many1 poRecord <?> "PO results"
main :: IO ()
main = do
    f <- B.readFile "mypot.pot"
    print $ parseOnly poFile f