如何从Pandoc中获取Pandoc？

Question

I started to write few programs using the pandoc api ,taking examples from , https://johnmacfarlane.net/BayHac2014/#/the-pandoc-types , But I am unable to isolate Pandoc from m Pandoc.

module ImageAttr where
import Text.Pandoc.Builder
import Text.Pandoc.Generic
import Text.Pandoc.Definition
import Text.Pandoc.JSON
import qualified Data.Char as DataChar
import qualified Data.List as DataList
import Data.Either
import Text.Parsec as Parsec
import Text.Pandoc
import qualified Data.Text as T
import qualified Data.Text.IO as TIO

fromRight :: b -> Either a b -> b
fromRight _ (Right b) = b
fromRight b _         = b


fromLeft :: b -> Either b a -> b
fromLeft  _ (Left b)  = b
fromLeft  b _        = b


main :: IO ()
main = do
        result <- runIO $ do
            doc <- readMarkdown def (T.pack "[testing](url)")
            writeRST def doc
        x<-((handleError result))
        print x

output is String but I need pandoc ast previously it was Either PandocError Pandoc type , instead I am getting "m Pandoc". How to get Pandoc out of "m Pandoc".

any Thought? By the way , I am new and still learning haskell.

I am getting error message as :

ImageSizeModifer.hs:28:12:
    Couldn't match expected type ‘PandocIO a0’ with actual type ‘IO ()’
    In a stmt of a 'do' block: print ""
    In the second argument of ‘($)’, namely
      ‘do { doc <- readMarkdown def (T.pack "[testing](url)");
            print "" }’
Failed, modules loaded: none.
Prelude T>

My problem is I need Pandoc , but I have not much idea how to to get it ? There are new PandocIO and PandocPure , I believe these are new.

Answer 1

The new version of pandoc has readMarkdown :: PandocMonad m => ReaderOptions -> Text -> m Pandoc . This follows the usual mtl -style of specifying the constraints on the monad, instead of the actual monad.

That means that to use readMarkdown , you need to pick a particular m satisfying PandocMonad . You can find what instances PandocMonad has by scrolling to the "Instances" section of its doc . In this case, you find:

• PandocMonad PandocPure maybe interesting
• PandocMonad PandocIO maybe interesting
• (MonadTrans t, PandocMonad m, Functor (tm), MonadError PandocError (tm), Monad (tm), Applicative (tm)) => PandocMonad (tm) - not interesting, requires another PandocMonad m
• PandocMonad m => PandocMonad (ParsecT s st m) - not interesting, requires another PandocMonad m

Now, scrolling past where PandocPure and PandocIO are defined, you can find ways of running extracting information out of those:

• runIO :: PandocIO a -> IO (Either PandocError a)
• runPure :: PandocPure a -> Either PandocError a

Combining either of these functions with readMarkdown , you can get a concrete Either type.

runIO (readMarkdown def (T.pack "[testing](url)")) :: IO (Either PandocError Pandoc)
runPure (readMarkdown def (T.pack "[testing](url)")) :: Either PandocError Pandoc

Note that I know nothing about Pandoc or its API - all I did was walk through the above process.

Answer 2

请访问http://pandoc.org/using-the-pandoc-api.html ，以获取有关使用pandoc 2.x API的基本教程。