Add spine to file path resolving

app/EpubParser.hs

@@ -4,11 +4,12 @@ import Codec.Archive.Zip (toArchive, findEntryByPath, fromEntry, Archive)
 import qualified Data.ByteString.Lazy as BL
 import qualified Data.ByteString.Lazy.Char8 as C8
 import Text.HTML.TagSoup (parseTags, Tag)
-import Text.HTML.Scalpel (scrape, text, tagSelector, attr, attrs)
-import System.FilePath (takeDirectory)
+import Text.HTML.Scalpel (scrape, text, tagSelector, attr, attrs, Scraper, chroots, anySelector)
+import System.FilePath (takeDirectory, (</>))
 import Control.Monad.Reader
 import qualified Data.Text as T
 import Data.Maybe (fromMaybe)
+import qualified Data.Map as M
 
 type EpubAction a = ReaderT EpubEnv IO a
 
@@ -17,6 +18,7 @@ data EpubEnv = EpubEnv
   , author     :: T.Text
   , rootPrefix :: FilePath
   , spine      :: [T.Text]
+  , manifestMap :: M.Map T.Text T.Text
   } deriving (Show)
 
 myEnv :: Archive -> [Tag String] -> EpubEnv
@@ -27,10 +29,11 @@ myEnv archive tags =
 
   in
     EpubEnv 
-      { title      = getTagText "dc:title" tags
-      , author     = getTagText "dc:creator" tags
-      , rootPrefix = prefix
-      , spine      = getSpine tags
+      { title       = getTagText "dc:title" tags
+      , author      = getTagText "dc:creator" tags
+      , rootPrefix  = prefix
+      , spine       = getSpine tags
+      , manifestMap = getManifest tags
       }
 
 openEpub :: FilePath -> IO (Either String Archive)
@@ -63,3 +66,22 @@ getTagText tagName tags =
 getSpine :: [Tag String] -> [T.Text]
 getSpine tags =
   map T.pack . fromMaybe [] $ scrape (attrs "idref" (tagSelector "itemref")) tags
+
+-- This scraper will be run by chroot for every item
+itemScraper :: Scraper String (T.Text, T.Text)
+itemScraper =
+  liftA2 (,) (attr "id" anySelector) (attr "href" anySelector) >>= \(id, href) -> pure (T.pack id, T.pack href)
+
+getManifest :: [Tag String] -> M.Map T.Text T.Text
+getManifest tags =
+  M.fromList . fromMaybe [] $ scrape (chroots (tagSelector "item") itemScraper) tags
+
+resolvePath :: T.Text -> EpubAction (Maybe FilePath)
+resolvePath spineId = do
+  m <- asks manifestMap
+  p <- asks rootPrefix
+
+  let mHref = M.lookup spineId m
+
+
+  pure $ fmap (\href -> p </> T.unpack href) mHref

svitak.cabal

@@ -61,13 +61,14 @@ executable svitak
 
     build-depends:    base ^>=4.21.0.0,
                       bytestring >= 0.12.2.0,
+                      containers >= 0.7,
                       epub-metadata >=5.4,
                       filepath >= 1.5.4.0,
                       mtl >= 2.3.2,
+                      scalpel >= 0.6.2.2,
                       tagsoup,
                       text >= 2.1.2,
-                      zip-archive >= 0.4.3.2,
-                      scalpel >= 0.6.2.2
+                      zip-archive >= 0.4.3.2
 
     hs-source-dirs:   app
     default-language: Haskell2010