From 0f9a03822ef5beeccc0db0e0d092a1e7847c0221 Mon Sep 17 00:00:00 2001
From: Marko Andjelic <markoandjelic31@gmail.com>
Date: Fri, 23 Jan 2026 23:33:00 +0000
Subject: [PATCH] Add parsing and associated modules

---
 app/EpubParser.hs | 34 +++++++++++++++++++++++++++++++++-
 svitak.cabal      |  5 +++--
 2 files changed, 36 insertions(+), 3 deletions(-)

diff --git a/app/EpubParser.hs b/app/EpubParser.hs
index a83bfb9..0d71572 100644
--- a/app/EpubParser.hs
+++ b/app/EpubParser.hs
@@ -3,6 +3,14 @@ module EpubParser where
 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, fromAttrib, (~==), Tag, innerText, sections)
+import System.FilePath (takeDirectory)
+
+data BookInfo = BookInfo 
+  { title  :: String
+  , author :: String
+  } deriving (Show)
+
 
 openEpub :: FilePath -> IO (Either String Archive)
 openEpub path = toArchive <$> BL.readFile path >>= \archive -> maybe (pure $ Left "Error reading file") (\entry -> if fromEntry entry == C8.pack "application/epub+zip" then pure (Right archive) else pure (Left "Wrong mimetype")) (findEntryByPath "mimetype" archive)
@@ -10,5 +18,29 @@ openEpub path = toArchive <$> BL.readFile path >>= \archive -> maybe (pure $ Lef
 getOpfPath :: Archive -> Maybe FilePath
 getOpfPath archive = do
   entry <- findEntryByPath "META-INF/container.xml" archive
+  let content = C8.unpack (fromEntry entry)
 
-  return "OEBPS/content.opf"
+  let tags = parseTags content
+
+  let rootfileTags = filter (~== "<rootfile>") tags
+
+  -- If we found at least one, get the "full-path" attribute
+  case rootfileTags of
+    (t:_) -> Just (fromAttrib "full-path" t)
+    []    -> Nothing
+
+getRootPrefix :: FilePath -> FilePath
+getRootPrefix path = 
+    let dir = takeDirectory path
+    in if dir == "." 
+       then "" 
+       else dir ++ "/"
+
+tagify :: String -> String
+tagify item = "<" ++ item ++ ">"
+
+getTagText :: String -> [Tag String] -> String
+getTagText tagName tags =
+  case sections (~== tagify tagName) tags of
+    (x:_) -> innerText x
+    []    -> ""
diff --git a/svitak.cabal b/svitak.cabal
index 3aa8b40..2fc1242 100644
--- a/svitak.cabal
+++ b/svitak.cabal
@@ -62,8 +62,9 @@ executable svitak
     build-depends:    base ^>=4.21.0.0,
                       bytestring >= 0.12.2.0,
                       epub-metadata >=5.4,
-                      zip-archive >= 0.4.3.2,
-                      tagsoup
+                      filepath >= 1.5.4.0,
+                      tagsoup,
+                      zip-archive >= 0.4.3.2
 
     hs-source-dirs:   app
     default-language: Haskell2010