X-Git-Url: http://gitweb.michael.orlitzky.com/?p=dead%2Flwn-epub.git;a=blobdiff_plain;f=src%2FMain.hs;h=655f1bef40c0680bda2c98e2d4ae9bc6f06c2cd8;hp=373d423c9f322ba885db966fecff5b7e190c4dd4;hb=6789361c1ee570df06878535aff73d99b87528a5;hpb=6f0e6cbece7e1b1a3c6b43d19eb2f29088af981c

diff --git a/src/Main.hs b/src/Main.hs
index 373d423..655f1be 100644
--- a/src/Main.hs
+++ b/src/Main.hs
@@ -2,24 +2,44 @@
 module Main
 where
 
+import Prelude hiding (readFile)
+import System.Directory(doesFileExist)
 import System.IO (
   Handle,
   IOMode (WriteMode),
   openBinaryFile,
   stdout
   )
+import System.IO.UTF8 (readFile)
 import Text.XML.HXT.Core
 
-import CommandLine (Args(..), apply_args)
+import CommandLine (Args(..), apply_args, show_help)
 import LWN.Page
 
-my_read :: String -> IOStateArrow s b XmlTree
-my_read =
-  readDocument [ withValidate  no,
+
+my_read_opts :: SysConfigList
+my_read_opts = [ withValidate  no,
                  withParseHTML yes,
-                 withInputEncoding utf8,
                  withWarnings no ]
 
+-- | My version of HandsomeSoup's parseHTML.
+my_read :: String -> IOStateArrow s b XmlTree
+my_read = readString my_read_opts
+
+-- | Try to parse the given article using HXT. We try a few different
+--   methods; if none of them work, we return 'Nothing'.
+get_xml_from_article :: String -> IO (Maybe (IOStateArrow s b XmlTree))
+get_xml_from_article s = do
+  article <- real_article_path s
+  is_file <- doesFileExist article
+  case is_file of
+    True -> do
+      contents <- readFile article
+      return $ Just $ my_read contents
+    False -> do
+      -- Download the URL and try to parse it.
+      return Nothing
+
 -- | If we're given an empty path, return a handle to
 --   'stdout'. Otherwise, open the given file and return a read/write
 --   handle to that.
@@ -43,12 +63,18 @@ main :: IO ()
 main = do
   Args{..} <- apply_args
   output_handle <- get_output_handle output
-  input_path <- real_article_path article
-  let html = my_read input_path
-  result <- parse html
+  maybe_html <- get_xml_from_article article
+
+  case maybe_html of
+
+    Just html -> do
+      result <- parse html
+      case result of
+        Just stuff  -> epublish stuff output_handle
+        Nothing     -> return ()
 
-  case result of
-    Just stuff  -> epublish stuff output_handle
-    Nothing     -> return ()
+    Nothing -> do
+      _ <- show_help
+      return ()
 
   putStrLn "Done."