Docx tests: rewrite mediabag tests.

This will allow us to test the whole mediabag (making sure, for example,
that images are added with the correct keys) instead of just individual
extracted images. We compare each entry in the media bag to an image
extracted on the fly from the docx. As a result, we only need one file
to test with.

The image in the current tests was also replaced with a smaller one.

pandoc.cabal

@@ -386,7 +386,6 @@ Test-Suite test-pandoc
                   syb >= 0.1 && < 0.5,
                   pandoc,
                   pandoc-types >= 1.12.3.3 && < 1.13,
-                  base64-bytestring >= 0.1 && < 1.1,
                   bytestring >= 0.9 && < 0.11,
                   text >= 0.11 && < 1.2,
                   directory >= 1 && < 1.3,
@@ -400,7 +399,8 @@ Test-Suite test-pandoc
                   QuickCheck >= 2.4 && < 2.8,
                   HUnit >= 1.2 && < 1.3,
                   containers >= 0.1 && < 0.6,
-                  ansi-terminal >= 0.5 && < 0.7
+                  ansi-terminal >= 0.5 && < 0.7,
+                  zip-archive >= 0.2.3.2 && < 0.3
   Other-Modules:  Tests.Old
                   Tests.Helpers
                   Tests.Arbitrary

tests/Tests/Readers/Docx.hs

@@ -5,14 +5,15 @@ import Text.Pandoc.Readers.Native
 import Text.Pandoc.Definition
 import Tests.Helpers
 import Test.Framework
-import qualified Data.ByteString as BS
+import Test.HUnit (assertBool)
+import Test.Framework.Providers.HUnit
 import qualified Data.ByteString.Lazy as B
-import qualified Data.ByteString.Char8 as B8
-import qualified Data.ByteString.Base64 as B64
 import Text.Pandoc.Readers.Docx
 import Text.Pandoc.Writers.Native (writeNative)
 import qualified Data.Map as M
-import Text.Pandoc.MediaBag (lookupMedia)
+import Text.Pandoc.MediaBag (MediaBag, lookupMedia, mediaDirectory)
+import Codec.Archive.Zip
+import System.FilePath (combine)
 
 -- We define a wrapper around pandoc that doesn't normalize in the
 -- tests. Since we do our own normalization, we want to make sure
@@ -56,22 +57,44 @@ testCompareWithOpts opts name docxFile nativeFile =
 testCompare :: String -> FilePath -> FilePath -> Test
 testCompare = testCompareWithOpts def
 
-testCompareMediaIO :: String -> FilePath -> FilePath -> FilePath -> IO Test
+getMedia :: FilePath -> FilePath -> IO (Maybe B.ByteString)
-testCompareMediaIO name docxFile mediaPath mediaFile = do
+getMedia archivePath mediaPath = do
-    df <- B.readFile docxFile
+  zf <- B.readFile archivePath >>= return . toArchive
-    mf <- B.readFile mediaFile
+  return $ findEntryByPath (combine "word" mediaPath) zf >>= (Just . fromEntry)
-    let (_, mb) = readDocx def df
+
-        dBytes = case lookupMedia mediaPath mb of
+compareMediaPathIO :: FilePath -> MediaBag -> FilePath -> IO Bool
+compareMediaPathIO mediaPath mediaBag docxPath = do
+  docxMedia <- getMedia docxPath mediaPath
+  let mbBS   = case lookupMedia mediaPath mediaBag of
                  Just (_, bs) -> bs
-          Nothing -> error "Media file not found"
+                 Nothing      -> error ("couldn't find " ++
-        d64 = B8.unpack $ B64.encode $ BS.concat $ B.toChunks dBytes
+                                        mediaPath ++
-        m64 = B8.unpack $ B64.encode $ BS.concat $ B.toChunks mf
+                                        " in media bag")
-    return $ test id name (d64, m64)
+      docxBS = case docxMedia of
+                 Just bs -> bs
+                 Nothing -> error ("couldn't find " ++
+                                   mediaPath ++
+                                   " in media bag")
+  return $ mbBS == docxBS
 
-testCompareMedia :: String -> FilePath -> FilePath -> FilePath -> Test
+compareMediaBagIO :: FilePath -> IO Bool
-testCompareMedia name docxFile mediaPath mediaFile =
+compareMediaBagIO docxFile = do
-  buildTest $ testCompareMediaIO name docxFile mediaPath mediaFile
+    df <- B.readFile docxFile
+    let (_, mb) = readDocx def df
+    bools <- mapM
+             (\(fp, _, _) -> compareMediaPathIO fp mb docxFile) 
+             (mediaDirectory mb)
+    return $ and bools
 
+testMediaBagIO :: String -> FilePath -> IO Test
+testMediaBagIO name docxFile = do
+  outcome <- compareMediaBagIO docxFile
+  return $ testCase name (assertBool 
+                          ("Media didn't match media bag in file " ++ docxFile)
+                          outcome)
+
+testMediaBag :: String -> FilePath -> Test
+testMediaBag name docxFile = buildTest $ testMediaBagIO name docxFile
 
 tests :: [Test]
 tests = [ testGroup "inlines"
@@ -186,11 +209,9 @@ tests = [ testGroup "inlines"
             "docx.track_changes_deletion_all.native"
           ]
         , testGroup "media"
-          [ testCompareMedia
+          [ testMediaBag
             "image extraction"
             "docx.image.docx"
-            "media/image1.jpeg"
-            "docx.image1.jpeg"
           ]
         , testGroup "metadata"
           [ testCompareWithOpts def{readerStandalone=True}

tests/docx.image_no_embed.native

@@ -1,2 +1,2 @@
-[Header 2 ("an-image",[],[]) [Str "An",Space,Str "image"]
+[Para [Str "An",Space,Str "image:"]
-,Para [Image [] ("media/image1.jpeg","")]]
+,Para [Image [] ("media/image1.jpg","")]]