Better fix for #2187.

* Reverted kludgy change to make-windows-installer.bat.
* Removed make-reference-fiels.hs.
* Moved the individual ingredients of reference.docx and
  reference.odt to the data directory.
* Removed reference.docx and reference.odt from data directory.
* We now build the reference archives from their ingredient pieces
  in the docx and odt writers, instead of having a reference.docx
  or reference.odt intermediary.

This should fix #2187.

It also simplifies the bulid procedure.

The one thing users may notice is different is that you can
no longer get the reference.docx or reference.odt using
`--print-default-data-file`.  Instead, simply generate a
docx or odt using pandoc with a blank or minimal input,
and use that (or a customized version) with `--reference-docx`
or `--reference-odt`.

Setup.hs

@@ -38,12 +38,11 @@ main = defaultMainWithHooks $ simpleUserHooks {
     , instHook = \pkgdescr ->
          instHook simpleUserHooks pkgdescr{ executables =
             [x | x <- executables pkgdescr, exeName x `notElem` noInstall] }
-    , postBuild = \args bf pkgdescr lbi -> do
+    , postBuild = \args bf pkgdescr lbi ->
             makeManPages args bf pkgdescr lbi
-            makeReferenceFiles args bf pkgdescr lbi
     }
   where
-    noInstall = ["make-pandoc-man-pages","make-reference-files"]
+    noInstall = ["make-pandoc-man-pages"]
 
 ppBlobSuffixHandler :: PPSuffixHandler
 ppBlobSuffixHandler = ("hsb", \_ _ ->
@@ -64,13 +63,3 @@ makeManPages _ bf _ LocalBuildInfo{buildDir=buildDir}
   where
     verbosity = fromFlagOrDefault normal $ buildVerbosity bf
     progPath = buildDir </> "make-pandoc-man-pages" </> "make-pandoc-man-pages"
-
-makeReferenceFiles :: Args -> BuildFlags -> PackageDescription -> LocalBuildInfo -> IO ()
-makeReferenceFiles _ bf _ LocalBuildInfo{buildDir=buildDir}
-  = mapM_
-      (rawSystemExit verbosity progPath . return)
-      referenceFormats
-  where
-    verbosity = fromFlagOrDefault normal $ buildVerbosity bf
-    progPath = buildDir </> "make-reference-files" </> "make-reference-files"
-    referenceFormats = ["docx", "odt"]

pandoc.cabal

@@ -69,28 +69,6 @@ Data-Files:
                  data/templates/default.org
                  data/templates/default.epub
                  data/templates/default.epub3
-                 -- data for ODT writer
-                 data/reference.odt
-                 -- data for docx writer
-                 data/reference.docx
-                 -- stylesheet for EPUB writer
-                 data/epub.css
-                 -- data for LaTeXMathML writer
-                 data/LaTeXMathML.js
-                 data/MathMLinHTML.js
-                 -- data for dzslides writer
-                 data/dzslides/template.html
-                 -- sample lua custom writer
-                 data/sample.lua
-                 -- documentation
-                 README, COPYRIGHT
-Extra-Source-Files:
-                 -- documentation
-                 INSTALL, BUGS, CONTRIBUTING.md, changelog
-                 -- code to create pandoc.1 man page
-                 Makefile
-                 man/man1/pandoc.1.template
-                 man/man5/pandoc_markdown.5.template
                  -- source files for reference.docx
                  data/docx/[Content_Types].xml
                  data/docx/_rels/.rels
@@ -116,6 +94,24 @@ Extra-Source-Files:
                  data/odt/Configurations2/accelerator/current.xml
                  data/odt/Thumbnails/thumbnail.png
                  data/odt/META-INF/manifest.xml
+                  -- stylesheet for EPUB writer
+                 data/epub.css
+                 -- data for LaTeXMathML writer
+                 data/LaTeXMathML.js
+                 data/MathMLinHTML.js
+                 -- data for dzslides writer
+                 data/dzslides/template.html
+                 -- sample lua custom writer
+                 data/sample.lua
+                 -- documentation
+                 README, COPYRIGHT
+Extra-Source-Files:
+                 -- documentation
+                 INSTALL, BUGS, CONTRIBUTING.md, changelog
+                 -- code to create pandoc.1 man page
+                 Makefile
+                 man/man1/pandoc.1.template
+                 man/man5/pandoc_markdown.5.template
                 -- trypandoc
                  trypandoc/Makefile
                  trypandoc/index.html
@@ -439,16 +435,6 @@ Executable make-pandoc-man-pages
   Default-Language: Haskell98
   Buildable:   True
 
-Executable make-reference-files
-  Main-Is:       make-reference-files.hs
-  Hs-Source-Dirs: data
-  Build-Depends: zip-archive >= 0.2.3.4 && < 0.3,
-                 base >= 4.2 && < 5,
-                 filepath >= 1.1 && < 1.5,
-                 directory >= 1 && < 1.3,
-                 bytestring >= 0.9 && < 0.11
-  Default-Language: Haskell2010
-
 Test-Suite test-pandoc
   Type:           exitcode-stdio-1.0
   Main-Is:        test-pandoc.hs

src/Text/Pandoc/Writers/Docx.hs

@@ -211,11 +211,10 @@ writeDocx opts doc@(Pandoc meta _) = do
   let doc' = stripInvalidChars . walk fixDisplayMath $ doc
   username <- lookup "USERNAME" <$> getEnvironment
   utctime <- getCurrentTime
-  refArchive <- liftM (toArchive . toLazy) $
+  distArchive <- getDefaultReferenceDocx datadir
-       case writerReferenceDocx opts of
+  refArchive <- case writerReferenceDocx opts of
-             Just f  -> B.readFile f
+                     Just f  -> liftM (toArchive . toLazy) $ B.readFile f
-             Nothing -> readDataFile datadir "reference.docx"
+                     Nothing -> return distArchive
-  distArchive <- liftM (toArchive . toLazy) $ readDataFile datadir "reference.docx"
 
   parsedDoc <- parseXml refArchive distArchive "word/document.xml"
   let wname f qn = qPrefix qn == Just "w" && f (qName qn)
@@ -474,7 +473,7 @@ writeDocx opts doc@(Pandoc meta _) = do
   settingsEntry <- copyChildren refArchive distArchive settingsPath epochtime settingsList
 
   let entryFromArchive arch path =
-         maybe (fail $ path ++ " corrupt or missing in reference docx")
+         maybe (fail $ path ++ " missing in reference docx")
                return
                (findEntryByPath path arch `mplus` findEntryByPath path distArchive)
   docPropsAppEntry <- entryFromArchive refArchive "docProps/app.xml"
@@ -1204,11 +1203,12 @@ defaultFootnotes = [ mknode "w:footnote"
 
 parseXml :: Archive -> Archive -> String -> IO Element
 parseXml refArchive distArchive relpath =
-  case ((findEntryByPath relpath refArchive `mplus`
+  case findEntryByPath relpath refArchive `mplus`
-         findEntryByPath relpath distArchive)
+         findEntryByPath relpath distArchive of
-         >>= parseXMLDoc . UTF8.toStringLazy . fromEntry) of
+            Nothing -> fail $ relpath ++ " missing in reference docx"
+            Just e  -> case parseXMLDoc . UTF8.toStringLazy . fromEntry $ e of
+                       Nothing -> fail $ relpath ++ " corrupt in reference docx"
                        Just d  -> return d
-            Nothing -> fail $ relpath ++ " corrupt or missing in reference docx"
 
 -- | Scales the image to fit the page
 -- sizes are passed in emu
@@ -1219,3 +1219,28 @@ fitToPage (x, y) pageWidth
     (pageWidth, round $
       ((fromIntegral pageWidth) / ((fromIntegral :: Integer -> Double) x)) * (fromIntegral y))
   | otherwise = (x, y)
+
+getDefaultReferenceDocx :: Maybe FilePath -> IO Archive
+getDefaultReferenceDocx datadir = do
+  let paths = ["[Content_Types].xml",
+               "_rels/.rels",
+               "docProps/app.xml",
+               "docProps/core.xml",
+               "word/document.xml",
+               "word/fontTable.xml",
+               "word/footnotes.xml",
+               "word/numbering.xml",
+               "word/settings.xml",
+               "word/webSettings.xml",
+               "word/styles.xml",
+               "word/_rels/document.xml.rels",
+               "word/_rels/footnotes.xml.rels",
+               "word/theme/theme1.xml"]
+  let pathToEntry path = do epochtime <- (floor . utcTimeToPOSIXSeconds) <$>
+                                          getCurrentTime
+                            contents <- toLazy <$> readDataFile datadir
+                                                       ("docx/" ++ path)
+                            return $ toEntry path epochtime contents
+  entries <- mapM pathToEntry paths
+  let archive = foldr addEntryToArchive emptyArchive entries
+  return archive

src/Text/Pandoc/Writers/ODT.hs

@@ -60,11 +60,10 @@ writeODT :: WriterOptions  -- ^ Writer options
 writeODT opts doc@(Pandoc meta _) = do
   let datadir = writerUserDataDir opts
   let title = docTitle meta
-  refArchive <- liftM toArchive $
+  refArchive <-
        case writerReferenceODT opts of
-             Just f -> B.readFile f
+             Just f -> liftM toArchive $ B.readFile f
-             Nothing -> (B.fromChunks . (:[])) `fmap`
+             Nothing -> getDefaultReferenceODT datadir
-                           readDataFile datadir "reference.odt"
   -- handle formulas and pictures
   picEntriesRef <- newIORef ([] :: [Entry])
   doc' <- walkM (transformPicMath opts picEntriesRef) $ walk fixDisplayMath doc
@@ -178,3 +177,21 @@ transformPicMath _ entriesRef (Math t math) = do
                                         , ("xlink:actuate", "onLoad")]
 
 transformPicMath _ _ x = return x
+
+getDefaultReferenceODT :: Maybe FilePath -> IO Archive
+getDefaultReferenceODT datadir = do
+  let paths = ["mimetype",
+               "manifest.rdf",
+               "styles.xml",
+               "content.xml",
+               "meta.xml",
+               "settings.xml",
+               "Configurations2/accelerator/current.xml",
+               "Thumbnails/thumbnail.png",
+               "META-INF/manifest.xml"]
+  let pathToEntry path = do epochtime <- floor `fmap` getPOSIXTime
+                            contents <- (B.fromChunks . (:[])) `fmap`
+                                          readDataFile datadir ("odt/" ++ path)
+                            return $ toEntry path epochtime contents
+  entries <- mapM pathToEntry paths
+  return $ foldr addEntryToArchive emptyArchive entries

windows/make-windows-installer.bat

@@ -5,10 +5,6 @@ cabal sandbox init
 cabal clean
 cabal install hsb2hs
 if %errorlevel% neq 0 exit /b %errorlevel%
-REM We do it once to regenrate data/reference.docx/odt
-cabal install -v1 --force --reinstall
-cabal clean
-REM then again for real, because otherwise it won't work with embed_data_files:
 cabal install -v1 --force --reinstall --flags="embed_data_files" . pandoc-citeproc
 if %errorlevel% neq 0 exit /b %errorlevel%
 strip .\.cabal-sandbox\bin\pandoc.exe