add: Prevent (most) modifications from being made to a file while it is being added to the annex.

Anything that tries to open the file for write, or delete the file,
or replace it with something else, will not affect the add.

Only if a process has the file open for write before add starts
can it still change it while (or after) it's added to the annex.
(fsck will catch this later of course)

2 files changed, 30 insertions, 6 deletions

diff --git a/Command/Add.hs b/Command/Add.hs
index 7029a9c16..2c671eea2 100644
--- a/Command/Add.hs
+++ b/Command/Add.hs
@@ -15,7 +15,9 @@ import qualified Annex.Queue
 import Backend
 import Logs.Location
 import Annex.Content
+import Annex.Perms
 import Utility.Touch
+import Utility.FileMode
 
 def :: [Command]
 def = [command "add" paramPaths seek "add files to annex"]
@@ -44,23 +46,38 @@ start file = notBareRepo $ ifAnnexed file fixup add
 			liftIO $ removeFile file
 			next $ next $ cleanup file key =<< inAnnex key
 
+{- The file that's being added is locked down before a key is generated,
+ - to prevent it from being modified in between. It's hard linked into a
+ - temporary location, and its writable bits are removed. It could still be
+ - written to by a process that already has it open for writing. -}
 perform :: FilePath -> CommandPerform
 perform file = do
-	let source = KeySource { keyFilename = file, contentLocation = file}
+	liftIO $ preventWrite file
+	tmp <- fromRepo gitAnnexTmpDir
+	createAnnexDirectory tmp
+	pid <- liftIO getProcessID
+	let tmpfile = tmp </> "add" ++ show pid ++ "." ++ takeFileName file
+	nuke tmpfile
+	liftIO $ createLink file tmpfile
+	let source = KeySource { keyFilename = file, contentLocation = tmpfile }
 	backend <- chooseBackend file
-	genKey source backend >>= go
+	genKey source backend >>= go tmpfile
 	where
-		go Nothing = stop
-		go (Just (key, _)) = do
-			handle (undo file key) $ moveAnnex key file
+		go _ Nothing = stop
+		go tmpfile (Just (key, _)) = do
+			handle (undo file key) $ moveAnnex key tmpfile
+			nuke file
 			next $ cleanup file key True
 
+nuke :: FilePath -> Annex ()
+nuke file = liftIO $ whenM (doesFileExist file) $ removeFile file
+
 {- On error, put the file back so it doesn't seem to have vanished.
  - This can be called before or after the symlink is in place. -}
 undo :: FilePath -> Key -> IOException -> Annex a
 undo file key e = do
 	whenM (inAnnex key) $ do
-		liftIO $ whenM (doesFileExist file) $ removeFile file
+		nuke file
 		handle tryharder $ fromAnnex key file
 		logStatus key InfoMissing
 	throw e
diff --git a/debian/changelog b/debian/changelog
index fd4f7b98b..9a010327d 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -1,3 +1,10 @@
+git-annex (3.20120606) UNRELEASED; urgency=low
+
+  * add: Prevent (most) modifications from being made to a file while it
+    is being added to the annex.
+
+ -- Joey Hess <joeyh@debian.org>  Tue, 05 Jun 2012 20:25:51 -0400
+
 git-annex (3.20120605) unstable; urgency=low
 
   * sync: Show a nicer message if a user tries to sync to a special remote.