allow for union merges between a tree and the content in the index

This is needed for robust handling of the git-annex branch. Since changes
are staged to its index as git-annex runs, and committed at the end,
it's possible that git-annex is interrupted, and leaves a dirty index.

When it next runs, it needs to be able to merge the git-annex branch
as necessary, without losing the existing changes in the index.

Note that this assumes that the git-annex branch is only modified by
git-annex. Any changes to it will be lost when git-annex updates the
branch. I don't see a good, inexpensive way to find changes in
the git-annex branch that arn't in the index, and union merging the
git-annex branch into the index every time would likewise be expensive.

3 files changed, 70 insertions, 43 deletions

diff --git a/Branch.hs b/Branch.hs
index 82ff5fcad..9d7b1b094 100644
--- a/Branch.hs
+++ b/Branch.hs
@@ -12,12 +12,13 @@ module Branch (
 	commit
 ) where
 
-import Control.Monad (unless)
+import Control.Monad (unless, liftM)
 import Control.Monad.State (liftIO)
 import System.FilePath
 import System.Directory
 import Data.String.Utils
 import System.Cmd.Utils
+import Data.Maybe
 
 import qualified GitRepo as Git
 import qualified GitUnionMerge
@@ -72,14 +73,18 @@ update = do
 	updated <- Annex.getState Annex.updated
 	unless updated $ withIndex $ do
 		g <- Annex.gitRepo
-		refs <- liftIO $ Git.pipeRead g [Param "show-ref", Param name]
-		mapM_ updateRef $ map (last . words) (lines refs)
+		r <- liftIO $ Git.pipeRead g [Param "show-ref", Param name]
+		let refs = map (last . words) (lines r)
+		updated <- catMaybes `liftM` mapM updateRef refs
+		unless (null updated) $ liftIO $
+			GitUnionMerge.commit g "update" fullname
+				(fullname:updated)
 		Annex.changeState $ \s -> s { Annex.updated = True }
 
-{- Ensures that a given ref has been merged into the local git-annex branch. -}
-updateRef :: String -> Annex ()
+{- Ensures that a given ref has been merged into the index. -}
+updateRef :: String -> Annex (Maybe String)
 updateRef ref
-	| ref == fullname = return ()
+	| ref == fullname = return Nothing
 	| otherwise = do
 		g <- Annex.gitRepo
 		diffs <- liftIO $ Git.pipeRead g [
@@ -87,9 +92,15 @@ updateRef ref
 			Param (name++".."++ref),
 			Params "--oneline -n1"
 			]
-		unless (null diffs) $ do
-			showSideAction $ "merging " ++ ref ++ " into " ++ name ++ "..."
-			liftIO $ unionMerge g fullname ref fullname True
+		if (null diffs)
+			then return Nothing
+			else do
+				showSideAction $ "merging " ++ ref ++ " into " ++ name ++ "..."
+				-- By passing only one ref, it is actually
+				-- merged into the index, preserving any
+				-- changes that may already be staged.
+				liftIO $ GitUnionMerge.merge g [ref]
+				return $ Just ref
 
 {- Stages the content of a file into the branch's index. -}
 change :: FilePath -> String -> Annex ()
@@ -104,7 +115,7 @@ change file content = do
 commit :: String -> Annex ()
 commit message = withIndex $ do
 	g <- Annex.gitRepo
-	liftIO $ GitUnionMerge.commit g message branch []
+	liftIO $ GitUnionMerge.commit g message fullname []
 
 {- Gets the content of a file on the branch, or content staged in the index
  - if it's newer. Returns an empty string if the file didn't exist yet. -}
diff --git a/GitUnionMerge.hs b/GitUnionMerge.hs
index bc12cbe27..82f01cc0f 100644
--- a/GitUnionMerge.hs
+++ b/GitUnionMerge.hs
@@ -7,7 +7,6 @@
 
 module GitUnionMerge (
 	merge,
-	stage,
 	commit
 ) where
 
@@ -19,38 +18,54 @@ import Data.String.Utils
 import qualified GitRepo as Git
 import Utility
 
-{- Performs a union merge. Should be run with a temporary index file
- - configured by Git.useIndex.
+{- Performs a union merge between two branches, staging it in the index.
+ - Any previously staged changes in the index will be lost.
  -
- - Use indexpopulated only if the index file already contains exactly the
- - contents of aref.
+ - When only one branch is specified, it is merged into the index.
+ - In this case, previously staged changes in the index are preserved.
+ -
+ - Should be run with a temporary index file configured by Git.useIndex.
  -}
-merge :: Git.Repo -> String -> String -> String -> Bool -> IO ()
-merge g aref bref newref indexpopulated = do
-	stage g aref bref indexpopulated
-	commit g "union merge" newref [aref, bref]
+merge :: Git.Repo -> [String] -> IO ()
+merge g (x:y:[]) = do
+	a <- ls_tree g x
+	b <- merge_trees g x y
+	update_index g (a++b)
+merge g [x] = merge_tree_index g x >>= update_index g
+merge _ _ = error "wrong number of branches to merge"
 
-{- Stages the content of both refs into the index. -}
-stage :: Git.Repo -> String -> String -> Bool -> IO ()
-stage g aref bref indexpopulated = do
-	-- Get the contents of aref, as a starting point, unless
-	-- the index is already populated with it.
-	ls <- if indexpopulated
-		then return []
-		else fromgit ["ls-tree", "-z", "-r", "--full-tree", aref]
-	-- Identify files that are different between aref and bref, and
-	-- inject merged versions into git.
-	diff <- fromgit
-		["diff-tree", "--raw", "-z", "-r", "--no-renames", "-l0", aref, bref]
-	ls' <- mapM mergefile (pairs diff)
-	-- Populate the index file. Later lines override earlier ones.
-	togit ["update-index", "-z", "--index-info"]
-		(join "\0" $ ls++catMaybes ls')
+{- Feeds a list into update-index. Later items in the list can override
+ - earlier ones, so the list can be generated from any combination of
+ - ls_tree, merge_trees, and merge_tree. -}
+update_index :: Git.Repo -> [String] -> IO ()
+update_index g l =  togit ["update-index", "-z", "--index-info"] (join "\0" l)
 	where
-		fromgit l = Git.pipeNullSplit g (map Param l)
-		togit l content = Git.pipeWrite g (map Param l) content
+		togit ps content = Git.pipeWrite g (map Param ps) content
 			>>= forceSuccess
 
+{- Gets the contents of a tree in a format suitable for update_index. -}
+ls_tree :: Git.Repo -> String -> IO [String]
+ls_tree g x = Git.pipeNullSplit g $ 
+	map Param ["ls-tree", "-z", "-r", "--full-tree", x]
+
+{- For merging two trees. -}
+merge_trees :: Git.Repo -> String -> String -> IO [String]
+merge_trees g x y = calc_merge g
+	["diff-tree", "--raw", "-z", "-r", "--no-renames", "-l0", x, y]
+
+{- For merging a single tree into the index. -}
+merge_tree_index :: Git.Repo -> String -> IO [String]
+merge_tree_index g x = calc_merge g
+	["diff-index", "--raw", "-z", "-r", "--no-renames", "-l0", x]
+
+{- Calculates how to perform a merge, using git to get a raw diff,
+ - and returning a list suitable for update_index. -}
+calc_merge :: Git.Repo -> [String] -> IO [String]
+calc_merge g differ = do
+	diff <- Git.pipeNullSplit g $ map Param differ
+	l <- mapM mergefile (pairs diff)
+	return $ catMaybes l
+	where
 		pairs [] = []
 		pairs (_:[]) = error "parse error"
 		pairs (a:b:rest) = (a,b):pairs rest
@@ -62,7 +77,7 @@ stage g aref bref indexpopulated = do
 		mergefile (info, file) = do
 			let [_colonamode, _bmode, asha, bsha, _status] = words info
 			if bsha == nullsha
-				then return Nothing -- already staged from aref
+				then return Nothing -- already staged
 				else mergefile' file asha bsha
 		mergefile' file asha bsha = do
 			let shas = filter (/= nullsha) [asha, bsha]
@@ -70,10 +85,10 @@ stage g aref bref indexpopulated = do
 			sha <- Git.hashObject g $ unionmerge content
 			return $ Just $ ls_tree_line sha file
 
-{- Commits the index into the specified branch. If refs are specified,
- - commits a merge. -}
+{- Commits the index into the specified branch, 
+ - with the specified parent refs. -}
 commit :: Git.Repo -> String -> String -> [String] -> IO ()
-commit g message newref mergedrefs = do
+commit g message newref parentrefs = do
 	tree <- Git.getSha "write-tree" $ ignorehandle $
 		pipeFrom "git" ["write-tree"]
 	sha <- Git.getSha "commit-tree" $ ignorehandle $ 
@@ -81,4 +96,4 @@ commit g message newref mergedrefs = do
 	Git.run g "update-ref" [Param newref, Param sha]
 	where
 		ignorehandle a = return . snd =<< a
-		ps = concatMap (\r -> ["-p", r]) mergedrefs
+		ps = concatMap (\r -> ["-p", r]) parentrefs
diff --git a/git-union-merge.hs b/git-union-merge.hs
index f02db6be3..7c0c1cd84 100644
--- a/git-union-merge.hs
+++ b/git-union-merge.hs
@@ -44,5 +44,6 @@ main = do
 	g <- Git.configRead =<< Git.repoFromCwd
 	Git.useIndex (tmpIndex g)
 	setup g
-	GitUnionMerge.merge g aref bref newref False
+	GitUnionMerge.merge g [aref, bref]
+	GitUnionMerge.commit g "union merge" newref [aref, bref]
 	cleanup g