Merge branch 'master' into watch

Annex/Branch.hs

@@ -33,6 +33,7 @@ import qualified Git.Command
 import qualified Git.Ref
 import qualified Git.Branch
 import qualified Git.UnionMerge
+import qualified Git.UpdateIndex
 import Git.HashObject
 import qualified Git.Index
 import Annex.CatFile
@@ -258,8 +259,8 @@ files = withIndexUpdate $ do
  - in changes from other branches.
  -}
 genIndex :: Git.Repo -> IO ()
-genIndex g = Git.UnionMerge.stream_update_index g
-	[Git.UnionMerge.ls_tree fullname g]
+genIndex g = Git.UpdateIndex.stream_update_index g
+	[Git.UpdateIndex.ls_tree fullname g]
 
 {- Merges the specified refs into the index.
  - Any changes staged in the index will be preserved. -}
@@ -335,13 +336,13 @@ stageJournal = do
 	g <- gitRepo
 	withIndex $ liftIO $ do
 		h <- hashObjectStart g
-		Git.UnionMerge.stream_update_index g
+		Git.UpdateIndex.stream_update_index g
 			[genstream (gitAnnexJournalDir g) h fs]
 		hashObjectStop h
 	where
 		genstream dir h fs streamer = forM_ fs $ \file -> do
 			let path = dir </> file
 			sha <- hashFile h path
-			_ <- streamer $ Git.UnionMerge.update_index_line
+			_ <- streamer $ Git.UpdateIndex.update_index_line
 				sha (fileJournal file)
 			removeFile path

Git/UnionMerge.hs

@@ -7,11 +7,7 @@
 
 module Git.UnionMerge (
 	merge,
-	merge_index,
-	update_index,
-	stream_update_index,
-	update_index_line,
-	ls_tree
+	merge_index
 ) where
 
 import System.Cmd.Utils
@@ -24,8 +20,7 @@ import Git
 import Git.Sha
 import Git.CatFile
 import Git.Command
-
-type Streamer = (String -> IO ()) -> IO ()
+import Git.UpdateIndex
 
 {- Performs a union merge between two branches, staging it in the index.
  - Any previously staged changes in the index will be lost.
@@ -47,38 +42,6 @@ merge_index :: CatFileHandle -> Repo -> [Ref] -> IO ()
 merge_index h repo bs =
 	stream_update_index repo $ map (\b -> merge_tree_index b h repo) bs
 
-{- Feeds content 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_index. -}
-update_index :: Repo -> [String] -> IO ()
-update_index repo ls = stream_update_index repo [(`mapM_` ls)]
-
-{- Streams content into update-index. -}
-stream_update_index :: Repo -> [Streamer] -> IO ()
-stream_update_index repo as = do
-	(p, h) <- hPipeTo "git" (toCommand $ gitCommandLine params repo)
-	fileEncoding h
-	forM_ as (stream h)
-	hClose h
-	forceSuccess p
-	where
-		params = map Param ["update-index", "-z", "--index-info"]
-		stream h a = a (streamer h)
-		streamer h s = do
-			hPutStr h s
-			hPutStr h "\0"
-
-{- Generates a line suitable to be fed into update-index, to add
- - a given file with a given sha. -}
-update_index_line :: Sha -> FilePath -> String
-update_index_line sha file = "100644 blob " ++ show sha ++ "\t" ++ file
-
-{- Gets the current tree for a ref. -}
-ls_tree :: Ref -> Repo -> Streamer
-ls_tree (Ref x) repo streamer = mapM_ streamer =<< pipeNullSplit params repo
-	where
-		params = map Param ["ls-tree", "-z", "-r", "--full-tree", x]
-
 {- For merging two trees. -}
 merge_trees :: Ref -> Ref -> CatFileHandle -> Repo -> Streamer
 merge_trees (Ref x) (Ref y) h = calc_merge h $ "diff-tree":diff_opts ++ [x, y]

Git/UpdateIndex.hs

@@ -0,0 +1,49 @@
+{- git-update-index library
+ -
+ - Copyright 2011, 2012 Joey Hess <joey@kitenet.net>
+ -
+ - Licensed under the GNU GPL version 3 or higher.
+ -}
+
+module Git.UpdateIndex (
+	Streamer,
+	stream_update_index,
+	update_index_line,
+	ls_tree
+) where
+
+import System.Cmd.Utils
+
+import Common
+import Git
+import Git.Command
+
+{- Streamers are passed a callback and should feed it lines in the form
+ - read by update-index, and generated by ls-tree. -}
+type Streamer = (String -> IO ()) -> IO ()
+
+{- Streams content into update-index from a list of Streamers. -}
+stream_update_index :: Repo -> [Streamer] -> IO ()
+stream_update_index repo as = do
+	(p, h) <- hPipeTo "git" (toCommand $ gitCommandLine params repo)
+	fileEncoding h
+	forM_ as (stream h)
+	hClose h
+	forceSuccess p
+	where
+		params = map Param ["update-index", "-z", "--index-info"]
+		stream h a = a (streamer h)
+		streamer h s = do
+			hPutStr h s
+			hPutStr h "\0"
+
+{- Generates a line suitable to be fed into update-index, to add
+ - a given file with a given sha. -}
+update_index_line :: Sha -> FilePath -> String
+update_index_line sha file = "100644 blob " ++ show sha ++ "\t" ++ file
+
+{- Gets the current tree for a ref. -}
+ls_tree :: Ref -> Repo -> Streamer
+ls_tree (Ref x) repo streamer = mapM_ streamer =<< pipeNullSplit params repo
+	where
+		params = map Param ["ls-tree", "-z", "-r", "--full-tree", x]

doc/design/assistant/blog/day_2__races.mdwn

@@ -0,0 +1,45 @@
+Last night I got `git annex watch` to also handle deletion of files.
+This was not as tricky as feared; the key is using `git rm --ignore-unmatch`,
+which avoids most problimatic situations (such as a just deleted file
+being added back before git is run).
+
+Also fixed some races when `git annex watch` is doing its startup scan of
+the tree, which might be changed as it's being traversed. Now only one
+thread performs actions at a time, so inotify events are queued up during
+the scan, and dealt with once it completes. It's worth noting that inotify
+can only buffer so many events .. Which might have been a problem except
+for a very nice feature of Haskell's inotify interface: It has a thread
+that drains the limited inotify buffer and does its own buffering.
+
+----
+
+Right now, `git annex watch` is not as fast as it could be when doing
+something like adding a lot of files, or deleting a lot of files.
+For each file, it currently runs a git command that updates the index.
+I did some work toward coalescing these into one command (which `git annex`
+already does normally). It's not quite ready to be turned on yet,
+because of some races involving `git add` that become much worse
+if it's delayed by event coalescing.
+
+----
+
+And races were the theme of today. Spent most of the day really
+getting to grips with all the fun races that can occur between
+modification happening to files, and `git annex watch`. The [[inotify]]
+page now has a long list of known races, some benign, and several,
+all involving adding files, that are quite nasty.
+
+I fixed one of those races this evening. The rest will probably involve
+moving away from using `git add`, which necessarily examines the file
+on disk, to directly shoving the symlink into git's index.
+
+BTW, it turns out that `dvcs-autosync` has grappled with some of these same
+races: <http://comments.gmane.org/gmane.comp.version-control.home-dir/665>
+I hope that `git annex watch` will be in a better place to deal with them,
+since it's only dealing with git, and with a restricted portion of it
+relevant to git-annex.
+
+It's important that `git annex watch` be rock solid. It's the foundation
+of the git annex assistant. Users should not need to worry about races
+when using it. Most users won't know what race conditions are. If only I
+could be so lucky!

doc/design/assistant/inotify.mdwn

@@ -58,12 +58,19 @@ Many races need to be dealt with by this code. Here are some of them.
 * File is added and then replaced with another file before the annex add
   moves its content into the annex.
 
-  **Currently unfixed**; The new content will be moved to the annex under the
-  old checksum, and fsck will later catch this inconsistency.
+  Fixed this problem; Now it hard links the file to a temp directory and
+  operates on the hard link, which is also made unwritable.
 
-  Possible fix: Move content someplace before doing checksumming. Perhaps
-  using a hard link and removing the write bit to prevent modification 
-  while checksumming.
+* A process has a file open for write, another one closes it, and so it's
+  added. Then the first process modifies it.
+
+  **Currently unfixed**; This changes content in the annex, and fsck will
+  later catch the inconsistency.
+
+  Possible fixes: Somehow track or detect if a file is open for write
+  by any processes. Or, when possible, making a copy on write copy
+  before adding the file would avoid this.  Or, as a last resort, make
+  an expensive copy of the file and add that.
 
 * File is added and then replaced with another file before the annex add
   makes its symlink.