dropunused

9 files changed, 84 insertions, 46 deletions

diff --git a/Backend.hs b/Backend.hs
index 14af56bbf..2f0f71d74 100644
--- a/Backend.hs
+++ b/Backend.hs
@@ -25,7 +25,8 @@ module Backend (
 	hasKey,
 	fsckKey,
 	lookupFile,
-	chooseBackends
+	chooseBackends,
+	keyBackend
 ) where
 
 import Control.Monad.State
@@ -111,8 +112,8 @@ removeKey backend key = (Internals.removeKey backend) key
 {- Checks if a key is present in its backend. -}
 hasKey :: Key -> Annex Bool
 hasKey key = do
-	bs <- Annex.supportedBackends
-	(Internals.hasKey (lookupBackendName bs $ backendName key)) key
+	backend <- keyBackend key
+	(Internals.hasKey backend) key
 
 {- Checks a key's backend for problems. -}
 fsckKey :: Backend -> Key -> Annex Bool
@@ -154,3 +155,9 @@ chooseBackends fs = do
 	bs <- Annex.supportedBackends
 	pairs <- liftIO $ Git.checkAttr g "git-annex-backend" fs
 	return $ map (\(f,b) -> (f, maybeLookupBackendName bs b)) pairs
+
+{- Returns the backend to use for a key. -}
+keyBackend :: Key -> Annex Backend
+keyBackend key = do
+	bs <- Annex.supportedBackends
+	return $ lookupBackendName bs $ backendName key
diff --git a/CmdLine.hs b/CmdLine.hs
index cc163fff5..35e889d7d 100644
--- a/CmdLine.hs
+++ b/CmdLine.hs
@@ -26,6 +26,7 @@ import qualified Command.Fix
 import qualified Command.Init
 import qualified Command.Fsck
 import qualified Command.Unused
+import qualified Command.DropUnused
 import qualified Command.Unlock
 import qualified Command.Lock
 import qualified Command.PreCommit
@@ -65,6 +66,8 @@ subCmds =
 		"check for problems"
 	, SubCommand "unused" nothing	Command.Unused.seek
 		"look for unused file content"
+	, SubCommand "dropunused" number Command.DropUnused.seek
+		"drop unused file content"
 	, SubCommand "find" maybepath	Command.Find.seek
 		"lists available files"
 	]
@@ -73,6 +76,7 @@ subCmds =
 		maybepath = "[PATH ...]"
 		key = "KEY ..."
 		desc = "DESCRIPTION"
+		number = "NUMBER ..."
 		nothing = ""
 
 -- Each dashed command-line option results in generation of an action
diff --git a/Command.hs b/Command.hs
index 09c935b3b..c6d2d0d5d 100644
--- a/Command.hs
+++ b/Command.hs
@@ -129,6 +129,8 @@ backendPairs a files = do
 	return $ map a pairs
 withString :: SubCmdSeekStrings
 withString a params = return [a $ unwords params]
+withStrings :: SubCmdSeekStrings
+withStrings a params = return $ map a params
 withFilesToBeCommitted :: SubCmdSeekStrings
 withFilesToBeCommitted a params = do
 	repo <- Annex.gitRepo
diff --git a/Command/Fsck.hs b/Command/Fsck.hs
index 02b66d01a..a72d753fa 100644
--- a/Command/Fsck.hs
+++ b/Command/Fsck.hs
@@ -7,12 +7,7 @@
 
 module Command.Fsck where
 
-import qualified Data.Map as M
-
 import Command
-import Types
-import Core
-import Messages
 import qualified Command.FsckFile
 import qualified Command.Unused
 
diff --git a/Command/Unused.hs b/Command/Unused.hs
index ed3de5d57..7a34d393c 100644
--- a/Command/Unused.hs
+++ b/Command/Unused.hs
@@ -7,12 +7,15 @@
 
 module Command.Unused where
 
+import Control.Monad.State (liftIO)
 import qualified Data.Map as M
 
 import Command
 import Types
 import Core
 import Messages
+import Locations
+import qualified Annex
 
 seek :: [SubCmdSeek]
 seek = [withNothing start]
@@ -37,13 +40,24 @@ checkUnused = do
 	if (null unused)
 		then return True
 		else do
-			showLongNote $ w unused
+			let list = number 1 unused
+			g <- Annex.gitRepo
+			liftIO $ writeFile (annexUnusedLog g) $ unlines $ 
+				map (\(n, k) -> show n ++ " " ++ show k) list
+			showLongNote $ w list
 			return False
 	where
-		w u = unlines $ [
-			"Some annexed data is no longer pointed to by any files in the repository.",
-			"If this data is no longer needed, it can be removed using git-annex dropkey:"
-			] ++ map (\k -> "  " ++ show k) u
+		w u = unlines $
+			["Some annexed data is no longer pointed to by any files in the repository:",
+			 "  NUMBER  KEY"]
+			++ (map (\(n, k) -> "  " ++ (pad 6 $ show n) ++ "  " ++ show k) u) ++
+			["(To see where data was previously used, try: git log --stat -S'KEY')",
+			 "(To remove unwanted data: git-annex dropunused NUMBER)"]
+		pad n s = s ++ replicate (n - length s) ' '
+
+number :: Integer -> [a] -> [(Integer, a)]
+number _ [] = []
+number n (x:xs) = (n, x):(number (n+1) xs)
 
 {- Finds keys whose content is present, but that do not seem to be used
  - by any files in the git repo. -}
diff --git a/Locations.hs b/Locations.hs
index c3bab285d..24ccc75c6 100644
--- a/Locations.hs
+++ b/Locations.hs
@@ -14,6 +14,7 @@ module Locations (
 	annexLocationRelative,
 	annexTmpLocation,
 	annexBadLocation,
+	annexUnusedLog,
 	annexDir,
 	annexObjectDir,
 
@@ -56,14 +57,18 @@ annexDir r = Git.workTree r ++ "/.git/annex"
 annexObjectDir :: Git.Repo -> FilePath
 annexObjectDir r = annexDir r ++ "/objects"
 
-{- .git-annex/tmp is used for temp files -}
+{- .git-annex/tmp/ is used for temp files -}
 annexTmpLocation :: Git.Repo -> FilePath
 annexTmpLocation r = annexDir r ++ "/tmp/"
 
-{- .git-annex/bad is used for bad files found during fsck -}
+{- .git-annex/bad/ is used for bad files found during fsck -}
 annexBadLocation :: Git.Repo -> FilePath
 annexBadLocation r = annexDir r ++ "/bad/"
 
+{- .git/annex/unused is used to number possibly unused keys -}
+annexUnusedLog :: Git.Repo -> FilePath
+annexUnusedLog r = annexDir r ++ "/unused"
+
 {- Converts a key into a filename fragment.
  -
  - Escape "/" in the key name, to keep a flat tree of files and avoid
diff --git a/debian/changelog b/debian/changelog
index dcdbe15e2..cffa7bb26 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -2,6 +2,8 @@ git-annex (0.07) UNRELEASED; urgency=low
 
   * find: New subcommand.
   * unused: New subcommand, finds unused data (the global part of fsck).
+  * dropunused: New subcommand, provides for easy dropping of unused keys
+    by number, as listed by unused subcommand.
 
  -- Joey Hess <joeyh@debian.org>  Sun, 14 Nov 2010 12:34:49 -0400
 
diff --git a/doc/git-annex.mdwn b/doc/git-annex.mdwn
index a522534da..618ddf203 100644
--- a/doc/git-annex.mdwn
+++ b/doc/git-annex.mdwn
@@ -112,6 +112,32 @@ Many git-annex subcommands will stage changes for later `git commit` by you.
   Use this to undo an unlock command if you don't want to modify
   the files, or have made modifications you want to discard.
 
+* fsck [path ...]
+
+  With no parameters, this subcommand checks the whole annex for consistency,
+  and warns about any problems found.
+
+  With parameters, only the specified files are checked.
+
+* unused
+
+  Checks the annex for data that is not used by any files currently
+  in the annex, and prints a numbered list of the data. 
+
+  (This is run as part of `git annex fsck`.)
+
+* dropunused [number ...]
+
+  Drops the data corresponding to the numbers, as listed by the last
+  `git annex unused` or `git annex fsck`
+
+* find [path ...]
+
+  Outputs a list of annexed files whose content is currently present.
+
+  With no parameters, defaults to finding all files in the current directory
+  and its subdirectories.
+
 * unannex [path ...]
 
   Use this to undo an accidental add command. This is not the command you
@@ -159,25 +185,6 @@ Many git-annex subcommands will stage changes for later `git commit` by you.
 
 	git annex setkey --key=1287765018:3 /tmp/file
 
-* fsck [path ...]
-
-  With no parameters, this subcommand checks the whole annex for consistency,
-  and warns about any problems found.
-
-  With parameters, only the specified files are checked.
-
-* unused
-
-  Checks the annex for data that is not used by any files currently
-  in the annex, and prints a report.
-
-* find [path ...]
-
-  Outputs a list of annexed files whose content is currently present.
-
-  With no parameters, defaults to finding all files in the current directory
-  and its subdirectories.
-
 # OPTIONS
 
 * --force
@@ -194,14 +201,6 @@ Many git-annex subcommands will stage changes for later `git commit` by you.
 
   Enable verbose logging.
 
-* --backend=name
-
-  Specifies the key-value backend to use when adding a file.
-
-* --key=name
-
-  Specifies a key to operate on, for use with the addkey subcommand.
-
 * --from=repository
 
   Specifies a repository that content will be retrieved from.
@@ -212,6 +211,14 @@ Many git-annex subcommands will stage changes for later `git commit` by you.
   Specifies a git repository that content will be sent to. 
   It should be specified using the name of a configured git remote.
 
+* --backend=name
+
+  Specifies which key-value backend to use.
+
+* --key=name
+
+  Specifies a key to operate on.
+
 # CONFIGURATION
 
 Like other git commands, git-annex is configured via `git-config`.
diff --git a/doc/walkthrough.mdwn b/doc/walkthrough.mdwn
index 9a6ee220c..1e07053ae 100644
--- a/doc/walkthrough.mdwn
+++ b/doc/walkthrough.mdwn
@@ -287,7 +287,7 @@ setting is satisfied for all files, and it warns about any dangling values
 in `.git/annex/objects/`. 
 
 	# git annex fsck
-	fsck  (checking for unused data...) ok
+	unused  (checking for unused data...) ok
 	fsck my_cool_big_file (checksum...) ok
 	......
 
@@ -304,10 +304,12 @@ Fsck never deletes possibly bad data; instead it will be moved to
 might say about a badly messed up annex:
 
 	# git annex fsck
-	fsck  (checking for unused data...) 
+	unused  (checking for unused data...) 
 	  Some annexed data is no longer pointed to by any files in the repository.
-	  If this data is no longer needed, it can be removed using git-annex dropkey:
-	    WORM:1289672605:3:file
+	    NUMBER  KEY
+	    1       WORM:1289672605:3:file
+	  (To see where data was previously used, try: git log --stat -S'KEY')
+	  (To remove unwanted data: git-annex dropunused NUMBER)
 	failed
 	fsck my_cool_big_file (checksum...)
 	  Bad file content; moved to .git/annex/bad/SHA1:7da006579dd64330eb2456001fd01948430572f2