From: Joey Hess <joeyh@joeyh.name>
Date: Mon, 22 Feb 2021 17:35:00 +0000 (-0400)
Subject: fix unannex data overwrite bug
X-Git-Tag: archive/raspbian/10.20250416-2+rpi1~1^2~98^2~2
X-Git-Url: https://dgit.raspbian.org/?a=commitdiff_plain;h=530e96b80ebd9671ab63fdd20d0d25faf556d4d7;p=git-annex.git

fix unannex data overwrite bug

unannex, uninit: When an annexed file is modified, don't overwrite the
modified version with an older version from the annex

This commit was sponsored by Mark Reidenbach on Patreon.
---

diff --git a/CHANGELOG b/CHANGELOG
index cd27e8a44d..fc99e6e64e 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -28,7 +28,9 @@ git-annex (8.20210128) UNRELEASED; urgency=medium
     rather than doing it in a second pass.
   * Bugfix: fsck --from a ssh remote did not actually check that the
     content on the remote is not corrupted.
-  * unannex, uninit: Avoid running git rm once per annexed file,
+  * unannex, uninit: When an annexed file is modified, don't overwrite
+    the modified version with an older version from the annex.
+  * unannex, uninit: Don't run git rm once per annexed file, 
     for a large speedup.
 
  -- Joey Hess <id@joeyh.name>  Thu, 28 Jan 2021 12:34:32 -0400
diff --git a/Command/Unannex.hs b/Command/Unannex.hs
index 690a946be6..94a89f1502 100644
--- a/Command/Unannex.hs
+++ b/Command/Unannex.hs
@@ -10,6 +10,7 @@ module Command.Unannex where
 import Command
 import qualified Annex
 import Annex.Perms
+import Annex.Link
 import qualified Annex.Queue
 import Utility.CopyFile
 import qualified Database.Keys
@@ -41,7 +42,6 @@ start si file key =
 
 perform :: RawFilePath -> Key -> CommandPerform
 perform file key = do
-	liftIO $ removeFile (fromRawFilePath file)
 	Annex.Queue.addCommand [] "rm"
 		[ Param "--cached"
 		, Param "--force"
@@ -49,19 +49,34 @@ perform file key = do
 		, Param "--"
 		]
 		[fromRawFilePath file]
-	next $ cleanup file key
+	isAnnexLink file >>= \case
+		-- If the file is locked, it needs to be replaced with
+		-- the content from the annex. Note that it's possible
+		-- for key' (read from the symlink) to differ from key
+		-- (cached in git).
+		Just key' -> do
+			removeassociated
+			next $ cleanup file key'
+		-- If the file is unlocked, it can be unmodified or not and
+		-- does not need to be replaced either way.
+		Nothing -> do
+			removeassociated
+			next $ return True
+  where
+	removeassociated = 
+		Database.Keys.removeAssociatedFile key
+			=<< inRepo (toTopFilePath file)
 
 cleanup :: RawFilePath -> Key -> CommandCleanup
 cleanup file key = do
-	Database.Keys.removeAssociatedFile key =<< inRepo (toTopFilePath file)
+	liftIO $ removeFile (fromRawFilePath file)
 	src <- calcRepo (gitAnnexLocation key)
 	ifM (Annex.getState Annex.fast)
 		( do
 			-- Only make a hard link if the annexed file does not
-			-- already have other hard links pointing at it.
-			-- This avoids unannexing (and uninit) ending up
-			-- hard linking files together, which would be
-			-- surprising.
+			-- already have other hard links pointing at it. This
+			-- avoids unannexing (and uninit) ending up hard
+			-- linking files together, which would be surprising.
 			s <- liftIO $ R.getFileStatus src
 			if linkCount s > 1
 				then copyfrom src
diff --git a/doc/bugs/unannex_of_modified_file_loses_modification.mdwn b/doc/bugs/unannex_of_modified_file_loses_modification.mdwn
index 28a0f1d182..54a6d81190 100644
--- a/doc/bugs/unannex_of_modified_file_loses_modification.mdwn
+++ b/doc/bugs/unannex_of_modified_file_loses_modification.mdwn
@@ -4,3 +4,5 @@ This is a data loss bug.
 
 Command.Unannex.cleanup just overwrites whatever's there without checking.
 Happens with both locked and unlocked files. --[[Joey]]
+
+> [[fixed|done]] --[[Joey]]