annex.maxextensionlength for view

view: Support annex.maxextensionlength when generating filenames for the
view branch.

Note that refining an existing view will reuse the extension length that was
configured when initially constructing the view. This is necessarily the case
because it reuses the filenames.

Also view files used to have all extensions at the end, no matter how
many there were. Since annex.maxextensionlength's documentation includes
that it's limited to 2 extensions, I made it consistent with that.

Sponsored-by: k0ld on Patreon

Annex/View.hs

@@ -387,7 +387,7 @@ prop_view_roundtrips (AssociatedFile Nothing) _ _ = True
 prop_view_roundtrips (AssociatedFile (Just f)) metadata visible = or
 	[ B.null (P.takeFileName f) && B.null (P.takeDirectory f)
 	, viewTooLarge view
-	, all hasfields (viewedFiles view viewedFileFromReference (fromRawFilePath f) metadata)
+	, all hasfields (viewedFiles view (viewedFileFromReference' Nothing) (fromRawFilePath f) metadata)
 	]
   where
 	view = View (Git.Ref "foo") $
@@ -421,7 +421,9 @@ getViewedFileMetaData = getDirMetaData . dirFromViewedFile . takeFileName
  - branch for the view.
  -}
 applyView :: View -> Maybe Adjustment -> Annex Git.Branch
-applyView = applyView' viewedFileFromReference getWorkTreeMetaData
+applyView v ma = do
+	gc <- Annex.getGitConfig
+	applyView' (viewedFileFromReference gc) getWorkTreeMetaData v ma
 
 {- Generates a new branch for a View, which must be a more narrow
  - version of the View originally used to generate the currently
@@ -553,7 +555,8 @@ updateView view madj = do
 		Git.LsTree.LsTreeRecursive
 		(Git.LsTree.LsTreeLong True)
 		(viewParentBranch view)
-	applyView'' viewedFileFromReference getWorkTreeMetaData view madj l clean $
+	gc <- Annex.getGitConfig
+	applyView'' (viewedFileFromReference gc) getWorkTreeMetaData view madj l clean $
 		\ti -> do
 			let ref = Git.Ref.branchFileRef (viewParentBranch view)
 				(getTopFilePath (Git.LsTree.file ti))

Annex/View/ViewedFile.hs

@@ -1,6 +1,6 @@
 {- filenames (not paths) used in views
  -
- - Copyright 2014 Joey Hess <id@joeyh.name>
+ - Copyright 2014-2023 Joey Hess <id@joeyh.name>
  -
  - Licensed under the GNU AGPL version 3 or higher.
  -}
@@ -11,6 +11,7 @@ module Annex.View.ViewedFile (
 	ViewedFile,
 	MkViewedFile,
 	viewedFileFromReference,
+	viewedFileFromReference',
 	viewedFileReuse,
 	dirFromViewedFile,
 	prop_viewedFile_roundtrips,
@@ -35,17 +36,27 @@ type MkViewedFile = FilePath -> ViewedFile
  -
  - So, from dir/subdir/file.foo, generate file_%dir%subdir%.foo
  -}
-viewedFileFromReference :: MkViewedFile
-viewedFileFromReference f = concat $
-	[ escape (fromRawFilePath base)
+viewedFileFromReference :: GitConfig -> MkViewedFile
+viewedFileFromReference g = viewedFileFromReference' (annexMaxExtensionLength g)
+
+viewedFileFromReference' :: Maybe Int -> MkViewedFile
+viewedFileFromReference' maxextlen f = concat $
+	[ escape (fromRawFilePath base')
 	, if null dirs then "" else "_%" ++ intercalate "%" (map escape dirs) ++ "%"
-	, escape $ fromRawFilePath $ S.concat extensions
+	, escape $ fromRawFilePath $ S.concat extensions'
 	]
   where
 	(path, basefile) = splitFileName f
 	dirs = filter (/= ".") $ map dropTrailingPathSeparator (splitPath path)
-	(base, extensions) = splitShortExtensions (toRawFilePath basefile')
-	
+	(base, extensions) = case maxextlen of
+		Nothing -> splitShortExtensions (toRawFilePath basefile')
+		Just n -> splitShortExtensions' (n+1) (toRawFilePath basefile')
+	{- Limit to two extensions maximum. -}
+	(base', extensions')
+		| length extensions <= 2 = (base, extensions)
+		| otherwise = 
+			let (es,more) = splitAt 2 (reverse extensions)
+			in (base <> mconcat (reverse more), reverse es)
 	{- On Windows, if the filename looked like "dir/c:foo" then
 	 - basefile would look like it contains a drive letter, which will
 	 - not work. There cannot really be a filename like that, probably,
@@ -90,7 +101,7 @@ prop_viewedFile_roundtrips tf
 	-- Relative filenames wanted, not directories.
 	| any (isPathSeparator) (end f ++ beginning f) = True
 	| isAbsolute f || isDrive f = True
-	| otherwise = dir == dirFromViewedFile (viewedFileFromReference f)
+	| otherwise = dir == dirFromViewedFile (viewedFileFromReference' Nothing f)
   where
 	f = fromTestableFilePath tf
 	dir = joinPath $ beginning $ splitDirectories f

CHANGELOG

@@ -5,6 +5,8 @@ git-annex (10.20230322) UNRELEASED; urgency=medium
     drop when annex.adjustedbranchrefresh=1
   * Avoid leaving repo with a detached head when there is a failure
     checking out an updated adjusted branch.
+  * view: Support annex.maxextensionlength when generating filenames for
+    the view branch.
 
  -- Joey Hess <id@joeyh.name>  Thu, 23 Mar 2023 15:04:41 -0400
 

Utility/Path.hs

@@ -20,6 +20,7 @@ module Utility.Path (
 	runSegmentPaths',
 	dotfile,
 	splitShortExtensions,
+	splitShortExtensions',
 	relPathDirToFileAbs,
 	inSearchPath,
 	searchPath,

doc/git-annex-view.mdwn

@@ -44,6 +44,12 @@ into the `_` directory and committing will unset the metadata.
 The name of the `_` directory can be changed using the annex.viewunsetdirectory
 git config.
 
+Filenames in the view branch include their path within the original branch, to
+ensure that they are unique. The path comes after the main filename, and
+before any extensions. For example, "foo/bar.baz" will have a name
+like "bar_%foo%.baz". annex.maxextensionlength can be used to configure
+what is treated as an extension.
+
 # OPTIONS
 
 * The [[git-annex-common-options]](1) can be used.

doc/git-annex.mdwn

@@ -825,9 +825,11 @@ repository, using [[git-annex-config]]. See its man page for a list.)
 
 * `annex.maxextensionlength`
 
-  Maximum length, in bytes, of what is considered a filename extension when
-  adding a file to a backend that preserves filename extensions. The
-  default length is 4, which allows extensions like "jpeg". The dot before
+  Maximum length, in bytes, of what is considered a filename extension.
+  This is used when adding a file to a backend that preserves filename extensions,
+  and also when generating a view branch.
+
+  The default length is 4, which allows extensions like "jpeg". The dot before
   the extension is not counted part of its length. At most two extensions
   at the end of a filename will be preserved, e.g. .gz or .tar.gz .
 

doc/todo/Configuring_metadata_view_filenames/comment_8_2bcfc677da72637f34904b84fdd95c10._comment

@@ -0,0 +1,9 @@
+[[!comment format=mdwn
+ username="joey"
+ subject="""comment 8"""
+ date="2023-03-24T17:48:43Z"
+ content="""
+I've made git-annex view use `annex.maxextensionlength`. Note that refining
+an existing view will reuse the extension length that was configured when
+initially constructing the view.
+"""]]