better fix for slash in view metadata

The homomorphs are back, just encoded such that it doesn't crash in LANG=C

However, I noticed a bug in the old escaping; [pseudoSlash] was escaped the
same as ['/','/']. Fixed by using '%' to escape pseudoSlash. Which requires
doubling '%' to escape it, but that's already done in the escaping of
worktree filenames in a view, so is probably ok.

3 files changed, 30 insertions, 14 deletions

diff --git a/Annex/View.hs b/Annex/View.hs
index 92a7ffc48..060301fd5 100644
--- a/Annex/View.hs
+++ b/Annex/View.hs
@@ -222,23 +222,37 @@ viewComponentMatcher viewcomponent = \metadata ->
 		| S.null s = Nothing
 		| otherwise = Just (S.toList s)
 
+-- This is '∕', a unicode character that displays the same as '/' but is
+-- not it. It is encoded using the filesystem encoding, which allows it
+-- to be used even when not in a unicode capable locale.
+pseudoSlash :: String
+pseudoSlash = "\56546\56456\56469"
+
+-- And this is '╲' similarly.
+pseudoBackslash :: String
+pseudoBackslash = "\56546\56469\56498"
+
 toViewPath :: MetaValue -> FilePath
-toViewPath = concatMap escapeslash . fromMetaValue
+toViewPath = escapeslash [] . fromMetaValue
   where
-	escapeslash c
-		| c == '/' = "%_"
-		| c == '\\' = "%."
-		| c == '%' = "%%"
-		| otherwise = [c]
+	escapeslash s ('/':cs) = escapeslash (pseudoSlash:s) cs
+	escapeslash s ('\\':cs) = escapeslash (pseudoBackslash:s) cs
+	escapeslash s ('%':cs) = escapeslash ("%%":s) cs
+	escapeslash s (c1:c2:c3:cs)
+		| [c1,c2,c3] == pseudoSlash = escapeslash ("%":pseudoSlash:s) cs
+		| [c1,c2,c3] == pseudoBackslash = escapeslash ("%":pseudoBackslash:s) cs
+		| otherwise = escapeslash ([c1]:s) (c2:c3:cs)
+	escapeslash s cs = concat (reverse (cs:s))
 
 fromViewPath :: FilePath -> MetaValue
 fromViewPath = toMetaValue . deescapeslash []
   where
-	deescapeslash s [] = reverse s
-	deescapeslash s ('%':'_':cs) = deescapeslash ('/':s) cs
-	deescapeslash s ('%':'.':cs) = deescapeslash ('\\':s) cs
-	deescapeslash s ('%':'%':cs) = deescapeslash ('%':s) cs
-	deescapeslash s (c:cs) = deescapeslash (c:s) cs
+	deescapeslash s ('%':escapedc:cs) = deescapeslash ([escapedc]:s) cs
+	deescapeslash s (c1:c2:c3:cs)
+		| [c1,c2,c3] == pseudoSlash = deescapeslash ("/":s) cs
+		| [c1,c2,c3] == pseudoBackslash = deescapeslash ("\\":s) cs
+		| otherwise = deescapeslash ([c1]:s) (c2:c3:cs)
+	deescapeslash s cs = concat (reverse (cs:s))
 
 prop_viewPath_roundtrips :: MetaValue -> Bool
 prop_viewPath_roundtrips v = fromViewPath (toViewPath v) == v
diff --git a/debian/changelog b/debian/changelog
index 8c7258ccb..49af4347f 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -48,9 +48,8 @@ git-annex (6.20151219) UNRELEASED; urgency=medium
   * rekey: No longer copies over urls from the old to the new key.
     It makes sense for migrate to do that, but not for this low-level
     (and little used) plumbing command to.
-  * view: Avoid using cute unicode homomorphs for '/' and '\' and instead
-    use ugly escaping, as the unicode method doesn't work on non-unicode
-    supporting systems.
+  * view: Fix crash in non-unicode capable locale when entering a view
+    of metadata containing a slash or backslash.
 
  -- Joey Hess <id@joeyh.name>  Sat, 19 Dec 2015 13:31:17 -0400
 
diff --git a/doc/bugs/slash_in_metadata_breaks_field__61____42___view/comment_1_249d786ce15ab0c91191987c0bab76ef._comment b/doc/bugs/slash_in_metadata_breaks_field__61____42___view/comment_1_249d786ce15ab0c91191987c0bab76ef._comment
index 2cd288951..f6fb82b51 100644
--- a/doc/bugs/slash_in_metadata_breaks_field__61____42___view/comment_1_249d786ce15ab0c91191987c0bab76ef._comment
+++ b/doc/bugs/slash_in_metadata_breaks_field__61____42___view/comment_1_249d786ce15ab0c91191987c0bab76ef._comment
@@ -21,4 +21,7 @@ unicode support.
 
 Sigh, 2016 and still can't have nice things.. Suppose it'll have to use an
 ugly encoding for them instead.
+
+Update: Can have nice things, just have to encode the characters using the
+FileSystemEncoding!
 """]]