addurl: Add --pathdepth option.

3 files changed, 36 insertions, 15 deletions

diff --git a/Command/AddUrl.hs b/Command/AddUrl.hs
index 981af2f7e..a6c89542e 100644
--- a/Command/AddUrl.hs
+++ b/Command/AddUrl.hs
@@ -22,34 +22,40 @@ import qualified Option
 import Types.Key
 
 def :: [Command]
-def = [withOptions [fileOption] $
+def = [withOptions [fileOption, pathdepthOption] $
 	command "addurl" (paramRepeating paramUrl) seek "add urls to annex"]
 
 fileOption :: Option
 fileOption = Option.field [] "file" paramFile "specify what file the url is added to"
 
+pathdepthOption :: Option
+pathdepthOption = Option.field [] "pathdepth" paramFile "number of path components to use in filename"
+
 seek :: [CommandSeek]
 seek = [withField fileOption return $ \f ->
-	withStrings $ start f]
+	withField pathdepthOption (return . maybe Nothing readish) $ \d ->
+	withStrings $ start f d]
 
-start :: Maybe FilePath -> String -> CommandStart
-start optfile s = notBareRepo $ go $ fromMaybe bad $ parseURI s
+start :: Maybe FilePath -> Maybe Int -> String -> CommandStart
+start optfile pathdepth s = notBareRepo $ go $ fromMaybe bad $ parseURI s
 	where
 		bad = fromMaybe (error $ "bad url " ++ s) $
 			parseURI $ escapeURIString isUnescapedInURI s
 		go url = do
-			let file = fromMaybe (url2file url) optfile
+			let file = fromMaybe (url2file url pathdepth) optfile
 			showStart "addurl" file
-			next $ perform s file
+			next $ perform s file pathdepth
 
-perform :: String -> FilePath -> CommandPerform
-perform url file = ifAnnexed file addurl geturl
+perform :: String -> FilePath -> Maybe Int -> CommandPerform
+perform url file pathdepth = ifAnnexed file addurl geturl
 	where
 		geturl = do
 			liftIO $ createDirectoryIfMissing True (parentDir file)
 			fast <- Annex.getState Annex.fast
 			if fast then nodownload url file else download url file
 		addurl (key, _backend) = do
+			when (pathdepth /= Nothing) $
+				error $ file ++ " already exists"
 			unlessM (liftIO $ Url.check url (keySize key)) $
 				error $ "failed to verify url: " ++ url
 			setUrlPresent key url
@@ -80,8 +86,17 @@ nodownload url file = do
 	setUrlPresent key url
 	next $ Command.Add.cleanup file key False
 
-url2file :: URI -> FilePath
-url2file url = take 255 $ escape $ uriRegName auth ++ uriPath url ++ uriQuery url
+url2file :: URI -> Maybe Int -> FilePath
+url2file url pathdepth = case pathdepth of
+	Nothing -> filesize $ escape fullurl
+	Just depth
+		| depth > 0 -> filesize $ join "/" $
+			fromend depth $ map escape $
+			filter (not . null) $ split "/" fullurl
+		| otherwise -> error "bad --pathdepth value"
 	where
-		escape = replace "/" "_" . replace "?" "_"
+		fullurl = uriRegName auth ++ uriPath url ++ uriQuery url
 		auth = fromMaybe (error $ "bad url " ++ show url) $ uriAuthority url
+		filesize = take 255
+		escape = replace "/" "_" . replace "?" "_"
+		fromend n = reverse . take n . reverse
diff --git a/debian/changelog b/debian/changelog
index dc792a6e7..9ca875217 100644
--- a/debian/changelog
+++ b/debian/changelog
@@ -29,6 +29,7 @@ git-annex (3.20120124) UNRELEASED; urgency=low
   * whereis: Prints the urls of files that the web special remote knows about.
   * Added a annex.queuesize setting, useful when adding hundreds of thousands
     of files on a system with plenty of memory.
+  * addurl: Add --pathdepth option.
 
  -- Joey Hess <joeyh@debian.org>  Tue, 24 Jan 2012 16:21:55 -0400
 
diff --git a/doc/git-annex.mdwn b/doc/git-annex.mdwn
index d4e62568f..d5814a2c2 100644
--- a/doc/git-annex.mdwn
+++ b/doc/git-annex.mdwn
@@ -149,10 +149,15 @@ subdirectories).
 
   To avoid immediately downloading the url, specify --fast.
 
-  To specify what file the url is added to, specify --file. This changes
-  the behavior; now all the specified urls are recorded as alternate
-  locations from which the file can be downloaded. In this mode, addurl
-  can be used both to add new files, or to add urls to existing files.
+  Normally the filename is based on the full url, so will look like
+  "www.example.com_subdir_bigfile". For a shorter filename, specify
+  --pathdepth=N. For example, --pathdepth=2 will use "subdir/bigfile",
+  while --parhdepth=1 will use "bigfile".
+
+  Or, to directly specify what file the url is added to, specify --file.
+  This changes the behavior; now all the specified urls are recorded as
+  alternate locations from which the file can be downloaded. In this mode,
+  addurl can be used both to add new files, or to add urls to existing files.
 
 # REPOSITORY SETUP COMMANDS