From 1b10a096181f1b53dc3f572d35d822d34e31240b Mon Sep 17 00:00:00 2001
From: rolux <rolux@rolux.org>
Date: Tue, 6 Nov 2012 19:35:30 +0100
Subject: [PATCH] don't leave orphaned extras/segments/versions directories

---
 pandoraclient | 52 +++++++++++++++++++++++++++++++++++++++++++--------
 1 file changed, 44 insertions(+), 8 deletions(-)

diff --git a/pandoraclient b/pandoraclient
index 155bfa1..738f360 100755
--- a/pandoraclient
+++ b/pandoraclient
@@ -21,6 +21,7 @@ FILES = {
     'organize': 'pandoraclient.organize.jsonc',
     'sync': 'pandoraclient.sync.jsonc'
 }
+DIRECTORIES = ['Extras', 'Segments', 'Versions']
 
 
 def copy():
@@ -64,10 +65,11 @@ def organize():
         if data['path'] == volume_path:
             for file in data['files']:
                 cache['%s %s %s' % (file['path'], str(file['size']), file['time'])] = file['hash']
-    directories = []
+    directories_by_path = {}
     files = []
     files_by_hash = {}
     files_by_item = {}
+    files_by_path = {}
     issues = {
         'duplicate files': [],
         'empty directories': [],
@@ -91,9 +93,11 @@ def organize():
     for absolute_path, dirnames, filenames in os.walk(volume_path, followlinks=True):
         absolute_path = unicodedata.normalize('NFD', absolute_path)
         relative_path = absolute_path[len(volume_path):]
+        parts = relative_path.split('/')
+        length = len(parts)
         for filename in ox.sorted_strings(filenames):
             filename = unicodedata.normalize('NFD', filename)
-            if relative_path != previous_path and len(relative_path.split('/')) == 3:
+            if relative_path != previous_path and length == 3:
                 print relative_path
                 previous_path = relative_path
             full_path = os.path.join(absolute_path, filename)
@@ -103,10 +107,9 @@ def organize():
             else:
                 # unexpected files, unexpected directories
                 parts = path.split('/')
-                length = len(parts)
-                if length < 4:
+                if length < 3:
                     issues['unexpected files'].append(path)
-                if length <= 4:
+                if length <= 3:
                     file = {}
                     file['path'] = path
                     file['size'] = os.path.getsize(full_path)
@@ -117,13 +120,23 @@ def organize():
                     if not file['hash'] in files_by_hash:
                         files_by_hash[file['hash']] = []
                     files_by_hash[file['hash']].append(file)
-                elif not parts[3] in ['Extras', 'Segments', 'Versions']:
+                    if not relative_path in files_by_path:
+                        files_by_path[relative_path] = []
+                    files_by_path[relative_path].append(path)
+                elif parts[3] in DIRECTORIES:
+                    path_key = '/'.join(parts[:3])
+                    if not path_key in directories_by_path:
+                        directories_by_path[path_key] = []
+                    directory = '/'.join(parts[:4])
+                    if not directory in directories_by_path[path_key]:
+                        directories_by_path[path_key].append(directory)
+                else:
                     issues['unexpected directories'].append(path)
         # empty directories, missing files
         if not filenames or not filter(lambda x: not is_system_file(x), filenames):
             if not dirnames:
                 issues['empty directories'].append(relative_path)
-            elif len(relative_path.split('/')) == 3:
+            elif length == 3:
                 issues['missing files'].append(relative_path)
 
     print 'Scanning for duplicates'
@@ -212,6 +225,29 @@ def organize():
                 if not paths in issues[exists_key]:
                     issues[exists_key].append(paths)
 
+    # orphaned directories
+    new_path = {}
+    rename_key = {}
+    for key in ['incorrect filenames', 'non-canonical filenames']:
+        for i, paths in enumerate(issues[key]):
+            new_path[paths[0]] = '/'.join(paths[1].split('/')[:-1])
+            if not new_path[paths[0]] in rename_key:
+                rename_key[new_path[paths[0]]] = key
+    for path in [path for path in files_by_path if path in directories_by_path]:
+        new_paths = []
+        for path_file in files_by_path[path]:
+            if path_file in new_path:
+                new_paths.append(new_path[path_file])
+        if len(new_paths) == len(files_by_path[path]) and len(set(new_paths)) == 1 and new_paths[0] != path:
+            for directory in directories_by_path[path]:
+                new_directory = os.path.join(new_paths[0], directory.split('/')[-1])
+                key = rename_key[new_paths[0]]
+                exists_key = '%s (target exists)' % key
+                issues[exists_key if os.path.exists(new_directory) else key].append([directory, new_directory])
+    for key in ['incorrect filenames', 'non-canonical filenames']:
+        for key in [key, '%s (target exists)' % key]:
+            issues[key] = sorted(issues[key], key=lambda x: x[0].lower())
+
     for issue in issues:
         if issues[issue]:
             if isinstance(issues[issue][0], str):
@@ -221,8 +257,8 @@ def organize():
     keys = {
         'automatic': {
             'remove': [
-                'system files',
                 'empty directories',
+                'system files',
                 'unexpected files',
                 'unknown extensions'
             ],