From c8ee253ba67c03fbaeea37dfe29081ba4ced48a5 Mon Sep 17 00:00:00 2001
From: Greg Von Kuster <greg@bx.psu.edu>
Date: Sun, 27 Jan 2008 16:41:29 +0000
Subject: [PATCH] Missed committing similar fixes to cleanup_datasets script
 and admin controller.

---
 lib/galaxy/web/controllers/admin.py | 13 ++++++++++---
 scripts/cleanup_datasets.py         |  9 ++++++++-
 2 files changed, 18 insertions(+), 4 deletions(-)

diff --git a/lib/galaxy/web/controllers/admin.py b/lib/galaxy/web/controllers/admin.py
index 0e0da443f84..c0b5215a348 100644
--- a/lib/galaxy/web/controllers/admin.py
+++ b/lib/galaxy/web/controllers/admin.py
@@ -265,9 +265,16 @@ class Admin( BaseController ):
                 for row in dt.select( dt.c.purged=='t' ).execute():
                     last = time.mktime( time.strptime( row.update_time.strftime( '%a %b %d %H:%M:%S %Y' ) ) )
                     diff = (now-last)/3600/24 # days
-                    if diff > days:
-                        data = self.app.model.Dataset.get( row.id )
-                        purged_file_name = data.file_name + "_purged"
+                    if diff > days:                        
+                        data = app.model.Dataset.get( row.id )
+                        # First try filename directly under file_path
+                        purged_file_name = os.path.join( data.file_path, "dataset_%d.dat_purged" % data.id )
+                        # Only use that filename if it already exists (backward compatibility),
+                        # otherwise construct hashed path
+                        if not os.path.exists( purged_file_name ):
+                            dir = os.path.join( data.file_path, *directory_hash_id( data.id ) )
+                            # Look for file inside hashed directory
+                            purged_file_name = os.path.abspath( os.path.join( dir, "dataset_%d.dat_purged" % data.id ) )
                         if os.path.isfile( purged_file_name ):
                             dataset_count += 1
                             try:
diff --git a/scripts/cleanup_datasets.py b/scripts/cleanup_datasets.py
index 103c66e135f..25e100e14ea 100644
--- a/scripts/cleanup_datasets.py
+++ b/scripts/cleanup_datasets.py
@@ -234,7 +234,14 @@ def info_remove_datasets( app, days ):
         diff = (now-last)/3600/24 # days
         if diff > days:
             data = app.model.Dataset.get( row.id )
-            purged_file_name = data.file_name + "_purged"
+            # First try filename directly under file_path
+            purged_file_name = os.path.join( data.file_path, "dataset_%d.dat_purged" % data.id )
+            # Only use that filename if it already exists (backward compatibility),
+            # otherwise construct hashed path
+            if not os.path.exists( purged_file_name ):
+                dir = os.path.join( data.file_path, *directory_hash_id( data.id ) )
+                # Look for file inside hashed directory
+                purged_file_name = os.path.abspath( os.path.join( dir, "dataset_%d.dat_purged" % data.id ) )
             if os.path.isfile( purged_file_name ):
                 print "Dataset id %s with file_size %s will be removed from disk." %( str( row.id ), str( row.file_size ) )
                 dataset_count += 1