Encode unicode strings as utf-8

3 files changed, 19 insertions, 3 deletions

diff --git a/src/olpc/datastore/indexstore.py b/src/olpc/datastore/indexstore.py
index ec97b4a..bd689b0 100644
--- a/src/olpc/datastore/indexstore.py
+++ b/src/olpc/datastore/indexstore.py
@@ -116,7 +116,11 @@ class IndexStore(object):
             if key not in _PROPERTIES_NOT_TO_INDEX:
                 if text:
                     text += ' '
-                text += str(value)
+                if isinstance(value, unicode):
+                    value = value.encode('utf-8')
+                elif not isinstance(value, basestring):
+                    value = str(value)
+                text += value
         return text
 
     def find(self, query):
diff --git a/src/olpc/datastore/metadatastore.py b/src/olpc/datastore/metadatastore.py
index 8d1e377..5cd97ba 100644
--- a/src/olpc/datastore/metadatastore.py
+++ b/src/olpc/datastore/metadatastore.py
@@ -20,7 +20,15 @@ class MetadataStore(object):
 
         metadata['uid'] = uid
         for key, value in metadata.items():
-            open(os.path.join(metadata_path, key), 'w').write(str(value))
+            f = open(os.path.join(metadata_path, key), 'w')
+            try:
+                if isinstance(value, unicode):
+                    value = value.encode('utf-8')
+                elif not isinstance(value, basestring):
+                    value = str(value)
+                f.write(value)
+            finally:
+                f.close()
 
     def retrieve(self, uid, properties=None):
         dir_path = layoutmanager.get_instance().get_entry_path(uid)
diff --git a/src/olpc/datastore/migration.py b/src/olpc/datastore/migration.py
index b94cd93..2cc1bd8 100644
--- a/src/olpc/datastore/migration.py
+++ b/src/olpc/datastore/migration.py
@@ -65,7 +65,11 @@ def _migrate_metadata(root_path, old_root_path, uid):
     for key, value in metadata.items():
         f = open(os.path.join(metadata_path, key), 'w')
         try:
-            f.write(str(value))
+            if isinstance(value, unicode):
+                value = value.encode('utf-8')
+            if not isinstance(value, basestring):
+                value = str(value)
+            f.write(value)
         finally:
             f.close()