implement the extract method + finalize the moulinette loop

1 files changed, 40 insertions, 12 deletions

diff --git a/datastore_reader/_datastore_reader.py b/datastore_reader/_datastore_reader.py
index 35cde1e..9417581 100644
--- a/datastore_reader/_datastore_reader.py
+++ b/datastore_reader/_datastore_reader.py
@@ -1,11 +1,21 @@
 # -*- coding: utf-8 -*-
 # python import
-import json, os
+import json, os, tarfile
 from datetime import datetime
 
 # ds reader import
 from datastore_reader.utils import _csv, config as c
 
+# common relative store path
+STORE_PATH = os.path.join(
+                'home',
+                'olpc',
+                '.sugar',
+                'default',
+                'datastore',
+                'store'
+                )
+
 
 def parse_meta_file(meta_path):
     """Returns an info dict from matadata content parsed.
@@ -41,18 +51,28 @@ def list_meta_files(datastore_path):
             yield os.path.join(datastore_path, file_name)
 
 
-def extract_backup(backup_path):
+def extract_backup(backup_serial, backup_path):
     """Extracts backup archive in the tmp folder and the the corresponding
     datastore path.
 
     :param backup: a path, ex.: '/somewhere/here'
     :return backup: a path, ex.: '/somewhere/here/tmp/home/.../store'
 
-    >>> extract_backup('data/demo.bz2')
-    './data/tmp/demo/home/olpc/.sugar/default/datastore/store'
+    >>> extract_backup('demo', 'data/demo.tar.bz2')
+    'data/tmp/demo/home/olpc/.sugar/default/datastore/store'
 
     """
-    pass
+    tmp_path = os.path.join(c.main.working_dir, 'tmp', backup_serial)
+    with tarfile.open(backup_path) as t:
+        t.extractall(tmp_path)
+    # datastore should be as follow
+    datastore_path = os.path.join(tmp_path, STORE_PATH)
+    # check path or raise an error
+    if os.path.exists(datastore_path):
+        return datastore_path
+    # Oups
+    else:
+        raise Exception('bad data store path for serial: %s!' % backup_serial)
 
 
 def list_backups(working_dir):
@@ -79,21 +99,25 @@ def read_backups():
     # working dir check
     if not os.path.exists(c.main.working_dir):
         raise Exception('no working_dir found!')
+    # out dir
+    csv_dir = os.path.join(c.main.working_dir, 'out')
+    if not os.path.exists(csv_dir):
+        os.mkdir(csv_dir)
     # csv path
     csv_name = datetime.now().strftime('%Y-%m-%d_%H:%M_moulinette_result.csv')
-    csv_path = os.path.join(c.main.working_dir, 'out', csv_name)
-    if not os.path.exists(csv_path):
-        os.mkdir(csv_path)
+    csv_path = os.path.join(csv_dir, csv_name)
+    # columns shortcut
+    columns = c.moulinette.columns.as_list()
     # init the csv writer
     with open(csv_path, 'wb') as f:
         # init csv writer
-        writer = csv.Writer(f, delimiter=';')
+        writer = _csv.Writer(f, delimiter=';')
         # header line
-        writer.writerow(c.moulinette.columns)
+        writer.writerow(columns)
         # list backups
         for backup_serial, backup_path in list_backups(c.main.working_dir):
             # extract backup
-            datastore_path = extract_backup(backup_path)
+            datastore_path = extract_backup(backup_serial, backup_path)
             # list metadata
             for meta_path in list_meta_files(datastore_path):
                 # parse the metadata file
@@ -101,7 +125,11 @@ def read_backups():
                 # add serial do dict
                 meta_dict['serial'] = backup_serial
                 # write meta info in a new csv row according config
-                writer.writerow([meta_dict[k] for k in c.moulinette.columns])
+                row = list()
+                for k in columns:
+                    row.append(meta_dict[k] if k in meta_dict else '')
+                # write row
+                writer.writerow(row)
 
 
 if __name__ == "__main__":