select schedule field from database when select_task for auto_recrawl

fix #205
binux · May 22, 2015 · a46dd6a · a46dd6a · horrower · May 22, 2015
1 parent a2c845f
commit a46dd6a
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 2 deletions.
diff --git a/pyspider/database/redis/taskdb.py b/pyspider/database/redis/taskdb.py
@@ -53,9 +53,9 @@ def _parse(self, data):
         if 'status' in data:
             data['status'] = int(data['status'])
         if 'lastcrawltime' in data:
-            data['lastcrawltime'] = float(data['lastcrawltime'])
+            data['lastcrawltime'] = float(data['lastcrawltime'] or 0)
         if 'updatetime' in data:
-            data['updatetime'] = float(data['updatetime'])
+            data['updatetime'] = float(data['updatetime'] or 0)
         return data
 
     def _stringify(self, data):

diff --git a/pyspider/scheduler/scheduler.py b/pyspider/scheduler/scheduler.py
@@ -304,6 +304,7 @@ def _check_cronjob(self):
         'project',
         'url',
         'status',
+        'schedule',
         'fetch',
         'process',
         'track',