Force len in avg calc to be a float too
[ckanext-ga-report.git] / ckanext / ga_report / controller.py
blob:a/ckanext/ga_report/controller.py -> blob:b/ckanext/ga_report/controller.py
--- a/ckanext/ga_report/controller.py
+++ b/ckanext/ga_report/controller.py
@@ -21,7 +21,7 @@
     return '%s %s' % (calendar.month_name[d.tm_mon], d.tm_year)
 
 
-def _month_details(cls):
+def _month_details(cls, stat_key=None):
     '''
     Returns a list of all the periods for which we have data, unfortunately
     knows too much about the type of the cls being passed as GA_Url has a
@@ -32,9 +32,12 @@
     months = []
     day = None
 
-    vals = model.Session.query(cls.period_name,cls.period_complete_day)\
-        .filter(cls.period_name!='All').distinct(cls.period_name)\
-        .order_by("period_name desc").all()
+    q = model.Session.query(cls.period_name,cls.period_complete_day)\
+        .filter(cls.period_name!='All').distinct(cls.period_name)
+    if stat_key:
+        q=  q.filter(cls.stat_name==stat_key)
+
+    vals = q.order_by("period_name desc").all()
     if vals and vals[0][1]:
         day = int(vals[0][1])
         ordinal = 'th' if 11 <= day <= 13 \
@@ -52,7 +55,7 @@
     def csv(self, month):
         import csv
 
-        q = model.Session.query(GA_Stat)
+        q = model.Session.query(GA_Stat).filter(GA_Stat.stat_name!='Downloads')
         if month != 'all':
             q = q.filter(GA_Stat.period_name==month)
         entries = q.order_by('GA_Stat.period_name, GA_Stat.stat_name, GA_Stat.key').all()
@@ -68,6 +71,26 @@
                              entry.stat_name.encode('utf-8'),
                              entry.key.encode('utf-8'),
                              entry.value.encode('utf-8')])
+
+    def csv_downloads(self, month):
+        import csv
+
+        q = model.Session.query(GA_Stat).filter(GA_Stat.stat_name=='Downloads')
+        if month != 'all':
+            q = q.filter(GA_Stat.period_name==month)
+        entries = q.order_by('GA_Stat.period_name, GA_Stat.key').all()
+
+        response.headers['Content-Type'] = "text/csv; charset=utf-8"
+        response.headers['Content-Disposition'] = str('attachment; filename=downloads_%s.csv' % (month,))
+
+        writer = csv.writer(response)
+        writer.writerow(["Period", "Resource URL", "Count"])
+
+        for entry in entries:
+            writer.writerow([entry.period_name.encode('utf-8'),
+                             entry.key.encode('utf-8'),
+                             entry.value.encode('utf-8')])
+
 
     def index(self):
 
@@ -179,6 +202,35 @@
 
         return render('ga_report/site/index.html')
 
+    def downloads(self):
+
+        # Get the month details by fetching distinct values and determining the
+        # month names from the values.
+        c.months, c.day = _month_details(GA_Stat, "Downloads")
+
+        # Work out which month to show, based on query params of the first item
+        c.month_desc = 'all months'
+        c.month = request.params.get('month', '')
+        if c.month:
+            c.month_desc = ''.join([m[1] for m in c.months if m[0]==c.month])
+
+        c.downloads = []
+        q = model.Session.query(GA_Stat).filter(GA_Stat.stat_name=='Downloads')
+        q = q.filter(GA_Stat.period_name==c.month) if c.month else q
+        q = q.order_by("ga_stat.value::int desc")
+
+        data = collections.defaultdict(int)
+        for entry in q.all():
+            r = model.Session.query(model.Resource).filter(model.Resource.url==entry.key).first()
+            if not r:
+                continue
+            data[r] += int(entry.value)
+
+        c.downloads = [(k,v,) for k,v in data.iteritems()]
+        c.downloads = sorted(c.downloads, key=operator.itemgetter(1), reverse=True)
+
+        return render('ga_report/site/downloads.html')
+
 
 class GaDatasetReport(BaseController):
     """