Added a helper for generating a list of all time popular datasets
[ckanext-ga-report.git] / ckanext / ga_report / helpers.py
blob:a/ckanext/ga_report/helpers.py -> blob:b/ckanext/ga_report/helpers.py
--- a/ckanext/ga_report/helpers.py
+++ b/ckanext/ga_report/helpers.py
@@ -1,40 +1,115 @@
 import logging
 import operator
+
 import ckan.lib.base as base
 import ckan.model as model
+from ckan.logic import get_action
 
+from ckanext.ga_report.ga_model import GA_Url, GA_Publisher
+from ckanext.ga_report.controller import _get_publishers
 _log = logging.getLogger(__name__)
 
+def popular_datasets(count=10):
+    import random
+
+    publisher = None
+    publishers = _get_publishers(30)
+    total = len(publishers)
+    while not publisher or not datasets:
+        rand = random.randrange(0, total)
+        publisher = publishers[rand][0]
+        if not publisher.state == 'active':
+            publisher = None
+            continue
+        datasets = _datasets_for_publisher(publisher, 10)[:count]
+
+    ctx = {
+        'datasets': datasets,
+        'publisher': publisher
+    }
+    return base.render_snippet('ga_report/ga_popular_datasets.html', **ctx)
+
+def single_popular_dataset(top=20):
+    '''Returns a random dataset from the most popular ones.
+
+    :param top: the number of top datasets to select from
+    '''
+    import random
+
+    top_datasets = model.Session.query(GA_Url).\
+                   filter(GA_Url.url.like('/dataset/%')).\
+                   order_by('ga_url.pageviews::int desc')
+    num_top_datasets = top_datasets.count()
+
+    dataset = None
+    if num_top_datasets:
+        count = 0
+        while not dataset:
+            rand = random.randrange(0, min(top, num_top_datasets))
+            ga_url = top_datasets[rand]
+            dataset = model.Package.get(ga_url.url[len('/dataset/'):])
+            if dataset and not dataset.state == 'active':
+                dataset = None
+                count += 1
+                if count > 10:
+                    break
+    if not dataset:
+        # fallback
+        dataset = model.Session.query(model.Package)\
+                  .filter_by(state='active').first()
+        if not dataset:
+            return None
+    dataset_dict = get_action('package_show')({'model': model,
+                                               'session': model.Session},
+                                              {'id':dataset.id})
+    return dataset_dict
+
+def single_popular_dataset_html(top=20):
+    dataset_dict = single_popular_dataset(top)
+    groups = package.get('groups', [])
+    publishers = [ g for g in groups if g.get('type') == 'publisher' ]
+    publisher = publishers[0] if publishers else {'name':'', 'title': ''}
+    context = {
+        'dataset': dataset_dict,
+        'publisher': publisher_dict
+        }
+    return base.render_snippet('ga_report/ga_popular_single.html', **context)
+
+
 def most_popular_datasets(publisher, count=20):
-    from ckanext.ga_report.ga_model import GA_Url
 
     if not publisher:
         _log.error("No valid publisher passed to 'most_popular_datasets'")
         return ""
 
+    results = _datasets_for_publisher(publisher, count)
+
+    ctx = {
+        'dataset_count': len(results),
+        'datasets': results,
+
+        'publisher': publisher
+    }
+
+    return base.render_snippet('ga_report/publisher/popular.html', **ctx)
+
+def _datasets_for_publisher(publisher, count):
     datasets = {}
     entries = model.Session.query(GA_Url).\
         filter(GA_Url.department_id==publisher.name).\
         filter(GA_Url.url.like('/dataset/%')).\
-        order_by('ga_url.pageviews::int desc')[:count]
+        order_by('ga_url.pageviews::int desc').all()
     for entry in entries:
-        p = model.Package.get(entry.url[len('/dataset/'):])
-        if not p in datasets:
-            datasets[p] = {'views':0, 'visits': 0}
-        datasets[p]['views'] = datasets[p]['views'] + int(entry.pageviews)
-        datasets[p]['visits'] = datasets[p]['visits'] + int(entry.visitors)
+        if len(datasets) < count:
+            p = model.Package.get(entry.url[len('/dataset/'):])
+            if not p in datasets:
+                datasets[p] = {'views':0, 'visits': 0}
+            datasets[p]['views'] = datasets[p]['views'] + int(entry.pageviews)
+            datasets[p]['visits'] = datasets[p]['visits'] + int(entry.visits)
 
     results = []
     for k, v in datasets.iteritems():
         results.append((k,v['views'],v['visits']))
 
-    results = sorted(results, key=operator.itemgetter(1), reverse=True)
+    return sorted(results, key=operator.itemgetter(1), reverse=True)
 
-    ctx = {
-        'dataset_count': len(datasets),
-        'datasets': results,
-        'publisher': publisher
-    }
-
-    return base.render_snippet('ga_report/publisher/popular.html', **ctx)
-