Fixes to show All Time data
--- a/ckanext/ga_report/controller.py
+++ b/ckanext/ga_report/controller.py
@@ -1,6 +1,6 @@
import logging
import operator
-from ckan.lib.base import BaseController, c, render, request, response
+from ckan.lib.base import BaseController, c, render, request, response, abort
import sqlalchemy
from sqlalchemy import func, cast, Integer
@@ -34,7 +34,6 @@
filter(GA_Stat.period_name==month).\
order_by('GA_Stat.stat_name, GA_Stat.key').all()
- #response.headers['Content-disposition'] = 'attachment; filename=dgu_analytics_%s.csv' % (month,)
response.headers['Content-Type'] = "text/csv; charset=utf-8"
writer = csv.writer(response)
@@ -60,7 +59,19 @@
filter(GA_Stat.stat_name=='Totals').\
filter(GA_Stat.period_name==c.month).\
order_by('ga_stat.key').all()
- c.global_totals = [(s.key, s.value) for s in entries ]
+
+ c.global_totals = []
+ for e in entries:
+ val = e.value
+ if e.key in ['Average time on site', 'Pages per visit', 'Percent new visits']:
+ val = "%.2f" % round(float(e.value), 2)
+ if e.key == 'Average time on site':
+ mins, secs = divmod(float(val), 60)
+ hours, mins = divmod(mins, 60)
+ val = '%02d:%02d:%02d (%s seconds) ' % (hours, mins, secs, val)
+ e.key = '%s *' % e.key
+ c.global_totals.append((e.key, val))
+
keys = {
'Browser versions': 'browsers',
@@ -94,20 +105,31 @@
c.months = _month_details(GA_Url)
# Work out which month to show, based on query params of the first item
- c.month = request.params.get('month', c.months[0][0] if c.months else '')
- c.month_desc = ''.join([m[1] for m in c.months if m[0]==c.month])
+ c.month = request.params.get('month', '')
+ c.month_desc = 'all time'
+ if c.month:
+ c.month_desc = ''.join([m[1] for m in c.months if m[0]==c.month])
connection = model.Session.connection()
q = """
select department_id, sum(pageviews::int) views, sum(visitors::int) visits
from ga_url
- where department_id <> ''
- and not url like '/publisher/%%'
- and period_name=%s
- group by department_id order by views desc limit 20;
- """
+ where department_id <> ''"""
+ if c.month:
+ q = q + """
+ and period_name=%s
+ """
+ q = q + """
+ group by department_id order by views desc limit 20;
+ """
+
+ # Add this back (before and period_name =%s) if you want to ignore publisher
+ # homepage views
+ # and not url like '/publisher/%%'
+
c.top_publishers = []
res = connection.execute(q, c.month)
+
for row in res:
c.top_publishers.append((model.Group.get(row[0]), row[1], row[2]))
@@ -115,7 +137,11 @@
def read(self, id):
+ count = 20
+
c.publisher = model.Group.get(id)
+ if not c.publisher:
+ abort(404, 'A publisher with that name could not be found')
c.top_packages = [] # package, dataset_views in c.top_packages
# Get the month details by fetching distinct values and determining the
@@ -123,22 +149,50 @@
c.months = _month_details(GA_Url)
# Work out which month to show, based on query params of the first item
- c.month = request.params.get('month', c.months[0][0] if c.months else '')
- c.month_desc = ''.join([m[1] for m in c.months if m[0]==c.month])
+ c.month = request.params.get('month', '')
+ if not c.month:
+ c.month_desc = 'all time'
+ else:
+ c.month_desc = ''.join([m[1] for m in c.months if m[0]==c.month])
- entry = model.Session.query(GA_Url).\
- filter(GA_Url.url=='/publisher/%s' % c.publisher.name).\
- filter(GA_Url.period_name==c.month).first()
- c.publisher_page_views = entry.pageviews if entry else 0
+ c.publisher_page_views = 0
+ q = model.Session.query(GA_Url).\
+ filter(GA_Url.url=='/publisher/%s' % c.publisher.name)
+ if c.month:
+ entry = q.filter(GA_Url.period_name==c.month).first()
+ c.publisher_page_views = entry.pageviews if entry else 0
+ else:
+ for e in q.all():
+ c.publisher_page_views = c.publisher_page_views + int(e.pageviews)
- entries = model.Session.query(GA_Url).\
+
+ q = model.Session.query(GA_Url).\
filter(GA_Url.department_id==c.publisher.name).\
- filter(GA_Url.period_name==c.month).\
- order_by('ga_url.pageviews::int desc')[:20]
- for entry in entries:
- if entry.url.startswith('/dataset/'):
+ filter(GA_Url.url.like('/dataset/%'))
+ if c.month:
+ q = q.filter(GA_Url.period_name==c.month)
+ q = q.order_by('ga_url.pageviews::int desc')
+
+ if c.month:
+ for entry in q[:count]:
p = model.Package.get(entry.url[len('/dataset/'):])
c.top_packages.append((p,entry.pageviews,entry.visitors))
+ else:
+ ds = {}
+ for entry in q.all():
+ if len(ds) >= count:
+ break
+ p = model.Package.get(entry.url[len('/dataset/'):])
+ if not p in ds:
+ ds[p] = {'views':0, 'visits': 0}
+ ds[p]['views'] = ds[p]['views'] + int(entry.pageviews)
+ ds[p]['visits'] = ds[p]['visits'] + int(entry.visitors)
+
+ results = []
+ for k, v in ds.iteritems():
+ results.append((k,v['views'],v['visits']))
+
+ c.top_packages = sorted(results, key=operator.itemgetter(1), reverse=True)
return render('ga_report/publisher/read.html')
--- /dev/null
+++ b/ckanext/ga_report/helpers.py
@@ -1,1 +1,42 @@
+import logging
+import operator
+import ckan.lib.base as base
+import ckan.model as model
+_log = logging.getLogger(__name__)
+
+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 ""
+
+ 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').all()
+ for entry in entries:
+ 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.visitors)
+
+ results = []
+ for k, v in datasets.iteritems():
+ results.append((k,v['views'],v['visits']))
+
+ results = 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)
+
--- a/ckanext/ga_report/plugin.py
+++ b/ckanext/ga_report/plugin.py
@@ -2,23 +2,32 @@
import ckan.lib.helpers as h
import ckan.plugins as p
from ckan.plugins import implements, toolkit
-#import gasnippet
-#import commands
-#import dbutil
log = logging.getLogger('ckanext.ga-report')
class GAReportPlugin(p.SingletonPlugin):
implements(p.IConfigurer, inherit=True)
implements(p.IRoutes, inherit=True)
+ implements(p.ITemplateHelpers, inherit=True)
def update_config(self, config):
toolkit.add_template_directory(config, 'templates')
toolkit.add_public_directory(config, 'public')
+ def get_helpers(self):
+ """
+ A dictionary of extra helpers that will be available to provide
+ ga report info to templates.
+ """
+ from ckanext.ga_report.helpers import most_popular_datasets
+ return {
+ 'ga_report_installed': lambda: True,
+ 'most_popular_datasets': most_popular_datasets,
+ }
+
def after_map(self, map):
map.connect(
- '/data/analytics/publisher/',
+ '/data/analytics/publisher',
controller='ckanext.ga_report.controller:GaPublisherReport',
action='index'
)
--- a/ckanext/ga_report/templates/ga_report/publisher/index.html
+++ b/ckanext/ga_report/templates/ga_report/publisher/index.html
@@ -8,16 +8,19 @@
<py:match path="primarysidebar">
<li class="widget-container boxed widget_text">
<h4>Publishers</h4>
+ <p>The table shows the top 20 publishers as recorded by page views of datasets owned by that publisher, and the number of visits to each publisher's home page.</p>
</li>
</py:match>
<div py:match="content">
<h1>Publisher Analytics</h1>
- <h2>The top 20 publishers</h2>
+ <h2>The top 20 publishers of ${c.month_desc}</h2>
<form class="form-inline" action="${h.url_for(controller='ckanext.ga_report.controller:GaPublisherReport',action='index')}" method="get">
<div class="controls">
<select name="month">
+ <option value='' py:attrs="{'selected': 'selected' if not c.month else None}">All time</option>
+
<py:for each="val,desc in c.months">
<option value='${val}' py:attrs="{'selected': 'selected' if c.month == val else None}">${desc}</option>
</py:for>
--- /dev/null
+++ b/ckanext/ga_report/templates/ga_report/publisher/popular.html
@@ -1,1 +1,25 @@
+<html xmlns:py="http://genshi.edgewall.org/"
+ xmlns:i18n="http://genshi.edgewall.org/i18n"
+ xmlns:xi="http://www.w3.org/2001/XInclude"
+ py:strip="">
+ <py:if test="dataset_count == 0">
+ <p>We do not currently have analytics data for ${publisher.title}</p>
+ </py:if>
+ <py:if test="dataset_count > 0">
+ <div class="popular_datasets">
+ <ul>
+ <py:for each="dataset,views,visits in datasets">
+ <li>
+ ${h.link_to(dataset.title, h.url_for(controller='package', action='read', id=dataset.name))}
+ </li>
+ </py:for>
+ </ul>
+ <p class="">${h.link_to("More analytics for " + publisher.title, h.url_for(controller='ckanext.ga_report.controller:GaPublisherReport',action='read',id=publisher.name))}</p>
+ </div>
+ </py:if>
+</html>
+
+
+
+
--- a/ckanext/ga_report/templates/ga_report/publisher/read.html
+++ b/ckanext/ga_report/templates/ga_report/publisher/read.html
@@ -6,16 +6,17 @@
<py:def function="page_title">Analytics for ${g.site_title}</py:def>
<py:match path="primarysidebar">
+
<li class="widget-container boxed widget_text">
<h4>${c.publisher.title}</h4>
<p>
The table shows the top 20 most viewed datasets belonging to ${c.publisher.title}.
</p>
<p>
- As well as showing the number of views within ${c.month_desc}, it will also show the
+ As well as showing the number of views for ${c.month_desc}, it will also show the
number of visitors that viewed each dataset.
</p>
- <p>The dataset list page for <a href="${h.url_for(controller='ckanext.dgu.controllers.publisher:PublisherController', action='read', id=c.publisher.name)}">${c.publisher.title}</a> was viewed ${c.publisher_page_views} times during ${c.month_desc}</p>
+ <p>The dataset list page for <a href="${h.url_for(controller='ckanext.dgu.controllers.publisher:PublisherController', action='read', id=c.publisher.name)}">${c.publisher.title}</a> was viewed ${c.publisher_page_views} times during the selected time period</p>
<p>View the <a href="${h.url_for(controller='ckanext.ga_report.controller:GaPublisherReport', action='index')}">publishers</a> leaderboard</p>
</li>
</py:match>
@@ -23,12 +24,13 @@
<div py:match="content">
<h1>Analytics for ${c.publisher.title}</h1>
- <h2>Top 20 most viewed datasets</h2>
+ <h2>Top 20 most viewed datasets of ${c.month_desc}</h2>
<p><em>Note: this data does not include API calls</em></p>
<form class="form-inline" action="${h.url_for(controller='ckanext.ga_report.controller:GaPublisherReport',action='read',id=c.publisher.name)}" method="get">
<div class="controls">
<select name="month">
+ <option value='' py:attrs="{'selected': 'selected' if not c.month else None}">All time</option>
<py:for each="val,desc in c.months">
<option value='${val}' py:attrs="{'selected': 'selected' if c.month == val else None}">${desc}</option>
</py:for>
--- a/ckanext/ga_report/templates/ga_report/site/index.html
+++ b/ckanext/ga_report/templates/ga_report/site/index.html
@@ -9,6 +9,10 @@
<li class="widget-container boxed widget_text">
<h4>Statistics</h4>
<p>It is possible to <a href="${h.url_for(controller='ckanext.ga_report.controller:GaReport',action='csv',month=c.month)}">export the analytics data</a> as a CSV file, which contains all of the information for ${c.month_desc}</p>
+ </li>
+ <li class="widget-container boxed widget_text">
+ <h4>Publisher statistics</h4>
+ <p>You can view statistics about specific publishers at the <a href="${h.url_for(controller='ckanext.ga_report.controller:GaPublisherReport',action='index')}">Publisher Analytics</a> page</p>
</li>
</py:match>
@@ -50,6 +54,8 @@
</tr>
</py:for>
</table>
+
+ <p>* Values are rounded up to 2 decimal places.</p>
</div>
<div class="tab-pane" id="browsers">
<table class="table table-condensed table-bordered table-striped">