def test_document_translate_fallback(self):
d_en = document(locale='en-US',
title=u'How to delete Google Chrome?',
save=True)
invalid_translate = reverse('wiki.document', locale='tr',
args=[d_en.slug])
self.assertEqual(d_en, Document.from_url(invalid_translate))
def test_document_translate_fallback(self):
d_en = DocumentFactory(locale='en-US',
title=u'How to delete Google Chrome?')
invalid_translate = reverse('wiki.document',
locale='tr',
args=[d_en.slug])
self.assertEqual(d_en, Document.from_url(invalid_translate))
def test_document_translate_fallback(self):
d_en = DocumentFactory(locale="en-US",
title="How to delete Google Chrome?")
invalid_translate = reverse("wiki.document",
locale="tr",
args=[d_en.slug])
self.assertEqual(d_en, Document.from_url(invalid_translate))
def pageviews_by_document(start_date, end_date):
"""Return the number of pageviews by document in a given date range.
* Only returns en-US documents for now since that's what we did with
webtrends.
Returns a dict with pageviews for each document:
{<document_id>: <pageviews>,
1: 42,
7: 1337,...}
"""
counts = {}
request = _build_request()
start_index = 1
max_results = 10000
while True: # To deal with pagination
@retry_503
def _make_request():
return request.get(
ids='ga:' + profile_id,
start_date=str(start_date),
end_date=str(end_date),
metrics='ga:pageviews',
dimensions='ga:pagePath',
filters='ga:pagePathLevel2==/kb/;ga:pagePathLevel1==/en-US/',
max_results=max_results,
start_index=start_index).execute()
results = _make_request()
for result in results['rows']:
path = result[0]
pageviews = int(result[1])
doc = Document.from_url(path, id_only=True, check_host=False)
if not doc:
continue
# The same document can appear multiple times due to url params.
counts[doc.pk] = counts.get(doc.pk, 0) + pageviews
# Move to next page of results.
start_index += max_results
if start_index > results['totalResults']:
break
return counts
def test_document_translate_fallback(self):
d_en = document(locale="en-US", title=u"How to delete Google Chrome?", save=True)
invalid_translate = reverse("wiki.document", locale="tr", args=[d_en.slug])
self.assertEqual(d_en, Document.from_url(invalid_translate))
def pageviews_by_document(start_date, end_date, verbose=False):
"""Return the number of pageviews by document in a given date range.
* Only returns en-US documents for now since that's what we did with
webtrends.
Returns a dict with pageviews for each document:
{<document_id>: <pageviews>,
1: 42,
7: 1337,...}
"""
counts = {}
request = _build_request()
max_results = 10000
end_date_step = end_date
while True: # To reduce the size of result set request 3 months at a time
start_date_step = end_date_step - timedelta(90)
if start_date_step < start_date:
start_date_step = start_date
if verbose:
print 'Fetching data for {0!s} to {1!s}:'.format(start_date_step,
end_date_step)
start_index = 1
while True: # To deal with pagination
@retry_503
def _make_request():
return request.get(
ids='ga:' + profile_id,
start_date=str(start_date_step),
end_date=str(end_date_step),
metrics='ga:pageviews',
dimensions='ga:pagePath',
filters=('ga:pagePathLevel2==/kb/;'
'ga:pagePathLevel1==/en-US/'),
max_results=max_results,
start_index=start_index).execute()
results = _make_request()
if verbose:
d = (max_results - 1
if start_index + max_results - 1 < results['totalResults']
else results['totalResults'] - start_index)
print '- Got {0!s} of {1!s} results.'.format(start_index + d,
results['totalResults'])
for result in results.get('rows', []):
path = result[0]
pageviews = int(result[1])
doc = Document.from_url(path, id_only=True, check_host=False)
if not doc:
continue
# The same document can appear multiple times due to url params
counts[doc.pk] = counts.get(doc.pk, 0) + pageviews
# Move to next page of results.
start_index += max_results
if start_index > results.get('totalResults', 0):
break
end_date_step = start_date_step - timedelta(1)
if start_date_step == start_date or end_date_step < start_date:
break
return counts
def pageviews_by_document(start_date, end_date, verbose=False):
"""Return the number of pageviews by document in a given date range.
* Only returns en-US documents for now since that's what we did with
webtrends.
Returns a dict with pageviews for each document:
{<document_id>: <pageviews>,
1: 42,
7: 1337,...}
"""
counts = {}
request = _build_request()
max_results = 10000
end_date_step = end_date
while True: # To reduce the size of result set request 3 months at a time
start_date_step = end_date_step - timedelta(90)
if start_date_step < start_date:
start_date_step = start_date
if verbose:
print 'Fetching data for %s to %s:' % (start_date_step,
end_date_step)
start_index = 1
while True: # To deal with pagination
@retry_503
def _make_request():
return request.get(ids='ga:' + profile_id,
start_date=str(start_date_step),
end_date=str(end_date_step),
metrics='ga:pageviews',
dimensions='ga:pagePath',
filters=('ga:pagePathLevel2==/kb/;'
'ga:pagePathLevel1==/en-US/'),
max_results=max_results,
start_index=start_index).execute()
results = _make_request()
if verbose:
d = (max_results - 1 if start_index + max_results -
1 < results['totalResults'] else results['totalResults'] -
start_index)
print '- Got %s of %s results.' % (start_index + d,
results['totalResults'])
for result in results['rows']:
path = result[0]
pageviews = int(result[1])
doc = Document.from_url(path, id_only=True, check_host=False)
if not doc:
continue
# The same document can appear multiple times due to url params
counts[doc.pk] = counts.get(doc.pk, 0) + pageviews
# Move to next page of results.
start_index += max_results
if start_index > results['totalResults']:
break
end_date_step = start_date_step - timedelta(1)
if start_date_step == start_date or end_date_step < start_date:
break
return counts
