def test_build_query_with_multiple_dimensions_and_visualizations(self):
queries = (
mock_dataset.query()
.widget(f.ReactTable(mock_dataset.fields.votes, mock_dataset.fields.wins))
.widget(
f.HighCharts()
.axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
.axis(f.HighCharts.LineSeries(mock_dataset.fields.wins))
)
.dimension(f.day(mock_dataset.fields.timestamp))
.dimension(mock_dataset.fields.political_party)
.sql
)
self.assertEqual(len(queries), 1)
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'"political_party" "$political_party",'
'SUM("votes") "$votes",'
'SUM("is_winner") "$wins" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp","$political_party" '
'ORDER BY "$timestamp","$political_party"',
str(queries[0]),
)
def test_dod(self):
queries = (
mock_dataset.query.widget(
f.HighCharts().axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
)
.dimension(timestamp_daily)
.reference(f.DayOverDay(mock_dataset.fields.timestamp))
.sql
)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('day',1,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[1]),
)
def test_dimension_with_single_reference_wow(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.timestamp) \
.reference(f.WeekOverWeek(slicer.dimensions.timestamp)) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[0]))
with self.subTest('reference query is same as base query with reference dimension shifted'):
self.assertEqual('SELECT '
'TRUNC(TIMESTAMPADD(\'week\',1,"timestamp"),\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes_wow" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[1]))
def test_reference_with_dimension_using_display_definition_includes_it_in_all_queries(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.timestamp) \
.dimension(slicer.dimensions.candidate) \
.reference(f.YearOverYear(slicer.dimensions.timestamp)) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'DD\') "$d$timestamp",'
'"candidate_id" "$d$candidate",'
'"candidate_name" "$d$candidate_display",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp","$d$candidate","$d$candidate_display" '
'ORDER BY "$d$timestamp","$d$candidate_display"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'TRUNC(TIMESTAMPADD(\'year\',1,"timestamp"),\'DD\') "$d$timestamp",'
'"candidate_id" "$d$candidate",'
'"candidate_name" "$d$candidate_display",'
'SUM("votes") "$m$votes_yoy" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp","$d$candidate","$d$candidate_display" '
'ORDER BY "$d$timestamp","$d$candidate_display"', str(queries[1]))
def test_filters_on_other_dimensions_are_not_adapted(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.timestamp) \
.reference(f.DayOverDay(slicer.dimensions.timestamp)) \
.filter(slicer.dimensions.timestamp
.between(date(2018, 1, 1), date(2018, 1, 31))) \
.filter(slicer.dimensions.political_party
.isin(['d'])) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'WHERE "timestamp" BETWEEN \'2018-01-01\' AND \'2018-01-31\' '
'AND "political_party" IN (\'d\') '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'TRUNC(TIMESTAMPADD(\'day\',1,"timestamp"),\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes_dod" '
'FROM "politics"."politician" '
'WHERE TIMESTAMPADD(\'day\',1,"timestamp") BETWEEN \'2018-01-01\' AND \'2018-01-31\' '
'AND "political_party" IN (\'d\') '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[1]))
def test_adding_duplicate_reference_does_not_join_more_queries(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.timestamp) \
.reference(f.DayOverDay(slicer.dimensions.timestamp),
f.DayOverDay(slicer.dimensions.timestamp)) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'TRUNC(TIMESTAMPADD(\'day\',1,"timestamp"),\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[1]))
def test_use_same_nested_query_for_joining_references_with_same_period_and_dimension(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.timestamp) \
.reference(f.DayOverDay(slicer.dimensions.timestamp),
f.DayOverDay(slicer.dimensions.timestamp, delta=True),
f.DayOverDay(slicer.dimensions.timestamp, delta_percent=True)) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'TRUNC(TIMESTAMPADD(\'day\',1,"timestamp"),\'DD\') "$d$timestamp",'
'SUM("votes") "$m$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[1]))
def test_reference_on_dimension_with_weekly_interval(self):
weekly_timestamp = slicer.dimensions.timestamp(f.weekly)
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(weekly_timestamp) \
.reference(f.DayOverDay(weekly_timestamp)) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'IW\') "$d$timestamp",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'TRUNC(TIMESTAMPADD(\'day\',1,"timestamp"),\'IW\') "$d$timestamp",'
'SUM("votes") "$m$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[1]))
def test_single_reference_dod_with_dimension_but_not_reference_dimension_in_query_using_filter(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.political_party) \
.reference(f.DayOverDay(slicer.dimensions.timestamp)) \
.filter(slicer.dimensions.timestamp.between(date(2000, 1, 1), date(2000, 3, 1))) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'"political_party" "$d$political_party",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'WHERE "timestamp" BETWEEN \'2000-01-01\' AND \'2000-03-01\' '
'GROUP BY "$d$political_party" '
'ORDER BY "$d$political_party"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'"political_party" "$d$political_party",'
'SUM("votes") "$m$votes_dod" '
'FROM "politics"."politician" '
'WHERE TIMESTAMPADD(\'day\',1,"timestamp") BETWEEN \'2000-01-01\' AND \'2000-03-01\' '
'GROUP BY "$d$political_party" '
'ORDER BY "$d$political_party"', str(queries[1]))
def test_date_dim_with_weekly_interval(self):
weekly_timestamp = f.week(mock_dataset.fields.timestamp)
queries = (mock_dataset.query.widget(f.HighCharts().axis(
f.HighCharts.LineSeries(mock_dataset.fields.votes))).dimension(
weekly_timestamp).reference(
f.DayOverDay(weekly_timestamp)).sql)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'IW\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with filter on reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('day',1,TRUNC(\"timestamp\",'IW')),'IW') \"$timestamp\","
'SUM("votes") "$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[1]),
)
def test_yoy_with_monthly_interval(self):
queries = (mock_dataset.query.widget(f.HighCharts().axis(
f.HighCharts.LineSeries(mock_dataset.fields.votes))).dimension(
timestamp_monthly).reference(
f.YearOverYear(mock_dataset.fields.timestamp)).sql)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'MM\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('year',1,TRUNC(\"timestamp\",'MM')),'MM') \"$timestamp\","
'SUM("votes") "$votes_yoy" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[1]),
)
def test_adapt_dow_for_leap_year_for_yoy_reference_with_date_filter(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.dimension(slicer.dimensions.timestamp(f.weekly)) \
.reference(f.YearOverYear(slicer.dimensions.timestamp)) \
.filter(slicer.dimensions.timestamp.between(date(2018, 1, 1), date(2018, 1, 31))) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'TRUNC("timestamp",\'IW\') "$d$timestamp",'
'SUM("votes") "$m$votes" '
'FROM "politics"."politician" '
'WHERE "timestamp" BETWEEN \'2018-01-01\' AND \'2018-01-31\' '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[0]))
with self.subTest('reference query is same as base query with filter on reference dimension shifted'):
self.assertEqual('SELECT '
'TIMESTAMPADD(\'year\',-1,TRUNC(TIMESTAMPADD(\'year\',1,"timestamp"),\'IW\')) '
'"$d$timestamp",'
'SUM("votes") "$m$votes_yoy" '
'FROM "politics"."politician" '
'WHERE TIMESTAMPADD(\'year\',1,"timestamp") BETWEEN \'2018-01-01\' AND \'2018-01-31\' '
'GROUP BY "$d$timestamp" '
'ORDER BY "$d$timestamp"', str(queries[1]))
def test_build_query_for_chart_visualization_with_single_axis(self):
queries = mock_dataset.query \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))) \
.sql
self.assertEqual(len(queries), 1)
self.assertEqual(
'SELECT '
'SUM("votes") "$votes" '
'FROM "politics"."politician"', str(queries[0]))
def test_build_query_for_chart_visualization_with_single_axis(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.queries
self.assertEqual(len(queries), 1)
self.assertEqual(
'SELECT '
'SUM("votes") "$m$votes" '
'FROM "politics"."politician"', str(queries[0]))
def test_use_same_nested_query_for_joining_references_with_same_period_and_dimension_with_different_periods(
self,
):
queries = (
mock_dataset.query.widget(
f.HighCharts().axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
)
.dimension(timestamp_daily)
.reference(
f.DayOverDay(mock_dataset.fields.timestamp),
f.DayOverDay(mock_dataset.fields.timestamp, delta=True),
# also work with modified dimensions
f.YearOverYear(timestamp_daily),
f.YearOverYear(timestamp_daily, delta=True),
)
.sql
)
self.assertEqual(3, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[0]),
)
with self.subTest("second query for all DoD references"):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('day',1,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[1]),
)
with self.subTest("third query for all YoY references"):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('week',52,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_yoy" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[2]),
)
def test_dimension_with_multiple_references(self):
queries = (mock_dataset.query.widget(f.HighCharts().axis(
f.HighCharts.LineSeries(mock_dataset.fields.votes))).dimension(
timestamp_daily).reference(
f.DayOverDay(mock_dataset.fields.timestamp)).reference(
f.YearOverYear(mock_dataset.fields.timestamp,
delta_percent=True)).sql)
self.assertEqual(3, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with filter on reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('day',1,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_dod" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[1]),
)
with self.subTest(
"reference query is same as base query with filter on reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('week',52,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_yoy" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp" '
'LIMIT 200000',
str(queries[2]),
)
def test_build_query_for_chart_visualization_with_multiple_axes(self):
queries = mock_dataset.query.widget(f.HighCharts().axis(
f.HighCharts.LineSeries(mock_dataset.fields.votes)).axis(
f.HighCharts.LineSeries(mock_dataset.fields.wins))).sql
self.assertEqual(len(queries), 1)
self.assertEqual(
'SELECT '
'SUM("votes") "$votes",'
'SUM("is_winner") "$wins" '
'FROM "politics"."politician" '
'ORDER BY 1 '
'LIMIT 200000',
str(queries[0]),
)
def test_filters_on_other_dimensions_are_not_adapted(self):
queries = (
mock_dataset.query.widget(
f.HighCharts().axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
)
.dimension(timestamp_daily)
.reference(f.DayOverDay(mock_dataset.fields.timestamp))
.filter(
mock_dataset.fields.timestamp.between(
date(2018, 1, 1), date(2018, 1, 31)
)
)
.filter(mock_dataset.fields.political_party.isin(["d"]))
.sql
)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN '2018-01-01' AND '2018-01-31' "
"AND \"political_party\" IN ('d') "
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with filter on reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('day',1,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_dod" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN TIMESTAMPADD('day',-1,'2018-01-01') AND TIMESTAMPADD('day',-1,'2018-01-31') "
"AND \"political_party\" IN ('d') "
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(queries[1]),
)
def test_single_reference_dod_with_no_dimension_uses_multiple_from_clauses_instead_of_joins(self):
queries = slicer.data \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))) \
.reference(f.DayOverDay(slicer.dimensions.timestamp)) \
.queries
self.assertEqual(2, len(queries))
with self.subTest('base query is same as without reference'):
self.assertEqual('SELECT '
'SUM("votes") "$m$votes" '
'FROM "politics"."politician"', str(queries[0]))
with self.subTest('reference query is same as base query'):
self.assertEqual('SELECT '
'SUM("votes") "$m$votes_dod" '
'FROM "politics"."politician"', str(queries[1]))
def test_reference_with_no_dimensions_or_filters_creates_same_query(self):
# TODO reduce this to a single query
queries = (mock_dataset.query.widget(f.HighCharts().axis(
f.HighCharts.LineSeries(mock_dataset.fields.votes))).reference(
f.DayOverDay(mock_dataset.fields.timestamp)).sql)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
'SELECT SUM("votes") "$votes" FROM "politics"."politician" ORDER BY 1 LIMIT 200000',
str(queries[0]),
)
with self.subTest("reference query is same as base query"):
self.assertEqual(
'SELECT SUM("votes") "$votes_dod" FROM "politics"."politician" ORDER BY 1 LIMIT 200000',
str(queries[1]),
)
def test_reference_without_selecting_ref_dimension_using_date_range_filter(self):
queries = (
mock_dataset.query.widget(
f.HighCharts().axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
)
.dimension(mock_dataset.fields.political_party)
.reference(f.DayOverDay(mock_dataset.fields.timestamp))
.filter(
mock_dataset.fields.timestamp.between(
date(2000, 1, 1), date(2000, 3, 1)
)
)
.sql
)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'"political_party" "$political_party",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN '2000-01-01' AND '2000-03-01' "
'GROUP BY "$political_party" '
'ORDER BY "$political_party"',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with filter on reference dimension shifted"
):
self.assertEqual(
"SELECT "
'"political_party" "$political_party",'
'SUM("votes") "$votes_dod" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN TIMESTAMPADD('day',-1,'2000-01-01') "
"AND TIMESTAMPADD('day',-1,'2000-03-01') "
'GROUP BY "$political_party" '
'ORDER BY "$political_party"',
str(queries[1]),
)
def test_build_query_with_multiple_dimensions_and_visualizations(self):
queries = slicer.data \
.widget(f.DataTablesJS(slicer.metrics.votes, slicer.metrics.wins)) \
.widget(f.HighCharts()
.axis(f.HighCharts.LineSeries(slicer.metrics.votes))
.axis(f.HighCharts.LineSeries(slicer.metrics.wins))) \
.dimension(slicer.dimensions.timestamp) \
.dimension(slicer.dimensions.political_party) \
.queries
self.assertEqual(len(queries), 1)
self.assertEqual(
'SELECT '
'TRUNC("timestamp",\'DD\') "$d$timestamp",'
'"political_party" "$d$political_party",'
'SUM("votes") "$m$votes",'
'SUM("is_winner") "$m$wins" '
'FROM "politics"."politician" '
'GROUP BY "$d$timestamp","$d$political_party" '
'ORDER BY "$d$timestamp","$d$political_party"', str(queries[0]))
def test_reference_with_dimension_using_display_definition_includes_it_in_all_queries(
self,
):
queries = (
mock_dataset.query.widget(
f.HighCharts().axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
)
.dimension(timestamp_daily)
.dimension(mock_dataset.fields["candidate-name"])
.reference(f.YearOverYear(mock_dataset.fields.timestamp))
.sql
)
self.assertEqual(2, len(queries))
with self.subTest("base query is same as without reference"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'"candidate_name" "$candidate-name",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp","$candidate-name" '
'ORDER BY "$timestamp","$candidate-name"',
str(queries[0]),
)
with self.subTest(
"reference query is same as base query with filter on reference dimension shifted"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('week',52,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'"candidate_name" "$candidate-name",'
'SUM("votes") "$votes_yoy" '
'FROM "politics"."politician" '
'GROUP BY "$timestamp","$candidate-name" '
'ORDER BY "$timestamp","$candidate-name"',
str(queries[1]),
)
def test_highcharts_requires_at_least_one_axis(self):
with self.assertRaises(MetricRequiredException):
slicer.data \
.widget(f.HighCharts()) \
.dimension(slicer.dimensions.timestamp) \
.queries
def test_highcharts_requires_at_least_one_axis(self):
with self.assertRaises(MetricRequiredException):
mock_dataset.query.widget(f.HighCharts()).dimension(
mock_dataset.fields.timestamp).sql
def test_reference_with_rollup_dimension_and_date_range_filter(self):
queries = (
mock_dataset.query.widget(
f.HighCharts().axis(f.HighCharts.LineSeries(mock_dataset.fields.votes))
)
.dimension(Rollup(timestamp_daily))
.reference(f.WeekOverWeek(mock_dataset.fields.timestamp))
.filter(
mock_dataset.fields.timestamp.between(
date(2018, 1, 1), date(2018, 1, 31)
)
)
.sql
)
self.assertEqual(4, len(queries))
base, reference, base_rollup, reference_rollup = queries
with self.subTest("base query applies dimensions and date range filter"):
self.assertEqual(
"SELECT "
'TRUNC("timestamp",\'DD\') "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN '2018-01-01' AND '2018-01-31' "
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(base),
)
with self.subTest(
"reference query shifts timestamp dimension and date range filter by a week"
):
self.assertEqual(
"SELECT "
"TRUNC(TIMESTAMPADD('week',1,TRUNC(\"timestamp\",'DD')),'DD') \"$timestamp\","
'SUM("votes") "$votes_wow" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN TIMESTAMPADD('week',-1,'2018-01-01') "
"AND TIMESTAMPADD('week',-1,'2018-01-31') "
'GROUP BY "$timestamp" '
'ORDER BY "$timestamp"',
str(reference),
)
with self.subTest("totals query selects NULL for timestamp dimension"):
self.assertEqual(
"SELECT "
'NULL "$timestamp",'
'SUM("votes") "$votes" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN '2018-01-01' AND '2018-01-31' "
'ORDER BY "$timestamp"',
str(base_rollup),
)
with self.subTest(
"reference totals query selects NULL for timestamp dimension and shifts date range filter"
):
self.assertEqual(
"SELECT "
'NULL "$timestamp",'
'SUM("votes") "$votes_wow" '
'FROM "politics"."politician" '
"WHERE \"timestamp\" BETWEEN TIMESTAMPADD('week',-1,'2018-01-01') "
"AND TIMESTAMPADD('week',-1,'2018-01-31') "
'ORDER BY "$timestamp"',
str(reference_rollup),
)
