This is an automated email from the ASF dual-hosted git repository.
damccorm pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/beam.git
The following commit(s) were added to refs/heads/master by this push:
new 4b3dcf26142 Fix GroupBy snippet tests for issue #30778 (#37672)
4b3dcf26142 is described below
commit 4b3dcf26142c6fe756ac38c7f60b122a06effdb3
Author: Mansi Singh <[email protected]>
AuthorDate: Fri Feb 27 07:08:51 2026 -0800
Fix GroupBy snippet tests for issue #30778 (#37672)
* Fix GroupBy snippet tests for issue #30778
- Add missing strawberry entry to GROCERY_LIST in all snippet files
- Fix test() block scoping (moved inside with beam.Pipeline() context)
- Remove beam.Map(print) from pipeline, add to else branch for standalone
use
- Fix check_simple_aggregate_result to compare Row objects directly
- Set skip_due_to_30778 = False now that tests are passing
- Restore [START]/[END] snippet markers and 2-space indentation
* Fix yapf formatting issues in GroupBy snippet files
* Fix skip flag, restore markers, remove to_grocery_row, fix
check_simple_aggregate_result
* Remove skip_due_to_30778 entirely and restore groupby_table markers
* Fix pylint W0106 by assigning beam.Map(print) result to _ in else branches
---
.../snippets/transforms/aggregation/groupby_attr.py | 13 ++++++-------
.../transforms/aggregation/groupby_attr_expr.py | 10 ++++++----
.../snippets/transforms/aggregation/groupby_expr.py | 5 +++--
.../transforms/aggregation/groupby_expr_aggregate.py | 10 ++++++----
.../aggregation/groupby_global_aggregate.py | 6 ++++--
.../aggregation/groupby_simple_aggregate.py | 5 +++--
.../snippets/transforms/aggregation/groupby_test.py | 20 +-------------------
.../transforms/aggregation/groupby_two_exprs.py | 9 +++++----
8 files changed, 34 insertions(+), 44 deletions(-)
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr.py
index 2b3afb8feea..0842d65e094 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr.py
@@ -43,6 +43,7 @@ GROCERY_LIST = [
beam.Row(recipe='pie', fruit='blueberry', quantity=1, unit_price=2.00),
beam.Row(recipe='muffin', fruit='blueberry', quantity=2, unit_price=2.00),
beam.Row(recipe='muffin', fruit='banana', quantity=3, unit_price=1.00),
+ beam.Row(recipe='pie', fruit='strawberry', quantity=3, unit_price=1.50),
]
# [END groupby_table]
@@ -50,15 +51,13 @@ GROCERY_LIST = [
def groupby_attr(test=None):
with beam.Pipeline() as p:
# [START groupby_attr]
- grouped = (
- p
- | beam.Create(GROCERY_LIST)
- | beam.GroupBy('recipe')
- | beam.Map(print))
+ grouped = (p | beam.Create(GROCERY_LIST) | beam.GroupBy('recipe'))
# [END groupby_attr]
- if test:
- test(grouped)
+ if test:
+ test(grouped)
+ else:
+ _ = grouped | beam.Map(print)
if __name__ == '__main__':
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr_expr.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr_expr.py
index cef24534342..a3e44cf8b7c 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr_expr.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_attr_expr.py
@@ -43,6 +43,7 @@ GROCERY_LIST = [
beam.Row(recipe='pie', fruit='blueberry', quantity=1, unit_price=2.00),
beam.Row(recipe='muffin', fruit='blueberry', quantity=2, unit_price=2.00),
beam.Row(recipe='muffin', fruit='banana', quantity=3, unit_price=1.00),
+ beam.Row(recipe='pie', fruit='strawberry', quantity=3, unit_price=1.50),
]
# [END groupby_table]
@@ -53,12 +54,13 @@ def groupby_attr_expr(test=None):
grouped = (
p
| beam.Create(GROCERY_LIST)
- | beam.GroupBy('recipe', is_berry=lambda x: 'berry' in x.fruit)
- | beam.Map(print))
+ | beam.GroupBy('recipe', is_berry=lambda x: 'berry' in x.fruit))
# [END groupby_attr_expr]
- if test:
- test(grouped)
+ if test:
+ test(grouped)
+ else:
+ _ = grouped | beam.Map(print)
if __name__ == '__main__':
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr.py
index 1a62af8c4f6..7669a2f0bd1 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr.py
@@ -44,11 +44,12 @@ def groupby_expr(test=None):
p
| beam.Create(
['strawberry', 'raspberry', 'blueberry', 'blackberry', 'banana'])
- | beam.GroupBy(lambda s: s[0])
- | beam.Map(print))
+ | beam.GroupBy(lambda s: s[0]))
# [END groupby_expr]
if test:
test(grouped)
+ else:
+ _ = grouped | beam.Map(print)
if __name__ == '__main__':
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr_aggregate.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr_aggregate.py
index ed37360f720..0e7647297ea 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr_aggregate.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_expr_aggregate.py
@@ -43,6 +43,7 @@ GROCERY_LIST = [
beam.Row(recipe='pie', fruit='blueberry', quantity=1, unit_price=2.00),
beam.Row(recipe='muffin', fruit='blueberry', quantity=2, unit_price=2.00),
beam.Row(recipe='muffin', fruit='banana', quantity=3, unit_price=1.00),
+ beam.Row(recipe='pie', fruit='strawberry', quantity=3, unit_price=1.50),
]
# [END groupby_table]
@@ -55,12 +56,13 @@ def expr_aggregate(test=None):
| beam.Create(GROCERY_LIST)
| beam.GroupBy('recipe').aggregate_field(
'quantity', sum, 'total_quantity').aggregate_field(
- lambda x: x.quantity * x.unit_price, sum, 'price')
- | beam.Map(print))
+ lambda x: x.quantity * x.unit_price, sum, 'price'))
# [END expr_aggregate]
- if test:
- test(grouped)
+ if test:
+ test(grouped)
+ else:
+ _ = grouped | beam.Map(print)
if __name__ == '__main__':
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_global_aggregate.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_global_aggregate.py
index 876644483a5..c7b82e55d43 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_global_aggregate.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_global_aggregate.py
@@ -44,6 +44,7 @@ GROCERY_LIST = [
beam.Row(recipe='pie', fruit='blueberry', quantity=1, unit_price=2.00),
beam.Row(recipe='muffin', fruit='blueberry', quantity=2, unit_price=2.00),
beam.Row(recipe='muffin', fruit='banana', quantity=3, unit_price=1.00),
+ beam.Row(recipe='pie', fruit='strawberry', quantity=3, unit_price=1.50),
]
# [END groupby_table]
@@ -57,11 +58,12 @@ def global_aggregate(test=None):
| beam.GroupBy().aggregate_field(
'unit_price', min, 'min_price').aggregate_field(
'unit_price', MeanCombineFn(), 'mean_price').aggregate_field(
- 'unit_price', max, 'max_price')
- | beam.Map(print))
+ 'unit_price', max, 'max_price'))
# [END global_aggregate]
if test:
test(grouped)
+ else:
+ _ = grouped | beam.Map(print)
if __name__ == '__main__':
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_simple_aggregate.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_simple_aggregate.py
index 528159b4990..b32a11a1563 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_simple_aggregate.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_simple_aggregate.py
@@ -43,6 +43,7 @@ GROCERY_LIST = [
beam.Row(recipe='pie', fruit='blueberry', quantity=1, unit_price=2.00),
beam.Row(recipe='muffin', fruit='blueberry', quantity=2, unit_price=2.00),
beam.Row(recipe='muffin', fruit='banana', quantity=3, unit_price=1.00),
+ beam.Row(recipe='pie', fruit='strawberry', quantity=3, unit_price=1.50),
]
# [END groupby_table]
@@ -54,9 +55,9 @@ def simple_aggregate(test=None):
p
| beam.Create(GROCERY_LIST)
| beam.GroupBy('fruit').aggregate_field(
- 'quantity', sum, 'total_quantity')
- | beam.Map(print))
+ 'quantity', sum, 'total_quantity'))
# [END simple_aggregate]
+
if test:
test(grouped)
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_test.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_test.py
index a8ccf2b6308..0d2e5dfe25a 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_test.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_test.py
@@ -38,10 +38,6 @@ from .groupby_global_aggregate import global_aggregate
from .groupby_simple_aggregate import simple_aggregate
from .groupby_two_exprs import groupby_two_exprs
-#
-# TODO: Remove early returns in check functions
-# https://github.com/apache/beam/issues/30778
-skip_due_to_30778 = True
class UnorderedList(object):
@@ -80,8 +76,6 @@ NamedTuple = beam.Row
def check_groupby_expr_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
grouped | beam.MapTuple(normalize_kv),
equal_to([
@@ -94,8 +88,6 @@ def check_groupby_expr_result(grouped):
def check_groupby_two_exprs_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
grouped | beam.MapTuple(normalize_kv),
equal_to([
@@ -109,8 +101,6 @@ def check_groupby_two_exprs_result(grouped):
def check_groupby_attr_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
grouped | beam.MapTuple(normalize_kv),
equal_to([
@@ -157,8 +147,6 @@ def check_groupby_attr_result(grouped):
def check_groupby_attr_expr_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
grouped | beam.MapTuple(normalize_kv),
equal_to([
@@ -209,10 +197,8 @@ def check_groupby_attr_expr_result(grouped):
def check_simple_aggregate_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
- grouped | beam.MapTuple(normalize_kv),
+ grouped,
equal_to([
#[START simple_aggregate_result]
NamedTuple(fruit='strawberry', total_quantity=3),
@@ -225,8 +211,6 @@ def check_simple_aggregate_result(grouped):
def check_expr_aggregate_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
grouped | beam.Map(normalize),
equal_to([
@@ -238,8 +222,6 @@ def check_expr_aggregate_result(grouped):
def check_global_aggregate_result(grouped):
- if skip_due_to_30778:
- return
assert_that(
grouped | beam.Map(normalize),
equal_to([
diff --git
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_two_exprs.py
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_two_exprs.py
index 6b890fee9f9..0e6ca8390bf 100644
---
a/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_two_exprs.py
+++
b/sdks/python/apache_beam/examples/snippets/transforms/aggregation/groupby_two_exprs.py
@@ -44,12 +44,13 @@ def groupby_two_exprs(test=None):
p
| beam.Create(
['strawberry', 'raspberry', 'blueberry', 'blackberry', 'banana'])
- | beam.GroupBy(letter=lambda s: s[0], is_berry=lambda s: 'berry' in s)
- | beam.Map(print))
+ | beam.GroupBy(letter=lambda s: s[0], is_berry=lambda s: 'berry' in s))
# [END groupby_two_exprs]
- if test:
- test(grouped)
+ if test:
+ test(grouped)
+ else:
+ _ = grouped | beam.Map(print)
if __name__ == '__main__':
