Skip to content

Reading upper/lower bounds values with type promotions #3293

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
geruh merged 5 commits into from
May 1, 2026
+98 −2
Merged

Reading upper/lower bounds values with type promotions #3293

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
0bac388
Type promotion - metadata file reading
rambleraptor Apr 27, 2026
095fd76
linter fix
rambleraptor Apr 27, 2026
803ac61
PR comments
rambleraptor Apr 29, 2026
ef8e8c9
PR comments
rambleraptor Apr 30, 2026
7467c0a
PR comments
rambleraptor May 1, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 11 additions & 1 deletion pyiceberg/conversions.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -343,7 +343,6 @@ def _(_: PrimitiveType, b: bytes) -> int:
return _INT_STRUCT.unpack(b)[0]


@from_bytes.register(LongType)
@from_bytes.register(TimeType)
@from_bytes.register(TimestampType)
@from_bytes.register(TimestamptzType)
Expand All @@ -353,13 +352,24 @@ def _(_: PrimitiveType, b: bytes) -> int:
return _LONG_STRUCT.unpack(b)[0]


@from_bytes.register(LongType)
def _(_: PrimitiveType, b: bytes) -> int:
if len(b) == 4:
Copy link
Copy Markdown
Member

@geruh geruh Apr 30, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: just curious here when i linked the comment earlier to the java code I noticed they use < 8 any reason for the == 4 on these?

Copy link
Copy Markdown
Contributor Author

@rambleraptor rambleraptor May 1, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If it's not 8 bytes, it's going to be 4 bytes. I'm not sure what kind of data would be 5-7 bytes. I changed it nonetheless just to match the Java implementation.

# If the length is 4 bytes, it is a promoted IntegerType
return _INT_STRUCT.unpack(b)[0]
return _LONG_STRUCT.unpack(b)[0]


@from_bytes.register(FloatType)
def _(_: FloatType, b: bytes) -> float:
return _FLOAT_STRUCT.unpack(b)[0]


@from_bytes.register(DoubleType)
def _(_: DoubleType, b: bytes) -> float:
if len(b) == 4:
# If the length is 4 bytes, it is a promoted FloatType
return _FLOAT_STRUCT.unpack(b)[0]
return _DOUBLE_STRUCT.unpack(b)[0]


Expand Down
94 changes: 93 additions & 1 deletion tests/expressions/test_evaluator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,7 +41,14 @@
Or,
StartsWith,
)
from pyiceberg.expressions.visitors import _InclusiveMetricsEvaluator, _StrictMetricsEvaluator
from pyiceberg.expressions.visitors import (
ROWS_CANNOT_MATCH,
ROWS_MIGHT_MATCH,
ROWS_MIGHT_NOT_MATCH,
ROWS_MUST_MATCH,
_InclusiveMetricsEvaluator,
_StrictMetricsEvaluator,
)
from pyiceberg.manifest import DataFile, FileFormat
from pyiceberg.schema import Schema
from pyiceberg.typedef import Record
Expand All @@ -50,6 +57,7 @@
FloatType,
IcebergType,
IntegerType,
LongType,
NestedField,
PrimitiveType,
StringType,
Expand Down Expand Up @@ -1463,3 +1471,87 @@ def test_strict_integer_not_in(strict_data_file_schema: Schema, strict_data_file

should_read = _StrictMetricsEvaluator(strict_data_file_schema, NotIn("no_nulls", {"abc", "def"})).eval(strict_data_file_1)
assert not should_read, "Should not match: no_nulls field does not have bounds"


@pytest.mark.parametrize(
"file_type, evolved_type, lower_bound, upper_bound, op, lit, expected",
[
# Int -> Long
(IntegerType(), LongType(), 30, 79, GreaterThan, 100, ROWS_CANNOT_MATCH),
(IntegerType(), LongType(), 30, 79, LessThan, 50, ROWS_MIGHT_MATCH),
# Float -> Double
(FloatType(), DoubleType(), 30.0, 79.0, GreaterThan, 100.0, ROWS_CANNOT_MATCH),
(FloatType(), DoubleType(), 30.0, 79.0, LessThan, 50.0, ROWS_MIGHT_MATCH),
],
)
def test_inclusive_metrics_evaluator_with_type_promotion(
Comment thread
rambleraptor marked this conversation as resolved.
Outdated
file_type: PrimitiveType,
evolved_type: PrimitiveType,
lower_bound: Any,
upper_bound: Any,
op: Any,
lit: Any,
expected: bool,
) -> None:
# Schema defines 'col' with evolved state
Comment thread
rambleraptor marked this conversation as resolved.
Outdated
schema = Schema(NestedField(1, "col", evolved_type, required=True))

# Historical manifest contains file_type bounds
data_file = DataFile.from_args(
file_path="file_1.parquet",
file_format=FileFormat.PARQUET,
partition={},
record_count=100,
file_size_in_bytes=1024,
lower_bounds={1: to_bytes(file_type, lower_bound)},
upper_bounds={1: to_bytes(file_type, upper_bound)},
)

# Predicate refers to 'col'
evaluator = _InclusiveMetricsEvaluator(schema, op("col", lit))
assert evaluator.eval(data_file) == expected


@pytest.mark.parametrize(
"file_type, evolved_type, lower_bound, upper_bound, op, lit, expected",
[
# Int -> Long
(IntegerType(), LongType(), 30, 79, GreaterThan, 20, ROWS_MUST_MATCH),
(IntegerType(), LongType(), 30, 79, GreaterThan, 100, ROWS_MIGHT_NOT_MATCH),
(IntegerType(), LongType(), 30, 79, LessThan, 100, ROWS_MUST_MATCH),
(IntegerType(), LongType(), 30, 79, LessThan, 20, ROWS_MIGHT_NOT_MATCH),
# Float -> Double
(FloatType(), DoubleType(), 30.0, 79.0, GreaterThan, 20.0, ROWS_MUST_MATCH),
(FloatType(), DoubleType(), 30.0, 79.0, GreaterThan, 100.0, ROWS_MIGHT_NOT_MATCH),
(FloatType(), DoubleType(), 30.0, 79.0, LessThan, 100.0, ROWS_MUST_MATCH),
(FloatType(), DoubleType(), 30.0, 79.0, LessThan, 20.0, ROWS_MIGHT_NOT_MATCH),
],
)
def test_strict_metrics_evaluator_with_type_promotion(
file_type: PrimitiveType,
evolved_type: PrimitiveType,
lower_bound: Any,
upper_bound: Any,
op: Any,
lit: Any,
expected: bool,
) -> None:
# Schema defines 'col' with evolved state
schema = Schema(NestedField(1, "col", evolved_type, required=True))

# Historical manifest contains file_type bounds
data_file = DataFile.from_args(
file_path="file_1.parquet",
file_format=FileFormat.PARQUET,
partition={},
record_count=100,
file_size_in_bytes=1024,
lower_bounds={1: to_bytes(file_type, lower_bound)},
upper_bounds={1: to_bytes(file_type, upper_bound)},
null_value_counts={1: 0},
nan_value_counts={1: 0},
)

# Predicate refers to 'col'
evaluator = _StrictMetricsEvaluator(schema, op("col", lit))
assert evaluator.eval(data_file) == expected
Loading