lidavidm commented on code in PR #156:
URL: https://github.com/apache/iceberg-cpp/pull/156#discussion_r2253646204
##########
test/transform_test.cc:
##########
@@ -193,4 +187,316 @@ TEST(TransformResultTypeTest, NegativeCases) {
}
}
+TEST(TransformLiteralTest, IdentityTransform) {
+ struct Case {
+ std::shared_ptr<Type> source_type;
+ Literal source;
+ Literal expected;
+ };
+
+ const std::vector<Case> cases = {
+ {.source_type = iceberg::boolean(),
+ .source = Literal::Boolean(true),
+ .expected = Literal::Boolean(true)},
+ {.source_type = iceberg::int32(),
+ .source = Literal::Int(42),
+ .expected = Literal::Int(42)},
+ {.source_type = iceberg::int32(),
+ .source = Literal::Date(30000),
+ .expected = Literal::Date(30000)},
+ {.source_type = iceberg::int64(),
+ .source = Literal::Long(1234567890),
+ .expected = Literal::Long(1234567890)},
+ {.source_type = iceberg::timestamp(),
+ .source = Literal::Timestamp(1622547800000000),
+ .expected = Literal::Timestamp(1622547800000000)},
+ {.source_type = iceberg::timestamp_tz(),
+ .source = Literal::TimestampTz(1622547800000000),
+ .expected = Literal::TimestampTz(1622547800000000)},
+ {.source_type = iceberg::float32(),
+ .source = Literal::Float(3.14),
+ .expected = Literal::Float(3.14)},
+ {.source_type = iceberg::float64(),
+ .source = Literal::Double(1.23e-5),
+ .expected = Literal::Double(1.23e-5)},
+ {.source_type = iceberg::string(),
+ .source = Literal::String("Hello, World!"),
+ .expected = Literal::String("Hello, World!")},
+ {.source_type = iceberg::binary(),
+ .source = Literal::Binary({0x01, 0x02, 0x03}),
+ .expected = Literal::Binary({0x01, 0x02, 0x03})},
+ };
+
+ for (const auto& c : cases) {
+ auto transform = Transform::Identity();
+ auto transformPtr = transform->Bind(c.source_type);
+ ASSERT_TRUE(transformPtr.has_value()) << "Failed to bind identity
transform";
+
+ auto result = transformPtr.value()->Transform(c.source);
+ ASSERT_TRUE(result.has_value())
+ << "Failed to transform literal: " << c.source.ToString();
+
+ EXPECT_EQ(result.value(), c.expected)
+ << "Unexpected result for source: " << c.source.ToString();
+ }
+}
+
+TEST(TransformLiteralTest, BucketTransform) {
+ constexpr int32_t num_buckets = 4;
+ auto transform = Transform::Bucket(num_buckets);
+
+ struct Case {
+ std::shared_ptr<Type> source_type;
+ Literal source;
+ Literal expected;
+ };
+
+ const std::vector<Case> cases = {
+ {.source_type = iceberg::int32(),
+ .source = Literal::Int(42),
+ .expected = Literal::Int(3)},
+ {.source_type = iceberg::date(),
+ .source = Literal::Date(30000),
+ .expected = Literal::Int(2)},
+ {.source_type = iceberg::int64(),
+ .source = Literal::Long(1234567890),
+ .expected = Literal::Int(3)},
+ {.source_type = iceberg::timestamp(),
+ .source = Literal::Timestamp(1622547800000000),
+ .expected = Literal::Int(1)},
+ {.source_type = iceberg::timestamp_tz(),
+ .source = Literal::TimestampTz(1622547800000000),
+ .expected = Literal::Int(1)},
+ {.source_type = iceberg::string(),
+ .source = Literal::String("test"),
+ .expected = Literal::Int(3)},
+ };
+
+ for (const auto& c : cases) {
+ auto transformPtr = transform->Bind(c.source_type);
+ ASSERT_TRUE(transformPtr.has_value()) << "Failed to bind bucket transform";
+ auto result = transformPtr.value()->Transform(c.source);
+ ASSERT_TRUE(result.has_value())
+ << "Failed to transform literal: " << c.source.ToString();
+
+ EXPECT_EQ(result.value(), c.expected)
+ << "Unexpected result for source: " << c.source.ToString();
+ }
+}
+
+TEST(TransformLiteralTest, TruncateTransform) {
+ struct Case {
+ std::shared_ptr<Type> source_type;
+ int32_t width;
+ Literal source;
+ Literal expected;
+ };
+
+ const std::vector<Case> cases = {
+ {.source_type = iceberg::int32(),
+ .width = 5,
+ .source = Literal::Int(123456),
+ .expected = Literal::Int(123455)},
+ {.source_type = iceberg::string(),
+ .width = 5,
+ .source = Literal::String("Hello, World!"),
+ .expected = Literal::String("Hello")},
+ {.source_type = iceberg::string(),
+ .width = 5,
+ .source = Literal::String("😜🧐🤔🤪🥳"),
+ // Truncate to 5 bytes, the safe point should be four bytes which fits
the first
Review Comment:
I think this will be annoying to do in C++ without pulling in a library like
utf8proc (I suppose assuming UTF-8 you can implement just the bit logic to
detect a codepoint)
(FWIW: one emoji can be multiple codepoints, IIRC (due to e.g. flags, skin
color selectors) An emoji would be a grapheme cluster?)
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
