Use inplace normalization and add benchmark

Author: VianneyRuhlmann

tools/docker/Dockerfile.build

@@ -106,7 +106,7 @@ COPY "data-pipeline/Cargo.toml" "data-pipeline/"
 COPY "data-pipeline-ffi/Cargo.toml" "data-pipeline-ffi/"
 COPY "bin_tests/Cargo.toml"  "bin_tests/"
 RUN find -name "Cargo.toml" | sed -e s#Cargo.toml#src/lib.rs#g | xargs -n 1 sh -c 'mkdir -p $(dirname $1); touch $1; echo $1' create_stubs
-RUN echo profiling/benches/main.rs profiling/benches/interning_strings.rs trace-obfuscation/benches/trace_obfuscation.rs tools/src/bin/dedup_headers.rs tools/sidecar_mockgen/src/bin/sidecar_mockgen.rs ddtelemetry/examples/tm-worker-test.rs ipc/tarpc/tarpc/examples/compression.rs ipc/tarpc/tarpc/examples/custom_transport.rs ipc/tarpc/tarpc/examples/pubsub.rs ipc/tarpc/tarpc/examples/readme.rs ipc/tarpc/tarpc/examples/tracing.rs ipc/tarpc/tarpc/tests/compile_fail.rs ipc/tarpc/tarpc/tests/dataservice.rs ipc/tarpc/tarpc/tests/service_functional.rs bin_tests/src/bin/crashtracker_bin_test.rs bin_tests/src/bin/test_the_tests.rs | xargs -n 1 sh -c 'mkdir -p $(dirname $1); touch $1; echo $1' create_stubs
+RUN echo trace-normalization/benches/normalization_utils.rs profiling/benches/main.rs profiling/benches/interning_strings.rs trace-obfuscation/benches/trace_obfuscation.rs tools/src/bin/dedup_headers.rs tools/sidecar_mockgen/src/bin/sidecar_mockgen.rs ddtelemetry/examples/tm-worker-test.rs ipc/benches/ipc.rs ipc/tarpc/tarpc/examples/compression.rs ipc/tarpc/tarpc/examples/custom_transport.rs ipc/tarpc/tarpc/examples/pubsub.rs ipc/tarpc/tarpc/examples/readme.rs ipc/tarpc/tarpc/examples/tracing.rs ipc/tarpc/tarpc/tests/compile_fail.rs ipc/tarpc/tarpc/tests/dataservice.rs ipc/tarpc/tarpc/tests/service_functional.rs bin_tests/src/bin/crashtracker_bin_test.rs bin_tests/src/bin/test_the_tests.rs | xargs -n 1 sh -c 'mkdir -p $(dirname $1); touch $1; echo $1' create_stubs
 
 # cache dependencies
 RUN cargo fetch --locked

trace-normalization/Cargo.toml

@@ -14,3 +14,8 @@ datadog-trace-protobuf = { path = "../trace-protobuf" }
 [dev-dependencies]
 rand = "0.8.5"
 duplicate = "0.4.1"
+criterion = "0.5"
+
+[[bench]]
+name = "normalization_utils"
+harness = false

trace-normalization/benches/normalization_utils.rs

@@ -0,0 +1,89 @@
+// Copyright 2024-Present Datadog, Inc. https://www.datadoghq.com/
+// SPDX-License-Identifier: Apache-2.0
+
+use criterion::{criterion_group, criterion_main, BatchSize, BenchmarkId, Criterion};
+
+fn normalize_service_bench(c: &mut Criterion) {
+    let mut group = c.benchmark_group("normalization");
+    let cases = &[
+        ("#test_starting_hash", "test_starting_hash"),
+            ("TestCAPSandSuch", "testcapsandsuch"),
+            (
+                "Test Conversion Of Weird !@#$%^&**() Characters",
+                "test_conversion_of_weird_characters",
+            ),
+            ("$#weird_starting", "weird_starting"),
+            ("allowed:c0l0ns", "allowed:c0l0ns"),
+            ("1love", "love"),
+            ("ünicöde", "ünicöde"),
+            ("ünicöde:metäl", "ünicöde:metäl"),
+            ("Data🐨dog🐶 繋がっ⛰てて", "data_dog_繋がっ_てて"),
+            (" spaces   ", "spaces"),
+            (" #hashtag!@#spaces #__<>#  ", "hashtag_spaces"),
+            (":testing", ":testing"),
+            ("_foo", "foo"),
+            (":::test", ":::test"),
+            ("contiguous_____underscores", "contiguous_underscores"),
+            ("foo_", "foo"),
+            (
+                "\u{017F}odd_\u{017F}case\u{017F}",
+                "\u{017F}odd_\u{017F}case\u{017F}",
+            ), // edge-case
+            ("", ""),
+            (" ", ""),
+            ("ok", "ok"),
+            ("™Ö™Ö™™Ö™", "ö_ö_ö"),
+            ("AlsO:ök", "also:ök"),
+            (":still_ok", ":still_ok"),
+            ("___trim", "trim"),
+            ("12.:trim@", ":trim"),
+            ("12.:trim@@", ":trim"),
+            ("fun:ky__tag/1", "fun:ky_tag/1"),
+            ("fun:ky@tag/2", "fun:ky_tag/2"),
+            ("fun:ky@@@tag/3", "fun:ky_tag/3"),
+            ("tag:1/2.3", "tag:1/2.3"),
+            ("---fun:k####y_ta@#g/1_@@#", "fun:k_y_ta_g/1"),
+            ("AlsO:œ#@ö))œk", "also:œ_ö_œk"),
+            (
+                "A00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000 000000000000",
+                "a00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000_0"
+            ),
+        ];
+
+    group.bench_function("normalize_service", |b| {
+        b.iter_batched_ref(
+            || cases.iter().map(|(c, _)| c.to_string()).collect::<Vec<_>>(),
+            |cases| {
+                for c in cases {
+                    datadog_trace_normalization::normalize_utils::normalize_service(c);
+                }
+            },
+            BatchSize::NumIterations(100000),
+        )
+    });
+}
+
+fn normalize_name_bench(c: &mut Criterion) {
+    let mut group = c.benchmark_group("normalization");
+    let cases = &[
+        "good",
+        "bad-name",
+        "Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.Too-Long-.",
+    ];
+    for case in cases {
+        group.bench_with_input(
+            BenchmarkId::new("normalize_name", case),
+            *case,
+            |b, case| {
+                b.iter_batched_ref(
+                    || case.to_owned(),
+                    datadog_trace_normalization::normalize_utils::normalize_name,
+                    BatchSize::NumIterations(100000),
+                )
+            },
+        );
+    }
+}
+
+criterion_group!(benches, normalize_service_bench, normalize_name_bench);
+criterion_main!(benches);

trace-normalization/src/lib.rs

@@ -3,6 +3,6 @@
 
 #![deny(clippy::all)]
 
-pub mod normalizer;
-
 pub mod normalize_utils;
+pub mod normalizer;
+pub(crate) mod utf8_helpers;