Skip to content

perf: parse w3c traceparent using a custom parser #178

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
dmehala merged 2 commits into from
Jan 27, 2025
Merged

perf: parse w3c traceparent using a custom parser #178

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1e267a6
perf: parse w3c traceparent without regex
dmehala Jun 28, 2024
07b7518
fix bug
dmehala Jan 27, 2025
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
3 changes: 1 addition & 2 deletions fuzz/w3c-propagation/fuzz.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,8 +10,6 @@
#include <cstdlib>
#include <functional>
#include <memory>
#include <string>
#include <unordered_map>

namespace dd = datadog::tracing;

Expand All @@ -22,6 +20,7 @@ dd::Tracer& tracer_singleton() {
dd::TracerConfig config;
config.service = "fuzzer";
config.collector = std::make_shared<dd::NullCollector>();
config.extraction_styles = {dd::PropagationStyle::W3C};

const auto finalized_config = dd::finalize_config(config);
if (!finalized_config) {
Expand Down
130 changes: 75 additions & 55 deletions src/datadog/w3c_propagation.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,6 @@
#include <algorithm>
#include <cassert>
#include <cstddef>
#include <regex>
#include <utility>

#include "hex.h"
Expand All @@ -17,18 +16,6 @@ namespace datadog {
namespace tracing {
namespace {

// Note that match group 0 is the entire match.
constexpr StringView k_traceparent_pattern =
"([0-9a-f]{2})" // hex version number (match group 1)
"-"
"([0-9a-f]{32})" // hex trace ID (match group 2)
"-"
"([0-9a-f]{16})" // hex parent span ID (match group 3)
"-"
"([0-9a-f]{2})" // hex "trace-flags" (match group 4)
"($|-.*)"; // either the end, or a hyphen preceding further fields (match
// group 5)

// Return a predicate that returns whether its `char` argument is any of the
// following:
//
Expand All @@ -49,55 +36,82 @@ auto verboten(int lowest_ascii, int highest_ascii,
// entry of the specified `headers`. Return `nullopt` on success. Return a value
// for the `tags::internal::w3c_extraction_error` tag if an error occurs.
Optional<std::string> extract_traceparent(ExtractedData& result,
const DictReader& headers) {
const auto maybe_traceparent = headers.lookup("traceparent");
if (!maybe_traceparent) {
return nullopt;
StringView traceparent) {
enum class state : char {
version,
trace_id,
parent_span_id,
trace_flags
} internal_state = state::version;

if (traceparent.size() < 55) return "malformed_traceparent";

StringView version;
std::size_t beg = 0;
for (std::size_t i = 0; i < traceparent.size(); ++i) {
switch (internal_state) {
case state::version: {
if (i > 2) return "malformed_traceparent";
if (traceparent[i] == '-') {
version = StringView(traceparent.data() + beg, i - beg);
if (version == "ff") return "invalid_version";

beg = i + 1;
internal_state = state::trace_id;
}
} break;

case state::trace_id: {
if (i > 35) return "malformed_traceparent";
if (traceparent[i] == '-') {
auto maybe_trace_id =
TraceID::parse_hex(StringView(traceparent.data() + beg, i - beg));
if (maybe_trace_id.if_error() || *maybe_trace_id == 0)
return "malformed_traceid";

result.trace_id = *maybe_trace_id;

beg = i + 1;
internal_state = state::parent_span_id;
}
} break;

case state::parent_span_id: {
if (i > 52) return "malformed_traceparent";
if (traceparent[i] == '-') {
auto maybe_parent_id =
parse_uint64(StringView(traceparent.data() + beg, i - beg), 16);
if (maybe_parent_id.if_error() || *maybe_parent_id == 0)
return "malformed_parentid";

result.parent_id = *maybe_parent_id;

beg = i + 1;
internal_state = state::trace_flags;
goto handle_trace_flag;
}
} break;

default:
break;
}
}

const auto traceparent = trim(*maybe_traceparent);

static const std::regex regex{k_traceparent_pattern.data()};

std::cmatch match;
if (!std::regex_match(traceparent.data(),
traceparent.data() + traceparent.size(), match,
regex)) {
if (internal_state != state::trace_flags) {
return "malformed_traceparent";
}

assert(match.ready());
assert(match.size() == 6);

const auto to_string_view = [traceparent_beg = traceparent.data()](
const std::cmatch& match,
const std::size_t index) {
assert(index < match.size());
return StringView(traceparent_beg + match.position(index),
std::size_t(match.length(index)));
};

const auto version = to_string_view(match, 1);
if (version == "ff") {
return "invalid_version";
}

if (version == "00" && !to_string_view(match, 5).empty()) {
handle_trace_flag:
auto left = traceparent.size() - beg;
if (left < 2 ||
(left > 2 && (version == "00" || traceparent[beg + 2] != '-')))
return "malformed_traceparent";
}

result.trace_id = *TraceID::parse_hex(to_string_view(match, 2));
if (result.trace_id == 0) {
return "trace_id_zero";
}

result.parent_id = *parse_uint64(to_string_view(match, 3), 16);
if (*result.parent_id == 0) {
return "parent_id_zero";
}
auto maybe_trace_flags =
parse_uint64(StringView(traceparent.data() + beg, 2), 16);
if (maybe_trace_flags.if_error()) return "malformed_traceflags";

const auto flags = *parse_uint64(to_string_view(match, 4), 16);
result.sampling_priority = int(flags & 1);
result.sampling_priority = *maybe_trace_flags & 0x01;

return nullopt;
}
Expand Down Expand Up @@ -284,7 +298,13 @@ Expected<ExtractedData> extract_w3c(
ExtractedData result;
result.style = PropagationStyle::W3C;

if (auto error_tag_value = extract_traceparent(result, headers)) {
const auto maybe_traceparent = headers.lookup("traceparent");
if (!maybe_traceparent) {
return ExtractedData{};
}

if (auto error_tag_value =
extract_traceparent(result, trim(*maybe_traceparent))) {
span_tags[tags::internal::w3c_extraction_error] =
std::move(*error_tag_value);
return ExtractedData{};
Expand Down
16 changes: 14 additions & 2 deletions test/test_tracer.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -730,15 +730,27 @@ TEST_CASE("span extraction") {

{__LINE__, "invalid: trace ID zero",
"00-00000000000000000000000000000000-00f067aa0ba902b7-00", // traceparent
"trace_id_zero"}, // expected_error_tag_value
"malformed_traceid"}, // expected_error_tag_value

{__LINE__, "invalid: parent ID zero",
"00-4bf92f3577b34da6a3ce929d0e0e4736-0000000000000000-00", // traceparent
"parent_id_zero"}, // expected_error_tag_value
"malformed_parentid"}, // expected_error_tag_value

{__LINE__, "invalid: trailing characters when version is zero",
"00-4bf92f3577b34da6a3ce929d0e0e4736-00f067aa0ba902b7-00-foo", // traceparent
"malformed_traceparent"}, // expected_error_tag_value

{__LINE__, "invalid: non hex trace ID",
"00-abcdefghijklmnopqrstuvxyzabcdefg-00f067aa0ba902b7-00", // traceparent
"malformed_traceid"}, // expected_error_tag_value

{__LINE__, "invalid: non hex parent ID",
"00-4bf92f3577b34da6a3ce929d0e0e4736-abcdefghijklmnop-00", // traceparent
"malformed_parentid"}, // expected_error_tag_value

{__LINE__, "invalid: non hex trace tag ID",
"00-4bf92f3577b34da6a3ce929d0e0e4736-00f067aa0ba902b7-xy", // traceparent
"malformed_traceflags"}, // expected_error_tag_value
}));
// clang-format on

Expand Down
Loading