feat: Add CakePHP 3+ Support

[PROFeNoM]

Description

Add support for CakePHP 3+ by replacing the hook arguments with the new CakePHP v3+ classes.

Documented in DataDog/documentation#22569

Reviewer checklist

• Test coverage seems ok.
• Appropriate labels assigned.

[codecov-commenter]

Codecov Report

Merging #2618 (dcd2715) into master (e721e34) will decrease coverage by 10.81%.
The diff coverage is 1.19%.

Additional details and impacted files

@@              Coverage Diff              @@
##             master    #2618       +/-   ##
=============================================
- Coverage     77.64%   66.83%   -10.81%     
- Complexity     2226     2240       +14     
=============================================
  Files           226      202       -24     
  Lines         25942    22036     -3906     
  Branches        986        0      -986     
=============================================
- Hits          20142    14728     -5414     
- Misses         5274     7308     +2034     
+ Partials        526        0      -526     

Flag	Coverage Δ
appsec-extension	?
tracer-extension	78.55% <100.00%> (+<0.01%)	⬆️
tracer-php	50.43% <0.00%> (-29.73%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files	Coverage Δ
ext/integrations/integrations.c	91.11% <100.00%> (+0.09%)	⬆️
...DTrace/Integrations/CakePHP/CakePHPIntegration.php	0.00% <0.00%> (ø)
...tegrations/CakePHP/V2/CakePHPIntegrationLoader.php	0.00% <0.00%> (ø)
...tegrations/CakePHP/V3/CakePHPIntegrationLoader.php	0.00% <0.00%> (ø)

... and 44 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e721e34...dcd2715. Read the comment docs.

[pr-commenter]

Benchmarks

Benchmark execution time: 2024-04-10 09:13:54

Comparing candidate commit 1f3d417 in PR branch alex/feat/cakephp-310 with baseline commit c067f72 in branch master.

Found 1 performance improvements and 3 performance regressions! Performance is the same for 178 metrics, 0 unstable metrics.

scenario:ComposerTelemetryBench/benchTelemetryParsing

• 🟥 execution_time [+1.953µs; +4.647µs] or [+3.500%; +8.328%]

scenario:LogsInjectionBench/benchLogsInfoInjection-opcache

• 🟥 execution_time [+207.329ns; +484.271ns] or [+2.279%; +5.324%]

scenario:PDOBench/benchPDOOverheadWithDBM-opcache

• 🟩 mem_peak [-111.585KB; -111.582KB] or [-4.708%; -4.708%]

scenario:TraceSerializationBench/benchSerializeTrace

• 🟥 execution_time [+12.967µs; +16.433µs] or [+7.200%; +9.124%]

[bwoebi]

I suppose you will want to continue checking for these for CakePHP V2 (i.e. if the class_exists check fails)?

[bwoebi]

I suppose you copied from V2, but shouldn't we just use proper hook_method for these? I mean return false just unconditionally drops the span...

I think we can improve that one in all instances here. Then this instrument_when_limited thing also isn't necessary.

[PROFeNoM]

Absolutely :) Copy-pasted as you said and didn't pay attention to this; I'll change these

[bwoebi]

Can we please move away from caching rootSpan and rather refetch \DDTrace\root_span(), as more and more frameworks are also working under Swoole (where the root span isn't global, but may change with the executing fiber).
For CakePHP v2 this didn't matter, but modern versions of this don't have that much global state anymore (e.g. https://github.com/rochamarcelo/cakephp-api-swoole-sample)

[bwoebi]

Looks good to me :-)