[NO-TICKET] Improve tracing Rails route computation

[Strech]

What does this PR do?

Drastically improves performance of the code (less allocations, better performance)

Motivation:

In AppSec we need to compute http.route-like value of the request and I found that code I can improve.

Change log entry

Yes. Tracing: Improve performance of http.route tag computation for rails contrib.

Additional Notes:

I've replaced 1 string allocation with mutation and have used faster suffix removal method.

w: - there is a match in the string
w/o - there is no match in the string

ruby 3.3.5 (2024-09-03 revision ef084cc8f4) [arm64-darwin23] Warming up --------------------------------------
             w: gsub    72.280k i/100ms
              w: sub    81.689k i/100ms
    w: delete_suffix   502.700k i/100ms
           w/o: gsub   116.580k i/100ms
            w/o: sub   195.557k i/100ms
  w/o: delete_suffix   225.558k i/100ms
Calculating -------------------------------------
             w: gsub      1.609M (±68.8%) i/s  (621.46 ns/i) -     10.914M in  30.723566s
              w: sub      3.185M (±65.6%) i/s  (313.95 ns/i) -     17.481M in  30.198337s
    w: delete_suffix      4.903M (±118.8%) i/s  (203.97 ns/i) -     45.746M in  31.428342s
           w/o: gsub      4.935M (±64.8%) i/s  (202.61 ns/i) -     46.632M in  30.271821s
            w/o: sub      4.876M (±63.5%) i/s  (205.09 ns/i) -     46.934M in  30.233153s
  w/o: delete_suffix     10.564M (±78.8%) i/s   (94.66 ns/i) -     48.495M in  30.639072s

How to test the change?

CI + ST

[datadog-datadog-prod-us1]

Datadog Report

Branch report: appsec-57809-make-http-route-available-within-request
Commit report: af17fe4
Test service: dd-trace-rb

✅ 0 Failed, 21681 Passed, 1302 Skipped, 4m 52.24s Total Time

[pr-commenter]

Benchmarks

Benchmark execution time: 2025-06-14 12:34:05

Comparing candidate commit af17fe4 in PR branch appsec-57809-make-http-route-available-within-request with baseline commit 482c04e in branch master.

Found 0 performance improvements and 0 performance regressions! Performance is the same for 41 metrics, 6 unstable metrics.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.63%. Comparing base (482c04e) to head (af17fe4).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #4688      +/-   ##
==========================================
- Coverage   97.63%   97.63%   -0.01%     
==========================================
  Files        1474     1474              
  Lines       88022    88026       +4     
  Branches     4569     4568       -1     
==========================================
+ Hits        85942    85944       +2     
- Misses       2080     2082       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[Strech]

I'm going to withdraw the part with environment variable, but keep the performance part intact. I've found that we can leverage a little bit of hacks to achieve the same

[Strech]

@marcotc Will be happy to merge it. I've found tests that covers the change. 🤲🏼