refactor(workflow-operator): extract shared Sklearn descriptor base

[Ma77Ball]

What changes were proposed in this PR?

• Add abstract SklearnModelOpDesc extends PythonOperatorDescriptor holding the four shared config fields (target, countVectorizer, text, tfidfTransformer) with their Jackson/JSON-schema annotations, the two @JsonIgnore hooks (getImportStatements, getUserFriendlyModelName), and the shared getOutputSchemas.
• Change SklearnClassifierOpDesc and SklearnTrainingOpDesc to extend the new base, so each keeps only what genuinely differs (generatePythonCode, operatorInfo), removing ~50 lines of copy-pasted, drift-prone code.
• The 51 leaf subclasses inherit the fields unchanged, so the generated operator schema is identical (no behavior change).

Any related issues, documentation, discussions?

Closes: #6038

How was this PR tested?

• Pure refactor with no behavior change; covered by existing SklearnOpDescRegistrySpec, which instantiates a leaf classifier, sets the inherited target/countVectorizer fields, and asserts on generatePythonCode() output. Run it in CI via sbt "WorkflowOperator/jacoco", expect the Sklearn registry specs green.
• Verified locally with sbt "WorkflowOperator/Test/compile" (compiles clean; both descriptors resolve against SklearnModelOpDesc).
• Note: sbt "WorkflowOperator/testOnly *SklearnOpDescRegistrySpec" could not run locally because sbt-jacoco offline-instruments the Test classpath and JaCoCo 0.8.11 fails to instrument the vendored JsonSchemaDraft.class on this JDK (pre-existing tooling issue, unrelated to this change); the spec runs in CI under the WorkflowOperator/jacoco job.

Was this PR authored or co-authored using generative AI tooling?

Co-authored with Claude Opus 4.8 in compliance with ASF

[github-actions]

Automated Reviewer Suggestions

Based on the git blame history of the changed files, we recommend the following reviewers:

• Contributors with relevant context: @aglinxinyuan, @carloea2
  You can notify them by mentioning @aglinxinyuan, @carloea2 in a comment.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 86.66667% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 56.89%. Comparing base (104bcc4) to head (fd17af0).
⚠️ Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
...ber/operator/sklearn/SklearnClassifierOpDesc.scala	33.33%	2 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff              @@
##               main    #6048      +/-   ##
============================================
- Coverage     56.90%   56.89%   -0.01%     
+ Complexity     3024     3021       -3     
============================================
  Files          1129     1130       +1     
  Lines         43794    43787       -7     
  Branches       4743     4743              
============================================
- Hits          24920    24913       -7     
  Misses        17399    17399              
  Partials       1475     1475              

Flag	Coverage Δ		*Carryforward flag
access-control-service	70.00% <ø> (ø)
agent-service	44.59% <ø> (ø)		Carriedforward from 104bcc4
amber	58.73% <86.66%> (-0.02%)	⬇️
computing-unit-managing-service	0.00% <ø> (ø)
config-service	52.30% <ø> (ø)
file-service	62.81% <ø> (ø)
frontend	50.06% <ø> (ø)		Carriedforward from 104bcc4
notebook-migration-service	78.57% <ø> (ø)
pyamber	90.20% <ø> (ø)		Carriedforward from 104bcc4
python	90.76% <ø> (ø)		Carriedforward from 104bcc4
workflow-compiling-service	55.14% <ø> (ø)

*This pull request uses carry forward flags. Click here to find out more.

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

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[github-actions]

⚠️ Benchmark changes need a look

🟢 2 better · 🔴 6 worse · ⚪ 7 noise (<±5%) · 0 without baseline

Compared against main baca3f9 benchmarked on this same runner, so the delta is largely free of cross-runner hardware noise. The "7d avg" column still reflects the gh-pages dashboard. Treat <±5% as noise unless repeated.

Dashboard · Run

	config	throughput	MB/s	latency	max Δ latest / 7d
🔴	bs=10 sw=10 sl=64	452	0.276	22,114/28,815/28,815 us	🔴 +10.9% / 🔴 +94.9%
🔴	bs=100 sw=10 sl=64	935	0.57	104,162/158,675/158,675 us	🔴 +26.4% / 🔴 +48.4%
⚪	bs=1000 sw=10 sl=64	1,105	0.675	907,912/949,980/949,980 us	⚪ within ±5% / 🟢 -9.9%

Baseline details

Latest main baca3f9 from same runner

config	metric	PR	latest main	7d avg	Δ latest	Δ 7d
bs=10 sw=10 sl=64	throughput	452 tuples/sec	466 tuples/sec	786.27 tuples/sec	-3.0%	-42.5%
bs=10 sw=10 sl=64	MB/s	0.276 MB/s	0.284 MB/s	0.48 MB/s	-2.8%	-42.5%
bs=10 sw=10 sl=64	p50	22,114 us	19,942 us	12,495 us	+10.9%	+77.0%
bs=10 sw=10 sl=64	p95	28,815 us	30,514 us	14,784 us	-5.6%	+94.9%
bs=10 sw=10 sl=64	p99	28,815 us	30,514 us	18,468 us	-5.6%	+56.0%
bs=100 sw=10 sl=64	throughput	935 tuples/sec	994 tuples/sec	991.49 tuples/sec	-5.9%	-5.7%
bs=100 sw=10 sl=64	MB/s	0.57 MB/s	0.607 MB/s	0.605 MB/s	-6.1%	-5.8%
bs=100 sw=10 sl=64	p50	104,162 us	97,253 us	100,929 us	+7.1%	+3.2%
bs=100 sw=10 sl=64	p95	158,675 us	125,580 us	106,894 us	+26.4%	+48.4%
bs=100 sw=10 sl=64	p99	158,675 us	125,580 us	114,085 us	+26.4%	+39.1%
bs=1000 sw=10 sl=64	throughput	1,105 tuples/sec	1,114 tuples/sec	1,023 tuples/sec	-0.8%	+8.0%
bs=1000 sw=10 sl=64	MB/s	0.675 MB/s	0.68 MB/s	0.624 MB/s	-0.7%	+8.1%
bs=1000 sw=10 sl=64	p50	907,912 us	898,247 us	983,835 us	+1.1%	-7.7%
bs=1000 sw=10 sl=64	p95	949,980 us	999,338 us	1,023,777 us	-4.9%	-7.2%
bs=1000 sw=10 sl=64	p99	949,980 us	999,338 us	1,053,883 us	-4.9%	-9.9%

Raw CSV

config_idx,batch_size,schema_width,string_len,num_batches,total_ms,total_tuples,total_bytes,tuples_per_sec,mb_per_sec,lat_p50_us,lat_p95_us,lat_p99_us
0,10,10,64,20,442.28,200,128000,452,0.276,22114.42,28814.86,28814.86
1,100,10,64,20,2140.04,2000,1280000,935,0.570,104161.56,158674.63,158674.63
2,1000,10,64,20,18093.38,20000,12800000,1105,0.675,907911.68,949980.25,949980.25

[aglinxinyuan]

This string should be empty since it gets overridden anyway. The placeholder was introduced by mistake in the first place.

[aglinxinyuan]

ditto