fix(csharp): add precision fallbacks to JsonElementComparer for float/double

[Swimburger]

Description

Fixes mock server test failures (e.g. GetJobPredictionsTest.MockServerTest in the Hume C# SDK) caused by a serialization format mismatch when fields use format: float.

Root cause: When a JSON response contains a float value like 0.10722749680280685, C# deserializes it into System.Single, then re-serializes it using System.Text.Json which outputs the shortest roundtrip representation: 0.1072275. The test comparator (JsonElementComparer) compares via GetDecimal(), which is string-based, so 0.10722749680280685 != 0.1072275 and the test fails — even though both represent the same float32 bits.

Changes Made

• In JsonElementComparer.Template.cs, replaced the direct GetDecimal() fail with a 3-tier fallback chain: GetDecimal() → GetDouble() → GetSingle(). Only reports a mismatch if all three levels disagree. The intermediate GetDouble() check prevents GetSingle() from masking genuine differences in double-precision fields (e.g. integers 16777216 vs 16777217 are equal as Single but differ as Double)
• Added version entry 2.59.4 in versions.yml
• Updated all 287 seed snapshot JsonElementComparer.cs files

⚠️ Human Review Checklist

• The fallback chain is applied unconditionally to all JSON numbers, not scoped to fields known to be format: float. The GetDouble() intermediate check should prevent false positives for double-precision and integer fields, but consider edge cases: if two decimal values differ only beyond double precision (>15 significant digits), GetDouble() would consider them equal and skip the GetSingle() check entirely, which is the correct behavior. However, no test currently exercises any of these fallback paths.
• No seed test exercises this code path — no existing seed test fixtures use float primitives in mock server responses, so this change produces no behavioral snapshot diffs beyond the template update. The fix was validated externally against the Hume C# SDK (2 failures → 1, with the remaining failure being an unrelated empty path parameter issue).

Testing

• Lint passes (pnpm run check)
• Validated against Hume C# SDK — GetJobPredictionsTest.MockServerTest passes with this change

Link to Devin session: https://app.devin.ai/sessions/1dcf1274fbe74a7cbac9f76e9a65ed28
Requested by: @Swimburger

---

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment and CI monitoring

[claude]

Claude Code Review

This repository is configured for manual code reviews. Comment @claude review to trigger a review and subscribe this PR to future pushes, or @claude review once for a one-time review.

_{Tip: disable this comment in your organization's Code Review settings.}

[github-actions]

🌱 Seed Test Selector

Select languages to run seed tests for:

• Python
• TypeScript
• Java
• Go
• Ruby
• C#
• PHP
• Swift
• Rust
• OpenAPI

---

How to use: Click the ⋯ menu above → "Edit" → check the boxes you want → click "Update comment". Tests will run automatically and snapshots will be committed to this PR.

[github-actions]

SDK Generation Benchmark Results

Comparing PR branch against latest nightly baseline on main (2026-04-09T04:46:50Z).

Full benchmark table (click to expand)

Generator	Spec	main (generator)	main (E2E)	PR (generator)	Delta
csharp-sdk	square	95s	138s	87s	-8s (-8.4%)

main (generator): generator-only time via --skip-scripts (includes Docker image build, container startup, IR parsing, and code generation — this is the same Docker-based flow customers use via fern generate). main (E2E): full customer-observable time including build/test scripts (nightly baseline, informational). Delta is computed against generator-only baseline.
⚠️ = generation exited with a non-zero exit code (timing may not reflect a successful run).
Baseline from nightly runs on main (latest: 2026-04-09T04:46:50Z). Trigger benchmark-baseline to refresh.

[devin-ai-integration]

✅ Devin Review: No Issues Found

Devin Review analyzed this PR and found no potential bugs to report.

View in Devin Review to see 3 additional findings.

[github-actions]

SDK Generation Benchmark Results

Comparing PR branch against latest nightly baseline on main (2026-04-09T04:46:50Z).

Full benchmark table (click to expand)

Generator	Spec	main (generator)	main (E2E)	PR (generator)	Delta
csharp-sdk	square	95s	138s	92s	-3s (-3.2%)

main (generator): generator-only time via --skip-scripts (includes Docker image build, container startup, IR parsing, and code generation — this is the same Docker-based flow customers use via fern generate). main (E2E): full customer-observable time including build/test scripts (nightly baseline, informational). Delta is computed against generator-only baseline.
⚠️ = generation exited with a non-zero exit code (timing may not reflect a successful run).
Baseline from nightly runs on main (latest: 2026-04-09T04:46:50Z). Trigger benchmark-baseline to refresh.