fix(ruby): fix RuboCop Lint/Void and Naming/VariableNumber offenses in generated SDKs

[fern-support]

Description

Fixes RuboCop violations in generated Ruby SDKs that cause CI lint failures (e.g., fern-demo/schematic-test-ruby).

Changes Made

1. Fix Lint/Void offenses in template files

• boolean.Template.rb: Added return before bare value in the when TrueClass, FalseClass branch (line 20). Other branches already used explicit return, making this one fall into void context.
• utils.Template.rb: Added return before bare value on the else branches of the pattern-matching case (lines 84, 87). Without return, these are dead expressions since the method continues to the raise and final value below.

2. Fix Naming/VariableNumber offenses in field name generation

• The generated .rubocop.yml enforces Naming/VariableNumber: normalcase (since v1.0.0-rc86), which disallows underscores before digits (e.g., account_last_4 should be account_last4).
• The IR's snakeCase.safeName can produce names like account_last_4 when the casing generator doesn't use smart casing, or the input triggers word-boundary splitting before digits.
• Added normalizeVariableNumber() using regex /([a-zA-Z\d])_(?=\d)/g to strip underscores before digit sequences, applied in:
  • generateFields.ts — model field symbols
  • DynamicSnippetsGeneratorContext.ts — property names in generated snippets
  • DynamicToLiteralMapper.ts — object property names in snippet literal conversion
• Wire values (api_name) are unaffected, preserving correct serialization.
• The regex uses a lookahead (?=\d) to correctly handle consecutive digit groups (e.g., test_1_2_3 → test123) and preserves leading underscores and digit-preceded underscores (e.g., _3_d stays _3_d).

3. Version bump

• Added v1.1.4 entry to generators/ruby-v2/sdk/versions.yml with changelog entries for both fixes.

Human Review Checklist

• Regex /([a-zA-Z\d])_(?=\d)/g breadth: This strips underscores before digits when preceded by a letter or digit. This matches RuboCop's normalcase expectation, but verify edge cases (e.g., v_2 → v2, x_1_2 → x12).
• normalizeVariableNumber is duplicated 3 times across packages. Consider whether it should be extracted into a shared utility (e.g., @fern-api/ruby-base).
• getMethodName in DynamicSnippetsGeneratorContext.ts was intentionally NOT updated — only getPropertyName was. Confirm method names don't need the same normalization.
• No unit tests were added for normalizeVariableNumber. Consider adding tests for edge cases (_3_d, account_last_4, test_1_2_3, names with no digits).

Testing

• Unit tests added/updated
• pnpm run check (biome lint) passes
• Seed test snapshots may need regeneration — initial run showed reserved-keywords fixture failure due to overly aggressive regex (fixed in subsequent commit)

Link to Devin session: https://app.devin.ai/sessions/d1bd30a5218f438db2821cdb6222fcd0
Requested by: @cdonel707

[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.

_{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
• Postman

---

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.

[graphite-app]

The normalizeVariableNumber regex only removes the underscore before the first digit in a sequence, not all underscores before digits. For example:

• test_1_2_3 becomes test1_2_3 (only first _1 normalized, _2_3 remain)
• account_id_123_test becomes account_id123_test (only d_1 normalized)

This creates inconsistent normalization and may still trigger RuboCop violations. The regex needs to handle multiple digit groups:

return name.replace(/([a-zA-Z])_(?=\d)/g, "$1");

This uses a lookahead (?=\d) instead of capturing the digit, allowing the regex engine to continue matching subsequent letter-underscore-digit patterns in the same string.

Suggested change

	function normalizeVariableNumber(name: string): string {
	return name.replace(/([a-zA-Z])_(\d)/g, "$1$2");
	}
	function normalizeVariableNumber(name: string): string {
	return name.replace(/([a-zA-Z\d])_(?=\d)/g, "$1");
	}

Spotted by Graphite



Is this helpful? React 👍 or 👎 to let us know.

[devin-ai-integration]

Good catch! Updated the regex to ([a-zA-Z\d])_(?=\d) in all three files — this handles consecutive digit groups like test_1_2_3 → test123 correctly. Pushed in cc5483a.

[graphite-app]

Version chronology is incorrect. Version 1.1.7 shows createdAt: "2026-03-26" while version 1.1.6 shows createdAt: "2026-03-27". Since 1.1.7 is a higher version number than 1.1.6, it should have been created after, not before.

This will cause issues:

• Semantic versioning violation (newer version created earlier)
• Potential confusion in release ordering and changelogs
• Breaking assumptions about version progression

Fix: Change 1.1.7's createdAt to "2026-03-27" or later, or reconsider the version numbering if 1.1.6 was meant to be the newer release.

Suggested change

	- version: 1.1.7
	changelogEntry:
	- summary: |
	Fix RuboCop `Lint/Void` offenses in generated `boolean.rb` and `utils.rb` internal
	type files. Bare `value` expressions in `case/when` branches and fallthrough paths
	now use explicit `return value` to avoid void-context warnings.
	type: fix
	- summary: |
	Fix RuboCop `Naming/VariableNumber` offenses for model field names containing digits.
	Field names like `account_last_4` are now normalized to `account_last4` to match
	the `normalcase` style enforced by the generated `.rubocop.yml`. The wire value
	is preserved via the `api_name` parameter for correct serialization.
	type: fix
	createdAt: "2026-03-27"
	irVersion: 61
	- version: 1.1.6
	changelogEntry:
	- summary: |
	type: chore
	createdAt: "2026-03-27"
	irVersion: 61

Spotted by Graphite



Is this helpful? React 👍 or 👎 to let us know.