Extract rule: template-no-valueless-arguments

[NullVoxPopuli]

Split from #2371.

[NullVoxPopuli-ai-agent]

Review: template-no-valueless-arguments

Compared against the original ember-template-lint rule no-valueless-arguments.

What's done well

• Clean, minimal implementation. The rule is simple and the port keeps it that way -- just 35 lines of rule code.
• Correct detection approach. Checking for @ prefix on the attribute name and then verifying no = in the source text is a pragmatic workaround since the ESLint Glimmer AST may not expose an isValueless property directly.
• All original test cases are ported -- both good and bad cases match the original exactly.
• Both gjs and hbs parser modes are tested.

Observations and potential issues

1. Detection mechanism differs from the original. The original rule relies on node.isValueless -- a boolean property provided by the Handlebars AST. The new rule checks !text.includes('=') on the source text. This is a reasonable heuristic but could theoretically produce false positives if an attribute name itself contains = (extremely unlikely for @-prefixed names) or false negatives in edge cases where the AST representation differs from source. In practice this should be fine, but it's worth noting the approach is source-text-based rather than AST-property-based.

2. context.sourceCode || context.getSourceCode() -- the fallback to getSourceCode() is good for backwards compatibility with older ESLint versions. However, getSourceCode() is deprecated in ESLint 9+. Since this plugin likely targets a range of ESLint versions, the || pattern is appropriate.

3. No fixable property in meta. The original also doesn't provide autofix, so this is consistent.

4. The error message matches exactly: "Named arguments should have an explicitly assigned value." -- identical to the original.

5. Missing a Foo @bar={{true}} gjs valid test. Wait, actually it is there. The test coverage looks complete.

Summary

Straightforward, clean port. The source-text heuristic for valueless detection is the only notable divergence from the original's approach, and it works correctly for all test cases. This is a simple rule and the migration handles it well.

---

Automated review comparing with ember-template-lint source