Add language directive validation for multi-language scenario calls

[Copilot]

When using non-English Gherkin dialects, missing or mismatched # language: directives between calling and called feature files cause scenario call expansion to fail silently, as the parser defaults to English keywords.

Changes

Validation

• Added ValidateLanguageDirectives() to detect missing directives in called features and language mismatches
• Added HasExplicitLanguageDirective() to distinguish explicit directives from defaults
• Warnings are injected inline where scenario calls are expanded

Robustness

• Enhanced FindFeatureFileContent() to try common language dialects (en, nl, de, fr, es) when feature name extraction fails
• Enables feature discovery even when directives are missing, allowing validation warnings to display

Example output

When a Dutch feature calls another without a language directive:

Scenario: Nieuw gebruikersaccount aanmaken
    # WARNING: Language directive missing in called feature 'Authenticatie'. Add '# language: nl' at the top of the feature file.
    # Warning: Could not expand scenario call (likely due to language directive issue)

When languages mismatch:

# WARNING: Language mismatch: calling feature uses 'nl' but called feature 'Authentifizierung' uses 'de'. Both feature files should use the same language directive.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Language comment must be present at the top of both feature files for non-English dialects</issue_title>
<issue_description>When using a language other than English (for example, Dutch), the language comment (e.g. # language: nl) must be placed at the top of the feature file.

Current situation:

• If the # language: directive is missing, parsing and/or keyword matching can fall back to English expectations, which can break scenario call detection and expansion.
• This affects both:
  1. the feature file that contains the scenario call, and
  2. the feature file that defines the called scenario.

Why this matters for scenario calls:

• The plugin parses/rewrites feature text (pre-processing) and needs consistent dialect recognition in both files involved in the call chain. A missing or mismatched dialect line can lead to scenario calls not being recognized/expanded.

References (code/tests/docs):

• Feature generator (parsing/rewriting):
  https://github.com/jgerits/Reqnroll.ScenarioCall.Generator/blob/main/Reqnroll.ScenarioCall.Generator/ScenarioCallFeatureGenerator.cs
• Multi-language tests (examples of dialect behavior):
  https://github.com/jgerits/Reqnroll.ScenarioCall.Generator/blob/main/Reqnroll.ScenarioCall.Generator.Tests/MultiLanguageTests.cs
• README multi-language usage:
  https://github.com/jgerits/Reqnroll.ScenarioCall.Generator/blob/main/README.md

Expected functionality:

• Either:
  • Automatically ensure the correct # language: <code> directive is present at the top of both the calling and called feature files, OR
  • Provide a clear warning/error explaining that the directive is required (and in which file(s) it is missing) when a scenario call involves non-English feature files.

Implementation hint:

• Add validation during preprocessing: if detected language/dialect differs or is missing in either side of a scenario call, emit an explicit diagnostic comment (or fail the build, depending on configuration).</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Language comment must be present at the top of both feature files for non-English dialects #29

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.