Preserve selected source bodies in codegraph output

[GaoSSR]

Summary

• Return the selected symbol body from codegraph_node includeCode=true for container symbols instead of replacing classes/interfaces with member outlines
• Trim long context code blocks from the middle so both the opening context and tail lines remain visible
• Add regression coverage for container bodies and long code blocks with important tail content

Testing

• npx vitest run __tests__/context.test.ts __tests__/symbol-lookup.test.ts
• npm run build
• git diff --check

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 8cf3a56186

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Trim large container bodies before formatting

When includeCode is used for a large class/interface/etc. (common now that containers no longer use the member outline), this fetches the entire body and handleNode then applies truncateOutput to the already-formatted section. For container bodies over ~15 KB, the result only contains the prefix of the class and can cut off later methods, the closing fence, and the trail, so users still cannot see the selected symbol body reliably. Consider applying the same middle/tail-preserving trim to the code body before formatNodeDetails (or otherwise bounding it) instead of relying on whole-output truncation.

Useful? React with 👍 / 👎.