Fix codespell CI false positives on PDF artifacts

[Copilot]

What type of PR is this (这是什么类型的PR)

• Typo(勘误)
• BUG(程序错误)
• Improvement(提升)
• Feature(新特性)

Which issue(s) this PR fixes(Optional) (这个PR 修复了什么问题 (可选择))

What this PR does / why we need it (这个PR 做了什么/ 我们为什么需要这个PR)

Codespell CI was failing because it scanned binary audit PDFs and treated extracted garbage text as misspellings. This change narrows scan scope to text sources so the codespell job reflects actionable spelling issues.

• Workflow scope update

  • Extended codespell skip patterns in .github/workflows/codespell.yml to exclude all *.pdf files.
  • Keeps existing language/path exclusions intact.

• Effective config change

  with:
    check_filenames: true
    ignore_words_file: .codespellignore
    skip: "./.git,package-lock.json,./Languages/es,./Languages/pt-br,*.pdf"