Improve binary file check to better handle PDFs

[mmrwoods]

PDF files can have a very long text header before the first NUL byte

The check now reads the first 2KB rather than 128 bytes, and uses the
built-in index() function rather than looping which is very fast. That
still didn't catch all PDFs, even increasing it to 8000 bytes like git
diff didn't catch all PDFs, so also check for %PDF- signature in file.

The increased number of bytes to check remains even with the %PDF-
signature check because it costs very little (page size is min 4KB),
and will potentially catch other binary files with long text headers.

[Copilot]

Copilot review overview

🟢 Approval recommended

The implementation is sound; only a minor comment wording issue remains.

Review effort: Balanced
Findings: 1

Open (1)

• Fix spacing in “2 KB” unit · New

What changed in this PR

Improves binary detection for PDFs and files with long textual prefixes.

Changes:

• Expands NUL-byte scanning from 128 bytes to 2 KB.
• Detects the %PDF- signature directly.
• Uses index() instead of a byte loop.

File	Description
autoload/​fuzzbox/​internal/​previewer.vim	Enhances binary and PDF detection.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.