Magic Byte File Sniffer

#[derive(Debug, PartialEq, Eq, Clone, Copy)]
enum FileKind {
    Png,
    Jpeg,
    Gif,
    Pdf,
    Zip,
    Gzip,
    Unknown,
}

fn sniff(bytes: &[u8]) -> FileKind {
    const PNG: &[u8] = &[0x89, b'P', b'N', b'G', 0x0d, 0x0a, 0x1a, 0x0a];
    const JPEG: &[u8] = &[0xff, 0xd8, 0xff];
    const GIF87: &[u8] = b"GIF87a";
    const GIF89: &[u8] = b"GIF89a";
    const PDF: &[u8] = b"%PDF-";
    const ZIP: &[u8] = &[0x50, 0x4b, 0x03, 0x04];
    const GZIP: &[u8] = &[0x1f, 0x8b];

    if bytes.starts_with(PNG) {
        return FileKind::Png;
    }
    if bytes.starts_with(JPEG) {
        return FileKind::Jpeg;
    }
    if bytes.starts_with(GIF87) || bytes.starts_with(GIF89) {
        return FileKind::Gif;
    }
    if bytes.starts_with(PDF) {
        return FileKind::Pdf;
    }
    if bytes.starts_with(ZIP) {
        return FileKind::Zip;
    }
    if bytes.starts_with(GZIP) {
        return FileKind::Gzip;
    }
    FileKind::Unknown
}

fn extension_matches(kind: FileKind, extension: &str) -> bool {
    let normalized = extension.to_ascii_lowercase();
    matches!(
        (kind, normalized.as_str()),
        (FileKind::Png, "png")
            | (FileKind::Jpeg, "jpg" | "jpeg")
            | (FileKind::Gif, "gif")
            | (FileKind::Pdf, "pdf")
            | (FileKind::Zip, "zip")
            | (FileKind::Gzip, "gz" | "tgz")
    )
}

fn main() {
    let samples: [(&[u8], &str); 4] = [
        (&[0x89, b'P', b'N', b'G', 0x0d, 0x0a, 0x1a, 0x0a, 0x00][..], "logo.png"),
        (&b"%PDF-1.7"[..], "invoice.pdf"),
        (&[0xff, 0xd8, 0xff, 0xe0][..], "photo.jpeg"),
        (&[0x50, 0x4b, 0x03, 0x04, 0x14][..], "archive.zip"),
    ];
    for (bytes, name) in samples {
        let kind = sniff(bytes);
        let extension = name.rsplit('.').next().unwrap_or_default();
        println!(
            "{name}: {:?} extension_ok={}",
            kind,
            extension_matches(kind, extension)
        );
    }
    println!("unknown: {:?}", sniff(b"plain text"));
}