Add recursive file discovery, performance improvements, and usability fixes

[ssstonebraker]

What this does

Adds a --recursive (or --r) flag that uses find to discover .gnmap files in subdirectories so you don't have to pass them all manually on the command line. Works with all existing features (csv, summary, web, ssl, hostports, etc).

Performance

Replaced the shell while read loops in makecsv, summary, web, webdirty, ssl, and hostports with single-pass POSIX awk. This is a huge speedup when working with hundreds or thousands of scan files. The old code was spawning 5-7 subprocesses per line per file. Now it's one awk process per function.

Recursive file loading uses batched xargs cat instead of a per-file cat/sort loop.

Usability

• Single-dash switches now work (-all, -csv, etc) by normalizing to double-dash internally
• Added smb-ips.txt output with plain IPs for SMB hosts
• Output directory path printed at the end so you know where files landed
• Guard on summay-log so it doesn't error when no .log files exist
• Added recursive usage example to the help menu

Bug fixes

• Exclude temp.gnmap from recursive find (was catting into itself)
• mkdir before realpath so it works on macOS (BSD realpath fails on non-existent paths)
• Split gnmap port entries on /, instead of comma to preserve commas in version strings (e.g. LDAP domain info was getting truncated)
• Fixed double slash in summay-log file paths
• Fixed stray dollar sign in summay-log redirect

Compatibility

• All awk code uses POSIX features only, tested on both macOS (BSD awk) and Linux (gawk)
• find syntax is portable across BSD and GNU find
• No changes to existing behavior when --recursive is not used