Dynamically adjust table column widths in zfs-info output

[weak-head]

Description

This PR improves the readability of the zfs-info command output by replacing hardcoded column widths with dynamic, data-driven sizing.

Motivation and Context

Previously, the awk script used fixed-width formatting (e.g., %-30s) defined in the BEGIN block. If a dataset name, AWS bucket, or replication target string exceeded these fixed limits, the table layout would break or become misaligned. Conversely, short values left excessive empty space. This update ensures the table is always perfectly aligned based on the actual data being presented.

Changes Proposed

• Moved the header and separator generation to the END block so the script can evaluate the entire dataset first.
• The script now calculates the maximum string length for each specific column, factoring in the header lengths as the baseline.
• Automatically adds a 4-space padding to the maximum calculated width of each column for consistent spacing.
• Dynamic separator line: The horizontal separator (---) is now programmatically generated to match the exact total width of the new dynamic table.