is a command-line utility designed for efficiently storing large lists of IP addresses and subnets using a compact custom binary format, with additional support for merging, converting, and transforming IP data from various formats.

via go install:

$ go install github.com/anatoly-kussul/ipbin/cmd/ipbin@latest

This installs ipbin to $GOPATH/bin/ipbin

ipbin [options] <output-file>

  -i, --input string      Input file path
  -B                      Read input as binary
  -Z                      Read input as gzip
  -b                      Write output as binary
  -z                      Write output as gzip
  -s, --sep string        Separator for text output (default: \n)
  -f, --format int        Text output format (1=subnets+ips, 2=ranges+ips, 3=subnets, 4=ranges)
  -h, --help              Show this help message

If -b is specified, output is written in a compact binary format:

• Each prefix is encoded as follows:
  • The first byte encodes both the address family and prefix length:
    • Values 0–32 represent an IPv4 prefix length.
    • Values 33–161 represent an IPv6 prefix length (actual length = byte - 33).
  • The following bytes contain only the minimum number of bytes required to represent the prefix address, i.e., ceil(prefixLen / 8) bytes.
• Example:
  • IPv4 /24 → b[0] = 24, b[1:4] = first 3 bytes of IPv4 address
  • IPv6 /64 → b[0] = 97 (64 + 33), b[1:9] = first 8 bytes of IPv6 address
• The file is a concatenation of such encoded prefixes.

• 1 (default): subnets+ips — single IPs as IPs, others as subnets
• 2: ranges+ips — single IPs as IPs, others as start-end
• 3: subnets — everything in subnet format
• 4: ranges — everything in ranges format as start-end

• Text input: one IP, subnet, or range per line (e.g., 1.2.3.4, 10.0.0.0/8, 192.168.1.1-192.168.1.255)
• Binary input: compact encoded prefixes as described above

MIT