#!/bin/bash

# Extract gTLDs CDNs

# File paths
CSV_FILE="cdn_uniq.txt"
DOMAIN_FILE="tranco_V99PN_gTLDs_resolved_final.csv"
TMP_FILE="tranco_V99PN_gTLDs_resolved_final_cdn.tmp"
OUTPUT_FILE="tranco_V99PN_gTLDs_resolved_final_cdn.csv"

rm $TMP_FILE
rm $OUTPUT_FILE


# Read the CSV file line by line
while IFS=' ' read -r ip cdn cdn2 ; do
    grep -E "${ip}" "$DOMAIN_FILE" | sed "s/$/,-,-,$cdn $cdn2/" >> $TMP_FILE
    
    # Print the result
    #echo "Processed $ip ($cdn $cdn2)"
done < "$CSV_FILE"

# Remove duplicates
cat "$TMP_FILE" | sort | uniq > "$OUTPUT_FILE"