Merge pull request #9 from binpash/uniq-ips

Add uniq-ips benchmark

uniq-ips/.gitignore

@@ -0,0 +1 @@
+*.txt

uniq-ips/clean.sh

@@ -0,0 +1 @@
+rm *.txt

uniq-ips/inputs.sh

@@ -0,0 +1,3 @@
+N=40000000 # 40M
+
+./scripts/gen_data.py "$N" > logs-popcount-org.txt

uniq-ips/run.sh

@@ -0,0 +1 @@
+cat "logs-popcount-org.txt" | sort | uniq > "out.txt"

uniq-ips/scripts/gen_data.py

@@ -0,0 +1,28 @@
+#!/usr/bin/env python3
+import sys
+import random
+
+# Function to generate a random IP address
+def generate_ip():
+    return ".".join(str(random.randint(0, 255)) for _ in range(4))
+
+# Function to generate random data with IP addresses and datacenter numbers
+def generate_data():
+    ip = generate_ip()
+    num = random.randint(1, 200)
+    line = f"{ip} {num}"
+    return line
+
+def main():
+    if len(sys.argv) != 2:
+        print(f"Usage: {sys.argv[0]} <number of data points>", file=sys.stderr)
+        sys.exit(1)
+
+    random.seed(42)
+    n = sys.argv[1]
+    n = int(n)
+    for _ in range(n):
+        print(generate_data())
+
+if __name__ == "__main__":
+    main()