#!/usr/bin/env bash

# Nginx Popular Paths
# Just a simple script to mine Nginx logs to find the most popular paths.
#
# Usage:
# ./nginx-popular-paths.sh [-L|--logfile <file>] [-D|--days <n>] [-C|--count <n>]
#
# Author: Josh Sherman (joshtronic.com)
# Link: https://git.sherver.org/joshtronic/nginx-popular-paths
# License: MIT

# Argument defaults
LOGFILE='/var/log/nginx/access.log'
DAYS=30
COUNT=10

# Parse any arguments
while [[ "$#" -gt 0 ]]; do
  case "$1" in
    -C|--count)
      COUNT="$2"
      shift 2
      ;;
    -D|--days)
      DAYS="$2"
      shift 2
      ;;
    -L|--logfile)
      LOGFILE="$2"
      shift 2
      ;;
    -*)
      echo "Unknown option: $1" >&2
      exit 1
      ;;
    *)
      break
      ;;
  esac
done

# Stand up and tear down our temp file
TMPFILE=$(mktemp)
trap 'rm -f "$TMPFILE"' EXIT

# Start with the base logfile
[[ -f "$LOGFILE" ]] && cat "$LOGFILE" >> "$TMPFILE"

# Look for additional logfiles and archives
for i in $(seq 1 99); do
  RAWFILE="$LOGFILE.$i"
  GZFILE="$LOGFILE.$i.gz"

  if [[ -f "$RAWFILE" ]]; then
    cat "$RAWFILE" >> "$TMPFILE"
  elif [[ -f "$GZFILE" ]]; then
    zcat "$GZFILE" >> "$TMPFILE"
  else
    break
  fi
done

# DEBUG: Log the new file
cat $TMPFILE