Cacher is the code snippet organizer for pro developers

#!/bin/bash
#
# Author: bwangel<bwangel.me@gmail.com>
# Date:Oct,10,2017 14:16

LOG_FILE="./2017_10_10uwsgi-webapp.log"

echo "" | awk '
END {
    FMT = "%-33s\t%-5s\t%-15s%-10s\n"
    printf(FMT, "url", "count", "total-time(ms)", "average-time(ms)")
}'

cat ${LOG_FILE} | egrep '^{address' | awk '

function get_timestamp(datetime_fmt) {
    split(datetime_fmt, datetime_group, " ")

    year = datetime_group[4]
    month = sprintf("%02d",(match("JanFebMarAprMayJunJulAugSepOctNovDec",datetime_group[1])+2)/3)
    day = datetime_group[2]

    split(datetime_group[3], time, ":")
    hour = time[1]
    minute = time[2]
    second = time[3]
    timestamp = mktime(sprintf("%s %s %s %s %s %s", year, month, day, hour, minute, second))

    return timestamp
}

BEGIN {
    total = 0
    max_timestamp = 0
    min_timestamp = 99999999999
}
{
    timestamp = get_timestamp(sprintf("%s %s %s %s", $22, $23, $24, substr($25, 1, length($25) - 1)))
    if (timestamp > max_timestamp) {
        max_timestamp = timestamp
    }
    if (timestamp < min_timestamp) {
        min_timestamp = timestamp
        # print(timestamp, NR)
    }

    split($27, query_url, "?")
    url_count[query_url[1]] += 1
    url_time[query_url[1]] += $33
    total += 1
}
END {
    FMT = "%-33s\t%-5s\t%-15s%-10s\n"
    for (url in url_count)
        printf(FMT, url, url_count[url], url_time[url], url_time[url] / url_count[url])

    print("日志开始时间:", strftime("%Y-%m-%d %H:%M:%S", min_timestamp))
    print("日志结束时间:", strftime("%Y-%m-%d %H:%M:%S", max_timestamp))
    print("total:", total)

}
' | sort -n -r -t'	' -k 2