PW Log Viewer

[DaMadBoy]

I have been trying to use the PWlogviewer by Kenpachi but infortunately it doesn't show any information when I load the world2.log into it.

I've tried running the program as adminstrator with no luck and I also tried to get it working using capability mode and again with no luck.

Anyone got any suggestions on how to get it working?

[MentaL]

[Romulan]

I think it reads from formatlog and not world2.log

[DaMadBoy]

Just gave the formatlog a go but it still didn't read it. I'll have a go at opening all the logs, mabe I'll get lucky.

Gave it a try with the other world2 files but again none of them opened.

[das7002]

It's just plain text... Open it in Notepad or something...

[DaMadBoy]

I have O.o but was hoping to use the log viewer as it puts it all in a nice neat order.

[das7002]

I have O.o but was hoping to use the log viewer as it puts it all in a nice neat order.

So does Notepad++ because of it's nifty tabs

[Romulan]

I agree that notepad is great but not as a log viewer. Try searching some values within a log file of 150mb ... It will take a while and maybe notepad will crash.

Btw better to do a log viewer in php or java. So you can read the log directly on the server via your browser without downloading everytime the log file to check something.

[das7002]

I agree that notepad is great but not as a log viewer. Try searching some values within a log file of 150mb ... It will take a while and maybe notepad will crash.

Btw better to do a log viewer in php or java. So you can read the log directly on the server via your browser without downloading everytime the log file to check something.

Then use nano and Ctrl-W (which is search in nano)

[Romulan]

Command line is not very convenient for editing/viewing large text files. I still believe in doing a php web panel for every log files would be cool...

PS: I sucks with stripos and substr in php, that's why I can't do my own panel for that -.-

[DaMadBoy]

A web based log viewer would be very handy actually, especially for those using dedicated servers instead of home servers. Do you think anyone would put one together?

I would do myself but I know nothing about php at this time, although I plan to learn it at some point, once I'm done learning what need to for my server.

[tbnanubis]

anyone ever considered using grep, awk/sed and pipes on the server console? :D its much faster than any viewer because you dont have to download the full log, but a small snippet that is of interest.

[DaMadBoy]

I'm still a complete novice when it comes to linux programs but now you've mentioned it, it's something worth learning.

[Romulan]

I found this script made by Robb for his pvp ranking:

#!/bin/sh

cd /PWServer/logs
grep -e ':type=2:' world2.formatlog > kills.formatlog
chmod 755 kills.formatlog
sed -i 's/ /,/g' kills.formatlog
sed -i 's/=/,/g' kills.formatlog
sed -i 's/:/,/g' kills.formatlog
sed -i 's/,,,/,/g' kills.formatlog
sed -i 's/,,/,/g' kills.formatlog

I know what does the 3 first lines but wtf does the others lines? xD

What is the meaning of: sed -i 's/ /,/g' ?

Oh here is a sample of the line in world2.formatlog:

2011-10-15 18:42:28 pwserver gamed: notice : formatlog:die:roleid=32:type=2:attacker=48

Maybe it will help you to find out what sed does on that line.

[tbnanubis]

in-place edit
-i means instead of taking streams/files as input and output, only one file is read and changed

s/y/x/g is a replace function, which is used to replace all occurences of y with x
it supports regular expressions

what the script does is replace delimiter chars like " ", "=", ":" with "," and then delete all multiple occurences

when that is done you can use simple line.split(",") with any program and select the array index you want without any further processing to display the data fields in a page or program

In short sed is a stream editor which means it ignore line breaks...
when you want to do per-line-manipulation you can use line based editors like awk scripts, which are also very powerful, but limited to whole lines..

when viewing logs for exploit/cheat discovery or something that you server rules don't want to happen, its easier to just grep a bit of info directly to a file and look into it, like:

grep "2011-10-23 16:1" > x.txt

to get everything that has happened within 10 minutes, or use

grep -e "2011-10-23.*task:roleid=1024" world2.log

to get all quest stuff a certain character has done in one day... feel free to use

grep -e "2011-10-23.*task:roleid=1024" world2.log|last -n x> test.txt

to only get the last x actions/lines

the possibilities are countless as most unix tools support regular expressions in some way


Edit: read pages like this when you want to do something special and after that you should know what stuff to google to get all information you need....

Just keep in mind the internet-knowledge-gather-rule: search engines are you friend, use them whenever possible before asking anything anywhere

[das7002]

Spoiler:

in-place edit
-i means instead of taking streams/files as input and output, only one file is read and changed

s/y/x/g is a replace function, which is used to replace all occurences of y with x
it supports regular expressions

what the script does is replace delimiter chars like " ", "=", ":" with "," and then delete all multiple occurences

when that is done you can use simple line.split(",") with any program and select the array index you want without any further processing to display the data fields in a page or program

In short sed is a stream editor which means it ignore line breaks...
when you want to do per-line-manipulation you can use line based editors like awk scripts, which are also very powerful, but limited to whole lines..

when viewing logs for exploit/cheat discovery or something that you server rules don't want to happen, its easier to just grep a bit of info directly to a file and look into it, like:

grep "2011-10-23 16:1" > x.txt

to get everything that has happened within 10 minutes, or use

grep -e "2011-10-23.*task:roleid=1024" world2.log

to get all quest stuff a certain character has done in one day... feel free to use

grep -e "2011-10-23.*task:roleid=1024" world2.log|last -n x> test.txt

to only get the last x actions/lines

the possibilities are countless as most unix tools support regular expressions in some way


Edit: read pages like this when you want to do something special and after that you should know what stuff to google to get all information you need....

Just keep in mind the internet-knowledge-gather-rule: search engines are you friend, use them whenever possible before asking anything anywhere

All commands | commandlinefu.com is a great resource for useful little tidbits. I've also found sed useful for so much... I spend quite a bit of time in command line and to me it is just easier to explain to computer exactly what I want to do than figure out what the developer of an application thought would be a good place to put something in a GUI