grep masters wanted

[plexer]

I'm trying to create a plugin for nagios to get the consumables status for our Konica K5440dl printers.

 

I tried a snmpwalk on one and it only goes so far so I was hoping to try and copy the HP 2600 one which uses wget to download the supplies status webpage to a temporary file and then extracts the details from that.

 

	
	
	
	
	<br />
		 <!--<br />
		 a { color: #000000; text-decoration:none}<br />
		 body  { background-color: #FFFFFF }<br />
<br />
		 .menu     { background-color: #6666FF; color: #FFFFFF }<br />
		 .menu2    { background-color: #9999FF; color: #FFFFFF }<br />
		 .menu3    { background-color: #CCCCFF; color: #000000 }<br />
		 .bar      { background-color: #6666FF }<br />
		 --><br />
		 
	System - Detail - Consumable
	
	
	 </pre><table width="760" cellpadding="0" cellspacing="2">
	  
	   space.gif
	  
	  
	   
		
		 space.gif
		
		
		 
		  
		   head_s2.gif
		   Summary
		  
		 
		 
		  
		   head_a1.gif
		   Detail
		  
		 
   
		 
		  
		   ./space.gifhead_s2.gif
		   Input Tray
		  
		 

		 
		  	
		   ./space.gifhead_s2.gif
		   Network
		  
		 
		 
		  
		   ./space.gifhead_a2.gif
		   Consumables
		  
		 
		 
		  
		   head_s2.gif
		   Counter
		  
		 
		 
		  
		   head_s2.gif
		   Online Assistance
		  
		 
		

		 
		  space.gif
		 
		
		
		space.gif
	   
		
		 
		  
		  
		   
			
			 Consumables
			 
			  
Status
			 
			 
			 
Max Life
			

			 Black Toner Cartridge
			 
			  
				
				  
					
					  
						
					  
					
				  
				
			  
			 
			 53%
			  12000 
			

			 Cyan Toner Cartridge
			 
			  
				
				  
					
					  
						
					  
					
				  
				
			  
			 
			 1%
			 12000
			

			 Magenta Toner Cartridge
			 
			  

				
				  
					
					  
						
					  
					
				  
				
			  
			 
			 100%
			 12000
			

			 Yellow Toner Cartridge
			 
			  
				
				  
					
					  
						
					  
					
				  
				
			  
			 
			 100%
			 12000
			

			 Transfer Roller Unit
			 Ready
			 
			 
			  -
			

			 Transfer Belt Unit
			 Ready
			 
			 -
			

			 Waste Toner Bottle
			 Ready
			 
			 -
			

		   

		  
		 
		
	   
	  
	  
	   space.gif
	  
	 </table><br>	<br

 

Thats what the downloaded page looks like.

 

If anyone could come up with an idea to split that into values for each toner that would be cool.

 

Ben

[RabbieBurns]

wget http://blah.html && cat blah.html | grep % > output.txt

 

will put the following into an output.txt file.

 

53
1%
100%
100%

 

Not sure how to strip the html out though, will have a play

[plexer]

I then need to do a split on those lines then I suppose?

 

Ben

[CyberNerd]

pipe the output to awk, then $1, $2, $3 etc to get each column

awk '{ print $1, $2 }'

 

so

wget http://blah.html && cat blah.html | grep % | awk '{ print $1, $2 }'

[RabbieBurns]

pipe the output to awk, then $1, $2, $3 etc to get each column

awk '{ print $1, $2 }'

 

so

wget http://blah.html && cat blah.html | grep % | awk '{ print $1, $2 }'

 

that just displays

 

 
changing it to just $5 gives
 
[code]size="2">53%
size="2">1%
size="2">100%
size="2">100%

 

which is nearly there

[RabbieBurns]

wget http://temp.html && cat temp.html | grep % | awk '{ print $5 }' > test.txt && sed -e 's/size="2">/ /' test.txt > test2.txt

 

strips out the first bit of the html, but I cant get sed to remove the latter part. Not sure if its to do with the \ in the html tag

 

 53%
1%
100%
100%

[CyberNerd]

that just displays

 

 

Sure, I was just giving an example
 
 
how about replacing the < with whitespace, then running awk again
[code]sed 's/ /g' test.txt | awk '{print $1}' 

[plexer]

Thanks for all the suggestions guys looking good.

 

I just want to end up with some variables that contain the cartridge colour and the corresponding % or just % as long as I know variable1 is black, 2 is cyan I suppose.

 

Ben

[tom_newton]

I would suggest learning a bit of perl.. much easier.. then use a few match lines.. but...

 

try:

wget | grep % | cut -d\> -f3 | cut -d\< -f1

 

If you really want it doing nicely I will perl it for you and post here as an example?

[plexer]

Yes please Tom then once I can get the check working I'll add that.

 

Ben

[RabbieBurns]

Cracked it.

 

Probably not the cleverest or neatest way, but it works.

 

wget http://temp.html && cat temp.html | grep % | awk -F\> '{print $3}' | awk -F\< '{print $1}' 

 

add > filename.txt to output to a file.

 

Will list 1 per line, black, cyan, magenta, yellow.

 

Saving to: `temp.html'

100%[============================================>] 6,315   in 0.04s   

2008-10-16 18:26:37 (167 KB/s) - `temp.html' saved [6315/6315]

53%
1%
100%
100%

[RabbieBurns]

try:

wget | grep % | cut -d\> -f3 | cut -d\< -f1

 

works when you change it to

 

wget && cat | grep % | cut -d\> -f3 | cut -d\< -f1

[tom_newton]

works when you change it to

 

wget && cat | grep % | cut -d\> -f3 | cut -d\< -f1

 

Actually whatya need is wget -O - | grep %.....

 

no need for temp files

I was assuming plexer had the "file grabby bit" down already.

[tom_newton]

tom@white-elephant:/tmp$ cat plex.txt | ./plex.pl

Black: 53

Cyan: 1

Magenta: 100

Yellow: 100

 

where "plex.txt" is your HTML.

 

Could be done on the commandline but this is neater IMO.

 

Next step: remove wget, and go with LWP::Simple

plex.pl.gz

[plexer]

Thanks Tom.

 

I can't seem to open that it just comes out gobbledegook when I un compress it.

 

Ben

[CyberNerd]

Thanks Tom.

 

I can't seem to open that it just comes out gobbledegook when I un compress it.

 

Ben

 

It's perl, what did you expect

[plexer]

Haha not in that way

 

Ben

[tom_newton]

tom@white-elephant:/tmp$ zcat plex.pl.gz

#!/usr/bin/perl -Tw

#

 

use strict;

 

my $colour = "Polka Dot";

my $level = 0;

 

while(<>)

{

if (m/

(\w+) Toner Cartridge<\/td>/)

{

$colour = $1;

}

if (m/

(\d+)%<\/font><\/td>/)

{

$level = $1;

print "$colour: $level\n";

}

}

tom@white-elephant:/tmp$

[plexer]

That works for me.

 

Now to do the rest of the check

 

Ben

[plexer]

Ok here's a working script, many thanks to Tom for sorting this out for me.

 

Ben

k5440.txt

[tom_newton]

Looks like you've got it working well Ben - thanks for reposting the perl we worked on via email.. hope it helps someone else

 

Now go get "Learning Perl" (O'Reilly) it's good, it is.

[plexer]

Yeah the only thing I've commented out is trying to get it to return a value as well as printing the output to stdout as it does now.

 

Ben

[tom_newton]

Yeah the only thing I've commented out is trying to get it to return a value as well as printing the output to stdout as it does now.

 

Ben

 

Did you not get that to work? TBH I got as far as working out what you were trying to do, but didn't check the code

[plexer]

nagios takes the single line of text from stdout as a return but I'd seen some perl plugins that also returned the value as well.

 

Such as exit $state;

 

Allthough it may not be needed in my case as it works as we have it now.

 

Just need to do add in some error checking like you mentioned becuase if I run it against a non k5440 printer for instance I presume it will baulk.

 

Ben

[tom_newton]

Would have thought so - though you might adapt it to detect the type of printer and find the correct URL automatically