60 lines
		
	
	
		
			2.4 KiB
		
	
	
	
		
			Plaintext
		
	
	
	
			
		
		
	
	
			60 lines
		
	
	
		
			2.4 KiB
		
	
	
	
		
			Plaintext
		
	
	
	
| Here are some CGI scripts I was using to view and compare RTS style output.
 | |
| It will probably need a little more tweaking to be useful to LTP.
 | |
| 
 | |
| Expectations:
 | |
| 
 | |
|   A directory that contains all output.  It expects file names in the format
 | |
| 
 | |
|   <host>.<ISO Date>.<suite>.<type>
 | |
| 
 | |
|  	<host>		The hostname the tests were run on
 | |
| 	<ISO Date>	ISO standard format date. i.e. YYYYmmDDHHMM
 | |
| 	<suite>		The name of the pan file that was used
 | |
| 	<type>		One of driver, scanner, or summary
 | |
| 
 | |
| 			driver - the raw output from pan with <<<test_end>>>
 | |
| 			scanner - the output from scanner
 | |
|             summary - a very brief table listing how many tests passed,
 | |
|                       failed, didn't run, etc.  This wasn't released.
 | |
| 
 | |
| Scripts:
 | |
| 
 | |
|     results.cgi
 | |
| 
 | |
|         This lists out all of the results that are the results directory.  It
 | |
|         provides links to the driver output, scanned results, and summary for
 | |
|         each suite.  The sort order is host, date, suite.  On the results
 | |
|         page, which looks at the .scanner files, there are links that extract
 | |
|         the test tag output from the .driver file.
 | |
| 
 | |
|     browse.cgi
 | |
| 
 | |
|         This is a more complex form that I started working on.  It allows you
 | |
|         to compare as many results as you want, side by side.  Also, I started
 | |
|         working on sorting the results different ways, but didn't get too far.
 | |
|         The side by side comparison is done with in reconsile.cgi
 | |
| 
 | |
|     reconsile.cgi
 | |
| 
 | |
|         This script compares multiple scanner files and shows the differences
 | |
|         side by side in a table.  It expects to find .scanner files for the
 | |
|         results it is comparing.
 | |
| 
 | |
| Notes:
 | |
| 
 | |
|     The driver I was using with these scripts collects some system information
 | |
|     before running the tests.  I use this information to display the `uname
 | |
|     -a` output in browse.cgi and reconsile.cgi.  You will be missing this
 | |
|     information, but the scripts should still run.
 | |
| 
 | |
|     I apologize in advance for the use of Perl.  I've managed to steer clear
 | |
|     of Perl for five years until I started writing these scripts.  IMHO, the
 | |
|     people who learn programming with Perl write the ugliest code.  Luckily, I
 | |
|     learned programming mostly with Pascal.  Hopefully the code makes sense.
 | |
| 
 | |
| 
 | |
| --
 | |
| Nate Straz                                              nstraz@sgi.com
 | |
| sgi, inc                                           http://www.sgi.com/
 | |
| Linux Test Project                                  http://ltp.sf.net/
 |