[tor-bugs] #3233 [Metrics Utilities]: Make the VisiTor scripts accept more web log formats

[Tor Bug Tracker & Wiki]

#3233: Make the VisiTor scripts accept more web log formats
-------------------------------+--------------------------------------------
 Reporter:  karsten            |          Owner:     
     Type:  enhancement        |         Status:  new
 Priority:  minor              |      Milestone:     
Component:  Metrics Utilities  |        Version:     
 Keywords:                     |         Parent:     
   Points:                     |   Actualpoints:     
-------------------------------+--------------------------------------------
 The [https://metrics.torproject.org/tools.html#visitor VisiTor] scripts
 (both Java and Python) parse a web server log and the exit list archives
 to tell how many of the requests come from Tor users.  They currently
 require Apache's combined log format (`"%h %l %u %t \"%r\" %>s %b
 \"%{Referer}i\" \"%{User-agent}i\""`) as input.  If users have a different
 format, they need to transform their log to said format before passing it
 to VisiTor.

 We should also support Apache's common log format (`"%h %l %u %t \"%r\"
 %>s %b"`).  This format doesn't have the User-agent string and can
 therefore not detect Torbutton users.  But knowing whether users connect
 via the Tor network, with or without Torbutton, can still be a useful
 result.

 Are there further commonly used web log formats that we should support?

 Setting priority to minor, because I haven't heard of many VisiTor users
 lately.

-- 
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/3233>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online