X-Bogosity header?

Mon May 5 15:58:57 CEST 2003

Hi.  Why does the location of the X-Bogosity header have such a dramatic
impact on bogolexer?  In x1 it is the last header, but in x0 it has been
swapped with the header before it.  Otherwise the files are identical.
Yet, bogolexer finds over 3 times the number of tokens in x1.

Thanks
Fletcher

Script started on Fri May 02 13:22:19 2003

cs.utexas.edu$ bogolexer -q <x0
quiet mode.
52 tokens read.

cs.utexas.edu$ bogolexer -q <x1
quiet mode.
182 tokens read.

cs.utexas.edu$ diff -u x0 x1

--- x0	2003-05-02 13:16:38.673721000 -0500
+++ x1	2003-05-02 13:16:58.121721000 -0500
@@ -12,8 +12,8 @@
 Content-type: text/html;charset=big5
 Message-Id: <20030501051711.49DD61CB58 at libra>
 Date: Wed, 30 Apr 2003 22:17:11 -0700 (PDT)
-X-Bogosity: Unsure, tests=bogofilter, spamicity=0.946771, version=0.12.2
 Status: OR
+X-Bogosity: Unsure, tests=bogofilter, spamicity=0.946771, version=0.12.2
 
 <body bgcolor="#FFFFFF">
 <center>
cs.utexas.edu$ exit
exit

script done on Fri May 02 13:22:51 2003




