On 9 Nov 2017, at 11:15 (-0500), Emanuel wrote:
Hello,
I am working in a server with spamassassin 3.4.1, i need see the bayes
in a legible text
This is not possible, as others have already explained. The 5th column
is a one-way hash of each token, so you can't regenerate the original
tokens or even be absolutely certain that each hash represents only one
token.
The command sa-learn --dump all, show this info:
0.987 1 0 1491158923 936e45469a
0.987 1 0 1490742234 996c4b779f
0.997 5 0 1510240620 99bbb1343d
0.005 0 3 1510242056 9a19089c49
0.016 0 1 1509728119 a14e640e1d
0.987 1 0 1492521796 a3b7843c4a
0.987 1 0 1491153511 a4b1ec7417
0.995 3 0 1510164226 aac1930026
0.033 245 2686 1510242162 abee185aa7
0.993 2 0 1510241704 aec056d3b9
0.987 1 0 1490534092 af315bd372
0.005 0 3 1510237626 af79a72241
0.995 3 0 1510230300 b036d8c25c
0.987 1 0 1507076906 b2c21ed6b8
That's not the useful part of the dump. The useful part is the "magic"
metadata.
I am interested in seeing the bayes info in the database, because it
was created years ago
Not really relevant. The 4th column is the last time (in Unix Epoch
seconds) that token hash was seen. In the list above, the dates are all
between March 26th and November 9th of this year. By default, the Bayes
DB is limited to 150K tokens (bayes_expiry_max_db_size) and when it hits
that limit, SpamAssassin expires the oldest 37.5K tokens (25% of 150K.)
The 'magic' part of the dump will show the actual oldest and newest
'seen' times as well as information about how much spam and ham has been
learned and expiry information.
If you want to check tokens that match a particular regular expression,
you can try 'sa-learn --dump data --regexp <RE>' but to be candid, I'm
not sure how or whether that actually works, since I can't get anything
longer than an anchored 2-letter RE to match.
--
Bill Cole
b...@scconsult.com or billc...@apache.org
(AKA @grumpybozo and many *@billmail.scconsult.com addresses)
Currently Seeking Steady Work: https://linkedin.com/in/billcole