According to the docs...:

Option: botnet_clientwords
Space delimited list of regexps that are indicate an end client or
dynamic host which should not directly connect to other mail servers
besides its own provider's. Multiple entries are ORed together. Multiple
entries may be space delimited or made with multiple lines. Defaults
to empty (no client word check will be done). The example cf file comes
with a basic entry, however. The expressions will not match against the
top two domains (the TLD and usually the registed domain). All word
expressions have (\b|\d) added to the beginning and end, to ensure they
are not sub-words of larger words.

#####

I do not understand very well. Please allow me an example:

AAnnecy-256-1-92-172.w90-10.abo.wanadoo.fr

What is the term against which one the regex will test ?
Is it "AAnnecy-256-1-92-172.w90-10.abo." or perhaps
"AAnnecy-256-1-92-172.w90-10.abo" ?

If I'm writing
"\w+-\d{1,3}-\d{1,3}-\d{1,3}-\d{1,3}\.w\d{1,3}-\d{1,3}\.abo\."
will this really be prefixed and suffixed by "(\b|\d)" before applying
the test ? I do not understand why ? Explain me please !

Thanks a lot !
Claude