You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Chris <cp...@embarqmail.com> on 2007/05/08 02:57:14 UTC

[Possible SPAM] Malformed UTF-8 character (unexpected non-continuation byte 0x00

The entire error line reads:

Malformed UTF-8 character (unexpected non-continuation byte 0x00, immediately 
after start byte 0xce) in pattern match (m//) 
at /var/lib/spamassassin/3.002000/72_sare_bml_post25x_cf_sare_sa-update_dostech_net/200506020000.cf, 
rule SARE_OBFUMONEY1, line 1. 

I saw the same thing earlier this weekend but passed it off to possibly 
something I didn't have configured right. 

-- 
Chris
KeyID 0xE372A7DA98E6705C

Re: [Possible SPAM] Re: [Possible SPAM] Malformed UTF-8 character (unexpectednon-continuation byte 0x00

Posted by "Daryl C. W. O'Shea" <sp...@dostech.ca>.
Chris wrote:

> Thanks Loren, I doubt then that this accounts for the strange sa-update run I 
> had this morning which I've posted the cron output here: 
> http://mediasafe.embarq.com/chris1948/Hosted/saupdate0507.tar.bz2

I don't see anything strange at all in that output.


> Yesterdays update was a normal one.

None of the channels you are using had new updates yesterday (or any for 
a while actually).


Daryl




[Possible SPAM] Re: [Possible SPAM] Malformed UTF-8 character (unexpectednon-continuation byte 0x00

Posted by Chris <cp...@embarqmail.com>.
On Monday 07 May 2007 8:51 pm, Loren Wilton wrote:
> > rule SARE_OBFUMONEY1, line 1.
> >
> > I saw the same thing earlier this weekend but passed it off to possibly
> > something I didn't have configured right.
>
> No.  Its a combination of a perl bug and a change in SA to allow rules in
> other than the ascii character set.
> Previous versions of SA had 'use bytes' in the rule processing, so the
> regexs containing things like /\xA8/i worked fine.  Without use bytes, the
> combination of high bit characters and an uncased regex fall intp a perl
> bug, and at the very least result in this error message.
>
> We have to rewrite a fairly large number of the regexes, which ain't
> trivial to do, as they are large and hard to read.  And none of us have
> managed to get the time to do this since it showed up as a problem.  I'm
> hoping aI can make some time in a few days to fix this.
>
>         Loren

Thanks Loren, I doubt then that this accounts for the strange sa-update run I 
had this morning which I've posted the cron output here: 
http://mediasafe.embarq.com/chris1948/Hosted/saupdate0507.tar.bz2

Todays update run had such oddities as:

[1536] dbg: config: using "/tmp/.spamassassin1536V11iKbtmp/doesnotexist" for 
sys rules pre files
[1536] dbg: config: using "/tmp/.spamassassin1536V11iKbtmp/doesnotexist" for 
default rules dir
[1536] dbg: config: using 
"/tmp/.spamassassin1536V11iKbtmp/doesnotexist/doesnotexist" for user prefs 
file
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::RelayCountry from @INC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::URIDNSBL from @INC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::Hashcash from @INC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::SPF from @INC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::DCC from @INC
[1536] dbg: dcc: local tests only, disabling DCC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::Pyzor from @INC
[1536] dbg: pyzor: local tests only, disabling Pyzor
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::Razor2 from @INC
[1536] dbg: razor2: local tests only, skipping Razor
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::SpamCop from @INC
[1536] dbg: reporter: local tests only, disabling SpamCop
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::AWL from @INC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::AutoLearnThreshold 
from @INC
[1536] dbg: plugin: loading Mail::SpamAssassin::Plugin::TextCat from @INC

Yesterdays update was a normal one.
 
-- 
Chris
KeyID 0xE372A7DA98E6705C

Re: [Possible SPAM] Malformed UTF-8 character (unexpectednon-continuation byte 0x00

Posted by Loren Wilton <lw...@earthlink.net>.
> rule SARE_OBFUMONEY1, line 1.

> I saw the same thing earlier this weekend but passed it off to possibly
> something I didn't have configured right.

No.  Its a combination of a perl bug and a change in SA to allow rules in 
other than the ascii character set.
Previous versions of SA had 'use bytes' in the rule processing, so the 
regexs containing things like /\xA8/i worked fine.  Without use bytes, the 
combination of high bit characters and an uncased regex fall intp a perl 
bug, and at the very least result in this error message.

We have to rewrite a fairly large number of the regexes, which ain't trivial 
to do, as they are large and hard to read.  And none of us have managed to 
get the time to do this since it showed up as a problem.  I'm hoping aI can 
make some time in a few days to fix this.

        Loren