debugging BlameJason mailtrainer.crm

Hi

As reported before Ger H's release barfs in mailtrainer.crm and I'm trying 
to debug this. I am a complete novice in crm114 coding, so please bare 
with me if the following is complete nonsense.

Lines 271/272 seem to be broken.


###########################################################
#
#       set gooddir and spamdir to the directory parts of the spec
#
match [:good:] (:gooddir: :goodwildcard:) /^.*[\\\/]([^\\\/]*)/
match [:spam:] (:spamdir: :spamwildcard:) /^.*[\\\/]([^\\\/]*)/


Here :gooddir: will match the whole string of :good: but we want just the
directory path.

I fix this:

###########################################################
#
#       set gooddir and spamdir to the directory parts of the spec
#
match [:good:] (:: :gooddir: :goodwildcard:) /^(.*[\\\/])([^\\\/]*)/
match [:spam:] (:: :spamdir: :spamwildcard:) /^(.*[\\\/])([^\\\/]*)/

Now it works as expected. But then I tried to run it from a small test
script with BillY binary (BlameTheSegfault) and it barfs!

/usr/bin/crm114-bill: *WARNING*
  This operand doesn't seem to end.  Bug?
 -->

I'll try to keep working.
This happened at line 5 of file ./bugtest.crm
The line was:
--> match [:good:] ( :: :gooddir: :goodwildcard:) /^(.*[\\\/])([^\\\/]*)/
<--

/usr/bin/crm114-bill: *ERROR*
 Regular Expression Compilation Problem: Missing ']'
 Sorry, but this program is very sick and probably should be killed off.
This happened at line 5 of file ./bugtest.crm
The line was:
--> match [:good:] ( :: :gooddir: :goodwildcard:) /^(.*[\\\/])([^\\\/]*)/
<--

Interesting enough it runs ok with both binaries when we change the order in
the regex:

###########################################################
#
#       set gooddir and spamdir to the directory parts of the spec
#
match [:good:] (:: :gooddir: :goodwildcard:) /^(.*[\/\\])([^\/\\]*)/
match [:spam:] (:: :spamdir: :spamwildcard:) /^(.*[\/\\])([^\/\\]*)/


so that's what I did. (Mumble: very odd, should work the other way round 
as well).

We're not done yet. Syscalled 'find' still barfs, because lines 300/311 will
append an '*' to overlapped :spamdir: or :gooddir: as well. Maybe that was
introduced by my changes. Anyway, let's isolate them. It now reads:

###########################################################
#
#       set gooddir and spamdir to the directory parts of the spec
#
match [:good:] (:: :gooddir: :goodwildcard:) /^(.*[\/\\])([^\/\\]*)/
match [:spam:] (:: :spamdir: :spamwildcard:) /^(.*[\/\\])([^\/\\]*)/
isolate (:gooddir:) /:*:gooddir:/
isolate (:spamdir:) /:*:spamdir:/

...and lines 300 / 311 have been amended to:

alter (:goodwildcard:) /'*'/

and

alter (:spamwildcard:) /'*'/

to make find happy.

No errors any more, but (sigh) it doesn't learn. Number of messages learnt
count up (one by one) but 'Total buckets in use' remains zero and I don't
see the familiar message 'LEARNED AND CACHED SPAM/GOOD' but instead:

X-CRM114-Version: 20080502-BlameJason ( TRE 0.7.5 (LGPL) ) MR-A8493A25
X-CRM114-CacheID: sfid-20080818_125940_574421_8940C2CE
Message-Id: (sfid-20080818_125940_574421_8940C2CE)
X-CRM114-Status: UNSURE (   0.00 )
X-CRM114-Notice: Please train this message.

Giving up for now awaiting your comments
Thomas Spahni

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/