#98 token.c Segmentation Fault

[Rooyama]

Bogofilter 1.1.1 died(Segmentation fault)by scanning
this file.
I researched and found it bug.
src/token.c line 355 was wrong.

-if (text[leng-1] == '>') {
+if (leng > 0 && text[leng-1] == '>') {

[Rooyama]

SPAMofSegFault

[David Relson]

Logged In: YES
user_id=30510

Your example works fine on my machine. However, the "id"
line looks strange. In the downloaded file, I see "id
+GDA46-9PZ'A@-S.". Bogofilter's parsing rule for QUEUE_ID
_requires_ angle brackets, so the id (as shown earlier)
cannot produce the problem you reported.

What locale are you in and what character encoding are you
using? Can you zip the message and email it to me (relson
at bogofilter dot org)??

Thanks.

David

[David Relson]

parser fix

[David Relson]

Logged In: YES
user_id=30510

File patch.1107.lexer_v3_l.txt modifies the parser's
definition of a QUEUE_ID to fix a segmentation fault on 64
bit processors.