#17 Bug in invalid number format

[lnx85]

There is a bug in the mel2.tcl. It crashes when it try
to execute a [ expr ... ] with an improper value. When
an argument passed is, for example, 08 instead of 8, it
is interpreted as an improper octal value and the tcl
shows an error.

A possible solution is to apply the following function
to the arguments passed to each [ expr ... ]

proc strip:zero { num } {
if { [ regexp -- {^0*([1-9]*[0-9]*)$} ${num} all num
] } {
return ${num}
} else {
return 0
}
}

There are 39 [ lindex $t 3 ] that must be replaced with
[ strip:zero [ lindex $t 3 ] ]

Sorry for my bad school english :D

    Lnx85

[R081N]

Logged In: YES
user_id=1268649

The same thing happened to me. I'm not sure why because
the script worked fine at first. If someone finds out why this
started happening all of a sudden could you please let me
know?

My fix was very similar to lnx85's fix. I replaced every
instance of:
expr [lindex $t 3] -1
to
expr [string trimleft [lindex $t 3] "0"] -1

[Razoola]

Logged In: YES
user_id=1181869

I have the same problem and tried the fix as you mention
above but it diddent work, I get a 'invalid command
name "strip:zero08" while executing'....

The script is no longer working until this bug mentioned
above is fixed for me.

Raz

[Nobody/Anonymous]

Logged In: NO

Just an update to say I tried the fix r081n mentions in this
thread also and that one worked for me.

Raz

[gulu-gulu]

Logged In: YES
user_id=1106665

just a crappy workaround:
you can start the script after removing the "settings.mel"-file ..
but then you have to login to your bot and set all the settings
again. they all work fine till you set ".mel rotation ". after that a
".rehash" is impossible, but script still works..

[Nobody/Anonymous]

Logged In: NO

Thank you for your note. When I insert the strip-zero
procedure and replace the [ lindex $t 3 ] code with the new
one I get following error:

[13:45] unable to convert date-time string "2005808T000000"
while executing
"clock scan "[lindex $t 4][ strip:zero [ lindex $t 3 ]
][lindex $t 2]T000000""
("1" arm line 3)
invoked from within
"switch -- $stamplogs {
0 {
foreach search [binds time] {
if {[string compare
[lindex $search 4] "::mel::stamptimer"] == 0} {
unbind..."
(procedure "::mel::checkstamptimer" line 4)
invoked from within
"::mel::checkstamptimer"
invoked from within
"if {[::mel::starttests]} {
putlog $::mel::starterror
die "Fatal error - Can not continue"
} else {
::mel::arrayinit
::mel::logfile datestamp all
..."
(file "scripts/mel2.tcl" line 1114)
invoked from within
"source scripts/mel2.tcl"
(file "eggdrop.conf" line 1070)

[lnx85]

Logged In: YES
user_id=1245390

I've made an error when posting the bug-fig, sorry :P

In the TCL you must replace all the [expr [lindex $t 3] with
[expr [strip:zero [lindex $t 3]] because only the [lindex $t
3] passed to each expr function is affected by the bug and
not all [lindex $t 3]..

Sorry for my mistake and for my bad school english :P

[Nobody/Anonymous]

Logged In: NO

It stopped working for me too! I have no ideia why! It was
sudden...The scritp worked fine for months, but suddendly
gave this error.
It needs urgent fix...
I'll try this Lnx85 fix now... hope it works!

[Nobody/Anonymous]

Logged In: NO

it works! ;D

[Nobody/Anonymous]

Logged In: NO

thx lnx85 after I was able to understand your school english
your bugfix worked for me :)

[Nobody/Anonymous]

Logged In: NO

i thank you very much lnx85
your bug fix works proper

Gomorrha
-----
gomorrha@schwarzes-leipzig.com
www.schwarzes-leipzig.net

[Anonymous]

Logged In: YES
user_id=848992

the fix by r081n worked for me.
Thanks

[Nobody/Anonymous]

Logged In: NO

I don't know much programming. Could you provide a new mel.tcl with this patch included?

[Nobody/Anonymous]

Coud somebody attach a file with the mel2.tcl code corrected?

[Zangune]

I replaced

expr [lindex $t 3] -1

with

expr [string trimleft [lindex $t 3] "0"] -1

and I will attach the modified file.