3 p.m.
Dear Allin,
I found a potential bug using multiple loops. Please take a look at the
following Hansl code:
<hansl>
open australia.gdt
loop i = 1..6
loop j = i+1..i+2
print "Hello World!"
endloop
endloop
</hansl>
In that case I get this error message:
"Erro de sintaxe na expressão genr" ("Syntax error in genr
expression")
I do not know if there is something wrong in my code, but I did not get
this error before Gretl. 1.9.10cvs. When I try to fix this error by putting
"$" or "()" inside the second loop Gretl crashes:
<hansl>
open australia.gdt
loop i = 1..6
loop j = $i+1..$i+2
print "Hello World!"
endloop
endloop
</hansl>
The most intriguing fact is that these problems disappear when I use Gretl
in English (instead of Brazilian Portuguese).
Best regards,
Henrique Andrade*
*
3:20 p.m.
On 11/15/2012 04:00 PM, Henrique Andrade wrote:
Dear Allin,
I found a potential bug using multiple loops. Please take a look at the
following Hansl code:
<hansl>
open australia.gdt
loop i = 1..6
loop j = i+1..i+2
print "Hello World!"
endloop
endloop
</hansl>
In that case I get this error message:
"Erro de sintaxe na expressão genr" ("Syntax error in genr
expression")
I do not know if there is something wrong in my code, but I did not get
this error before Gretl. 1.9.10cvs. When I try to fix this error by
putting "$" or "()" inside the second loop Gretl crashes:
<hansl>
open australia.gdt
loop i = 1..6
loop j = $i+1..$i+2
print "Hello World!"
endloop
endloop
</hansl>
The most intriguing fact is that these problems disappear when I use
Gretl in English (instead of Brazilian Portuguese).
Yes that's really weird...
I guess that Allin will spot the bug cause anyway, but out of curiosity,
what happens if you use 'loop j=(i+1)..(i+2)'?
cheers,
sven
3:30 p.m.
Em 15 de novembro de 2012, Sven Schreiber escreveu:
(...)
Yes that's really weird...
I guess that Allin will spot the bug cause anyway, but out of curiosity,
what happens if you use 'loop j=(i+1)..(i+2)'?
Dear Sven, thanks for your answer. When I use "loop j=(i+1)..(i+2)" the
same problem occurs: Gretl crashes.
Best regards,
Henrique Andrade*
*
3:35 p.m.
Am 15.11.2012 16:30, schrieb Henrique Andrade:
Em 15 de novembro de 2012, Sven Schreiber escreveu:
(...)
Yes that's really weird...
I guess that Allin will spot the bug cause anyway, but out of
curiosity,
what happens if you use 'loop j=(i+1)..(i+2)'?
Dear Sven, thanks for your answer. When I use "loop j=(i+1)..(i+2)" the
same problem occurs: Gretl crashes.
This is working:
<hansl>
loop i = 1..6 -q
loop j = i..(i+2) -q
print "Hello World!"
endloop
endloop
<hansl>
And in terms of 'the loop line' string-substitution generates also an error:
(my challenge solution of the interaction stuff)
<hansl>
loop for i=1.. nelem(Flist) -q
scalar start2=i+1
loop for j=start2.. nelem(Flist) -q
loop foreach k Dum_$i -q # Message: Dum_ unknown
instead of Dum_1 unknown
<hansl>
Best regards,
Henrique Andrade*
*
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
5:56 p.m.
On Thu, 15 Nov 2012, Henrique Andrade wrote:
I found a potential bug using multiple loops. Please take a look at
the
following Hansl code:
<hansl>
open australia.gdt
loop i = 1..6
loop j = i+1..i+2
print "Hello World!"
endloop
endloop
</hansl>
In that case I get this error message:
"Erro de sintaxe na expressão genr" ("Syntax error in genr
expression")
Thanks for the report. This is now fixed in CVS.
Allin
6:11 p.m.
On Thu, 15 Nov 2012, Henrique Andrade wrote:
Em 15 de novembro de 2012, Sven Schreiber escreveu:
(...)
>
> Yes that's really weird...
>
> I guess that Allin will spot the bug cause anyway, but out of curiosity,
> what happens if you use 'loop j=(i+1)..(i+2)'?
Dear Sven, thanks for your answer. When I use "loop j=(i+1)..(i+2)" the
same problem occurs: Gretl crashes.
Do you mean a true crash (segmentation fault)? I haven't been
able to reproduce that, even before fixing the bug in
question.
Allin
6:39 p.m.
Am 15.11.2012 19:11, schrieb Allin Cottrell:
On Thu, 15 Nov 2012, Henrique Andrade wrote:
> Em 15 de novembro de 2012, Sven Schreiber escreveu:
>
> (...)
>> Yes that's really weird...
>>
>> I guess that Allin will spot the bug cause anyway, but out of curiosity,
>> what happens if you use 'loop j=(i+1)..(i+2)'?
> Dear Sven, thanks for your answer. When I use "loop j=(i+1)..(i+2)" the
> same problem occurs: Gretl crashes.
Do you mean a true crash (segmentation fault)? I haven't been
able to reproduce that, even before fixing the bug in
question.
Hi Allin,
it's an 'implicit crash' (at least on windows). After using the 'loop
j=(i+1)..(i+2)' GRETL does only prints new script commands, but does not
process them anymore.
Allin
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
6:42 p.m.
2012/11/15 Allin Cottrell <cottrell(a)wfu.edu>
On Thu, 15 Nov 2012, Henrique Andrade wrote:
> Em 15 de novembro de 2012, Sven Schreiber escreveu:
>
> (...)
>>
>> Yes that's really weird...
>>
>> I guess that Allin will spot the bug cause anyway, but out of curiosity,
>> what happens if you use 'loop j=(i+1)..(i+2)'?
>
> Dear Sven, thanks for your answer. When I use "loop j=(i+1)..(i+2)" the
> same problem occurs: Gretl crashes.
Do you mean a true crash (segmentation fault)? I haven't been
able to reproduce that, even before fixing the bug in
question.
Gretl unexpectedly quits (without any error message).
Best,
Henrique Andrade*
*
6:44 p.m.
On Thu, 15 Nov 2012, Pindar wrote:
And in terms of 'the loop line' string-substitution generates
also an error:
(my challenge solution of the interaction stuff)
True, up till now $-string substitution has not been available
in the line that starts a loop. This is now relaxed. In
current CVS both of the following should work OK, regardless
of the language in which gretl is running:
<hansl>
# Using numerical values
loop i=1..4 -q
printf "i=%d\n", i
loop j=i+1..i+2 -q
printf " j=%d\n", j
endloop
endloop
# Using string substitution
loop i=1..4 -q
printf "i=$i\n"
loop j=$i+1..$i+2 -q
printf " j=$j\n"
endloop
endloop
</hansl>
Allin Cottrell
6:50 p.m.
Am 15.11.2012 19:44, schrieb Allin Cottrell:
On Thu, 15 Nov 2012, Pindar wrote:
> And in terms of 'the loop line' string-substitution generates also an error:
> (my challenge solution of the interaction stuff)
True, up till now $-string substitution has not been available
in the line that starts a loop. This is now relaxed. In
current CVS both of the following should work OK, regardless
of the language in which gretl is running:
That's nice!
Cheers
Leon
<hansl>
# Using numerical values
loop i=1..4 -q
printf "i=%d\n", i
loop j=i+1..i+2 -q
printf " j=%d\n", j
endloop
endloop
# Using string substitution
loop i=1..4 -q
printf "i=$i\n"
loop j=$i+1..$i+2 -q
printf " j=$j\n"
endloop
endloop
</hansl>
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
6:53 p.m.
On Thu, 15 Nov 2012, Henrique Andrade wrote:
2012/11/15 Allin Cottrell <cottrell(a)wfu.edu>
> On Thu, 15 Nov 2012, Henrique Andrade wrote:
>
>> Em 15 de novembro de 2012, Sven Schreiber escreveu:
>>
>> (...)
>>>
>>> Yes that's really weird...
>>>
>>> I guess that Allin will spot the bug cause anyway, but out of curiosity,
>>> what happens if you use 'loop j=(i+1)..(i+2)'?
>>
>> Dear Sven, thanks for your answer. When I use "loop j=(i+1)..(i+2)"
the
>> same problem occurs: Gretl crashes.
>
> Do you mean a true crash (segmentation fault)? I haven't been
> able to reproduce that, even before fixing the bug in
> question.
Gretl unexpectedly quits (without any error message).
OK, that's a "true crash". Could you check that this doesn't
happen with current CVS/snapshot? Thanks!
Allin Cottrell
4178
days inactive
4178
days old
11 comments
4 participants
participants (4)
-
Allin Cottrell -
Henrique Andrade -
Pindar -
Sven Schreiber