1. Euphoria v4.03 bug in "append" ?
- Posted by alanjohnoxley Jan 22, 2012
- 1371 views
I discovered that missing the left hand side of a assignment where the right hand side was "append", does not give any error, it just continues...
I had a quick look in the bug index, but did not see this.
Is this "working as designed"?
Eui - v gives:
Euphoria Interpreter v4.0.3 development
Windows, Using Managed Memory
Revision Date: 2011-06-24 13:43:18, Id: 4985:26b65e77e3d9
-- Eu v4.0.3 issue with assignment missing in a append statement with warning with trace with type_check without profile procedure test() sequence s1, s2 s1 = {} s2 = {"dummy data"} append(s1,s2) -- which should have been -- s1 = append(s1,s2) end procedure test()
2. Re: Euphoria v4.03 bug in "append" ?
- Posted by mattlewis (admin) Jan 22, 2012
- 1350 views
I discovered that missing the left hand side of a assignment where the right hand side was "append", does not give any error, it just continues...
I had a quick look in the bug index, but did not see this.
Is this "working as designed"?
Yes. It is legal to ignore the return values of functions starting with Euphoria 4.0
Matt
3. Re: Euphoria v4.03 bug in "append" ?
- Posted by alanjohnoxley Jan 22, 2012
- 1343 views
Thanks Matt,
I did get caught for a while with this. I guess I will need to be more careful :)
Cheers,
Alan
4. Re: Euphoria v4.03 bug in "append" ?
- Posted by petelomax Jan 23, 2012
- 1275 views
Yes. It is legal to ignore the return values of functions starting with Euphoria 4.0
Matt
Of course I've known this for a while, but there are a large number of builtins for which this makes no sense:
open
allocate (/allocate_string)
call_back
sort (/custom_sort)
equal
compare
find
match
length
remainder
floor
where
rand
and_bits
or_bits
xor_bits
not_bits
time
power
log
instance
cos
sin
sqrt
tan
arctan
append
prepend
repeat
get_position
peek
peek4s
peek4u
define_c_func
define_c_proc
routine_id
platform
arccos
arcsin
define_c_var
open_dll
current_dir
peek_string
sprintf
sprint
command_line
date
db_table_list
reverse
get_screen_char
db_record_data
db_record_key
dir
getenv
upper
lower
peek2s
peek2u
The top four are pretty bad programming mistakes:
opening a file you cannot read, write, or close
allocating memory you cannot use or free
defining a call_back you cannot call
sorting a table then throwing the sorted result away and leaving the original unsorted
Pete
EDIT: Perhaps it should be illegal to ignore the result of a function call if that function has no side-effects.
5. Re: Euphoria v4.03 bug in "append" ?
- Posted by mattlewis (admin) Jan 23, 2012
- 1267 views
EDIT: Perhaps it should be illegal to ignore the result of a function call if that function has no side-effects.
I'm not sure I'd want it to be illegal, but a warning makes sense to me.
Matt
6. Re: Euphoria v4.03 bug in "append" ?
- Posted by alanjohnoxley Jan 26, 2012
- 1226 views
Yes, if there was a warning I would have saved some debugging time. :)
My actual program was much more complex of course, the supplied example was for simplicity's sake.
Looking at Pete's list, there is a lot of places for the issue to bite programmers.
Can this request be added to the "nice to have" list please?
Regards, Alan
7. Re: Euphoria v4.03 bug in "append" ?
- Posted by mattlewis (admin) Jan 26, 2012
- 1213 views
Yes, if there was a warning I would have saved some debugging time. :)
My actual program was much more complex of course, the supplied example was for simplicity's sake.
Looking at Pete's list, there is a lot of places for the issue to bite programmers.
Can this request be added to the "nice to have" list please?
Yes, it can. Please enter a ticket.
Matt
8. Re: Euphoria v4.03 bug in "append" ?
- Posted by alanjohnoxley Jan 26, 2012
- 1205 views
OK, ticket #743 opened.
Thanks!
