"but this doesn't get us all the way" ... what doesn't defined? cover?
Jeremy Wood
···
On 9/9/20 1:27 PM, Sean Felipe Wolfe wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there to be *only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this doesn't get us all the way. So I'm considering monkeypatching Object like so:
classObject
defdefined_true?(const)
begin
returneval(const) == true
rescueNameError
returnfalse
end
end
end
Any thoughts ? Thanks
--
A musician must make music, an artist must paint, a poet must write, if he is to be ultimately at peace with himself.
- Abraham Maslow
On Wed, 9 Sep 2020, 19:27 Sean Felipe Wolfe, <ether.joe@gmail.com> wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there to be
*only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this doesn't get us
all the way. So I'm considering monkeypatching Object like so:
class Object
def defined_true?(const)
begin
return eval(const) == true
rescue NameError
return false
end
end
end
Any thoughts ? Thanks
--
A musician must make music, an artist must paint, a poet must write, if he
is to be ultimately at peace with himself.
- Abraham Maslow
On Wed, Sep 9, 2020 at 2:36 PM Jeremy Wood <jwood@pearlcomm.com> wrote:
Hi Sean,
"but this doesn't get us all the way" ... what doesn't defined? cover?
Jeremy Wood
On 9/9/20 1:27 PM, Sean Felipe Wolfe wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there to be
*only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this doesn't get us
all the way. So I'm considering monkeypatching Object like so:
class Object
def defined_true?(const)
begin
return eval(const) == true
rescue NameError
return false
end
end
end
Any thoughts ? Thanks
--
A musician must make music, an artist must paint, a poet must write, if he
is to be ultimately at peace with himself.
- Abraham Maslow
if I have TEST_MODE=false and use only defined? TEST_MODE, then I get a
true / non-false response from defined ?
I would like to say *only* TEST_MODE=true, or be able to delete the
reference entirely.
···
On Wed, Sep 9, 2020 at 11:36 AM Jeremy Wood <jwood@pearlcomm.com> wrote:
Hi Sean,
"but this doesn't get us all the way" ... what doesn't defined? cover?
Jeremy Wood
On 9/9/20 1:27 PM, Sean Felipe Wolfe wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there to be
*only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this doesn't get us
all the way. So I'm considering monkeypatching Object like so:
class Object
def defined_true?(const)
begin
return eval(const) == true
rescue NameError
return false
end
end
end
Any thoughts ? Thanks
--
A musician must make music, an artist must paint, a poet must write, if he
is to be ultimately at peace with himself.
- Abraham Maslow
That looks like it could work also. I didn't think about using the &&
operator. I would rather have something like defined_true? in my code
though rather than the && chain.
I'll dig a bit deeper to see what I can use instead of eval. Thanks Kevin ~~
···
On Wed, Sep 9, 2020 at 11:48 AM Kevin Deisz <kevin.deisz@gmail.com> wrote:
It is not entirely clear what you're doing, but avoid `eval`, there's
plenty of stdlib reflection that can help you here.
If I understand correctly, what you actually want is closer to:
On Wed, Sep 9, 2020 at 2:36 PM Jeremy Wood <jwood@pearlcomm.com> wrote:
Hi Sean,
"but this doesn't get us all the way" ... what doesn't defined? cover?
Jeremy Wood
On 9/9/20 1:27 PM, Sean Felipe Wolfe wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there to be
*only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this doesn't get us
all the way. So I'm considering monkeypatching Object like so:
class Object
def defined_true?(const)
begin
return eval(const) == true
rescue NameError
return false
end
end
end
Any thoughts ? Thanks
--
A musician must make music, an artist must paint, a poet must write, if
he is to be ultimately at peace with himself.
- Abraham Maslow
I could, but I'm really looking for syntactic sugar here for aesthetic
reasons, "beautiful code" reasons. Less characters, less to type, less
symbol characters.
Now if I just add defined_true? to my syntax highlighting I think that gets
me over
···
On Wed, Sep 9, 2020 at 11:48 AM Tom Lord <lord.thom@gmail.com> wrote:
Why not use an environment variable?
Or if you really must, a global variable.
On Wed, 9 Sep 2020, 19:27 Sean Felipe Wolfe, <ether.joe@gmail.com> wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there to be
*only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this doesn't get us
all the way. So I'm considering monkeypatching Object like so:
class Object
def defined_true?(const)
begin
return eval(const) == true
rescue NameError
return false
end
end
end
Any thoughts ? Thanks
--
A musician must make music, an artist must paint, a poet must write, if
he is to be ultimately at peace with himself.
- Abraham Maslow
Please forgive me if I am misunderstanding something.
In the event that TEST_MODE is undefined #defined? returns nil, which evaluates to false. In the case,
that TEST_MODE is defined #defined returns true. So you should be able to delete the value just fine. The only adjustment I see is that your check code would need to look like:
if(defined?(TEST_MODE) and TEST_MODE)
#do test code
else
#do production code
end
Here is demo script:
def check_debug_mode if(defined?(DEBUG_MODE))
puts"Yup" else puts"Nope" end end #first check - nothing defined check_debug_mode
DEBUG_MODE=true #second check - debug set check_debug_mode
#third check - DEBUG_MODE=nil # this will give a warning for redefining a constant, but other than that shouldn't effect the results DEBUG_MODE=nil check_debug_mode
Jeremy Wood
···
On 9/9/20 2:01 PM, Sean Felipe Wolfe wrote:
Hello Jeremy,
So I'm trying to avoid code like this:
#TEST_MODE=true
TEST_MODE=false
if I have TEST_MODE=false and use only defined? TEST_MODE, then I get a true / non-false response from defined ?
I would like to say *only* TEST_MODE=true, or be able to delete the reference entirely.
On Wed, Sep 9, 2020 at 11:36 AM Jeremy Wood <jwood@pearlcomm.com > <mailto:jwood@pearlcomm.com>> wrote:
Hi Sean,
"but this doesn't get us all the way" ... what doesn't defined?
cover?
Jeremy Wood
On 9/9/20 1:27 PM, Sean Felipe Wolfe wrote:
Hello Rubyists, I've been working with this scenario:
* I want to set a constant to true to enable test mode
* otherwise, if the constant does not appear, enable normal mode
Rather than have TEST_MODE=true and TEST_MODE=off, I want there
to be *only* TEST_MODE=true, or nothing at all.
The closest thing i can find is Object.defined? , but this
doesn't get us all the way. So I'm considering
monkeypatching Object like so:
classObject
defdefined_true?(const)
begin
returneval(const) == true
rescueNameError
returnfalse
end
end
end
Any thoughts ? Thanks
-- A musician must make music, an artist must paint, a poet must
write, if he is to be ultimately at peace with himself.
- Abraham Maslow
Ah, yes, what I didn't mention at the beginning. I am looking to do this
for potentially multiple constants. Trying to simplify the problem to
communicate it better to the group and ... oversimplified.
As I think the problem through though ... these test parameters are
actually going to potentially be on or off depending on parameters we call
the server with. So, having a single constant evaluated once upon startup
doesn't work anyhow. I guess I'll have to use a variable.
Thanks everyone for your input, at least I learned a bit more about
monkeypatching
···
On Wed, Sep 9, 2020 at 12:10 PM Jeremy Wood <jwood@pearlcomm.com> wrote:
Sean,
Please forgive me if I am misunderstanding something.
In the event that TEST_MODE is undefined #defined? returns nil, which
evaluates to false. In the case,
that TEST_MODE is defined #defined returns true. So you should be able to
delete the value just fine. The only adjustment I see is that your check
code would need to look like:
DEBUG_MODE=true#second check - debug setcheck_debug_mode #third check - DEBUG_MODE=nil# this will give a warning for redefining a constant, but other than that shouldn't effect the resultsDEBUG_MODE=nilcheck_debug_mode
That's cool, I learned something too, #defined? actually responds to a defined value with a string describing the value passed which will evaluate as true but is not the bool value true.
asdf='asdf'
defined?(asdf) #=> "local-variable"
ASDF='asdf'
defined?(ASDF) #=> 'constant'
def asdf;end
defined?(asdf) #=> "method"
Cheers!
···
On 9/9/20 2:28 PM, Sean Felipe Wolfe wrote:
On Wed, Sep 9, 2020 at 12:10 PM Jeremy Wood <jwood@pearlcomm.com > <mailto:jwood@pearlcomm.com>> wrote:
Sean,
Please forgive me if I am misunderstanding something.
In the event that TEST_MODE is undefined #defined? returns nil,
which evaluates to false. In the case,
that TEST_MODE is defined #defined returns true. So you should be
able to delete the value just fine. The only adjustment I see is
that your check code would need to look like:
if(defined?(TEST_MODE) and TEST_MODE)
#do test code
else
#do production code
end
Here is demo script:
def check_debug_mode if(defined?(DEBUG_MODE))
puts"Yup" else puts"Nope" end end #first check - nothing defined check_debug_mode
DEBUG_MODE=true #second check - debug set check_debug_mode
#third check - DEBUG_MODE=nil # this will give a warning for
redefining a constant, but other than that shouldn't effect the
results DEBUG_MODE=nil check_debug_mode
Jeremy Wood
Ah, yes, what I didn't mention at the beginning. I am looking to do this for potentially multiple constants. Trying to simplify the problem to communicate it better to the group and ... oversimplified.
As I think the problem through though ... these test parameters are actually going to potentially be on or off depending on parameters we call the server with. So, having a single constant evaluated once upon startup doesn't work anyhow. I guess I'll have to use a variable.
Thanks everyone for your input, at least I learned a bit more about monkeypatching
If you have multiple parameters that change at runtime, it might be simpler
to use a single global openstruct with the params as keys than multiple
variables.
martin
···
On Wed, Sep 9, 2020 at 12:29 PM Sean Felipe Wolfe <ether.joe@gmail.com> wrote:
On Wed, Sep 9, 2020 at 12:10 PM Jeremy Wood <jwood@pearlcomm.com> wrote:
Sean,
Please forgive me if I am misunderstanding something.
In the event that TEST_MODE is undefined #defined? returns nil, which
evaluates to false. In the case,
that TEST_MODE is defined #defined returns true. So you should be able to
delete the value just fine. The only adjustment I see is that your check
code would need to look like:
DEBUG_MODE=true#second check - debug setcheck_debug_mode #third check - DEBUG_MODE=nil# this will give a warning for redefining a constant, but other than that shouldn't effect the resultsDEBUG_MODE=nilcheck_debug_mode
Jeremy Wood
Ah, yes, what I didn't mention at the beginning. I am looking to do this
for potentially multiple constants. Trying to simplify the problem to
communicate it better to the group and ... oversimplified.
As I think the problem through though ... these test parameters are
actually going to potentially be on or off depending on parameters we call
the server with. So, having a single constant evaluated once upon startup
doesn't work anyhow. I guess I'll have to use a variable.
Thanks everyone for your input, at least I learned a bit more about
monkeypatching
I did some more digging and found out about const_get as well ...
···
On Wed, Sep 9, 2020 at 1:37 PM Martin DeMello <martindemello@gmail.com> wrote:
If you have multiple parameters that change at runtime, it might be
simpler to use a single global openstruct with the params as keys than
multiple variables.
