Hash or bidimensional Array when key is case insensitive?

Hi, I’ve to take a decission about using an bidimensional Array or a
Hash to store parameters. Of course I prefer Hash since the order
doesn’t matter and I know the param name I’ll need.
The problem is that the param name is case insensitive.

Example:

tag = qweqweqwe
Method = INVITE
NAT=yes

is the same as:

Tag = qweqweqwe
METHOD = INVITE
Nat=yes

If I use an bidimensional Array and need to get “method” param I willl
need to do:

method = params_array.find { |param| param[0] =~ /^method$/i }[1]

And if I use a Hash I’ll do the same:

method = params_hash.find { |param| param[0] =~ /^method$/i }[1]

Is there any advantage using Array or Hash in my case (key is case
insensitive)?

Thanks a lot for any suggestion.

Iñaki Baz C. wrote:

And if I use a Hash I’ll do the same:

method = params_hash.find { |param| param[0] =~ /^method$/i }[1]

Use a hash and lowercase the keys. If you need to preserve the original
casing, do that in the value, e.g. like this: {key.lowercase => [key,
value], …}
Another approach would be a special Key class which is like String but
has #hash and #eql? redefined to work as if the String was lowercased.
There probably are more approaches.

Regards
Stefan

On Tue, Apr 15, 2008 at 6:31 PM, Iñaki Baz C. [email protected] wrote:

And if I use a Hash I’ll do the same:

method = params_hash.find { |param| param[0] =~ /^method$/i }[1]

Is there any advantage using Array or Hash in my case (key is case insensitive)?

Another approach is to store it in a hash, but downcase the key before
storing and retreiving:

irb(main):001:0> h = {“get” => “value_for_get”, “post” =>
“value_for_post”}
=> {“get”=>“value_for_get”, “post”=>“value_for_post”}
irb(main):002:0> method = “gEt”
=> “gEt”
irb(main):003:0> h[method.downcase]
=> “value_for_get”

Jesus.

El Martes, 15 de Abril de 2008, Stefan R. escribió:

Iñaki Baz C. wrote:

And if I use a Hash I’ll do the same:

method = params_hash.find { |param| param[0] =~ /^method$/i }[1]

Use a hash and lowercase the keys. If you need to preserve the original
casing, do that in the value, e.g. like this: {key.lowercase => [key,
value], …}

Ah ok, it’s a good idea :slight_smile:

Another approach would be a special Key class which is like String but
has #hash and #eql? redefined to work as if the String was lowercased.

Thanks but I don’t understnad this point. Do you mean extending the Hash
class? I don’t think so since there is not any Key class. Could you
please
explain in more detail you suggestion?

Thanks a lot for all.

El Martes, 15 de Abril de 2008, Jesús Gabriel y
Galán escribió:> Another approach is to store it in a hash, but downcase the key before

storing and retreiving:

irb(main):001:0> h = {“get” => “value_for_get”, “post” => “value_for_post”}
=> {“get”=>“value_for_get”, “post”=>“value_for_post”}
irb(main):002:0> method = “gEt”
=> “gEt”
irb(main):003:0> h[method.downcase]
=> “value_for_get”

Thanks, but the problem is exactly the opposite: I receive a request
that can
contain parameter name in low or uppercase, and I need to process them
but
keeping them untouched to forward the request to other location.

Thanks a lot.

On Tue, Apr 15, 2008 at 10:50 PM, Iñaki Baz C. [email protected] wrote:

Thanks, but the problem is exactly the opposite: I receive a request that can
contain parameter name in low or uppercase, and I need to process them but
keeping them untouched to forward the request to other location.

If I understand you correctly, in my snippet above, the variable
method will contain what you receive. You downcase it for searching in
the hash, but still have the original string to pass it along. Am I
right? Maybe I am getting the order of your stuff wrong. Can you post
an example of what you have?

Jesus.

El Martes, 15 de Abril de 2008, Jesús Gabriel y
Galán escribió:

attr_reader :key
end
end

and then you populate the hash as:

h[Key.new(method)] = [method, values_for_method]

this way you locate the entry in the hash ignoring case and have the
original value in the value of the hash.

Hope I’ve explained myself…

Sure, it’s cool :slight_smile:

Thanks a lot to both.

On Tue, Apr 15, 2008 at 11:27 PM, Iñaki Baz C. [email protected] wrote:

Ah ok, it’s a good idea :slight_smile:
Hey, now I got your requirements !! Please ignore my other post…

Another approach would be a special Key class which is like String but
has #hash and #eql? redefined to work as if the String was lowercased.

Thanks but I don’t understnad this point. Do you mean extending the Hash
class? I don’t think so since there is not any Key class. Could you please
explain in more detail you suggestion?

What he means is that instead of lowercasing the key before inserting
in the hash, you create a Key class in a way that the hash will think
two keys are the same if they have the same string ignoring case.
Something like (not tested):

class Key
attr_reader :key
def initialize key
@key = key
end

def hash
@key.downcase.hash
end

def eql? other
@key.downcase.eql? other.downcase
end
end

and then you populate the hash as:

h[Key.new(method)] = [method, values_for_method]

this way you locate the entry in the hash ignoring case and have the
original value in the value of the hash.

Hope I’ve explained myself…

Jesus.

This forum is not affiliated to the Ruby language, Ruby on Rails framework, nor any Ruby applications discussed here.

| Privacy Policy | Terms of Service | Remote Ruby Jobs