Cloning arrays

[Len Lawrence]

Another newbie question. I have started recoding some of my homegrown
Tcl/Tk applications in Ruby/rubytk and have hit an unexpected snag when
attempting to clone arrays. The following code is snipped verbatim from
the program:

# Calculate coordinates for the Postscript page canvas and the
# label icons canvas in the setupInkjet window.

def User.calcxy ( disposition, f )

xygrid = { }

# Split the disposition string into x and y counts.
xy = Typefaces:isposition[disposition]
# Obtain the label dimensions.
labelsize = Typefaces::Labelsizes[disposition]
# Populate the coordinate data structure.
xygrid['x'] = nx = xy[0]
xygrid['y'] = ny = xy[1]
xygrid['dx'] = dx = labelsize[0] * f
xygrid['dy'] = dy = labelsize[1] * f
# Obtain an array of 4 element arrays from the Typefaces module.
# These represent opposing corners of the address spaces on the
# Postscript canvas.
xygrid['page'] = Typefaces.labelgrid( disposition )
puts 'page coordinates'
puts xygrid['page'][0]
puts '______________'
# Copy the page label coordinates.
e = xygrid['page'].clone
# Shrink the label icon coordinates.
e.each { |z| z.collect! { |d| d *= f } }
# Shrink the icons further to make separations visible.
e.each { |z| z[2] -= 2.0; z[3] -= 2.0 }
# Save the label icon coordinates.
xygrid['boxes'] = e
puts 'page again'
puts xygrid['page'][0]
puts 'boxes'
puts e[0]
puts '=============='
xygrid['max'] = nx * ny
return xygrid

end

The diagnostics at 'page again' and 'boxes' show that the copy of the
original array has been successfully processed but the original array now
reflects the same values. The 'page' and 'boxes' arrays are to be used in
different circumstances.

What am I missing? It is bound to be something simple or something
fundamental. I tried Object#dup also - same result.

Len

 

[Robert Klemme]

Another newbie question. I have started recoding some of my homegrown
Tcl/Tk applications in Ruby/rubytk and have hit an unexpected snag when
attempting to clone arrays. The following code is snipped verbatim from
the program:

# Calculate coordinates for the Postscript page canvas and the
# label icons canvas in the setupInkjet window.

def User.calcxy ( disposition, f )

xygrid = { }

# Split the disposition string into x and y counts.
xy = Typefaces:isposition[disposition]
# Obtain the label dimensions.
labelsize = Typefaces::Labelsizes[disposition]
# Populate the coordinate data structure.
xygrid['x'] = nx = xy[0]
xygrid['y'] = ny = xy[1]
xygrid['dx'] = dx = labelsize[0] * f
xygrid['dy'] = dy = labelsize[1] * f
# Obtain an array of 4 element arrays from the Typefaces module.
# These represent opposing corners of the address spaces on the
# Postscript canvas.
xygrid['page'] = Typefaces.labelgrid( disposition )
puts 'page coordinates'
puts xygrid['page'][0]
puts '______________'
# Copy the page label coordinates.
e = xygrid['page'].clone
# Shrink the label icon coordinates.
e.each { |z| z.collect! { |d| d *= f } }
# Shrink the icons further to make separations visible.
e.each { |z| z[2] -= 2.0; z[3] -= 2.0 }
# Save the label icon coordinates.
xygrid['boxes'] = e
puts 'page again'
puts xygrid['page'][0]
puts 'boxes'
puts e[0]
puts '=============='
xygrid['max'] = nx * ny
return xygrid

end

The diagnostics at 'page again' and 'boxes' show that the copy of the
original array has been successfully processed but the original array now
reflects the same values. The 'page' and 'boxes' arrays are to be used in
different circumstances.

What am I missing? It is bound to be something simple or something
fundamental. I tried Object#dup also - same result:

Your issue is caused by the fact that dup and clone to shallow copies
only, i.e. both copies of a clone Array contain the exact same
instances. Now you modify those instances and of course changes are
seen via both arrays:

# Shrink the label icon coordinates.
e.each { |z| z.collect! { |d| d *= f } }
# Shrink the icons further to make separations visible.
e.each { |z| z[2] -= 2.0; z[3] -= 2.0 }

A solution would be to do

e = xygrid['page'].map do |z|
z = z.map { |d| d *= f }
z[2] -= 2.0; z[3] -= 2.0
z
end

Or you can do a deep copy via the usual Marshal idiom

e = Marshal.load(Marshal.dump(xygrid['page']))

Kind regards

robert

 

[Stefano Crocco]

Another newbie question. I have started recoding some of my homegrown
Tcl/Tk applications in Ruby/rubytk and have hit an unexpected snag when
attempting to clone arrays. The following code is snipped verbatim from
the program:

# Calculate coordinates for the Postscript page canvas and the
# label icons canvas in the setupInkjet window.

def User.calcxy ( disposition, f )

xygrid = { }

# Split the disposition string into x and y counts.
xy = Typefaces:isposition[disposition]
# Obtain the label dimensions.
labelsize = Typefaces::Labelsizes[disposition]
# Populate the coordinate data structure.
xygrid['x'] = nx = xy[0]
xygrid['y'] = ny = xy[1]
xygrid['dx'] = dx = labelsize[0] * f
xygrid['dy'] = dy = labelsize[1] * f
# Obtain an array of 4 element arrays from the Typefaces module.
# These represent opposing corners of the address spaces on the
# Postscript canvas.
xygrid['page'] = Typefaces.labelgrid( disposition )
puts 'page coordinates'
puts xygrid['page'][0]
puts '______________'
# Copy the page label coordinates.
e = xygrid['page'].clone
# Shrink the label icon coordinates.
e.each { |z| z.collect! { |d| d *= f } }
# Shrink the icons further to make separations visible.
e.each { |z| z[2] -= 2.0; z[3] -= 2.0 }
# Save the label icon coordinates.
xygrid['boxes'] = e
puts 'page again'
puts xygrid['page'][0]
puts 'boxes'
puts e[0]
puts '=============='
xygrid['max'] = nx * ny
return xygrid

end

The diagnostics at 'page again' and 'boxes' show that the copy of the
original array has been successfully processed but the original array now
reflects the same values. The 'page' and 'boxes' arrays are to be used in
different circumstances.

What am I missing? It is bound to be something simple or something
fundamental. I tried Object#dup also - same result.

Len

I don't know rubytk, so I couldn't completely follow your code, but I think
your problem is related to the fact that clone and dup both perform shallow
copies of the array. This means that, while the array itself is duplicated,
its contents are not. For example, running this code

a = %w[a b c]
b = a.dup
puts "The ID of a is #{a.object_id}"
puts "The ID of b is #{b.object_id}"
puts "a is the same object as b? #{a.equal? b}"

puts "The ID of the contents of a: #{a.map{|i| i.object_id}.join(', ')}"
puts "The ID of the contents of b: #{b.map{|i| i.object_id}.join(', ')}"
a.each_index do |i|
puts "a[#{i}] is the same object as b[#{i}]? #{a.equal? b}"
end

produces:
The ID of a is -605837488
The ID of b is -605837528
a is the same object as b? false
The ID of the contents of a: -605837498, -605837508, -605837518
The ID of the contents of b: -605837498, -605837508, -605837518
a[0] is the same object as b[0]? true
a[1] is the same object as b[1]? true
a[2] is the same object as b[2]? true

This means that while a and b are not the same object (and so, you can modify
one, for example adding or removing an item, leaving the other as is), they
contain the same objects. So, calling 'destructive' methods (that is, methods
which change the receiver) on the elements of one of the array will also
change the corresponding element of the other array (because they contain the
same object).

To achieve what you want, you need deep copy. As someone on this list recently
pointed out, you can usually achieve this using marshal:

b = Marshal.load(Marshal.dump(a))

The problem is that with rubytk you're dealing with a C extension, and with
objects created in C, which may, or may not, work correctly with marshal. If
you find out they don't, you can try to manually deep-duplicate the array. In
the simple case of my example above, you can do:

b = a.map{|i| a.dup}

If your array is a nested array, things become more complicated, as you'd need
nested loops. You also need to keep in mind that object created from C
extensions may not be clonable. In this case, there's nothing to do.

I hope this helps

Stefano

 

[Len Lawrence]

A solution would be to do

e = xygrid['page'].map do |z|
z = z.map { |d| d *= f }
z[2] -= 2.0; z[3] -= 2.0
z
end

Or you can do a deep copy via the usual Marshal idiom

e = Marshal.load(Marshal.dump(xygrid['page']))

Kind regards

robert

Thanks Robert. So the issue was fundamental. I had not yet come to
appreciate the meaning of 'shallow copy' or come to terms with map or
Marshal. The Tk connection here is actually irrelevant.

That has been a great help.

Len

 

[Len Lawrence]

On Saturday 22 March 2008, Len Lawrence wrote: 0
a = %w[a b c]
b = a.dup
puts "The ID of a is #{a.object_id}"
puts "The ID of b is #{b.object_id}"
puts "a is the same object as b? #{a.equal? b}"

puts "The ID of the contents of a: #{a.map{|i| i.object_id}.join(', ')}"
puts "The ID of the contents of b: #{b.map{|i| i.object_id}.join(', ')}"
a.each_index do |i|
puts "a[#{i}] is the same object as b[#{i}]? #{a.equal? b}"
end

produces:
The ID of a is -605837488
The ID of b is -605837528
a is the same object as b? false
The ID of the contents of a: -605837498, -605837508, -605837518
The ID of the contents of b: -605837498, -605837508, -605837518
a[0] is the same object as b[0]? true
a[1] is the same object as b[1]? true
a[2] is the same object as b[2]? true

This means that while a and b are not the same object (and so, you can modify
one, for example adding or removing an item, leaving the other as is), they
contain the same objects. So, calling 'destructive' methods (that is, methods
which change the receiver) on the elements of one of the array will also
change the corresponding element of the other array (because they contain the
same object).

To achieve what you want, you need deep copy. As someone on this list

recently pointed out, you can usually achieve this using marshal:

b = Marshal.load(Marshal.dump(a))

The problem is that with rubytk you're dealing with a C extension, and
with objects created in C, which may, or may not, work correctly with
marshal. If you find out they don't, you can try to manually
deep-duplicate the array. In the simple case of my example above, you
can do:

b = a.map{|i| a.dup}

If your array is a nested array, things become more complicated, as

you'd need nested loops. You also need to keep in mind that object
created from C extensions may not be clonable. In this case, there's
nothing to do.

I hope this helps

Thanks Stefano; it sure does. I missed the earlier references to
Marshalling - not enough time to keep up with the newsgroup. I can cope
with nested arrays. Anyhow, many thanks for the education.

Len