New API doc - typos and questions

Dan Rathbun

@jim said:

Oh, I think there is a completely undocumented UTM class in there somewhere under Geom.

hm.. all I can find out about that one is:
Geom::UTM.instance_methods.sort.join("\n")... snip ...

from Wikipedia search on "UTM":
Universal Transverse Mercator coordinate system, a grid-based method of mapping locations on the surface of the Earth.
http://en.wikipedia.org/wiki/Universal_Transverse_Mercator_coordinate_system
_

Dan Rathbun

Model.get_product_family
http://code.google.com/apis/sketchup/docs/ourdoc/model.html#get_product_family

@unknownuser said:

Returns a constant number which indicates the product family of the installed SketchUp application.

This is an Application related attribute, why is it part of the Model class?

This should be a method of the Sketchup module as in: Sketchup.get_product_family
_

Dan Rathbun

Model.export
http://code.google.com/apis/sketchup/docs/ourdoc/model.html#export

@unknownuser said:

options_hash______ A hash containing the COLLADA (.dae) export options.

The options listed in the hash example is missing the option for hidden geometry.
The Registry Attribute Name is "ExportHiddenGeometry", but what is the options hash key?

hiddengeometry ?* hidden_geometry ?
_

Dan Rathbun

Set class
http://code.google.com/apis/sketchup/docs/ourdoc/set.html

This class is defined in the Objectspace and should be listed under "Core Ruby Classes" (along with Array, Length, Numeric and String.)

Instead it's buried under "App Level Classes" and listed with all those classes that are defined inside the Sketchup module.
_

Dan Rathbun

OptionsProviderObserver.onOptionsProviderChanged
http://code.google.com/apis/sketchup/docs/ourdoc/optionsproviderobserver.html#onOptionsProviderChanged

@unknownuser said:

The onOptionsProviderChanged method is invoked when an element is added to an Pages object.

Should read, "The onOptionsProviderChanged method is invoked when an option is changed for an OptionsProvider object."
_

thomthom

Entities.add_curve
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/entities.html#add_curve

Returns an array of edges - not a Curve object.

TIG

@thomthom said:

Entities.add_curve
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/entities.html#add_curve

Returns an array of edges - not a Curve object.

Ita's the same as add_circle which returns an array od edges and to find the just made curves curve you use edges=entities.add_curve(array_of_points) curve=edges[0].curve
But strangely not with add_arc which returns an ArcCurve ?

It would be unadvisable to change what the API does with these add_ methods - rather it we should keep it as is and fix the API's Guide to reflect it properly - too many scripts already get the curve from the edges array and would fail if it were updated to return the curve not the curve's edges...

thomthom

@tig said:

But strangely not with add_arc which returns an ArcCurve ?

Nope, also returns an array.

@tig said:

It would be unadvisable to change what the API does with these add_ methods

Agree.

TIG

@thomthom said:

@tig said:

But strangely not with add_arc which returns an ArcCurve ?

Nope, also returns an array.

@tig said:

It would be unadvisable to change what the API does with these add_ methods

Agree.

Perhaps I meant to say, "... add_arc which says it returns an ArcCurve wrongly!"

thomthom

Sketchup.write_default
Sketchup.read_default
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/sketchup.html#write_default
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/sketchup.html#read_default

@unknownuser said:

[..] the string associated with a value within the specified sub-section section of a .INI file or registry (within the Software > @Last Software > SketchUp section).

Is it really stored under @Last any more?

TIG

@thomthom said:

Sketchup.write_default
Sketchup.read_default
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/sketchup.html#write_default
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/sketchup.html#read_default

@unknownuser said:

[..] the string associated with a value within the specified sub-section section of a .INI file or registry (within the Software > @Last Software > SketchUp section).

Is it really stored under @Last any more?

No, / Google / Sketchup 7 / etc...

thomthom

Group.move!
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/group.html#move!

.move! moves to an absolute point. Worth noting in the manual.

Dan Rathbun

Geom::Point3d.+
http://code.google.com/apis/sketchup/docs/ourdoc/point3d.html#+

@unknownuser said:

The '+' operator is a fast way to add to the current x, y and z values of a point, or to set the values of a point by adding to other points together.
The end of the sentence should read, "...by adding **two** other points together."

@unknownuser said:

Returns:
vector a Vector object
There is no defined object named 'vector'. If we follow the example the class of the result is given as Geom::Point3d.

pt2 = Geom;;Point3d.new(100,200,300)
pt3 = pt2 + [1,1,1]
>> Point3d(101, 201, 301)
pt3.class
>> Geom;;Point3d

@unknownuser said:

Arguments:
point2 A Point3d object.
An ArugemtError exception is raised if point2 is a Point3d object.
Valid argument classes are: Array or Vector3d. This means to use a Point3d as an argument, you must convert to Array. Example:

pt3 = pt1 + pt2.to_a

This method needs to be fixed to take a Point3d object and internally convert it to an Array...

# internal to + method
if arg[0].is_a? Point3d { arg[0]=arg[0].to_a }

...or a Vector3d:

# internal to + method
if arg[0].is_a? Point3d { arg[0]=[0,0,0].vector_to(arg[0]) }

either way. The current implementation is clumsy.

Dan Rathbun

**Geom::Point3d.-**
http://code.google.com/apis/sketchup/docs/ourdoc/point3d.html#-

@unknownuser said:

Returns:
vector a Vector object
There is no defined object named 'vector', but there IS one named ' Vector3d'. If we follow the example the class of the result is given as Geom::Vector3d.

pt2 = Geom;;Point3d.new(100,200,300)
pt3 = pt2 - [1,1,1]
>> Vector3d(99, 199, 299)
pt3.class
>> Geom;;Vector3d

It seems a bit strange that + returns Point3d, but - returns Vector3d.

Dan Rathbun

Geom::Point3d.offset
Code example has errors:
[line1]replace"point ="with"point1 ="
[line1]use parameter delimiters both lines
[line3]replace".offset**!**"with".offset"
add comments
Corrected code:


point1 = Geom;;Point3d.new(10,10,10)
vector = Geom;;Vector3d.new(0,0,1)
point2 = point1.offset vector
# point1 is unchanged
# point2 is Point3d(10, 10, 11)

Geom::Point3d.offset!
Code example has errors:
[line1]replace"point ="with"point1 ="
[line1]use parameter delimiters both lines
add comments


point1 = Geom;;Point3d.new(10,10,10)
vector = Geom;;Vector3d.new(0,0,1)
point2 = point1.offset! vector
# point1 now equals point2
# point2 is Point3d(10, 10, 11)

Dan Rathbun

Geom::Vector3d.x=
http://code.google.com/apis/sketchup/docs/ourdoc/vector3d.html#x=
Geom::Vector3d.y=
http://code.google.com/apis/sketchup/docs/ourdoc/vector3d.html#y=
Geom::Vector3d.z=
http://code.google.com/apis/sketchup/docs/ourdoc/vector3d.html#z=

#Anchors: '=' char should be replaced with 'equals' in links.

All 3 methods suffer from the same return class anomoly as the Vector3d.[]= method, in that they return the class of the argument (which may not always be Float.) In contrast, the .x, .y and .z methods always return class Float.
Propose that the .x=, .y= and .z= methods should be modified to return Float as the .x, .y and .z methods do.

Propose to add note: "See the **.set!** method to set all three coordinates in one statement." ( .. to the description sections, of each of the three methods.)

Dan Rathbun

Geom::Vector3d.set!
http://code.google.com/apis/sketchup/docs/ourdoc/vector3d.html#set!

@unknownuser said:

This is a shortcut for writing:
Clarify: "This is a shortcut for setting all coordinates of the vector, instead of setting them individually in three statements."

@unknownuser said:

You may also call this method with an array or another vector:
Confusing. (I first took this mean that the SU extended Array class had an Array.set! method; which is NOT the case.)
Clarify: "You may also call this method with an Array or a Vector3d object as the first argument,(and any other arguments will then be ignored.)"

Add: "If only one Numeric argument is supplied, or if a one element Array is supplied, then only the x value will be changed. IF only 2 values are supplied, in the same way, then the z value will remain unchanged.
All coordinate values must be implicitly convertable to Float class.(If you pass a nil value, a TypeError exception is raised.)"

@unknownuser said:

Arguments:
x ____ The x value for the vector.
y ____ The y value for the vector.
z ____ The z value for the vector.
vector2 ____ A Vector3d object.
Remove 4th argument. (There is no valid fourth argument for this method. All args beyond 3 are ignored.)
Clarify: (suggested Argument list)
Arguments:
x ____ The x value for the vector, a Vector3d object, or an Array.
y ____ The y value for the vector. (optional)
z ____ The z value for the vector. (optional)

Better Code Example:


vector = Geom;;Vector3d.new 0,0,1
vector.set! 1,0,0  # three Numeric arguments
vector2 = Geom;;Vector3d.new 1,2,3
vector.set! vector2  # one Vector3d argument
# vector now is (1,2,3)
array1 = Array.new(4,5,6)
vector.set! array1[2]  # one Numeric argument
# vector now is (6,2,3)
vector2.set! array1  # one Array argument
# vector2 now is (4,5,6)
vector.set! [5,4]  # one Array with 2 values
# vector now is (5,4,3)
# change x and z in one statement, this way;
vector.set! 9, vector.y, 8
# vector now is (9,4,8)

Dan Rathbun

Geom::Vector3d.[=]
[="]http://code.google.com/apis/sketchup/docs/ourdoc/vector3d.html#[]=[/url]
Again the HTML anchor may cause problems in some browsers, suggest '#subscriptEquals'

Obviously the #[]= anchor tag in the url cannot even be displayed properly in this forum.All of the #anchor links to methods using special chars need to be changed to comply with the HTML spec, as ThomThom noted:
@unknownuser said:

ID and NAME tokens must begin with a letter ([A-Za-z]) and may be followed by any number of letters, digits ([0-9]), hyphens ("-"), underscores ("_"), colons (":"), and periods ("."). [url:nfjuel6m]http://www.w3.org/TR/html4/types.html#type-name]("http)

@unknownuser said:

Returns:
value _______ the newly set coordinate value

the Geom::Vector3d.[] method returns class Float
BUT, the Geom::Vector3d.[]= method returns the class of the argument.

The code example may be misleading to the unwary,
@unknownuser said:

value = vector[i] = coordinate

..if the class of the coordinate variable is Integer then the class of the value variable will also be Integer.
This causes a coder to do a two-step expression if they want a Float, ie:


coordinate = 2
vector[i] = coordinate # returns Integer 2
value = vector[i]  # value.class >> Float

%(#FF4000)[Two Issues:

Is 'coordinate' the correct term, or would 'component' be a better term when refering to vectors?1. If all the rest of the methods that return subscript values from a Vector3d object are Float, then perhaps this method should as well, regardless of the class of the argument.]

Dan Rathbun

Geom::Point3d.<
http://code.google.com/apis/sketchup/docs/ourdoc/point3d.html#%3C

@unknownuser said:

Returns:
true if the point2 is closer to the origin.

Following the example, I get the opposite result.

TheReturns:should probably read,
" *true* if the point2 is **farther** from the origin; *false* if the point2 is closer to the origin (than the receiver.)

TIG

@thomthom said:

Group.move!
http://code.google.com/intl/nb/apis/sketchup/docs/ourdoc/group.html#move!

.move! moves to an absolute point. Worth noting in the manual.

It doesn't seem to though - we've been testing it in another thread this afternoon and it seem that it moves relative to the transformation-origin NOT the model's ORIGIN - you first have to do a transformation by vector from where it is to the ORIGIN and then off again to the absolute point ?