DRAFT
The Ayam scripting interface consists of a number of Tcl procedures
and Tcl commands that are also used internally by the application.
The main menu entry "File/New"
for instance calls the
scripting interface command "newScene"
(among other commands).
Using the scripting interface means to call these commands on
your own possibly in a mix with standard Tcl script code.
Furthermore, using Tcl and its introspection capabilities, you could easily modify the code Ayam consists of. This is, however, not recommended for good reasons (unless you read the Ayam source code and really know, what you are doing). So watch out for already existing procedures and commands when implementing your own. Using procedures and commands not listed in this documentation is dangerous too. Implementation and interfaces of those procedures and commands may change in future versions of Ayam without notice.
In Tcl, all variables, procedures, and commands are case sensitive,
it really is "sL"
and not "sl"
and not
"SL"
.
The scripting interface may be used directly from the console
of Ayam. You can, of course, also write scripts in your own
Tcl script files, that may be loaded at any time into Ayam using
the console and the Tcl command "source"
.
You can also arrange for a script file to be executed automatically
on every application startup using the preference setting
"Main/Scripts"
.
Moreover, on the X11 and Aqua window systems, Ayam is able to execute
script code sent via the Tk "send"
command or the AppleScript
"tell"
command from external applications.
In contrast to other modelling environments, in Ayam there is another way to run scripts. In Ayam, scripts may also be attached to Script objects and run when the notification mechanism updates the scene. See also section Script object. Even normal objects can trigger scripts upon notification using BNS or ANS tags. See also sections Before Notify Script and After Notify Script.
Note that most of the scripting interface commands listed in this
documentation work in the background, without changing anything to
the Ayam GUI and Ayam view windows, for the sake of execution speed.
If you want your changes to become visible you have to update the
various parts of the GUI (property GUIs, view windows) explicitly
(see also section
Updating the GUI).
However, since Ayam 1.13 it is also possible to automatically run GUI
updating commands in the console by using <Shift+Return>
instead of <Return>
when issuing scripting interface commands.
Note however that even though no updates of the GUI take place, all
notification processes are carried out immediately. The scene will
be consistent and up to date when the scripting interface command
returns.
If you want your changes to be recorded in the undo buffer, you have to arrange for this manually too (see the documentation of the undo command: Undo).
From scripts it may be necessary to check whether an error occurred
during the execution of a command. All commands return
TCL_OK in any case, so checking their return value avails
to nothing, but they set the global Tcl variable "ay_error"
to a value higher than 1 if an error occurred. You need to set
ay_error to zero before and check it after the operation in question
to see whether the operation performed successfully:
proc myProc { } { set ::ay_error 0 copOb if { $::ay_error > 1 } { ayError 2 "myProc" "Error copying object!" } }
Several global variables and arrays exist in the Ayam Tcl context, that may be useful for scripts.
"ay_error"
variable holds the current error state.
See also section
Reporting Errors."i"
is used by all "forAll"
command variants. See also section
Applying Commands to a Number of Objects."aymark"
is an array that contains the current
mark coordinates.[*]The global array "ay"
holds application state variables.
Furthermore, you can find the paths to important widgets
(e.g. the tree widget for the object hierarchy or the currently
active view) in this array.
Use "parray ay"
in the console to see what is there.
More documentation to come.
The global array "ayprefs"
holds preferences data.
The complete array is saved in the "ayamrc"
file upon exit,
so be careful when adding new elements to this array.
See also section
Ayamrc File.
Use "parray ayprefs"
in the console to see what is there.
More documentation to come.
Note that changes to this array on the Tcl side do not immediately
take effect as the data needs to be transferred to the C context
using the "setPrefs"
command.
See also section
Managing Preferences.
For every property, a corresponding global arrays exists, where the property is managed. For the Transformations property, this array looks like this:
Transformations { arr transfPropData sproc setTrafo gproc getTrafo w fTrafoAttr }
"arr"
, designates the name of the global property
data array (thus, transformation data is stored in an array called
"transfPropData"
).
The entries "sproc"
and "gproc"
designate the set-property and
get-property callbacks (procedures or commands) respectively.
If sproc or gproc are empty strings (""
), standard callbacks
named "setProp"
or "getProp"
should be used to get
or set the property values.
But for the transformations property, the "setTrafo"
and
"getTrafo"
commands should be used.
The last entry, "w"
, is the name of the main property GUI window.
To get the full widget path of this window, the current value of
ay(pca) needs to be prepended.
Note that the global property data array only holds useful data when the respective property GUI is active, or when it has been filled explicitly by the so called get-property callback.
The following global arrays and callbacks to get or set the data exist:
property | array | get-property callback | set-property callback |
Transformations | transfPropData | getTrafo | setTrafo |
Attributes | attrPropData | getAttr | setAttrp |
Material | matPropData | getMat | setMat |
Tags | tagsPropData | getTagsp | setTagsp |
Note that this list is pretty much incomplete, however you can always
infer such information using commands like "parray Tags"
in the Ayam
console.
See also section Manipulating Properties for more information on how to edit property values from the scripting interface.
Since Ayam 1.16, the global property management array may be created
easily using the new scripting interface command "addPropertyGUI"
.
This section provides documentation on the most important scripting interface procedures and commands of Ayam sorted by category.
Note that the "help"
command in the Ayam console can be used
to directly jump to the appropriate sub-section of this part of
the documentation.
All procedures and commands are documented in the following scheme:
"command param1 param2 [optionalparam1]"
(syntax of the command and its parameters),"command 1 2"
(example application of the command
with explanation of expected results).Since Ayam 1.8.2 a scripting interface command named "help"
is available,
that displays the help of scripting interface commands using a web
browser (similar to the "Help on Object"
feature):
"help command"
"help help"
displays the help of the help command.
To create new objects the "crtOb"
command can be used.
"crtOb type [args]"
"crtOb"
,
"type"
may be derived from the object type names, as displayed in the
tree view.
The new object will be created and linked to the scene as last object
in the current level, no part of the GUI (object selection widget, property
GUI, views) will be updated. Furthermore, the new object will not
be selected.
Depending on the type, further arguments may (or have to) be given; some object types expect other objects to be selected upon creation.
All arguments consist of a option name part and a value part (i.e. it is
"-center 1"
and not "-center"
and also not
"-center=1"
).
The option names can be abbreviated. Useful default and fallback
values exist (see below). The arguments can be mixed freely (their
order is not important) and repeated. If arguments are repeated, only the
last set value is used, even if this leads to errors and application of
fallback values later on.
Here is a comprehensive list of available arguments sorted by object type:
"NCurve"
: NURBS curves accept the following
arguments:
"-length"
: length of the new curve, the length defaults to 4."-order"
: order of the new curve, the order defaults to 4.
If a value greater than the length is specified, the order will be made
identical to the length value."-kt"
: the knot type of the new curve, must be one of
0 – Bezier, 1 – BSpline, 2 – NURB, 3 – Custom,
4 – Chordal, 5 – Centripetal.
A knot vector of specified type will automatically be created.
The knot type defaults to 2 – NURB.
If a custom knot vector is specified using the "-kv"
option below,
the knot type will always be set to 3 – Custom.
"-kv"
: the knot vector of the new curve. The value of this
option is a list of floating point numbers of length curve length plus
curve order, e.g. for a curve with 2 control points and order 2, specify
4 knots: "-kv {0.0 0.0 1.0 1.0}"
. The knot vector defaults to
an automatically created knot vector of the type specified by the
"-kt"
option above.
"-kn"
: the knot vector of the new curve. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-kv"
option
above.
"-cv"
: the control vector of the new curve. The value of this
option is a list of floating point numbers that describe the 4D euclidean
rational (weight not multiplied in) coordinates of the control
points."-cv {0.0 0.0 0.0 1.0 1.0 0.0 0.0 1.0 2.0 0.0 0.0 1.0}"
.
"-cn"
: the control vector of the new curve. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-dx"
: the value of this option specifies the distance
of automatically created control points in the x dimension, default is 0.25.
"-dy"
: the value of this option specifies the distance
of automatically created control points in the y dimension, default is 0.0.
"-dz"
: the value of this option specifies the distance
of automatically created control points in the z dimension, default is 0.0
"-center"
: If the value of the "-center"
option is 1,
the new curve will be centered. The default value is 0, no centering.
This option is only in effect if no "-cv"
option is specified.
"-createmp"
: The "-createmp"
option toggles creation
of multiple points. The default value is 0."crtOb NCurve"
"crtOb NCurve -center 1"
"crtOb NCurve -length 5 -center 1 -dx 0.5"
In versions of Ayam prior to 1.17, NURBS curves only accepted the
"-length"
argument.
"ICurve"
: Interpolating curves accept the following
arguments:
"-type"
: the type of the new curve, must be one of
0 – Open, 1 – Closed; default is 0.
"-length"
: length (number of data points to interpolate)
of the new curve; the length defaults to 4.
"-order"
: order of the new curve, the order defaults to 4.
If a value greater than the length is specified, the order will be made
identical to the length value.
"-pt"
: the parameter type of the new curve, must be one of
0 – Chordal, 1 – Centripetal, 2 – Uniform; default is 0.
"-cv"
: the control vector of the new curve. The value of this
option is a list of floating point numbers that describe the 3D
(non rational) coordinates of the control points. This list may also
only specify one point, which is then taken as starting point and
DX/DY/DZ (see below)
are used to create the missing control points automatically.
To specify a complete
control vector, this list should have curve length · 3
elements, e.g. for a curve of length 3, specify 9 values:
"-cv {0.0 0.0 0.0 1.0 0.0 0.0 2.0 0.0 0.0}"
.
"-cn"
: the control vector of the new curve. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-dx"
: the value of this option specifies the distance
of automatically created control points in the x dimension, default is 0.25.
"-dy"
: the value of this option specifies the distance
of automatically created control points in the y dimension, default is 0.0.
"-dz"
: the value of this option specifies the distance
of automatically created control points in the z dimension, default is 0.0
"-center"
: If the value of the "-center"
option is 1,
the new curve will be centered. The default value is 0, no centering.
This option is only in effect if no "-cv"
option is specified.
"-derivs"
: the value of this option controls whether
user defined end derivatives should be used: 0 – no, 1 – yes,
default is 0.
"-sdlen"
: the value of this option specifies the
relative length (in relation to the distance of the first and second
control point) of the start derivative, default is 0.125.
"-sderiv"
: is the start derivative, specified as
a list of three float values. The derivative is specified relative to the
first control point. The start derivative defaults to an automatically created
derivative of a direction taken from the first two control points
and length specified by the "-sdlen"
option.
"-edlen"
: the value of this option specifies the
relative length (in relation to the distance of the second to last and last
control point) of the end derivative, default is 0.125.
"-ederiv"
: is the end derivative, specified as
a list of three float values. The derivative is specified relative to the
last control point. The end derivative defaults to an automatically created
derivative of a direction taken from the last two control points
and length specified by the "-edlen"
option.
"crtOb ICurve"
"crtOb ICurve -l 5 -sderiv {0.0 -0.5 0.0} -ederiv {0.0 -0.5 0.0} -derivs 1 -center 1"
In versions of Ayam prior to 1.17, interpolating curves only accepted the
"-length"
argument.
"ACurve"
: Approximating curves accept the following
arguments:
"-type"
: the type of the new curve, must be one of
0 – Open, 1 – Closed; default is 0.
"-length"
: length (number of data points to approximate)
of the new curve, the length defaults to 4.
"-alength"
: number of control points to use for the
approximating curve, the alength defaults to 3.
"-order"
: order of the new curve, the order defaults to 3.
If a value greater than the length is specified, the order will be made
identical to the length value.
"-symmetric"
: toggles creation of symmetric curves,
must be one of 0 – Asymmetric, 1 – Symmetric; default is 0.
"-cv"
: the control vector of the new curve. The value of this
option is a list of floating point numbers that describe the 3D
(non rational) coordinates of the control points. This list may also
only specify one point, which is then taken as starting point and
DX/DY/DZ (see below)
are used to create the missing control points automatically.
To specify a complete
control vector, this list should have curve length · 3
elements, e.g. for a curve of length 3, specify 9 values:
"-cv {0.0 0.0 0.0 1.0 0.0 0.0 2.0 0.0 0.0}"
.
"-cn"
: the control vector of the new curve. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-dx"
: the value of this option specifies the distance
of automatically created control points in the x dimension, default is 0.25.
"-dy"
: the value of this option specifies the distance
of automatically created control points in the y dimension, default is 0.0.
"-dz"
: the value of this option specifies the distance
of automatically created control points in the z dimension, default is 0.0
"-center"
: If the value of the "-center"
option is 1,
the new curve will be centered. The default value is 0, no centering.
This option is only in effect if no "-cv"
option is specified.
"crtOb ACurve -length 6"
"crtOb ACurve -l 5 -center 1"
In versions of Ayam prior to 1.17, approximating curves only accepted the
"-length"
argument.
"NPatch"
: NURBS patches accept the following
arguments:
"-width"
: width of the new patch, the width defaults to 4."-height"
: height of the new patch, the height defaults to 4."-uorder"
: order of the new patch in U parametric dimension,
the order defaults to 4. If a value greater than the width is specified,
the order will be made identical to the width value."-ukt"
: the U knot type of the new patch, must be one of
0 – Bezier, 1 – BSpline, 2 – NURB, 3 – Custom,
4 – Chordal, 5 – Centripetal.
A knot vector of specified type will automatically be created.
The knot type defaults to 2 – NURB.
If a custom knot vector is specified using the "-ukv"
option below,
the knot type will always be set to 3 – Custom."-ukv"
: the U knot vector of the new patch. The value of this
option is a list of floating point numbers of length width plus
patch U order, e.g. for a patch with width 2 and U order 2, specify
4 knots: "-ukv {0.0 0.0 1.0 1.0}"
. The knot vector defaults to
an automatically created knot vector of the type specified by the
"-ukt"
option above.
"-un"
: the U knot vector of the new patch. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-ukv"
option
above.
"-vorder"
: order of the new patch in V parametric dimension,
the order defaults to 4. If a value greater than the height is specified,
the order will be made identical to the height value."-vkt"
: the V knot type of the new patch, must be one of
0 – Bezier, 1 – BSpline, 2 – NURB, 3 – Custom,
4 – Chordal, 5 – Centripetal.
A knot vector of specified type will automatically be created.
The knot type defaults to 2 – NURB.
If a custom knot vector is specified using the "-vkv"
option below,
the knot type will always be set to 3 – Custom."-vkv"
: the V knot vector of the new patch. The value of this
option is a list of floating point numbers of length height plus
patch V order, e.g. for a patch with height 2 and V order 2, specify
4 knots: "-vkv {0.0 0.0 1.0 1.0}"
. The knot vector defaults to
an automatically created knot vector of the type specified by the
"-vkt"
option above.
"-vn"
: the V knot vector of the new patch. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-vkv"
option
above.
"-cv"
: the control vector of the new patch. The value of this
option is a list of floating point numbers that describe the 4D euclidean
rational (weight not multiplied in) coordinates of the control
points.
This list may also only specify one point, which is then taken as starting
point and UDX/UDY/UDZ and VDX/VDY/VDZ (see below) are used to create the
missing control points automatically.
To specify a complete control vector, this list should have
width · height · 4 elements,
e.g. for a patch of width 2 and height 2, specify 16 values:
"-cv {0.0 0.0 0.0 1.0 1.0 0.0 0.0 1.0 2.0 0.0 0.0 1.0 2.0 1.0 0.0 1.0}"
.
"-cn"
: the control vector of the new patch. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-udx"
: the value of this option specifies the distance
of automatically created control points in the x dimension between points
in a row (U parametric dimension, along width), default is 0.25.
"-udy"
: the value of this option specifies the distance
of automatically created control points in the y dimension between points
in a row (U parametric dimension, along width), default is 0.0.
"-udz"
: the value of this option specifies the distance
of automatically created control points in the z dimension between points
in a row (U parametric dimension, along width), default is 0.0
"-vdx"
: the value of this option specifies the distance
of automatically created control points in the x dimension between points
in a column (V parametric dimension, along height), default is 0.0.
"-vdy"
: the value of this option specifies the distance
of automatically created control points in the y dimension between points
in a column (V parametric dimension, along height), default is 0.25.
"-vdz"
: the value of this option specifies the distance
of automatically created control points in the z dimension between points
in a column (V parametric dimension, along height), default is 0.0
"-center"
: If the value of the "-center"
option is 1,
the new patch will be centered. The default value is 0, no centering.
This option is only in effect if no "-cv"
option is specified.
"-createmp"
: The "-createmp"
option toggles creation
of multiple points. The default value is 0."crtOb NPatch"
"crtOb NPatch -vdy 0 -vdz 0.25"
"crtOb NPatch -udy 0.25"
"crtOb NPatch -udy 0.25 -vdz 0.25"
"crtOb NPatch -width 2 -height 2 -center 1 -udx 2 -vdy 2"
"crtOb NPatch -width 3 -height 2 -uorder 2 -cv {-1 0 1 1 1 0 1 1 -1 0 0 1 1 0 0 1 -1 1 0 1 1 1 0 1}"
-uorder 2
to get
a smooth shape).In versions of Ayam prior to 1.17, NURBS patches only accepted the
"-width"
and "-height"
argument.
"IPatch"
: Interpolating patches accept the following
arguments:
"-width"
: width of the new patch, the width defaults to 4."-height"
: height of the new patch, the height defaults to 4."-uorder"
: order of the new patch in U parametric dimension,
the order defaults to 4. If a value greater than the width is specified,
the order will be made identical to the width value.
A value of 0 switches off interpolation along U."-ukt"
: the U parameterisation type, must be one of
0 – Chordal (default), 1 – Centripetal, 2 – Uniform."-vorder"
: order of the new patch in V parametric dimension,
the order defaults to 4. If a value greater than the height is specified,
the order will be made identical to the height value.
A value of 0 switches off interpolation along V."-vkt"
: the V parameterisation type, must be one of
0 – Chordal (default), 1 – Centripetal, 2 – Uniform.
"-deriv_u"
: the end derivative mode for U, must be one of
0 – None (default), 1 – Automatic, or 2 – Manual.
In manual mode full derivative vectors must be provided via
"-ederiv_u"
and "-sderiv_u"
."-edlen_u"
: the length of automatically calculated end
derivatives at end of patch in U (default 0.125)."-sdlen_u"
: the length of automatically calculated end
derivatives at start of patch in U (default 0.125)."-ederiv_u"
: end derivatives for U at end of patch.
The value of this option is a list of 3 · height
floating point numbers.
There is no default value."-sderiv_u"
: end derivatives for U at start of patch.
The value of this option is a list of 3 · height
floating point numbers.
There is no default value.
"-deriv_v"
: the end derivative mode for V, must be one of
0 – None (default), 1 – Automatic, or 2 – Manual.
In manual mode full derivative vectors must be provided via
"-ederiv_v"
and "-sderiv_v"
."-edlen_v"
: the length of automatically calculated end
derivatives at end of patch in V (default 0.125)."-sdlen_v"
: the length of automatically calculated end
derivatives at start of patch in V (default 0.125)."-ederiv_v"
: end derivatives for V at end of patch.
The value of this option is a list of 3 · width
floating point numbers.
There is no default value."-sderiv_v"
: end derivatives for V at start of patch.
The value of this option is a list of 3 · width
floating point numbers.
There is no default value.
"-cv"
: the control vector of the new patch. The value of this
option is a list of floating point numbers that describe the 3D
non rational coordinates of the data points to be interpolated."-cv {0.0 0.0 0.0 1.0 0.0 0.0 2.0 0.0 0.0 2.0 1.0 0.0}"
.
"-cn"
: the control vector of the new patch. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-udx"
: the value of this option specifies the distance
of automatically created control points in the x dimension between points
in a row (U parametric dimension, along width), default is 0.25.
"-udy"
: the value of this option specifies the distance
of automatically created control points in the y dimension between points
in a row (U parametric dimension, along width), default is 0.0.
"-udz"
: the value of this option specifies the distance
of automatically created control points in the z dimension between points
in a row (U parametric dimension, along width), default is 0.0
"-vdx"
: the value of this option specifies the distance
of automatically created control points in the x dimension between points
in a column (V parametric dimension, along height), default is 0.0.
"-vdy"
: the value of this option specifies the distance
of automatically created control points in the y dimension between points
in a column (V parametric dimension, along height), default is 0.25.
"-vdz"
: the value of this option specifies the distance
of automatically created control points in the z dimension between points
in a column (V parametric dimension, along height), default is 0.0
"-center"
: If the value of the "-center"
option is 1,
the new patch will be centered. The default value is 0, no centering.
This option is only in effect if no "-cv"
option is specified.
"crtOb IPatch"
"crtOb IPatch -vdy 0 -vdz 0.25"
"crtOb IPatch -udy 0.25"
"crtOb IPatch -udy 0.25 -vdz 0.25"
"crtOb IPatch -width 3 -height 3 -center 1 -udx 2 -vdy 2"
"PolyMesh"
: Polymeshes accept the following arguments:
"-polys"
: the value of this option specifies the number of
polygons/faces in the mesh. The number of polygons defaults to 0.
"-loops"
: the value of this option specifies the number of
loops per polygon. It is therefore a list of positive integer values of a
length equal to the value of the "-polys"
option. The default value
of this option is a list of proper length with all elements set to 1
(only normal polygons, without holes, are specified).
"-nverts"
: the value of this option specifies the number of
vertices per loop. It is therefore a list of positive integer
values of a length equal to the sum of all elements of the "-loops"
option.
The default value of this option is a list of proper length with all
elements set to 3 (only triangles are in the mesh).
"-iverts"
: the value of this option specifies all the (zero
based) indices of the vertices of all loops. It is therefore a list of
integer values of a length equal to the sum of all elements of the
"-nverts"
option.
The default value of this option is a list of proper length with the
elements set to a sequence of integers so that the control points are
used in the same order as specified via the "-cv"
option
(0, 1, 2, 3, ...).
"-cv"
: the control points of the new mesh. The value of this
option is a list of floating point numbers that describe the 3D
(non rational) coordinates of the control
points. The indices specified via the "-iverts"
option point to
this list. If the "-vnormals"
option is 1, also vertex normals are
specified in this list (directly following the coordinate values of each
control point) and stride is 6, otherwise stride is 3.
This list must have a length of stride by the highest value in
the list provided via the "-iverts"
option.
The default value of this option is an empty list, this implies that this
option must be specified to create a non-empty PolyMesh object.
"-cn"
: the control points of the new mesh. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-vnormals"
: determines whether vertex normals are present.
The default value is 0 – no vertex normals are present.
"crtOb PolyMesh -p 1 -cv {0 0 0 1 0 0 0 1 0}"
"crtOb PolyMesh -p 2 -cv {0 0 0 1 0 0 1 1 0 0 1 0} -iv {0 1 2 0 2 3}"
"crtOb PolyMesh -p 3 -cv {0 0 0 1 0 0 1 1 0 0 1 0 1.5 0 0 1.5 1 0} -iv {0 1 2 0 2 3 1 4 5 2} -nv {3 3 4}"
"crtOb PolyMesh -p 1 -loops {2} -cv {0 0 0 1 0 0 0 1 0 .25 .25 0 .5 .25 0 .25 .5 0}"
"SDMesh"
: Subdivision meshes accept the following arguments:
"-scheme"
: the value of this option specifies the subdivision
scheme, it may be set to 0 – Catmull-Clark or 1 – Loop only.
Default is 0.
"-faces"
: the value of this option specifies the number of
faces in the mesh. The number of faces defaults to 0.
"-nverts"
: the value of this option specifies the number of
vertices per face. It is therefore a list of positive integer
values of a length equal to the number of faces.
The default value of this option is a list of proper length with all
elements set to 3 (only triangles are in the mesh).
"-verts"
: the value of this option specifies all the (zero
based) indices of the vertices of all faces. It is therefore a list of
integer values of a length equal to the sum of all elements of the
"-nverts"
option.
The default value of this option is a list of proper length with the
elements set to a sequence of integers so that the control points are
used in the same order as specified via the "-cv"
option
(0, 1, 2, 3, ...).
"-cv"
: The value of this option is a list of floating point
numbers that describe the 3D (non rational) coordinates of the control
points. The indices specified via the "-verts"
option point to
this list.
This list must have a length of 3 multiplied by the highest value in
the list provided via the "-verts"
option.
The default value of this option is an empty list, this implies that this
option must be specified to create a non-empty SDMesh object.
"-cn"
: the control points of the new mesh. The value of
this option is a variable name (with optional array and namespace specifier).
The value of this variable must be compatible to the "-cv"
option
above.
"-tags"
: the value of this option specifies a number of tags.
It is therefore a list of positive integer values of arbitrary length.
The only allowed values are 0 – hole, 1 – corner,
2 – crease, and 3 – interpolateboundary.
The default value of this option is an empty list: no tags.
"-args"
: the value of this option specifies the number of
integer and floating point arguments per tag.
It is therefore a list of positive integer values of length:
double number of tags. The even entries specify the number of integer and
the odd entries the number of floating point arguments per tag.
The content of this list is partially dictated by the "-tags"
option, e.g. a crease entry has at least two integer arguments and one
floating point argument.
The default value of this option is list of proper length, with all
elements set to zero (no tags have any arguments).
"-intargs"
: the value of this option specifies the
integer arguments of all tags.
It is therefore a list of integer values of length
sum of all even elements given by the "-args"
option.
"-doubleargs"
: the value of this option specifies the
floating point arguments of all tags.
It is therefore a list of double values of length
sum of all odd elements given by the "-args"
option.
"crtOb SDMesh -f 4 -v {0 1 3 1 2 3 0 3 2 0 2 1} -cv {0 0 0 1 0 0 0 0 -1 0.5 1 -0.5}"
"crtOb SDMesh -f 4 -v {0 1 3 1 2 3 0 3 2 0 2 1} -cv {0 0 0 1 0 0 0 0 -1 0.5 1 -0.5} -tags {1} -args {1 1} -intargs {0} -doubleargs {3.0}"
"crtOb SDMesh -f 4 -v {0 1 3 1 2 3 0 3 2 0 2 1} -cv {0 0 0 1 0 0 0 0 -1 0.5 1 -0.5} -tags {1} -args {2 1} -intargs {0 1} -doubleargs {10.0}"
"Level"
: Levels must be given an additional argument
determining the type of the new level, this argument may be one of:
"0"
(level), "1"
(union), "2"
(intersection),
"3"
(difference), or "4"
(primitive).
Examples:
"crtOb Level 0"
"crtOb Level 3"
"Material"
: Materials must be given an additional
argument giving the name of the new material.
Example:
"crtOb Material Wood"
"Instance"
: creates an instance of the selected object."crtOb Sphere; uS; rV"
.There are some helper commands, that create certain often used curves:
crtNCircle – create NURBS circle:
"crtNCircle [-r radius] [-a arc]"
-r
option and arc as
defined via the -a
option. The curve always starts on the
positive X axis. The radius defaults to 1.0 and
the arc to 360.0. The arc option supports negative values.
See also section NURB Circle Tool.
crtClosedBS – create closed (circular) B-Spline:
"crtClosedBS [-s sections] [-o order] [-a arc] [-r radius]"
See also section Circular B-Spline Tool.
crtNRect – create a rectangular NURBS curve:
"crtNRect [-w width] [-h height]"
crtTrimRect – create a rectangular bounding trim curve:
"crtTrimRect"
"CreateAtMark"
option is ignored.
See also section TrimRect Tool.
"delOb"
"getType varname"
"getName varname"
"nameOb name"
"hasChild"
These commands are probably the most important ones, because many other scripting interface commands operate on selected objects only:
"selOb [index]"
"selOb"
clears the current selection, "selOb 0"
selects the first object in the current level, "selOb 0 1"
selects
the first two objects in the current level."sL"
# create object crtOb NCurve # update tree uCR # select new object sL
"uCR"
command, as access to the GUI is blocked anyway in this
context:
# create object crtOb NCurve # select new object sL
hSL – hidden select last object:
"hSL"
"sL"
instead.This command manipulates the point selection.
"selPnts [(-count | -get) vname | -all | index1 index2 ...]"
"-count"
, this command puts the number
of all currently selected points into the variable specified by the
"vname"
argument."-get"
, this command puts the indices
of all currently selected points into the variable specified by the
"vname"
argument."-all"
, this command selects all points."selPnts -all"
selects all points of the curve, and"selPnts 0 2"
selects the first and third point of the curve.These procedures allow easy access to object properties from the scripting interface:[*]
getProperty – get single property value
"getProperty propname(elemname) varname"
"r"
easily using the command
getProperty SphereAttr(Radius) r
In contrast to using "getProperty"
, here is an equivalent example for the
direct (fast) access of property values:
getProp
set r $::SphereAttrData(Radius)
See also section
Global Property Management and Data Arrays.setProperty – set single property value
"setProperty propname(elemname) value"
"3.0"
easily using the command
setProperty SphereAttr(Radius) 3.0
In contrast to using "setProperty"
, here is an equivalent example for the
direct (fast) access of property values:
getProp
set SphereAttrData(Radius) 3.0
setProp
See also section
Global Property Management and Data Arrays.These commands operate the object clipboard:
"copOb [-append]"
"-append"
is used, the clipboard will not be cleared
before this operation."cutOb [-append]"
"-append"
is used, the clipboard will not be cleared
before this operation."pasOb [-move]"
"-move"
is given,
the objects are moved and not copied, i.e. after a move, the clipboard
is empty.repOb – replace clipboard content with selected object(s):
"repOb"
clearClip – clear object clipboard:
"clearClip"
The following procedures operate the property clipboard, which is totally independent from the object clipboard.
copyProp – copy a property to the property clipboard
"copyProp [mode]"
"pclip_omit"
."pasteProp"
These commands manipulate the current level of Ayam:
"goDown index"
"goUp"
"goTop"
The following commands move objects around in the hierarchy:
upOb – shuffle object(s) backward in the current level:
"upOb"
downOb – shuffle object(s) forward in the current level:
"downOb"
These commands transform objects or selected points of objects:
"movOb dx dy dz"
dx
in direction of the
objects X axis, by dy
in direction of the objects Y axis, and by
dz
in direction of the objects Z axis."rotOb dx dy dz"
dx
degrees around the
objects X axis, then by dy
degrees around objects Y axis, and finally by
dz
degrees around the objects Z axis. Note the order of the rotations."scalOb dx dy dz"
dx
in
direction of the objects X axis, by a factor of dy
in
direction of the objects Y axis, and by a factor of dz
in
direction of the objects Z axis.movPnts – move selected points:
"movPnts dx dy dz"
dx
in direction of the
objects X axis, by dy
in direction of the objects Y axis, and by
dz
in direction of the objects Z axis.rotPnts – rotate selected points:
"rotPnts dx dy dz"
dx
degrees around the
objects X axis, then by dy
degrees around objects Y axis, and finally
by dz
degrees around the objects Z axis. Note the order of the rotations.scalPnts – scale selected points:
"scalPnts dx dy dz"
dx
in
direction of the objects X axis, by a factor of dy
in
direction of the objects Y axis, and by a factor of dz
in
direction of the objects Z axis.delegTrafo – delegate transformations:
"delegTrafo"
This operation fails for complex setups (i.e. if the combination of parent and child transformation is a shear transformation).
applyTrafo – apply transformations:
"applyTrafo [-sel]"
"-sel"
is given) are modified.
It is no error, if an object has no points at all or if the points
are readonly.
Additionally, if any points of an object are modified, the transformations
of this object will be reset to the default values.normTrafos – normalize transformation values:
"normTrafos"
"NormalizeDigits"
(by default 6)."normPnts"
"NormalizeDigits"
(by default 6).These commands operate the shader properties:
"shaderSet shadertype [varname]"
"surface"
, "displacement"
,
"light"
, "imager"
, "atmosphere"
, "exterior"
or "interior"
. If varname is not given, the shader in question
is deleted from the object instead.
Otherwise, varname points to an associative array that contains
the data (arguments) of the shader. Example content may be created
with the "shaderGet"
command below. The data is not checked
against the internal shader database for correctness or completeness."shaderGet shadertype varname"
shadertype
for the selected
object. Type may be one of "surface"
, "displacement"
,
"light"
, "imager"
, "atmosphere"
, "exterior"
or "interior"
. The shader data will be written to an associative
array pointed to by varname
.These commands may be used to modify the tags of an object:
"setTag type value"
type
and value-string
value
to the currently selected object(s).""
as value parameter. This is
e.g. needed for the "NoExport"
tag type."WarnUnknownTag"
."setTag NoExport """
"NoExport"
tag to the selected objects."setTag RP Transformations"
"RP"
(remove property) tag to the selected objects
that hides the Transformations property GUI."addTag type value"
type
and value-string
value
to the currently selected objects.""
as value parameter. This is
e.g. needed for the "NoExport"
tag type."WarnUnknownTag"
."addTag NoExport """
"NoExport"
tag to the selected objects."addTag RP Transformations"
"RP"
(remove property) tag to the selected objects
that hides the Transformations property GUI."hasTag type"
"1"
if the selected
object has atleast one tag of the designated type. Otherwise "0"
is returned."delTags [type]"
"all"
, all tags are
deleted from the currently selected objects."delTags"
"delTags RP"
"RP"
tags from the selected objects."getTags tvname vvname"
"setTags tags"
"setTags {RP Transformations RP Attributes}"
"RP"
tags."getTag tagname vname"
""
."registerTag type"
These are more specialized commands to change NURBS curve and surface properties:
"clampNC [-s | -e]"
"Custom"
and
the knots will have o equal values at the desired side(s), where o
is the order of the curve.
If the side parameter is omitted, both sides are clamped.
If the side parameter is "-s"
only the start, and if it is "-e"
only the end is clamped.
In Ayam versions prior to 1.18 it was an error if the curve was already clamped at either side, this is no longer the case. Furthermore, curves with multiple knots in the end region(s) could not be clamped, this works ok now.
See also section Clamp Tool.
unclampNC – unclamp NURBS curve:
"unclampNC [-s | -e]"
"Custom"
.
If the side parameter is omitted, both sides are unclamped.
If the side parameter is "-s"
only the start, and if it is "-e"
only the end is unclamped.
See also section Unclamp Tool.
extendNC – extend NURBS curve:
"extendNC (x y z (w) | -vn varname | -m)"
See also section Extend Tool.
elevateNC – elevate NURBS curve:
"elevateNC n"
"Custom"
.
See also section Elevate Tool.
insknNC – insert knot into NURBS curve:
"insknNC u r"
u
r
times. The valid range for u
is determined by
the current knot vector U as follows: U[p] <= u <= U[n]
, where
p is the degree (order-1) of the curve and n is the length of the curve.
The knot type of the curves will always be changed to custom but
the shape of the curves will not change.
See also section Insert Knot Tool.
remknNC – remove knot from NURBS curve:
"remknNC (u | -i ind) r [tol]"
"remknNC -i 3 1"
instead of
"remknNC 0.5 1"
for the knot vector "0 0 0 0.5 1 1 1"
).
Note that the shape of the curve may be changed by this tool unless the parameter tol is specified. If tol is specified the new curve does not deviate from the original curve more than tol in any point on the curve. If the knot can not be removed r times due to the tolerance, an error is reported and the original curve is left unchanged.
See also section Remove Knot Tool.
"openC"
See also section Open Tool.
"closeC"
See also section Close Tool.
"refineC"
See also section Refine Tool.
refineknNC – refine knots of NURBS curve:
"refineknNC {u1 u2 un}"
See also section Refine Knots Tool.
coarsenNC – coarsen NURBS curve:
"coarsenNC"
See also section Coarsen Tool.
"revertC"
See also section Revert Tool.
tweenNC – interpolate (tween) curves:
"tweenNC [r]"
See also section Tween Curve Tool.
"revertuS"
See also section Revert U Tool.
"revertvS"
See also section Revert V Tool.
swapuvS – swap dimensions of surfaces:
"swapuvS"
See also section Swap UV Tool.
rescaleknNC – rescale knots of NURBS curves:
"rescaleknNC [-r rmin rmax | -d mindist]"
[0.0, 1.0]
(if no argument is present) or to the range
[rmin, rmax]
if the "-r"
argument is given or to the minimum
distance mindist if the "-d"
argument is used. Scaling to a minimum
distance ensures that all knots (except for multiple knots) have a distance
bigger than mindist afterwards.
Since Ayam 1.20 the knot type of the curve does
not have to be "Custom"
anymore. Furthermore, rescaling the knots
does not change the knot type.
This operation does not change the shape of the curve.
See also section Rescale Knots to Range Tool.
"splitNC u"
See also section Split Tool.
shiftC – shift control points of a (closed) curve:
"shiftC i"
See also section Shift Closed Curve Tool.
toXYC – move curve to XY plane
"toXYC"
See also section To XY Tool.
"trimNC umin umax"
See also section Curve Trim Tool.
estlenNC – estimate length of NURBS curve:
"estlenNC [-trafo] varname"
"-trafo"
is given, the transformation
attributes of the curve will be applied to the control points for the
length estimation.reparamNC – reparameterise a NURBS curve:
"reparamNC type"
"Custom"
."isCompNC [-l level]"
If the option "-l"
is 0, only the orders of the curves
are compared.
If the option "-l"
is 1, only the orders and lengths of the curves
are compared.
makeCompNC – make NURBS curves compatible
"makeCompNC [-f | -l level]"
If the option "-l"
is 0, only the orders will be adapted.
If the option "-l"
is 1, only the orders and lengths will be adapted.
See also section Make Compatible Tool.
clampuNP – clamp NURBS patch in U direction:
"clampuNP [-s | -e]"
"Custom"
and the knots will have o
equal values at start and end (where o is the order of the patch in
U direction).
If the side parameter is omitted, both sides are clamped.
If the side parameter is "-s"
only the start, and if it is "-e"
only the end is clamped.
In Ayam versions prior to 1.18 it was an error if the patch was already clamped at either side, this is no longer the case. Furthermore, patches with multiple knots in the end region(s) could not be clamped, this works ok now.
See also section Clamp Surface Tool.
clampvNP – clamp NURBS patch in V direction:
"clampvNP [-s | -e]"
"Custom"
and the knots will have o
equal values at start and end (where o is the order of the patch in
V direction).
If the side parameter is omitted, both sides are clamped.
If the side parameter is "-s"
only the start, and if it is "-e"
only the end is clamped.
In Ayam versions prior to 1.18 it was an error if the patch was already clamped at either side, this is no longer the case. Furthermore, patches with multiple knots in the end region(s) could not be clamped, this works ok now.
See also section Clamp Surface Tool.
unclampuNP – unclamp NURBS patch in U direction:
"unclampuNP [-s | -e]"
"Custom"
.
If the side parameter is omitted, both sides are unclamped.
If the side parameter is "-s"
only the start, and if it is "-e"
only the end is unclamped.
See also section Unclamp Surface Tool.
unclampvNP – unclamp NURBS patch in V direction:
"unclampvNP [-s | -e]"
"Custom"
.
If the side parameter is omitted, both sides are unclamped.
If the side parameter is "-s"
only the start, and if it is "-e"
only the end is unclamped.
See also section Unclamp Surface Tool.
rescaleknNP – rescale knots of NURBS patches:
"rescaleknNP [-r[u|v] rmin rmax | -d[u|v] mindist]"
[0.0, 1.0]
(if no argument is present) or to the range
[rmin, rmax]
if the "-r"
argument is given or to the minimum
distance mindist if the "-d"
argument is used. The "-ru"
, "-rv"
,
"-du"
, and "-dv"
variants scale only the designated dimension.
Scaling to a minimum distance ensures that all knots (except for multiple
knots) have a distance bigger than mindist afterwards. Trim curves, if
present, will also be scaled to match the new range.
Since Ayam 1.20 the knot type of the curve does
not have to be "Custom"
anymore. Furthermore, rescaling the knots
does not change the knot type.
This operation does not change the shape of the patch.
See also sections Rescale Knots to Range Surface Tool and Rescale Knots to Mindist Surface Tool.
"rescaleknNP -ru 0.2 0.3"
scales the u knot vector
of the selected NURBS patch objects to the new range (0.2, 0.3).insknuNP – insert knot into NURBS patch:
"insknuNP u r"
U[p] <= u <= U[n]
, where
p is the degree (order-1) of the patch in U direction and n is the
width of the patch.
The u knot type of the patches will always be changed to "Custom"
but
the shape of the patches will not change.
See also section Insert Knot Surface Tool.
insknvNP – insert knot into NURBS patch:
"insknvNP v r"
V[p] <= v <= V[n]
, where
p is the degree (order-1) of the patch in V direction and n is the
height of the patch.
The v knot type of the patches will always be changed to "Custom"
but
the shape of the patches will not change.
See also section Insert Knot Surface Tool.
remknuNP – remove u knot from NURBS surface:
"remknuNP (u | -i ind) r [tol]"
"remknuNP -i 3 1"
instead of
"remknuNP 0.5 1"
for the knot vector "0 0 0 0.5 1 1 1"
).
Note that the shape of the surface may be changed by this tool unless the
parameter tol is specified. If tol
is specified, the new surface does
not deviate from the original surface more than tol in any point.
If the knot can not be removed r times due to the tolerance, an
error is reported and the original surface is left unchanged.
See also section Remove Knot Surface Tool.
remknvNP – remove v knot from NURBS surface:
"remknvNP (v | -i ind) r [tol]"
"remknvNP -i 3 1"
instead of
"remknvNP 0.5 1"
for the knot vector "0 0 0 0.5 1 1 1"
).
Note that the shape of the surface may be changed by this tool unless
the parameter tol is specified. If tol
is specified, the new surface does
not deviate from the original surface more than tol in any point.
If the knot can not be removed r times due to the tolerance, an
error is reported and the original surface is left unchanged.
See also section Remove Knot Surface Tool.
refineuNP – refine NURBS surface in U direction:
"refineuNP [{u1 u2 un}]"
"Custom"
.
See also section Refine Knots Surface Tool.
refinevNP – refine NURBS surface in V direction:
"refinevNP [{v1 v2 vn}]"
"Custom"
.
See also section Refine Knots Surface Tool.
elevateuNP – elevate NURBS surface in U direction:
"elevateuNP n"
"Custom"
.
See also section Elevate Surface Tool.
elevatevNP – elevate NURBS surface in V direction:
"elevatevNP n"
"Custom"
.
See also section Elevate Surface Tool.
"splituNP u"
See also section Split Surface Tools.
"splitvNP v"
See also section Split Surface Tools.
"extrNP umin umax vmin vmax"
See also section Extract Patch Tool.
interpuNP – interpolate NURBS surface in U direction:
"interpuNP [-order order | -ktype type | -closed (0|1) | -sdlen length | -edlen length]"
Using the options "-sdlen"
and "-edlen"
(which both default
to 0.0) the length of automatically created start/end derivatives
can be adjusted. If any of these is not 0.0, a different interpolation
algorithm will be used, which increases the width of the resulting NURBS
surface.
The surface will interpolate all current control points after the interpolation and the position of certain control points will be changed in this process so that, after interpolation, the new control points will not be interpolated by the surface. The surface will interpolate the old control point positions.
The u knot type of the interpolated surfaces will be changed to "Custom"
.
interpvNP – interpolate NURBS surface in V direction:
"interpvNP [-order order | -ktype type | -closed (0|1) | -sdlen length | -edlen length]"
Using the options "-sdlen"
and "-edlen"
(which both default
to 0.0) the length of automatically created start/end derivatives
can be adjusted. If any of these is not 0.0, a different interpolation
algorithm will be used, which increases the height of the resulting NURBS
surface.
The surface will interpolate all current control points after the interpolation and the position of certain control points will be changed in this process so that, after interpolation, the new control points will not be interpolated by the surface. The surface will interpolate the old control point positions.
The v knot type of the interpolated surfaces will be changed to "Custom"
.
concatS – concatenate surfaces:
"concatS [-o order | -t type | -k knottype | -u uvselect]"
The option "-o"
determines the desired order
of the surface in U direction.
The option "-t"
allows to set a surface type
(0 – open, 1 – closed, 3 – periodic; default 0).
The option "-k"
allows to set a knot type (default 1 – NURB).
Finally, the "-u"
option allows to specify the uv-select-string.
See also section ConcatNPAttr Property for more information on these options.
tweenNP – interpolate (tween) surfaces:
"tweenNP [r]"
See also section Tween Surfaces Tool.
breakNP – break NURBS patch into curves:
"breakNP [-a | (-u | -v)]"
"-u"
or "-v"
is specified, default is u).
If the option "-a"
is specified, the transformations of the NPatch
objects will be applied to the control points and the NCurve objects
will be created with default transformation attributes, otherwise the
control points will be copied verbatim and the NCurve objects will get
the transformation attributes of the respective NPatch.
See also section Break into Curves Tool.
buildNP – build NURBS patch from curves:
"buildNP [-a (0|1) | -o order | -t type | -k knottype]"
The option "-a"
controls, whether the transformation attributes
of the NURBS curves shall be applied to the respective control points
before building the patch (default 1 – yes).
The option "-o"
determines the desired order of the surface in U
direction (default min(4,width)).
The option "-t"
allows to set a surface type
(0 – open, 1 – closed, 3 – periodic; default 0).
The option "-k"
allows to set a knot type (default 1 – NURB).
See also section Build from Curves Tool.
"isCompNP [(-u|-v) | -l level]"
If the option "-u"
is given, only the U dimension will be checked.
If the option "-v"
is given, only the V dimension will be checked.
If the option "-l"
is 0, only the orders of the surfaces
are compared.
If the option "-l"
is 1, width/height and orders of the surfaces are
compared.
makeCompNP – make NURBS surfaces compatible
"makeCompNP [-f | (-u|-v) | -l level]"
If the option "-u"
is given, only the U dimension will be adapted.
If the option "-v"
is given, only the V dimension will be adapted.
If the option "-l"
is 0, only the orders will be adapted.
If the option "-l"
is 1, only the orders and lengths will be adapted.
See also section Make Compatible Tool.
tobasisPM – convert PatchMesh to a different basis:
"tobasisPM [-t type | -s step | -b basis]"
The option "-t"
controls the new basis type
(0 - Bezier, 1 - B-Spline, 2 - CatmullRom, 3 - Hermite, 4 - Power,
5 - Custom), default is 1 (conversion to B-Spline).
The option "-s"
determines the new step size (1 to 4), it defaults
to the natural step size of the target basis type and thus can be
omitted safely, unless the target type is Custom, in which case the step
size must be specified.
The option "-b"
allows to convert to a custom basis and thus
is a list of 16 floating point values specifying a 4 by 4 basis matrix
in column major order. If "-b"
is given, the target type defaults
to Custom.
These are more specialized commands to change PolyMesh properties:
genfnPo – generate face normals:
"genfnPo"
The generated normals will be stored in a PV tag.
gensnPo – generate smooth normals:
"gensnPo"
Already existing vertex normals will be destroyed.
If face normals already exist, they will be used, otherwise, new face
normals will be generated using the same algorithm as implemented in the
"genfnPo"
command above.
remsnPo – remove smooth normals:
"remsnPo"
flipPo – flip normals or loops:
"flipPo [0|1|2]"
Use these two commands to read or manipulate the control points of objects that support point editing.
"getPnt [-trafo | -world | -eval] (index | indexu indexv | u | u v (varx vary varz [varw] | -vn varname) | -all varname)"
The index arguments needed depend on the type of the selected object, e.g. reading the points of a NURBS curve requires just one index parameter (index), whereas reading the points of a NURBS patch requires two index parameters (indexu and indexv) to be specified.
If the optional argument "-trafo"
is given, the coordinates
will additionally be transformed by the values given in the objects
Transformation property.
If the optional argument "-world"
is used, the coordinates
will additionally be transformed to world space.
If the optional argument "-eval"
is specified, the "indexu"
and "indexv"
values are interpreted as parametric values of a
NURBS curve or surface and the corresponding point on the curve or
surface is delivered in varx, vary, and varz.
If the alternative argument "-vn"
is given, the coordinate values
will be appended to the list variable specified by "varname"
.
If the alternative argument "-all"
is used, all coordinate values
of the selected objects will be appended to the list variable
specified by "varname"
.
"getPnt 1 x y z w"
"x y z w"
."getPnt -eval 0.5 x y z w"
"0.5"
and writes the
values to the variables "x y z w"
."setPnt [-world] (index | indexu indexv) (x y z [w] | -vn varname) | -all varname)"
The index arguments needed depend on the type of the selected object, e.g. manipulating the points of a NURBS curve requires just one index parameter (index), whereas manipulating the points of a NURBS patch requires two index parameters (indexu and indexv) to be specified.
If the optional parameter "-world"
is given, the coordinate values are
expressed in world space and will be transformed to appropriate
object space coordinates before setting.
If the optional parameter "w"
is omitted,
but the selected object has rational points, a default value of 1.0 will
be used for the weight.
If the alternative parameter "-vn"
is used, the coordinate
values will be read from the variable specified by "varname"
which must be a list of double values.
If the alternative parameter "-all"
is provided, all control
points of the selected objects will be set and the coordinate
values will be read from the variable specified by "varname"
which must be a list of double values.
When reading data from list variables, no precision will be lost as there are no double-string-double conversions involved.
"setPnt 1 0.0 0.2 0.3 1.0"
"0.0 0.2 0.3 1.0"
."setPnt -world 0 0 0 0"
"setPnt 2 1 0.0 0.2 0.3"
"0.0 0.2 0.3 1.0"
.These procedures update various parts of the Ayam user interface:
"rV"
"uS [update_prop maintain_selection]"
If update_prop is 0 no update of the property GUIs will take place.
If maintain_selection is 1 the old selection will be established again.
If both arguments are omitted update_prop defaults to 1 and maintain_selection to 0.
"ay(ul)"
(UpdateLevel) may be set
to the current level before calling "uS"
. This will not remove
and update the complete scene but just the part below "ay(ul)"
.
Example:
global ay; set ay(ul) $ay(CurrentLevel); uS;
"uCR"
may be used instead of "uS"
."uCL cl"
may be used instead of "uS"
."uCL mode [args]"
"uS"
above.
The parameter "mode"
may be "cl" or "cs", where "cl" is the normal
operation mode, and "cs" just clears the selection.uCR – update current level after create:
"uCR"
"uS"
above.plb_update – property listbox update:
"plb_update"
Since Ayam 1.13 it is also possible to automatically run GUI updating
commands in the console by using <Shift+Return>
instead
of <Return>
. The commands from the hidden preference setting
"AUCommands"
will be executed after the commands from the command
line, if the <Shift>
key is held down. <Shift+Return>
may also be used without commands on the command line.
By default, the "AUCommands"
are "uS; rV;"
, leading to
updated object tree, property GUI, and views.
These commands manage preferences data:
getPrefs – get preferences data:
"getPrefs"
setPrefs – set preferences data:
"setPrefs"
"ayprefs"
to let all changes take effect.This command manages custom objects (plugins):
"loadPlugin name"
name
is a complete
filename and the designated file exists, it will be loaded directly.
Otherwise, the file to load will be searched for in the list of configured
plugin directories (see "Plugins"
preference setting).
Note that it is currently not possible to unload a custom object from
Ayam.These procedures and commands help to apply arbitrary commands to a number of selected objects.
"forAll [(-recursive|-r) r | (-type|-t) t] command"
If r
is 1 (this is the default) then forAll will recurse into
every object (if it has child objects) before the execution of command.
If r
is 2, the recursion will happen after the execution of command.
If r
is 0, only objects from the current level will be processed.
If the "-type"
option is given, only objects of the specified type
t
will be processed.
Prior to Ayam 1.22, potential errors from the command(s) were suppressed and
processing continued regardless. But this made interactive usage and
debugging unnecessarily difficult.
From 1.22 on, errors are reported to the user and processing immediately
stops. Errors may still be suppressed using the "catch"
command
like this:
"forAll { catch { commands } }"
As the command will potentially be called multiple times, result
values of any kind can not be delivered using the "return"
command but should rather be collected in global variables.
The global variable "ay(CurrentLevel)"
will be maintained while
a recursive forAll browses the scene. In addition, the global variable
"i"
will be set to the index of the current object.
Note that forAll will run slowly if a property GUI is displayed. By de-selecting the property using e.g. the property context menu first, it will run much faster.
Furthermore, the current selection is correctly maintained.[*]
"-type"
option in this case."forAll { uplevel #0 { commands } }"
"forAll { global arrayname; commands }"
withOb – execute a command on certain selected object(s):
"withOb index [do] command"
"withOb 2 {movOb 0 1 0}"
moves the third
object from multiple selected objects. All objects stay selected.These commands help to load scenes from and save them to Ayam scene files:
"replaceScene filename"
"filename"
parameter."insertScene filename"
"filename"
parameter."saveScene filename [selected]"
"filename"
parameter. If the optional parameter "selected"
is 1,
only the selected objects will be saved."newScene"
This command allows to export the current scene to a RenderMan Interface Bytestream (RIB):
"wrib filename [-image imagename] [-smonly | -selonly | -objonly]"
"filename"
.
If the argument "-image"
is given, the RIB file
will create an image file named "imagename"
upon rendering. The export
will use the camera transformation from the currently selected Camera
object.
If the argument "-smonly"
is provided, a RIB to render shadow maps
will be created and the argument of "-image"
will be ignored.
If the argument "-selonly"
is used, only the selected (geometric)
objects will be exported, which will result in a RIB file not
suitable for rendering (no setup, camera transformation, or lights
are in it) but for inclusion into other scenes via RiArchive.
Likewise "-objonly"
leads to a RIB file containing all objects in the
scene but not suitable for rendering.
The "wrib"
command always needs a selected camera object
(unless the "-selonly"
or "-objonly"
options are given); if
there is none or if the camera transformations of the camera associated
with a view window shall be used, the corresponding Togl callback for
the view might be used like this instead:
.view1.f3D.togl wrib -file filename.rib
"wrib"
command.
"-filename "
, this is no longer the case.This command is for error reporting from scripts:
"ayError code place detail"
"ErrorLevel"
see section
Miscellaneous Preferences.These procedures help to manage property GUIs.
"addPropertyGUI name"
The array will be set up in a way that the data array of the property
will be named as the property with the string Data
appended,
i.e. for MyProperty
it will be MyPropertyData
.
The get/set procedure entries will be left empty.
After the creation of property GUI elements, "NP"
tags must
be used to make the new property visible to the user.
See also section
NP (New Property) Tag.
A complete example is available in section Script Object Examples.
"set w [addPropertyGUI MyProperty]"
addParam:
"addParam window arrayname paramname [defaults]"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter.
The "defaults"
parameter is a list
of default values. Those values will be presented to the user as an
additional drop down menu on the right side of the interface element.
"addParam $w MyPropertyData MyFloat {0.1 0.5 1.5}"
addString:
"addString window arrayname paramname [defaults]"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter.
The "defaults"
parameter is a list
of default values. Those values will be presented to the user as an
additional drop down menu on the right side of the interface element.
"addString $w MyPropertyData MyString {"a" "b" "abc"}"
addCheck:
"addCheck window arrayname paramname"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter.
"addCheck $w MyPropertyData MyBool"
addColor:
"addColor window arrayname paramname [defaults]"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter.
The "defaults"
parameter is a list
of default values. Those values will be presented to the user as an
additional drop down menu on the right side of the interface element.
"addColor $w MyPropertyData MyColor"
addMenu:
"addMenu window arrayname paramname choices"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter.
The "choices"
parameter is a list
of strings that will be presented in the menu.
In contrast to the other user interface element generating procedures, the corresponding entry in the property data array must exist before this procedure is called.
"addMenu $w MyPropertyData MyMenu {Choice1 Choice2}"
addFile:
"addFile window arrayname paramname [defaults]"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter.
The "defaults"
parameter is a list
of default values. Those values will be presented to the user as an
additional drop down menu on the right side of the interface element.
"addSFile"
"addFile $w MyPropertyData MyFile {"/tmp/file1" "/tmp/file2"}"
addCommand:
"addCommand window name text command"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "name"
parameter is the name of the corresponding
button widget. The names must be unique in each property GUI.
The "text"
parameter is the string to be put on the button.
The "command"
parameter is the command to be executed when the
button is pushed.
"addCommand $w b1 PushMe {puts pushed}"
addText:
"addText window name text"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "name"
parameter is the name of the corresponding
label widget. The names must be unique in each property GUI.
The "text"
parameter is the string to be displayed.
"addText $w t1 "Angular Parameters:""
addInfo:
"addInfo window arrayname paramname"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter whose
value is to be displayed.
"addInfo $w MyPropertyData NumGeneratedElems"
addProgress:
"addProgress window arrayname paramname"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "arrayname"
parameter is the name of the corresponding
data array of the property.
The "paramname"
parameter is the name of the parameter where
the progress is stored in percent.
"addProgress $w MyPropertyData Progress"
addVSpace:
"addVSpace window name height"
The "window"
parameter should contain the window name as returned by
"addPropertyGUI"
above.
The "name"
parameter is the name of the corresponding
widget. The names must be unique in each property GUI.
The "height"
parameter is the desired height in pixels.
"addVSpace $w v1 20"
Miscellaneous commands:
"tmpGet tmpdir varname [ext]"
"undo [redo | save opname [0|1] | clear | rewind]"
"redo"
, this command performs the redo
operation."save"
, the currently selected objects
are saved to the undo buffer for future undo operations. The name
of the now following modelling operation has to be provided in
a second argument ("opname"
).
This name will be displayed in the default console prompt, to inform the
user about which operation would be undone/redone, if undo/redo would be
used (e.g. "[Undo:MoveObj/Redo:none].../bin>"
).
Since Ayam 1.13, a third argument may be given, that controls whether
all the children of the selected objects should also be saved. This may
be needed if the modelling action that follows the undo save is about to
change the selected objects and also their children.
Note: undo save does not fail if no objects are selected."clear"
, all currently saved states
will be cleared from the undo buffer."rewind"
is available since Ayam 1.14. With
this command you can undo the last undo save operation. This may be
necessary, if a modelling operation failed. Care should be taken,
however, to not rewind the undo state, when a modelling
operation only failed for some (not for all) of the selected objects.undo save "MovOb" set ay_error "" movOb 0 1 0 if { $ay_error > 1 } { undo rewind }
"convOb [-inplace | -check type]"
If the option "-inplace"
is
used, the new object(s) will replace the old object(s).
If the option "-check"
is given, the convOb command does not
convert but check, whether a conversion to a given object type would be
successful, the result of the check will be returned as 0 – no,
or 1 – yes.
"notifyOb [-all | -modified | -parent]"
If the "-modified"
parameter is used, only modified objects
will be notified.
If the "-all"
parameter is used, all objects
will be notified regardless of the selection.
If the "-parent"
parameter is used, only the current
parent object of the current level will be notified.
Prior to Ayam 1.20 this command was named "forceNot"
,
the old name is still available for compatibility but its
use is deprecated.
∗: Since 1.21 this command is also available in the safe interpreter with limited functionality: only the notification callbacks of the selected objects will be executed. No BNS or ANS tags will be considered, notification of parents will not be done, and the complete notification will also be omitted. Hence, in the safe interpreter, this command ignores all parameters.
"addToProc procedure addition"
"return;"
statement.Various entries of dialogs for object creation and modelling tools support Tcl variables and expressions.
It is e.g. possible to enter
$::u
u
(that
may have been set before using the find u modelling action) and
insert a knot at the picked point.
It is also possible to enter complex mathematical expressions:
[expr sin(45)]
or call into own procedures (that have to return appropriately typed values):
[myproc]
"myproc"
is defined elsewhere (e.g. in a Tcl script file
loaded via the Scripts preference setting) as follows:
proc myproc { } { return [expr sin(45)]; }
Repeated calling of the tool without opening the dialog (using
the keyboard shortcut <Ctrl+T>
), will execute the
provided expression again.
This means, a number of curves with increasing length can be created
by entering into the Ayam console
set ::myvar 1
[incr ::myvar]
<Ctrl+T>
multiple times.
Here are some complete example scripts for the Ayam Tcl scripting interface.
All examples may be copied from the documentation and pasted directly into the console of Ayam.
The following example script shows how to move a selected object to a specified position in space.
proc placeOb { x y z } { global transfPropData # copy Transformations-property data to # global array "transfPropData" getTrafo # set array values according to procedure parameters set transfPropData(Translate_X) $x set transfPropData(Translate_Y) $y set transfPropData(Translate_Z) $z # copy Transformations-property data from # global array "transfPropData" to selected object setTrafo } # placeOb
forAll -recursive 0 {placeOb 1 1 1}
"placeOb"
procedure (defined above) with them:
global ay $ay(cm) add command -label "Place Object" -command { runTool {x y z} {"X:" "Y:" "Z:"} "forAll -recursive 0 {placeOb %0 %1 %2}" plb_update; rV }
"plb_update; rV"
command ensures that the GUI is updated
properly and all views display the new position of the moved objects.
The following example script snippet shows how to move control points of a NURBS curve.
# first, we create a new NURBS curve with 30 control points set len 30 crtOb NCurve -length $len # update selection uS # select last object (the newly created curve) sL # prepare moving set i 0 set r 3.0 set angle 0 set angled [expr 3.14159265/2.0] while { $i < $len } { set x [expr $r*cos($angle)] set y [expr $r*sin($angle)] set z [expr $i/3.0] # move control point to new position setPnt $i $x $y $z 1.0 set angle [expr $angle + $angled] incr i } # redraw all views rV
The following example script shows how to easily create a sweep from a selected path curve (avoiding the manual and lengthy creation and parameterisation of a suitable cross section).
proc easySweep { } { # first, we create a sweep object crtOb Sweep # now, we need to move the selected curve (path) to # the sweep and create a cross-section curve there too # for that, we move the currently selected curve to the clipboard cutOb # enter the Sweep (the last object in the current level) goDown -1 # now, we create a new curve (a closed B-Spline suitable as cross section) crtClosedBS -s 8 # select the new object selOb 0 # now, we rotate and scale the curve rotOb 0 90 0 scalOb 0.25 0.25 1.0 # move trajectory back (we use "-move", because we # really want to move (and not copy) the curve object pasOb -move # go up to where we came from goUp # finally, update the GUI... uS sL # ...and redraw all views rV } # easySweep
Run this procedure by selecting a NURBS curve object, then type into the console:
» easySweep
This command may be added to the main menu as well:
global ay $ay(cm) add command -label "Easy Sweep" -command { easySweep }
"Custom/Easy Sweep"
that calls the easySweep
procedure.
Here is another example script that shows how you may add buttons to the toolbox. myImage should be an image created e.g. from a GIF file of the size 25 by 25 pixels.
global ay ayprefs # create an image from a GIF file: image create photo myImage -format gif -file /home/user/giffile set b $ay(tbw).mybutton # if the button does not already exist: if { ![winfo exists $b] } { # create it: button $b -padx 0 -pady 0 -image myImage -command myCommand # tell Ayam about the new button: # you can use "linsert", to insert the button in a specific # place or just append to the end of the list using "lappend" lappend ay(toolbuttons) mybutton # display the button: toolbox_layout # from now on, the button will be under the # automatic toolbox layout management }
This example shows that a) toolbox buttons have to be created in the
frame ".tbw.f"
for multi-window GUI configurations or
".fv.fTools.f"
for single-window GUI configurations,
b) Ayam manages a list of all toolbox buttons in the global array
"ay"
in "ay(toolbuttons)"
, the order in that list is the order
in which the buttons appear in the toolbox, c) automatic layout
management is carried out by the procedure "toolbox_layout"
.
Adding buttons with just text is a little bit more involved,
as the sizes of those buttons often do not fit well in the
icon button scheme with its constant button size. However,
the procedure "toolbox_add"
can be of
considerable help.[*]
See also the script "scripts/topoly.tcl"
for an example.
The following example script adds two buttons to the bottom of the toolbox spanning the whole window (this works best with the standard toolbox layout of 4 by 12 buttons used in the multi-window GUI configuration):
global ay # create a frame: set f [frame $ay(tbw).fcollex] # calculate the row number below the last row: set row [expr [lindex [grid size $ay(tbw)] 1] + 1] # now display the frame at calculated row, spanning the whole window: grid $f -row $row -column 0 -columnspan [lindex [grid size $ay(tbw)] 0]\ -sticky we # create two buttons inside the frame: button $f.b1 -width 5 -text "Coll." -command { collMP; rV; } button $f.b2 -width 5 -text "Expl." -command { explMP; rV; } pack $f.b1 $f.b2 -side left -fill x -expand yes
This sections contains the documentation of some helper scripts that are distributed with Ayam.
The helper scripts may be run via the context menu of the console, the
Tcl "source"
command, or the "Scripts"
preference
setting of Ayam on each start (the latter except for repairAyam.tcl and
bgconvert.tcl).
The external Tcl script "repairAyam.tcl"
may be used to repair the application state of Ayam,
should it be stuck e.g. in an endless loop of Tcl error
messages.[*]
On Unix systems "repairAyam"
may be started from any
shell simply by typing
» ./repairAyam.tcl
or
» wish repairAyam.tcl
on the command prompt; if the script detects that it is running on Unix
and not in Ayam it will send itself to the Tcl interpreter Ayam is
running in using the Tk send command. On Mac OS X Aqua (not X11!)
AppleScript events will be used instead of the Tk send command.
If this does not work as expected "repairAyam.tcl"
may still
be run via the Ayam console (as on Win32).
On Win32 "repairAyam.tcl"
has to be started from the Ayam console
using the command:
» source scripts/repairAyam.tcl
or via the consoles context menu: "Console/Load File"
.
The script "repairAyam.tcl"
should be considered a
last resort to help saving the current state of
modified objects.
The script will close all views, clean up the application state variables, reset the mouse cursor and the console prompt, and try to update important main window widgets.
Furthermore, the script will also clear the console and try to break potential endless loops running e.g. in the console or in Script objects.[*]
After running "repairAyam.tcl"
the scene (or the most
important objects currently worked on) should be immediately
saved to a new scene file, not the file currently loaded,
using "File/Save As"
or
"Special/Save Selected"
) and Ayam should be restarted
afterwards.
Simply saving the scene using "File/Save"
or <Ctrl+s>
should be avoided because views were possibly deleted.
The script "topoly.tcl"
recursively browses through the scene and converts everything
to a polygonal representation.[*]
After running the script, there is a new button in the toolbox
named "ToPolyMesh"
. Additionally, there is a corresponding
entry in the "Custom"
main menu.
Pressing the button or using the menu entry immediately starts the
conversion process.
Since the changes of the conversion can not be undone, the conversion will not run if the scene contains unsaved changes.
The conversion will use the current parameters from the preference settings
"SMethod"
, "SParamU"
, and "SParamV"
; "TP"
tags (if present) will override these parameters.
TP tags may be created easily using the tesselation tool, see also
section
Tesselation Tool.
The script "tonpatch.tcl"
recursively browses through the scene and converts everything
to a NURBS patch representation effectively flattening the tool object
hierarchy.[*]
After running the script, there is a new button in the toolbox
named "ToNPatch"
. Additionally, there is a corresponding
entry in the "Custom"
main menu.
Pressing the button or using the menu entry immediately starts the
conversion process.
Since the changes of the conversion can not be undone, the conversion will not run if the scene contains unsaved changes.
The script "2lcons.tcl"
(for two line console) may be
used to restrict the screen space occupied by the console.
Normally, the Ayam console is resized with the main window and occupies a varying amount of screen space. After running the script, the console will always resize to exactly two lines of text. Different values may be chosen easily by adapting the script.
The script "colfocus.tcl"
(for colored focus) may be
used to paint the focus ring in a more visible color.
After running the script, the focus ring will be painted in blue (instead of black): focused sub-windows (views, console, object tree) will be more easily recognizable. Other colors may be used by editing the script.
The script "aac.tcl"
(for automatic about center)
may be used to switch all modelling actions to their about variants with
the mark set to the center of the current selection automatically.
After running the script, invoking e.g. the scale 2D action using
the shortcut <s>
will:
<sac>
)<saC>
)The script modifies all rotate and scale actions (including their axis confined variants).
Note, that the mark is not reset to a new center, when the selection changes. After a selection change (e.g. by selecting points in a different view) simply restart the action to transform about the new center.
To rotate or scale about a different point than the center, the mark
may still be set manually using <a>
.
To temporarily disable the modified behavior, the global keyboard
shortcut <F11>
can be used.
The script "apnt.tcl"
(for automatic point) may be
used to switch the modelling mode to point modelling automatically
after a point selection.
After running the script, selecting (tagging) a point using the
select point action (shortcut <t>
) will automatically
switch the view to point modelling so that the next modelling
actions (e.g. move, via shortcut <m>
) will always
transform the points and not modify the objects transformations.
Note that currently the switch to point modelling will also occur,
if no points are actually selected, it is just the mouse click that
counts.
Selecting all points via the keyboard shortcut <A>
will switch to point modelling and de-selecting all points via
<N>
will switch to object
modelling.[*]
It is also still possible to switch back to object modelling anytime
via the keyboard shortcut <o>
.
To temporarily disable the modified behavior, the global keyboard
shortcut <F12>
can be used.
The script "ssp.tcl"
(for save selected points)
allows to save the point selection to tags of type
SP
.[*]
After running the script, two new buttons appear in the toolbox that allow to save and restore the point selection respectively. There are also two corresponding entries in the custom menu.
Note that the tags can be saved to scene files and also copied to different objects.
The script "rc.tcl"
(for revert cursor) may be
used to get more useful cursor key behavior in primary modelling
views (parallel views).
After running the script, the keyboard shortcuts for rotating
and panning in parallel views are swapped, e.g. just pressing
<Left>
key will then pan the view, instead of rotating it.
The shortcuts will be swapped again, when the view changes
type to "Perspective"
.
The external Tcl script "bgconvert.tcl"
converts scene files from one 3D file format to another,
with the help of Ayam which is running in the
background.[*]
In the most simple form, bgconvert may be used from a Unix command line (or shell script) like this:
» bgconvert.tcl infile.x3d outfile.dxf
The above command would load the X3D file "infile.x3d"
into Ayam and export the scene as DXF file to "outfile.dxf"
.
For a successful conversion Ayam has to run and the plugins required
for the import and export processes need to be available and properly
configured (check the "Plugins"
preference setting).
The plugins necessary for the conversion will be loaded automatically.
Import and export options may also be given like this:
» bgconvert.tcl "infile.rib -p 1" outfile.dxf
In the example above the "-p 1"
option switches on reading
of partial RIB files.
Available options and their syntax may be inquired from the
import and export plugin Tcl scripts (e.g. "plugins/rrib.tcl"
).
The script "zap.tcl"
demonstrates, how arbitrary core
functionality that is just available through a main menu entry
or the scripting interface might be accessed easily via the
toolbox window.
After running the script "zap.tcl"
, there will be a new
toolbox button, labeled "Zap!"
, that simply runs the zap
command (which iconifies the complete application).
The script "kdialog.tcl"
switches all file dialogs of
Ayam to use the kdialog application of the KDE project instead
of the native Tk file dialog.
The script "zdialog.tcl"
switches all file dialogs of
Ayam to use the zenity application of the Gnome project instead
of the native Tk file dialog.
The script "useaqsisapp.tcl"
sets up Ayam to use Aqsis
from the application directory structure ("/Applications/Aqsis.app"
)
on Mac OS X. This is the default installation location of Aqsis on Mac OS X.
The script adapts the executable and shader search paths. Furthermore, environment variables vital for Aqsis to work will be set up properly.
Note that the script does not change the "RIB-Export/Renderer"
preferences, you still have to switch to Aqsis using the main menu
"Special/Select Renderer"
once.
The script "usepixie.tcl"
sets up Ayam to use Pixie
from the "/Library/pixie"
directory on Mac OS X.
This is the default installation location of Pixie on Mac OS X.
The script adapts the executable, shared library, and shader search paths. Furthermore, environment variables vital for Pixie to work will be set up properly.
Note that the script does not change the "RIB-Export/Renderer"
preferences, you still have to switch to Pixie using the main menu
"Special/Select Renderer"
once.
The script "myicons.tcl"
allows to replace the icons of Ayam with a set of user defined ones.
The new icons must be GIF image files of size 25 by 25 and reside
in the "icons"
directory relative to the Ayam executable.
The names of the image files may be obtained from the script or by the following scripting interface command (in the Ayam console):
» image names
Action icon variants (e.g. for the scale about actions) can also be created
automatically by changing the "createVariants"
variable in
the script file.
The script "dtree.tcl"
(dynamic/fast tree) replaces some of
BWidgets tree code for faster interaction with many objects.
If activated, the tree widget no longer creates a canvas item for every node. Instead, there are only as many canvas items as there are nodes visible in the current scroll region. Therefore, working with many objects in long lists becomes much faster.
The script "cvview.tcl"
(control vertice view) allows to
view the control points of NURBS surface or curve objects as property,
see also the image above.
The script currently supports the following object types: NPatch, NCurve, IPatch, PatchMesh, ACurve, and ICurve.
If the script is loaded, there is a new entry in the "Custom"
menu
that allows to add a new property to an individual surface or curve object.
The "CVView"
property displays all control vertices in a
regular grid. The coordinates of each vertice will be shown when
the mouse pointer hovers over it. This also works if the control points
lump together in 3D space or have unusual values.
Additionally, the point selection may be adjusted by clicking on the circles.
The script "tcone.tcl"
creates truncated cones with arguments
similar to the cylinder primitive, see also the above image.
This script must be used in a Script object of type "Create"
(see section
Script Object).
For convenience, there is also a property GUI; one must add a "NP"
tag
of value "TConeAttr"
to the Script object to see it.
These are the parameters of the truncated cone:
"Closed"
toggles whether the object should
be automatically sealed (closed by matching cap surfaces)."ThetaMax"
is the sweeping angle of the cone in degrees,
default is 360.
"ZMin"
is the base of the cone, default is 0."ZMax"
is the peak of the cone, default is 1."RMin"
is the radius of the cone at the base, default is 1."RMax"
is the radius of the cone at the peak, default is 0.5.Internally, the script creates a Hyperboloid; further information about conversion capabilities and RIB export may be found in section Hyperboloid.
An example scene file containing such an object is distributed with Ayam,
see the file:
"ayam/scn/scripts/tcone.ay"
.
Since Ayam 1.18 there is a complete example script for the JavaScript
scripting interface distributed as "polyhedron.js"
which
creates polyhedrons from Conway notations.
The script is based on the online Polyhedron VRML generator by
George W. Hart:
http://www.georgehart.com/virtual-polyhedra/conway_notation.html
This script must be used in a Script object of type "Create"
(see section
Script Object).
For convenience, there is also a property GUI; to make this GUI visible
a "NP"
tag of value "PolyhedronAttr"
must be added to the Script
object.
The Conway notation defines a set of operations executed consecutively on a seed/basic shape. The script currently supports the following seeds and operations (information taken from George W. Harts fine web pages, see above).
Seeds:
The Platonic solids are denoted T, O, C, I, and D, according to their
first letter. Other polyhedra which are implemented here include prisms:
Pn, antiprisms: An, and pyramids: Yn, where n is a number (3 or greater)
which must be specified to indicate the size of the base, e.g.,
Y3=T, P4=C, and A3=O.
Operations:
Currently, d, t, k, a, j, s, g, e, b, o, m, r, and p are defined.
They are motivated by the operations needed to create the Archimedean
solids and their duals from the Platonic solids. The following tables
explain the operations in more detail:
Letter | Name | Description |
d | dual | The dual of a polyhedron has a vertex for each face, and a face for each vertex, of the original polyhedron, e.g. dC=O. |
t / t n | truncate all / just n-fold vertices | Truncating a polyhedron cuts off each vertex, producing a new n-sided face for each n-fold vertex. |
k / k n | kis all / just n-sided faces | The kis operation divides each n-sided face into n triangles. A new vertex is added in the center of each face. |
a | ambo | The ambo operation can be thought of as truncating to the edge midpoints. It produces a polyhedron, aX, with one vertex for each edge of X. |
j | join | The join operator is dual to ambo, so jX=dadX=daX. jX is like kX without the original edges of X. |
e | expand | Each face of X is separated from all its neighbors and reconnected with a new 4-sided face, corresponding to an edge of X. An n-gon is then added to connect the 4-sided faces at each n-fold vertex. |
s | snub | The snub operation can be thought of as eC followed by the operation of slicing each of the new 4-fold faces along a diagonal into two triangles. With a consistent handedness to these cuts, all the vertices of sX are 5-fold. |
g | gyro | The dual operation to s is g. g is like k but with the new edges connecting the face centers to the 1/3 points on the edges rather than the vertices. |
b | bevel | The bevel operation can be defined by bX=taX. |
o | ortho | Dual to e, oX=deX=jjX. oX has the effect of putting new vertices in the middle of each face of X and connecting them, with new edges, to the edge midpoints of X. |
m | meta | Dual to b, m is like k and o combined; new edges connect new vertices at the face centers to the old vertices and new vertices at the edge midpoints. |
Letter | Name | Description |
r | reflect | Changes a left-handed solid to right handed, or vice versa, but has no effect on a reflexible solid. So rC=C, but compare sC and rsC. |
p | propellor | Makes each n-gon face into a "propellor" of an n-gon surrounded by n quadrilaterals, e.g. pT is the tetrahedrally stellated icosahedron. Try pkD and pt6kT. p is a self-dual operation, i.e., dpdX=pX and dpX=pdX, and p also commutes with a and j, i.e. paX=apX. |
These abbreviated explanations were again taken from George W. Hart.
The "x3dom-nurbs"
script implements the
<NurbsPatchSurface>
and <NurbsTrimmedSurface>
X3D NURBS nodes for x3dom (see http://www.x3dom.org/
) in
JavaScript.
After loading of the scene into the web browser, the NURBS surfaces
from these nodes are tessellated into <IndexedTriangleSet>
nodes.
This allows to directly publish NURBS
models on the web without prior conversion to a polygonal representation
which is cumbersome, inflexible, and leads to higher bandwidth consumption.
The tessellator is based on idea and example code from A. J. Chung and A. J. Field: "A Simple Recursive Tessellator for Adaptive Surface Triangulation" in Journal of Graphics Tools Vol. 5, Iss. 3, 2000.
The implementation spans four script files:
x3dom-nurbs-nodes.js
– interface of the tessellator to x3domx3dom-nurbs-pool.js
– worker management (current pool size is 3)x3dom-nurbs-worker.js
– a workerx3dom-nurbs-tess.js
– the tessellatorIn order to use the tessellator just add the following to your XHTML file
after inclusion of "x3dom.js"
:
<script type="text/javascript" src="x3dom-nurbs-pool.js"/>
<script type="text/javascript" src="x3dom-nurbs-nodes.js"/>
As the tessellator is fully automatic, no further adjustments are needed.
Proper XHTML files can be created using the X3D export in x3dom-mode, see section X3D (Web3D) export options.
While the tessellator runs in the background an initial polygonal
representation that is directly derived from the control polygon of
the NURBS surface is shown. Additionally, a busy prompt is displayed,
see also the image below.
As the busy prompt is derived from the x3dom loading prompt, it may also
be styled with the "x3dom-progress"
style in "x3dom.css"
.
The tessellation runs in so called webworkers, processes that run in
parallel to the main browser thread, in order to not block user interaction
and employ multi-core CPUs.
The webworker pool is currently hard coded to only use three such
webworkers at any given time. This value may be adapted easily in
"x3dom-nurbs-pool.js"
(e.g. if your CPU has more cores).
Note that each webworker will tessellate a single NURBS surface, i.e. scenes
with only one surface will not benefit from the parallelism.
By default, the tessellator tries to create a tessellation that represents all important surface features in a way that a visual inspection from mid to close viewing distance does not reveal the nature of the underlying triangular representation. This may be too fine/slow for objects that are never viewed from close distance or too coarse for very detailed objects.
Therefore, the tessellation quality and speed may be adjusted using the
"uTessellation"
and "vTessellation"
attributes as
explained in the next sections.
If no "uTessellation"
attribute is specified, or its value
is positive, the tessellator uses the so called object space sampling
mode. In this mode the tessellator subdivides an initial set of triangles
recursively until all edges of those triangles are shorter than a given
threshold value (in object space). The threshold value is set automatically
so that an object spanning one unit by one unit in object space is subdivided
to about 15 by 15 by two triangles.
The "uTessellation"
attribute value is simply multiplied into
this automatically determined threshold.
Therefore "uTessellation"
values larger than 1.0 lead to a coarser
and faster tessellation, whereas values smaller than 1.0 lead to a
finer and slower tessellation.
Note that the value of the "vTessellation"
attribute is not considered.
Also note that in highly curved regions and at trim edges, even smaller triangles than determined by the edge length criterion may be created.
If the "uTessellation"
attribute value is negative, the
tessellator switches to parametric space sampling.
In this mode also the "vTessellation"
attribute is considered.
As in the object space sampling mode an automatic threshold value is
computed, but here the extension of the object in parametric space
(i.e. the number of control points) is used. This actually results
in two threshold values, one for each parametric dimension.
The "uTessellation"
and "vTessellation"
attributes are
multiplied into those thresholds.
Therefore, values larger than -1.0 (in absolute value) lead to a coarser and faster tessellation in the respective dimension. Values smaller than -1.0 (in absolute value) lead to a finer and slower tessellation in the respective dimension. This means a value of -2.0 leads to roughly half as many triangles compared to the default and a value of -0.5 leads to twice as many triangles when compared to the default in the respective parametric dimension.
As the edge length computation is simpler and no curvature analysis is taking place, the parametric space sampling is considerably faster than object space sampling.
Note that at trim curve edges even smaller triangles than determined by the edge length criterion may be created.
To aid in parameterisation of the "uTessellation"
and
"vTessellation"
attributes, a "normalPerVertex"
attribute
can be added to the respective <NurbsPatchSurface>
or
<NurbsTrimmedSurface>
node. This attribute will then also
be set for the corresponding <IndexedTriangleSet>
node
that is created by the tessellator.
If the value of this attribute is "false"
, x3dom will display
this surface in a flat shaded style and the tesselated triangles will be
visible allowing easier judgment and adjustment of the tesselation quality.
Due to memoization of surface points, parametric values must not exceed:
Texture coordinates are always directly derived from the parametric values.
Surface normals are not computed by the tessellator, but by x3dom. This can lead to normals that are off for very coarse tessellations but is much faster. Another benefit of this approach is, that normals in non-differentiable surface points (e.g. the poles of the standard NURBS sphere) do not flip.
There is no support for the following nodes:
<NurbsSet>
,<NurbsTextureCoordinate>
,<NurbsSweptSurface>
,<NurbsSwungSurface>
,<NurbsCurve>
, and all<Nurbs*Interpolator>
nodes.This sections contains the documentation of the JavaScript scripting
interface which is available after loading of the "jsinterp"
plugin.
The JavaScript scripting interface exists since Ayam 1.18 and is based on the Mozilla SpiderMonkey JavaScript engine.
Upon loading, the "jsinterp"
plugin creates one JavaScript
context that lives (with all variables and objects defined therein)
until Ayam exits.
The JavaScript functionality may be accessed from the Tcl scripting interface
via the "jsEval"
command. The command can be used either to
directly execute JavaScript code provided via the commands argument
(Tcl code in bold):
» jsEval {var a = 0; a = a + 5.5; tclset("a", a);}
or to execute JavaScript code from a file:
» jsEval -f scriptfile.js
Note, that this command is not available in the safe interpreter.
Furthermore, Script object scripts may also be implemented in JavaScript, provided the first line of the script is a comment that instructs Ayam to use the JavaScript interpreter:
/* Ayam, use: JavaScript */
var a = 0;
...
Note that the JavaScript scripting context inherits the limitations
of the calling Tcl context.
For example, when running in a Script object,
the following code fails:
tcleval("exit");
because the Tcl command "exit"
is not available in the safe interpreter.
The command will not fail, when the calling context is the main Tcl
interpreter; one can e.g. type into the Ayam console:
» jsEval {tcleval("exit");}
and Ayam quits (see also section:
Safe Interpreter).
This subsection informs about the global functions additionally available in the Ayam JavaScript interpreter.
Those are converted Tcl commands, "tcleval"
, "tclvar"
,
and "tclset"
.
The functionality of Ayam is accessible from JavaScript via a larger set of global functions, named as the corresponding Tcl commands. For instance, Ayam objects can be created in JavaScript using a function call like this:
crtOb("NCircle");
or, with additional arguments:
crtOb("NCircle", "-radius", 3.0);
In general, all commands available in the safe Ayam Tcl interpreter are also
available as converted function (refer to section
Procedures and Commands
for a more or less complete list of those commands).
Note that Tcl procedures are generally not available as global JavaScript
function, but they can be called using "tcleval"
as documented
in the next paragraph.
tcleval:
Apart from Tcl commands converted to JavaScript functions, there is the
global JavaScript function "tcleval"
, that allows to evaluate
arbitrary Tcl scripts, delivered as string argument:
var a = 42;
a = tcleval("puts " + a + "; return 5;");
tcleval("puts " + a);
/* expected output: 42 5 */
The "tcleval"
function provides access to all the
functionality of Ayam that is just available as a Tcl procedure.
Note that return values are properly transferred back to JavaScript
according to the rules for data conversion as documented below.
However, due to the conversion to string data, the overhead of
such a call is considerable and data transport should be arranged
by other means, see below.
tclvar:
Using the JavaScript function "tclvar"
a link between a Tcl
variable and a corresponding variable in the
JavaScript context may be established.
The "tclvar"
function essentially creates a write trace on the Tcl
variable, so that changes on the Tcl side are always automatically
reflected on the JavaScript side:
tclvar("a");
tcleval("set a 42");
tcleval("puts " + a);
/* expected output: 42 */
Mind that the corresponding variable on the JavaScript side does
not exist until the first write operation on the Tcl variable
occurs.
The Tcl variable, in turn, does not have to exist, when the
"tclvar"
function is called (i.e. all the work is done
in the trace callback). If the variable name contains a namespace
specifier, this namespace has to exist, when "tclvar"
is called.
Even though it looks a perfect fit, "tclvar"
can not be used to
manage a property data array (if the array contains components
to be saved to Ayam scene files). This is, because upon reading a scene
file with such saved array items, the items will be read (and put into
the Tcl context) before the script can establish the write trace using
"tclvar"
and the data from the scene file never arrives in the JavaScript
context.
There is no easy way to get around this.
A suggested way to manage a property data array is shown in the
complete examples section below.
tclset:
The third global JavaScript function is "tclset"
that
allows to efficiently set Tcl variables from the JavaScript context
avoiding conversion to string data and back.
For example:
var a = 3.3;
var b = new Array(1, 3, 5);
tclset("a", a);
tclset("b", b);
sets the Tcl variable "a"
to the floating point value 3.3,
and "b"
to a list of integer values{ 1 3 5 }
.
Note that the variable names may also point to Tcl array elements:
tclset("SphereAttrData(Radius)", 1.2);
or contain namespace specifiers
tclset("::MyNameSpace::Radius", 1.2);
.
When data is transferred from the Tcl to the JavaScript side
(e.g. while converting return values of "tcleval"
or variable
values linked via "tclvar"
),
the following conversions are in effect:
Scalar data types will be converted to their directly matching counterparts,
except for Booleans, which will be converted to integer values.
Lists will be converted to Array objects (nesting is allowed and will
produce accordingly nested arrays).
Associative arrays will be converted to objects with named properties.
Unicode strings are currently not supported.
See also the table below.
Tcl | JavaScript |
Boolean (true, false) | Integer (1, 0) |
Integer (2) | Integer (2) |
Double (3.14) | Double (3.14) |
String ("mystr") | String ("mystr") |
List ({0 1 2}) | Array ((0, 1, 2)) |
Array (mya(mye) = 0.1) | Object (mya.mye = 0.1) |
When data is transferred from the JavaScript side to the Tcl side (e.g. as function argument), the following conversions are in effect: Scalar data types will be converted to their directly matching counterparts, Array objects will be converted to lists (nesting is allowed and will produce accordingly nested lists). Unicode strings and objects of a type other than Array (e.g. Boolean) are currently not supported. See also the following table.
JavaScript | Tcl |
Integer (2) | Integer (2) |
Double (3.14) | Double (3.14) |
String ("mystr") | String ("mystr") |
Array ((0, 1, 2)) | List ({0 1 2}) |
The transport/conversion of object properties (to e.g. associative array elements) can be arranged manually like this:
var a = new Object();
a.b = 3.14;
tclset("a(b)", a.b);
This section contains two complete examples for Script objects written in JavaScript.
For the first example use Script object type "Modify"
and put a Sphere
as child object of the Script object.
/* Ayam, use: JavaScript */ tclvar("SphereAttrData"); getProp(); if(SphereAttrData) { tclset("SphereAttrData(ZMin)", -SphereAttrData.Radius); tclset("SphereAttrData(ZMax)", SphereAttrData.Radius); setProp(); }
First, a link from the original Sphere object property data array
"SphereAttrData"
is established, so that when "getProp()"
(a converted Tcl Ayam command) is called, also the JavaScript
object "SphereAttrData"
is filled with meaningful data.
The next line (the if) is a safety measure that prevents the script from
failing if the child object of the Script object is not a Sphere
object.
Now the radius value is transferred back to Tcl directly into
the property data array to the ZMin and ZMax entries respectively
with the help of "tclset"
.
Finally the modified property is transferred back to the Sphere object
again with a converted Tcl Ayam command "setProp()"
.
The next example shows, how to manage a property GUI in a JavaScript
implemented Script object script. Use Script object type "Create"
and add a tag "NP MyProp"
to see the property GUI.
/* Ayam, use: JavaScript, save array: MyPropData */ var MyPropData = new Object(); if(!tcleval("info exists MyPropData;")) { /* initial script run (but not when loaded from scene file!) */ MyPropData.MyItem = tcleval("set MyPropData(MyItem) 1.0;"); tcleval("set MyPropData(SP) {MyItem};"); } else { /* all following script runs (and also when loaded from scene file!) */ MyPropData.MyItem = tcleval("set MyPropData(MyItem);"); } if(!tcleval("info exists MyPropGUI;")) { tcleval("set ::phw [addPropertyGUI MyProp \"\" \"\"];"); tcleval("addParam $::phw MyPropData MyItem;"); } crtOb("Sphere"); sL(); getProp(); tclset("SphereAttrData(Radius)", MyPropData.MyItem); tclset("SphereAttrData(ZMin)", -MyPropData.MyItem); tclset("SphereAttrData(ZMax)", MyPropData.MyItem); setProp();
This example demonstrates how to manage property data using the
JavaScript object variable "MyPropData"
. The property data
can be saved to and read from Ayam scene files with the help of
a mirroring array variable on the Tcl side (also named
"MyPropData"
).
To make this work properly, the initialisation of the JavaScript
object must be constrained to the first script run: when the
property data was read from a scene file, initialisation must
not be run, instead the read data must be fetched from the Tcl
context.
This is what the first "if"
statement, checking for existence
of the mirroring Tcl array variable, in above example is all about.
Following this scheme of dual mirroring data structures on the
Tcl and JavaScript sides, now the property GUI is created, which is also
constrained to just one script run by a similar "if"
statement.
After the GUI, a Sphere object is created and parameterized according to the data in the property GUI, which is used as radius, zmin, and zmax value.
This sections contains the documentation of the Lua scripting
interface which is available after loading of the "luainterp"
plugin.[*]
Upon loading, the "luainterp"
plugin creates one Lua
context that lives (with all variables and objects defined therein)
until Ayam exits.
The Lua functionality may be accessed from the Tcl scripting interface
via the "luaEval"
command. The command can be used either to
directly execute Lua code provided via the commands argument
(Tcl code in bold):
» luaEval {a = 0; a = a + 5.5; tclset("a", a);}
or to execute Lua code from a file:
» luaEval -f scriptfile.lua
Note, that this command is not available in the safe interpreter.
Furthermore, Script object scripts may also be implemented in Lua, provided the first line of the script is a comment that instructs Ayam to use the Lua interpreter:
-- Ayam, use: Lua
a = 0;
...
Note that the Lua scripting context inherits the limitations
of the calling Tcl context.
For example, when running in a Script object,
the following code fails:
tcleval("exit");
because the Tcl command "exit"
is not available in the safe interpreter.
The command will not fail, when the calling context is the main Tcl
interpreter; one can e.g. type into the Ayam console:
» luaEval {tcleval("exit");}
and Ayam quits (see also section:
Safe Interpreter).
This subsection informs about the global functions additionally available in the Ayam Lua interpreter.
Those are converted Tcl commands, "tcleval"
, "tclvar"
,
and "tclset"
.
The functionality of Ayam is accessible from Lua via a larger set of global functions, named as the corresponding Tcl commands. For instance, Ayam objects can be created in Lua using a function call like this:
crtOb("NCircle");
or, with additional arguments:
crtOb("NCircle", "-radius", 3.0);
In general, all commands available in the safe Ayam Tcl interpreter are also
available as converted function (refer to section
Procedures and Commands
for a more or less complete list of those commands).
Note that Tcl procedures are generally not available as global Lua function,
but they can be called using "tcleval"
as documented
in the next paragraph.
tcleval:
Apart from Tcl commands converted to Lua functions, there is the
global Lua function "tcleval"
, that allows to evaluate
arbitrary Tcl scripts, delivered as string argument:
a = 42;
a = tcleval("puts " .. a .. "; return 5;");
tcleval("puts " .. a);
-- expected output: 42 5
The "tcleval"
function provides access to all the
functionality of Ayam that is just available as a Tcl procedure.
Note that return values are properly transferred back to Lua
according to the rules for data conversion as documented below.
However, due to the conversion to string data, the overhead of
such a call is considerable and data transport should be arranged
by other means, see below.
tclvar:
Using the Lua function "tclvar"
a link between a Tcl variable
and a corresponding variable in the Lua context can be established.
The "tclvar"
function creates a write trace on the Tcl
variable, so that changes on the Tcl side are always automatically
reflected on the Lua side:
tclvar("a");
tcleval("set a 42");
tcleval("puts " .. a);
-- expected output: 42
Mind that the corresponding variable on the Lua side does
not exist until the first write operation onto the Tcl
variable occurs.
The Tcl variable, in turn, does not have to exist, when the
"tclvar"
function is called (i.e. all the work is done
in the trace callback). If the variable name contains a namespace
specifier, this namespace has to exist, when "tclvar"
is called.
Even though it looks a perfect fit, "tclvar"
can not be used to
manage a property data array (if the array contains components
to be saved to Ayam scene files). This is, because upon reading a scene
file with such saved array items, the items will be read (and put into
the Tcl context) before the script can establish the write trace using
"tclvar"
and the data from the scene file never arrives in the Lua
context.
There is no easy way to get around this.
A suggested way to manage a property data array is shown in the
complete examples section below.
tclset:
The third global Lua function is "tclset"
that
allows to efficiently set Tcl variables from the Lua context
avoiding conversion to string data and back.
For example:
a = 3.3;
b = {1, 3, 5};
tclset("a", a);
tclset("b", b);
sets the Tcl variable "a"
to the floating point value 3.3,
and "b"
to a list of integer values{ 1 3 5 }
.
Note that the variable names may also point to Tcl array elements:
tclset("SphereAttrData(Radius)", 1.2);
or contain namespace specifiers
tclset("::MyNameSpace::Radius", 1.2);
.
When data is transferred from the Tcl to the Lua side
(e.g. while converting return values of "tcleval"
or variable
values linked via "tclvar"
),
the following conversions are in effect:
Scalar data types will be converted to their directly matching counterparts.
Lists will be converted to array-tables (nesting is allowed and will
produce accordingly nested tables).
Associative arrays will be converted to tables with properly named keys.
Unicode strings are currently not supported.
See also the table below.
Tcl | Lua |
Boolean (true, false) | Boolean (true, false) |
Integer (2) | Integer (2) |
Double (3.14) | Double (3.14) |
String ("mystr") | String ("mystr") |
List ({0 1 2}) | Array ({0, 1, 2}) |
Array (mya(mye) = 0.1) | Table (mya.mye = 0.1) |
When data is transferred from the Lua side to the Tcl side (e.g. as function argument), the following conversions are in effect: Scalar data types will be converted to their directly matching counterparts, array-tables will be converted to lists (nesting is allowed and will produce accordingly nested lists). Sparse and mixed tables are currently not supported. Unicode strings are also currently not supported. See also the following table.
Lua | Tcl |
Boolean (true, false) | Boolean (true, false) |
Integer (2) | Integer (2) |
Double (3.14) | Double (3.14) |
String ("mystr") | String ("mystr") |
Array ({0, 1, 2}) | List ({0 1 2}) |
The transport/conversion of table entries (to e.g. associative array elements) can be arranged manually like this:
a.b = 3.14;
tclset("a(b)", a.b)
This section contains two complete examples for Script objects written in Lua.
For the first example use Script object type "Modify"
and put a Sphere
as child object of the Script object.
-- Ayam, use: Lua tclvar("SphereAttrData") getProp() if SphereAttrData then tclset("SphereAttrData(ZMin)", -SphereAttrData.Radius) tclset("SphereAttrData(ZMax)", SphereAttrData.Radius) setProp() end
First, a link from the original Sphere object property data array
"SphereAttrData"
is established, so that when "getProp()"
(a converted Tcl Ayam command) is called, also the Lua
object "SphereAttrData"
is filled with meaningful data.
The next line (the if) is a safety measure that prevents the script from
failing if the child object of the Script object is not a Sphere
object.
Now the radius value is transferred back to Tcl directly into
the property data array to the ZMin and ZMax entries respectively
with the help of "tclset"
.
Finally the modified property is transferred back to the Sphere object
again with a converted Tcl Ayam command "setProp()"
.
The next example shows, how to manage a property GUI in a Lua
implemented Script object script. Use Script object type "Create"
and add a tag "NP MyProp"
to see the property GUI.
-- Ayam, use: Lua, save array: MyPropData if tcleval("info exists MyPropData;") == 0 then -- initial script run (but not when loaded from scene file!) MyPropData = {} MyPropData.MyItem = tcleval("set MyPropData(MyItem) 1.0;") tcleval("set MyPropData(SP) {MyItem};") else -- all following script runs (and also when loaded from scene file!) MyPropData = {} MyPropData.MyItem = tcleval("set MyPropData(MyItem);") end if tcleval("info exists MyPropGUI;") == 0 then -- create property GUI "MyProp" tcleval("set ::phw [addPropertyGUI MyProp \"\" \"\"];") tcleval("addParam $::phw MyPropData MyItem;") end crtOb("Sphere") sL() getProp() tclset("SphereAttrData(Radius)", MyPropData.MyItem) tclset("SphereAttrData(ZMin)", -MyPropData.MyItem) tclset("SphereAttrData(ZMax)", MyPropData.MyItem) setProp()
This example demonstrates how to manage property data using the
Lua object variable "MyPropData"
. The property data
can be saved to and read from Ayam scene files with the help of
a mirroring array variable on the Tcl side (also named
"MyPropData"
).
To make this work properly, the initialisation of the Lua
object must be constrained to the first script run: when the
property data was read from a scene file, initialisation must
not be run, instead the read data must be fetched from the Tcl
context.
This is what the first "if"
statement, checking for existence
of the mirroring Tcl array variable, in above example is all about.
Following this scheme of dual mirroring data structures on the
Tcl and Lua sides, now the property GUI is created, which is also
constrained to just one script run by a similar "if"
statement.
After the GUI, a Sphere object is created and parameterized according to the data in the property GUI, which is used as radius, zmin, and zmax value.